freebsd-skq/sys/dev/ie/if_ie.c

1811 lines
46 KiB
C
Raw Normal View History

> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
/*-
* Copyright (c) 1992, 1993, University of Vermont and State
* Agricultural College.
* Copyright (c) 1992, 1993, Garrett A. Wollman.
*
* Portions:
* Copyright (c) 1990, 1991, William F. Jolitz
* Copyright (c) 1990, The Regents of the University of California
*
* 3Com 3C507 support:
* Copyright (c) 1993, 1994, Charles M. Hannum
*
* EtherExpress 16 support:
* Copyright (c) 1993, 1994, 1995, Rodney W. Grimes
* Copyright (c) 1997, Aaron C. Smith
*
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the University of
* Vermont and State Agricultural College and Garrett A. Wollman, by
* William F. Jolitz, by the University of California, Berkeley,
* Lawrence Berkeley Laboratory, and their contributors, by
* Charles M. Hannum, by Rodney W. Grimes, and by Aaron C. Smith.
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
* 4. Neither the names of the Universities nor the names of the authors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE UNIVERSITY OR AUTHORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* MAINTAINER: Matthew N. Dodd <winter@jurai.net>
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
/*
* Intel 82586 Ethernet chip
* Register, bit, and structure definitions.
*
* Written by GAW with reference to the Clarkson Packet Driver code for this
* chip written by Russ Nelson and others.
*
1997-06-27 19:36:27 +00:00
* Intel EtherExpress 16 support from if_ix.c, written by Rodney W. Grimes.
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
*/
/*
* The i82586 is a very versatile chip, found in many implementations.
* Programming this chip is mostly the same, but certain details differ
1995-05-30 08:16:23 +00:00
* from card to card. This driver is written so that different cards
* can be automatically detected at run-time.
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
*/
/*
2003-03-29 01:53:13 +00:00
* Mode of operation:
*
* We run the 82586 in a standard Ethernet mode. We keep NFRAMES
* received frame descriptors around for the receiver to use, and
* NRXBUFS associated receive buffer descriptors, both in a circular
* list. Whenever a frame is received, we rotate both lists as
* necessary. (The 586 treats both lists as a simple queue.) We also
* keep a transmit command around so that packets can be sent off
* quickly.
*
* We configure the adapter in AL-LOC = 1 mode, which means that the
* Ethernet/802.3 MAC header is placed at the beginning of the receive
* buffer rather than being split off into various fields in the RFD.
* This also means that we must include this header in the transmit
* buffer as well.
*
* By convention, all transmit commands, and only transmit commands,
* shall have the I (IE_CMD_INTR) bit set in the command. This way,
* when an interrupt arrives at ieintr(), it is immediately possible
* to tell what precisely caused it. ANY OTHER command-sending routines
* should run at splimp(), and should post an acknowledgement to every
* interrupt they generate.
*
* The 82586 has a 24-bit address space internally, and the adaptor's
* memory is located at the top of this region. However, the value
* we are given in configuration is normally the *bottom* of the adaptor
* RAM. So, we must go through a few gyrations to come up with a
* kernel virtual address which represents the actual beginning of the
* 586 address space. First, we autosize the RAM by running through
* several possible sizes and trying to initialize the adapter under
* the assumption that the selected size is correct. Then, knowing
* the correct RAM size, we set up our pointers in the softc `iomem'
* represents the computed base of the 586 address space. `iomembot'
* represents the actual configured base of adapter RAM. Finally,
* `iosize' represents the calculated size of 586 RAM. Then, when
* laying out commands, we use the interval [iomembot, iomembot +
* iosize); to make 24-pointers, we subtract iomem, and to make
* 16-pointers, we subtract iomem and and with 0xffff.
*/
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/eventhandler.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/socket.h>
#include <sys/sockio.h>
#include <sys/syslog.h>
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#include <sys/module.h>
#include <sys/bus.h>
#include <machine/bus.h>
#include <machine/resource.h>
#include <sys/rman.h>
#include <net/ethernet.h>
#include <net/if.h>
#include <net/if_types.h>
#include <net/if_dl.h>
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#include <netinet/in.h>
#include <netinet/if_ether.h>
#include <dev/ic/i82586.h>
#include <dev/ie/if_ievar.h>
#include <dev/ie/if_iereg.h>
#include <dev/ie/if_ie507.h>
#include <dev/ie/if_iee16.h>
#include <i386/isa/elink.h>
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#include <net/bpf.h>
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#ifdef DEBUG
1997-06-27 19:36:27 +00:00
#define IED_RINT 0x01
#define IED_TINT 0x02
#define IED_RNR 0x04
#define IED_CNA 0x08
#define IED_READFRAME 0x10
1998-02-09 06:11:36 +00:00
static int ie_debug = IED_RNR;
1997-06-27 19:36:27 +00:00
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
1997-06-27 19:36:27 +00:00
#define IE_BUF_LEN ETHER_MAX_LEN /* length of transmit buffer */
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
/* XXX this driver uses `volatile' and `caddr_t' to a fault. */
typedef volatile char *v_caddr_t; /* core address, pointer to volatile */
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
/* Forward declaration */
struct ie_softc;
static void ieinit (void *);
static void ieinit_locked (struct ie_softc *);
static void ie_stop (struct ie_softc *);
static int ieioctl (struct ifnet *, u_long, caddr_t);
static void iestart (struct ifnet *);
static void iestart_locked (struct ifnet *);
static __inline void
ee16_interrupt_enable (struct ie_softc *);
static void ee16_eeprom_outbits (struct ie_softc *, int, int);
static void ee16_eeprom_clock (struct ie_softc *, int);
static u_short ee16_read_eeprom (struct ie_softc *, int);
static int ee16_eeprom_inbits (struct ie_softc *);
static __inline void
ie_ack (struct ie_softc *, u_int);
static void iereset (struct ie_softc *);
static void ie_readframe (struct ie_softc *, int);
static void ie_drop_packet_buffer (struct ie_softc *);
static void find_ie_mem_size (struct ie_softc *);
static int command_and_wait (struct ie_softc *,
int, void volatile *, int);
static void run_tdr (struct ie_softc *,
volatile struct ie_tdr_cmd *);
static int ierint (struct ie_softc *);
static int ietint (struct ie_softc *);
static int iernr (struct ie_softc *);
static void start_receiver (struct ie_softc *);
static __inline int
ieget (struct ie_softc *, struct mbuf **);
static v_caddr_t setup_rfa (struct ie_softc *, v_caddr_t);
static int mc_setup (struct ie_softc *);
static void ie_mc_reset (struct ie_softc *);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#ifdef DEBUG
static void print_rbd (volatile struct ie_recv_buf_desc * rbd);
1998-02-09 06:11:36 +00:00
static int in_ierint = 0;
static int in_ietint = 0;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
1995-12-10 13:40:44 +00:00
static const char *ie_hardware_names[] = {
"None",
1997-06-27 19:36:27 +00:00
"StarLAN 10",
"EN100",
"StarLAN Fiber",
"3C507",
"NI5210",
"EtherExpress 16",
"Unknown"
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
};
1995-05-30 08:16:23 +00:00
/*
2003-03-29 01:53:13 +00:00
* sizeof(iscp) == 1+1+2+4 == 8
* sizeof(scb) == 2+2+2+2+2+2+2+2 == 16
* NFRAMES * sizeof(rfd) == NFRAMES*(2+2+2+2+6+6+2+2) == NFRAMES*24 == 384
* sizeof(xmit_cmd) == 2+2+2+2+6+2 == 18
* sizeof(transmit buffer) == 1512
* sizeof(transmit buffer desc) == 8
* -----
* 1946
*
* NRXBUFS * sizeof(rbd) == NRXBUFS*(2+2+4+2+2) == NRXBUFS*12
* NRXBUFS * IE_RBUF_SIZE == NRXBUFS*256
*
* NRXBUFS should be (16384 - 1946) / (256 + 12) == 14438 / 268 == 53
*
* With NRXBUFS == 48, this leaves us 1574 bytes for another command or
* more buffers. Another transmit command would be 18+8+1512 == 1538
* ---just barely fits!
*
* Obviously all these would have to be reduced for smaller memory sizes.
* With a larger memory, it would be possible to roughly double the number
* of both transmit and receive buffers.
*/
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#define NFRAMES 4 /* number of receive frames */
#define NRXBUFS 24 /* number of buffers to allocate */
#define IE_RBUF_SIZE 256 /* size of each buffer, MUST BE POWER OF TWO */
#define NTXBUFS 1 /* number of transmit commands */
#define IE_TBUF_SIZE ETHER_MAX_LEN /* size of transmit buffer */
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#define MK_24(base, ptr) ((caddr_t)((uintptr_t)ptr - (uintptr_t)base))
#define MK_16(base, ptr) ((u_short)(uintptr_t)MK_24(base, ptr))
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
void
ee16_shutdown(struct ie_softc *sc)
{
ee16_reset_586(sc);
outb(PORT(sc) + IEE16_ECTRL, IEE16_RESET_ASIC);
outb(PORT(sc) + IEE16_ECTRL, 0);
}
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
/*
* Taken almost exactly from Bill's if_is.c, then modified beyond recognition.
*/
int
ie_attach(device_t dev)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
struct ie_softc * sc;
struct ifnet * ifp;
size_t allocsize;
int error, factor;
1997-06-27 19:36:27 +00:00
sc = device_get_softc(dev);
ifp = sc->ifp = if_alloc(IFT_ETHER);
if (ifp == NULL) {
device_printf(sc->dev, "can not if_alloc()\n");
return (ENOSPC);
}
sc->dev = dev;
mtx_init(&sc->lock, device_get_nameunit(dev), MTX_NETWORK_LOCK,
MTX_DEF);
1997-06-27 19:36:27 +00:00
/*
* based on the amount of memory we have, allocate our tx and rx
* resources.
*/
factor = rman_get_size(sc->mem_res) / 8192;
sc->nframes = factor * NFRAMES;
sc->nrxbufs = factor * NRXBUFS;
sc->ntxbufs = factor * NTXBUFS;
1997-06-27 19:36:27 +00:00
/*
* Since all of these guys are arrays of pointers, allocate as one
* big chunk and dole out accordingly.
*/
allocsize = sizeof(void *) * (sc->nframes
+ (sc->nrxbufs * 2)
+ (sc->ntxbufs * 3));
sc->rframes = (volatile struct ie_recv_frame_desc **) malloc(allocsize,
1997-06-27 19:36:27 +00:00
M_DEVBUF,
M_NOWAIT);
if (sc->rframes == NULL) {
mtx_destroy(&sc->lock);
return (ENXIO);
}
sc->rbuffs =
(volatile struct ie_recv_buf_desc **)&sc->rframes[sc->nframes];
sc->cbuffs = (volatile u_char **)&sc->rbuffs[sc->nrxbufs];
sc->xmit_cmds =
(volatile struct ie_xmit_cmd **)&sc->cbuffs[sc->nrxbufs];
sc->xmit_buffs =
(volatile struct ie_xmit_buf **)&sc->xmit_cmds[sc->ntxbufs];
sc->xmit_cbuffs = (volatile u_char **)&sc->xmit_buffs[sc->ntxbufs];
if (bootverbose)
device_printf(sc->dev, "hardware type %s, revision %d\n",
ie_hardware_names[sc->hard_type], sc->hard_vers + 1);
ifp->if_softc = sc;
if_initname(ifp, device_get_name(dev), device_get_unit(dev));
ifp->if_flags = IFF_BROADCAST | IFF_SIMPLEX | IFF_MULTICAST;
1997-06-27 19:36:27 +00:00
ifp->if_start = iestart;
ifp->if_ioctl = ieioctl;
ifp->if_init = ieinit;
IFQ_SET_MAXLEN(&ifp->if_snd, ifqmaxlen);
1997-06-27 19:36:27 +00:00
ether_ifattach(ifp, sc->enaddr);
error = bus_setup_intr(dev, sc->irq_res, INTR_TYPE_NET | INTR_MPSAFE,
NULL, ie_intr, sc, &sc->irq_ih);
if (error) {
device_printf(dev, "Unable to register interrupt handler\n");
mtx_destroy(&sc->lock);
return (error);
}
return (0);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
static __inline void
ie_ack(struct ie_softc *sc, u_int mask)
{
sc->scb->ie_command = sc->scb->ie_status & mask;
(*sc->ie_chan_attn) (sc);
}
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
/*
* What to do upon receipt of an interrupt.
*/
void
ie_intr(void *xsc)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
struct ie_softc *sc = (struct ie_softc *)xsc;
u_short status;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
IE_LOCK(sc);
1997-06-27 19:36:27 +00:00
/* Clear the interrupt latch on the 3C507. */
if (sc->hard_type == IE_3C507
&& (inb(PORT(sc) + IE507_CTRL) & EL_CTRL_INTL))
outb(PORT(sc) + IE507_ICTRL, 1);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
/* disable interrupts on the EE16. */
if (sc->hard_type == IE_EE16)
outb(PORT(sc) + IEE16_IRQ, sc->irq_encoded);
status = sc->scb->ie_status;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
loop:
1997-06-27 19:36:27 +00:00
/* Don't ack interrupts which we didn't receive */
ie_ack(sc, IE_ST_WHENCE & status);
1997-06-27 19:36:27 +00:00
if (status & (IE_ST_RECV | IE_ST_RNR)) {
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#ifdef DEBUG
1997-06-27 19:36:27 +00:00
in_ierint++;
if (ie_debug & IED_RINT)
if_printf(sc->ifp, "rint\n");
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
ierint(sc);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#ifdef DEBUG
1997-06-27 19:36:27 +00:00
in_ierint--;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
1997-06-27 19:36:27 +00:00
}
if (status & IE_ST_DONE) {
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#ifdef DEBUG
1997-06-27 19:36:27 +00:00
in_ietint++;
if (ie_debug & IED_TINT)
if_printf(sc->ifp, "tint\n");
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
ietint(sc);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#ifdef DEBUG
1997-06-27 19:36:27 +00:00
in_ietint--;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
1997-06-27 19:36:27 +00:00
}
if (status & IE_ST_RNR) {
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#ifdef DEBUG
1997-06-27 19:36:27 +00:00
if (ie_debug & IED_RNR)
if_printf(sc->ifp, "rnr\n");
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
iernr(sc);
1997-06-27 19:36:27 +00:00
}
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#ifdef DEBUG
if ((status & IE_ST_ALLDONE) && (ie_debug & IED_CNA))
if_printf(sc->ifp, "cna\n");
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
if ((status = sc->scb->ie_status) & IE_ST_WHENCE)
1997-06-27 19:36:27 +00:00
goto loop;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
/* Clear the interrupt latch on the 3C507. */
if (sc->hard_type == IE_3C507)
outb(PORT(sc) + IE507_ICTRL, 1);
1997-06-27 19:36:27 +00:00
/* enable interrupts on the EE16. */
if (sc->hard_type == IE_EE16)
outb(PORT(sc) + IEE16_IRQ, sc->irq_encoded | IEE16_IRQ_ENABLE);
IE_UNLOCK(sc);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Process a received-frame interrupt.
*/
static int
ierint(struct ie_softc *sc)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
1997-06-27 19:36:27 +00:00
int i, status;
static int timesthru = 1024;
i = sc->rfhead;
1997-06-27 19:36:27 +00:00
while (1) {
status = sc->rframes[i]->ie_fd_status;
1997-06-27 19:36:27 +00:00
if ((status & IE_FD_COMPLETE) && (status & IE_FD_OK)) {
sc->ifp->if_ipackets++;
1997-06-27 19:36:27 +00:00
if (!--timesthru) {
sc->ifp->if_ierrors +=
sc->scb->ie_err_crc +
sc->scb->ie_err_align +
sc->scb->ie_err_resource +
sc->scb->ie_err_overrun;
sc->scb->ie_err_crc = 0;
sc->scb->ie_err_align = 0;
sc->scb->ie_err_resource = 0;
sc->scb->ie_err_overrun = 0;
1997-06-27 19:36:27 +00:00
timesthru = 1024;
}
ie_readframe(sc, i);
1997-06-27 19:36:27 +00:00
} else {
if (status & IE_FD_RNR) {
if (!(sc->scb->ie_status & IE_RU_READY)) {
sc->rframes[0]->ie_fd_next =
MK_16(MEM(sc), sc->rbuffs[0]);
sc->scb->ie_recv_list =
MK_16(MEM(sc), sc->rframes[0]);
command_and_wait(sc, IE_RU_START, 0, 0);
1997-06-27 19:36:27 +00:00
}
}
break;
}
i = (i + 1) % sc->nframes;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
1997-06-27 19:36:27 +00:00
return (0);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Process a command-complete interrupt. These are only generated by
1997-06-27 19:36:27 +00:00
* the transmission of frames. This routine is deceptively simple, since
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
* most of the real work is done by iestart().
*/
1997-06-27 19:36:27 +00:00
static int
ietint(struct ie_softc *sc)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
struct ifnet *ifp = sc->ifp;
1997-06-27 19:36:27 +00:00
int status;
int i;
ifp->if_drv_flags &= ~IFF_DRV_OACTIVE;
1997-06-27 19:36:27 +00:00
for (i = 0; i < sc->xmit_count; i++) {
status = sc->xmit_cmds[i]->ie_xmit_status;
1997-06-27 19:36:27 +00:00
if (status & IE_XS_LATECOLL) {
if_printf(ifp, "late collision\n");
ifp->if_collisions++;
ifp->if_oerrors++;
1997-06-27 19:36:27 +00:00
} else if (status & IE_XS_NOCARRIER) {
if_printf(ifp, "no carrier\n");
ifp->if_oerrors++;
1997-06-27 19:36:27 +00:00
} else if (status & IE_XS_LOSTCTS) {
if_printf(ifp, "lost CTS\n");
ifp->if_oerrors++;
1997-06-27 19:36:27 +00:00
} else if (status & IE_XS_UNDERRUN) {
if_printf(ifp, "DMA underrun\n");
ifp->if_oerrors++;
1997-06-27 19:36:27 +00:00
} else if (status & IE_XS_EXCMAX) {
if_printf(ifp, "too many collisions\n");
ifp->if_collisions += 16;
ifp->if_oerrors++;
1997-06-27 19:36:27 +00:00
} else {
ifp->if_opackets++;
ifp->if_collisions += status & IE_XS_MAXCOLL;
1997-06-27 19:36:27 +00:00
}
}
sc->xmit_count = 0;
1997-06-27 19:36:27 +00:00
/*
* If multicast addresses were added or deleted while we were
* transmitting, ie_mc_reset() set the want_mcsetup flag indicating
* that we should do it.
*/
if (sc->want_mcsetup) {
mc_setup(sc);
sc->want_mcsetup = 0;
1997-06-27 19:36:27 +00:00
}
/* Wish I knew why this seems to be necessary... */
sc->xmit_cmds[0]->ie_xmit_status |= IE_STAT_COMPL;
1997-06-27 19:36:27 +00:00
iestart_locked(ifp);
1997-06-27 19:36:27 +00:00
return (0); /* shouldn't be necessary */
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Process a receiver-not-ready interrupt. I believe that we get these
* when there aren't enough buffers to go around. For now (FIXME), we
* just restart the receiver, and hope everything's ok.
*/
1997-06-27 19:36:27 +00:00
static int
iernr(struct ie_softc *sc)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
#ifdef doesnt_work
setup_rfa(sc, (v_caddr_t) sc->rframes[0]);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
sc->scb->ie_recv_list = MK_16(MEM(sc), sc->rframes[0]);
command_and_wait(sc, IE_RU_START, 0, 0);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#else
1997-06-27 19:36:27 +00:00
/* This doesn't work either, but it doesn't hang either. */
command_and_wait(sc, IE_RU_DISABLE, 0, 0); /* just in case */
setup_rfa(sc, (v_caddr_t) sc->rframes[0]); /* ignore cast-qual */
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
sc->scb->ie_recv_list = MK_16(MEM(sc), sc->rframes[0]);
command_and_wait(sc, IE_RU_START, 0, 0); /* was ENABLE */
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
ie_ack(sc, IE_ST_WHENCE);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
sc->ifp->if_ierrors++;
1997-06-27 19:36:27 +00:00
return (0);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Compare two Ether/802 addresses for equality, inlined and
1997-06-27 19:36:27 +00:00
* unrolled for speed. I'd love to have an inline assembler
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
* version of this...
*/
1998-04-15 17:47:40 +00:00
static __inline int
1997-06-27 19:36:27 +00:00
ether_equal(u_char * one, u_char * two)
{
if (one[0] != two[0])
return (0);
if (one[1] != two[1])
return (0);
if (one[2] != two[2])
return (0);
if (one[3] != two[3])
return (0);
if (one[4] != two[4])
return (0);
if (one[5] != two[5])
return (0);
return 1;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Determine quickly whether we should bother reading in this packet.
* This depends on whether BPF and/or bridging is enabled, whether we
* are receiving multicast address, and whether promiscuous mode is enabled.
* We assume that if IFF_PROMISC is set, then *somebody* wants to see
* all incoming packets.
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
*/
1998-04-15 17:47:40 +00:00
static __inline int
check_eh(struct ie_softc *sc, struct ether_header *eh)
1997-06-27 19:36:27 +00:00
{
/* Optimize the common case: normal operation. We've received
either a unicast with our dest or a multicast packet. */
if (sc->promisc == 0) {
int i;
1997-06-27 19:36:27 +00:00
/* If not multicast, it's definitely for us */
if ((eh->ether_dhost[0] & 1) == 0)
1997-06-27 19:36:27 +00:00
return (1);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
/* Accept broadcasts (loose but fast check) */
if (eh->ether_dhost[0] == 0xff)
1997-06-27 19:36:27 +00:00
return (1);
/* Compare against our multicast addresses */
for (i = 0; i < sc->mcast_count; i++) {
1997-06-27 19:36:27 +00:00
if (ether_equal(eh->ether_dhost,
(u_char *)&sc->mcast_addrs[i]))
1997-06-27 19:36:27 +00:00
return (1);
}
return (0);
}
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
/* Always accept packets when in promiscuous mode */
if ((sc->promisc & IFF_PROMISC) != 0)
1997-06-27 19:36:27 +00:00
return (1);
/* Always accept packets directed at us */
if (ether_equal(eh->ether_dhost, IF_LLADDR(sc->ifp)))
1997-06-27 19:36:27 +00:00
return (1);
/* Must have IFF_ALLMULTI but not IFF_PROMISC set. The chip is
actually in promiscuous mode, so discard unicast packets. */
return((eh->ether_dhost[0] & 1) != 0);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* We want to isolate the bits that have meaning... This assumes that
* IE_RBUF_SIZE is an even power of two. If somehow the act_len exceeds
* the size of the buffer, then we are screwed anyway.
*/
1998-04-15 17:47:40 +00:00
static __inline int
ie_buflen(struct ie_softc *sc, int head)
1997-06-27 19:36:27 +00:00
{
return (sc->rbuffs[head]->ie_rbd_actual
1997-06-27 19:36:27 +00:00
& (IE_RBUF_SIZE | (IE_RBUF_SIZE - 1)));
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
1998-04-15 17:47:40 +00:00
static __inline int
ie_packet_len(struct ie_softc *sc)
1997-06-27 19:36:27 +00:00
{
int i;
int head = sc->rbhead;
1997-06-27 19:36:27 +00:00
int acc = 0;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
do {
if (!(sc->rbuffs[sc->rbhead]->ie_rbd_actual & IE_RBD_USED)) {
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#ifdef DEBUG
print_rbd(sc->rbuffs[sc->rbhead]);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
1997-06-27 19:36:27 +00:00
log(LOG_ERR,
"%s: receive descriptors out of sync at %d\n",
sc->ifp->if_xname, sc->rbhead);
iereset(sc);
1997-06-27 19:36:27 +00:00
return (-1);
}
i = sc->rbuffs[head]->ie_rbd_actual & IE_RBD_LAST;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
acc += ie_buflen(sc, head);
head = (head + 1) % sc->nrxbufs;
1997-06-27 19:36:27 +00:00
} while (!i);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
return (acc);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Read data off the interface, and turn it into an mbuf chain.
*
* This code is DRAMATICALLY different from the previous version; this
* version tries to allocate the entire mbuf chain up front, given the
* length of the data available. This enables us to allocate mbuf
* clusters in many situations where before we would have had a long
* chain of partially-full mbufs. This should help to speed up the
* operation considerably. (Provided that it works, of course.)
*/
1998-04-15 17:47:40 +00:00
static __inline int
ieget(struct ie_softc *sc, struct mbuf **mp)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
struct ether_header eh;
1997-06-27 19:36:27 +00:00
struct mbuf *m, *top, **mymp;
int offset;
int totlen, resid;
int thismboff;
int head;
totlen = ie_packet_len(sc);
1997-06-27 19:36:27 +00:00
if (totlen <= 0)
return (-1);
/*
* Snarf the Ethernet header.
*/
bcopy(sc->cbuffs[sc->rbhead], &eh, sizeof(struct ether_header));
1997-06-27 19:36:27 +00:00
/* ignore cast-qual warning here */
/*
* As quickly as possible, check if this packet is for us. If not,
* don't waste a single cycle copying the rest of the packet in.
* This is only a consideration when FILTER is defined; i.e., when
* we are either running BPF or doing multicasting.
*/
if (!check_eh(sc, &eh)) {
ie_drop_packet_buffer(sc);
sc->ifp->if_ierrors--; /* just this case, it's not an
1997-06-27 19:36:27 +00:00
* error
*/
return (-1);
}
MGETHDR(m, M_DONTWAIT, MT_DATA);
if (!m) {
ie_drop_packet_buffer(sc);
/* XXXX if_ierrors++; */
return (-1);
}
*mp = m;
m->m_pkthdr.rcvif = sc->ifp;
m->m_len = MHLEN;
resid = m->m_pkthdr.len = totlen;
top = 0;
1997-06-27 19:36:27 +00:00
mymp = &top;
/*
* This loop goes through and allocates mbufs for all the data we
* will be copying in. It does not actually do the copying yet.
*/
do { /* while(resid > 0) */
/*
* Try to allocate an mbuf to hold the data that we have.
* If we already allocated one, just get another one and
* stick it on the end (eventually). If we don't already
* have one, try to allocate an mbuf cluster big enough to
* hold the whole packet, if we think it's reasonable, or a
* single mbuf which may or may not be big enough. Got that?
*/
if (top) {
MGET(m, M_DONTWAIT, MT_DATA);
1997-06-27 19:36:27 +00:00
if (!m) {
m_freem(top);
ie_drop_packet_buffer(sc);
1997-06-27 19:36:27 +00:00
return (-1);
}
m->m_len = MLEN;
}
if (resid >= MINCLSIZE) {
MCLGET(m, M_DONTWAIT);
1997-06-27 19:36:27 +00:00
if (m->m_flags & M_EXT)
m->m_len = min(resid, MCLBYTES);
} else {
if (resid < m->m_len) {
if (!top && resid + max_linkhdr <= m->m_len)
m->m_data += max_linkhdr;
m->m_len = resid;
}
}
resid -= m->m_len;
*mymp = m;
mymp = &m->m_next;
} while (resid > 0);
resid = totlen; /* remaining data */
offset = 0; /* packet offset */
thismboff = 0; /* offset in m */
m = top; /* current mbuf */
head = sc->rbhead; /* current rx buffer */
1997-06-27 19:36:27 +00:00
/*
* Now we take the mbuf chain (hopefully only one mbuf most of the
* time) and stuff the data into it. There are no possible failures
* at or after this point.
*/
while (resid > 0) { /* while there's stuff left */
int thislen = ie_buflen(sc, head) - offset;
1997-06-27 19:36:27 +00:00
/*
* If too much data for the current mbuf, then fill the
* current one up, go to the next one, and try again.
*/
if (thislen > m->m_len - thismboff) {
int newlen = m->m_len - thismboff;
bcopy((v_caddr_t) (sc->cbuffs[head] + offset),
mtod(m, caddr_t) +thismboff, (unsigned) newlen);
1997-06-27 19:36:27 +00:00
/* ignore cast-qual warning */
m = m->m_next;
thismboff = 0; /* new mbuf, so no offset */
1997-06-27 19:36:27 +00:00
offset += newlen; /* we are now this far into
* the packet */
resid -= newlen; /* so there is this much left
* to get */
continue;
}
/*
* If there is more than enough space in the mbuf to hold
* the contents of this buffer, copy everything in, advance
* pointers, and so on.
*/
if (thislen < m->m_len - thismboff) {
bcopy((v_caddr_t) (sc->cbuffs[head] + offset),
1997-06-27 19:36:27 +00:00
mtod(m, caddr_t) +thismboff, (unsigned) thislen);
thismboff += thislen; /* we are this far into the
* mbuf */
resid -= thislen; /* and this much is left */
goto nextbuf;
}
/*
* Otherwise, there is exactly enough space to put this
* buffer's contents into the current mbuf. Do the
* combination of the above actions.
*/
bcopy((v_caddr_t) (sc->cbuffs[head] + offset),
1997-06-27 19:36:27 +00:00
mtod(m, caddr_t) + thismboff, (unsigned) thislen);
m = m->m_next;
thismboff = 0; /* new mbuf, start at the beginning */
resid -= thislen; /* and we are this far through */
/*
* Advance all the pointers. We can get here from either of
* the last two cases, but never the first.
*/
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
nextbuf:
1997-06-27 19:36:27 +00:00
offset = 0;
sc->rbuffs[head]->ie_rbd_actual = 0;
sc->rbuffs[head]->ie_rbd_length |= IE_RBD_LAST;
sc->rbhead = head = (head + 1) % sc->nrxbufs;
sc->rbuffs[sc->rbtail]->ie_rbd_length &= ~IE_RBD_LAST;
sc->rbtail = (sc->rbtail + 1) % sc->nrxbufs;
1997-06-27 19:36:27 +00:00
}
/*
* Unless something changed strangely while we were doing the copy,
* we have now copied everything in from the shared memory. This
* means that we are done.
*/
return (0);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Read frame NUM from unit UNIT (pre-cached as IE).
*
* This routine reads the RFD at NUM, and copies in the buffers from
* the list of RBD, then rotates the RBD and RFD lists so that the receiver
* doesn't start complaining. Trailers are DROPPED---there's no point
1997-06-27 19:36:27 +00:00
* in wasting time on confusing code to deal with them. Hopefully,
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
* this machine will never ARP for trailers anyway.
*/
1997-06-27 19:36:27 +00:00
static void
ie_readframe(struct ie_softc *sc, int num/* frame number to read */)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
struct ifnet *ifp = sc->ifp;
1997-06-27 19:36:27 +00:00
struct ie_recv_frame_desc rfd;
struct mbuf *m = 0;
#ifdef DEBUG
struct ether_header *eh;
#endif
1997-06-27 19:36:27 +00:00
bcopy((v_caddr_t) (sc->rframes[num]), &rfd,
1997-06-27 19:36:27 +00:00
sizeof(struct ie_recv_frame_desc));
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
/*
* Immediately advance the RFD list, since we we have copied ours
* now.
*/
sc->rframes[num]->ie_fd_status = 0;
sc->rframes[num]->ie_fd_last |= IE_FD_LAST;
sc->rframes[sc->rftail]->ie_fd_last &= ~IE_FD_LAST;
sc->rftail = (sc->rftail + 1) % sc->nframes;
sc->rfhead = (sc->rfhead + 1) % sc->nframes;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
if (rfd.ie_fd_status & IE_FD_OK) {
if (ieget(sc, &m)) {
sc->ifp->if_ierrors++; /* this counts as an
1997-06-27 19:36:27 +00:00
* error */
return;
}
}
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#ifdef DEBUG
eh = mtod(m, struct ether_header *);
1997-06-27 19:36:27 +00:00
if (ie_debug & IED_READFRAME) {
if_printf(ifp, "frame from ether %6D type %x\n",
eh->ether_shost, ":", (unsigned) eh->ether_type);
1997-06-27 19:36:27 +00:00
}
if (ntohs(eh->ether_type) > ETHERTYPE_TRAIL
&& ntohs(eh->ether_type) < (ETHERTYPE_TRAIL + ETHERTYPE_NTRAILER))
1997-06-27 19:36:27 +00:00
printf("received trailer!\n");
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
1997-06-27 19:36:27 +00:00
if (!m)
return;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
/*
* Finally pass this packet up to higher layers.
*/
IE_UNLOCK(sc);
(*ifp->if_input)(ifp, m);
IE_LOCK(sc);
1997-06-27 19:36:27 +00:00
}
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
static void
ie_drop_packet_buffer(struct ie_softc *sc)
1997-06-27 19:36:27 +00:00
{
int i;
do {
/*
* This means we are somehow out of sync. So, we reset the
* adapter.
*/
if (!(sc->rbuffs[sc->rbhead]->ie_rbd_actual & IE_RBD_USED)) {
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#ifdef DEBUG
print_rbd(sc->rbuffs[sc->rbhead]);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
log(LOG_ERR, "%s: receive descriptors out of sync at %d\n",
sc->ifp->if_xname, sc->rbhead);
iereset(sc);
1997-06-27 19:36:27 +00:00
return;
}
i = sc->rbuffs[sc->rbhead]->ie_rbd_actual & IE_RBD_LAST;
1997-06-27 19:36:27 +00:00
sc->rbuffs[sc->rbhead]->ie_rbd_length |= IE_RBD_LAST;
sc->rbuffs[sc->rbhead]->ie_rbd_actual = 0;
sc->rbhead = (sc->rbhead + 1) % sc->nrxbufs;
sc->rbuffs[sc->rbtail]->ie_rbd_length &= ~IE_RBD_LAST;
sc->rbtail = (sc->rbtail + 1) % sc->nrxbufs;
1997-06-27 19:36:27 +00:00
} while (!i);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Start transmission on an interface.
*/
static void
1997-06-27 19:36:27 +00:00
iestart(struct ifnet *ifp)
{
struct ie_softc *sc = ifp->if_softc;
IE_LOCK(sc);
iestart_locked(ifp);
IE_UNLOCK(sc);
}
static void
iestart_locked(struct ifnet *ifp)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
struct ie_softc *sc = ifp->if_softc;
1997-06-27 19:36:27 +00:00
struct mbuf *m0, *m;
volatile unsigned char *buffer;
1997-06-27 19:36:27 +00:00
u_short len;
/*
* This is not really volatile, in this routine, but it makes gcc
* happy.
*/
volatile u_short *bptr = &sc->scb->ie_command_list;
1997-06-27 19:36:27 +00:00
if (!(ifp->if_drv_flags & IFF_DRV_RUNNING))
1997-06-27 19:36:27 +00:00
return;
if (ifp->if_drv_flags & IFF_DRV_OACTIVE)
1997-06-27 19:36:27 +00:00
return;
do {
IF_DEQUEUE(&sc->ifp->if_snd, m);
1997-06-27 19:36:27 +00:00
if (!m)
break;
buffer = sc->xmit_cbuffs[sc->xmit_count];
1997-06-27 19:36:27 +00:00
len = 0;
for (m0 = m; m && len < IE_BUF_LEN; m = m->m_next) {
bcopy(mtod(m, caddr_t), buffer, m->m_len);
buffer += m->m_len;
len += m->m_len;
}
m_freem(m0);
len = max(len, ETHER_MIN_LEN);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
/*
* See if bpf is listening on this interface, let it see the
* packet before we commit it to the wire.
*/
BPF_TAP(sc->ifp,
(void *)sc->xmit_cbuffs[sc->xmit_count], len);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
sc->xmit_buffs[sc->xmit_count]->ie_xmit_flags =
1997-06-27 19:36:27 +00:00
IE_XMIT_LAST|len;
sc->xmit_buffs[sc->xmit_count]->ie_xmit_next = 0xffff;
sc->xmit_buffs[sc->xmit_count]->ie_xmit_buf =
MK_24(sc->iomem, sc->xmit_cbuffs[sc->xmit_count]);
1997-06-27 19:36:27 +00:00
sc->xmit_cmds[sc->xmit_count]->com.ie_cmd_cmd = IE_CMD_XMIT;
sc->xmit_cmds[sc->xmit_count]->ie_xmit_status = 0;
sc->xmit_cmds[sc->xmit_count]->ie_xmit_desc =
MK_16(sc->iomem, sc->xmit_buffs[sc->xmit_count]);
1997-06-27 19:36:27 +00:00
*bptr = MK_16(sc->iomem, sc->xmit_cmds[sc->xmit_count]);
bptr = &sc->xmit_cmds[sc->xmit_count]->com.ie_cmd_link;
sc->xmit_count++;
} while (sc->xmit_count < sc->ntxbufs);
1997-06-27 19:36:27 +00:00
/*
* If we queued up anything for transmission, send it.
*/
if (sc->xmit_count) {
sc->xmit_cmds[sc->xmit_count - 1]->com.ie_cmd_cmd |=
1997-06-27 19:36:27 +00:00
IE_CMD_LAST | IE_CMD_INTR;
/*
* By passing the command pointer as a null, we tell
* command_and_wait() to pretend that this isn't an action
* command. I wish I understood what was happening here.
*/
command_and_wait(sc, IE_CU_START, 0, 0);
ifp->if_drv_flags |= IFF_DRV_OACTIVE;
1997-06-27 19:36:27 +00:00
}
return;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Check to see if there's an 82586 out there.
*/
int
check_ie_present(struct ie_softc *sc)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
1997-06-27 19:36:27 +00:00
volatile struct ie_sys_conf_ptr *scp;
volatile struct ie_int_sys_conf_ptr *iscp;
volatile struct ie_sys_ctl_block *scb;
u_long realbase;
1995-05-30 08:16:23 +00:00
realbase = (uintptr_t) sc->iomembot + sc->iosize - (1 << 24);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
scp = (volatile struct ie_sys_conf_ptr *) (uintptr_t)
(realbase + IE_SCP_ADDR);
bzero((volatile char *) scp, sizeof *scp);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
/*
* First we put the ISCP at the bottom of memory; this tests to make
* sure that our idea of the size of memory is the same as the
* controller's. This is NOT where the ISCP will be in normal
* operation.
*/
iscp = (volatile struct ie_int_sys_conf_ptr *) sc->iomembot;
bzero((volatile char *)iscp, sizeof *iscp);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
scb = (volatile struct ie_sys_ctl_block *) sc->iomembot;
bzero((volatile char *)scb, sizeof *scb);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
scp->ie_bus_use = sc->bus_use; /* 8-bit or 16-bit */
scp->ie_iscp_ptr = (caddr_t) (uintptr_t)
((volatile char *) iscp - (volatile char *) (uintptr_t) realbase);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
iscp->ie_busy = 1;
iscp->ie_scb_offset = MK_16(realbase, scb) + 256;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
(*sc->ie_reset_586) (sc);
(*sc->ie_chan_attn) (sc);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
DELAY(100); /* wait a while... */
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
if (iscp->ie_busy) {
return (0);
}
/*
* Now relocate the ISCP to its real home, and reset the controller
* again.
*/
iscp = (void *) Align((caddr_t) (uintptr_t)
(realbase + IE_SCP_ADDR -
sizeof(struct ie_int_sys_conf_ptr)));
bzero((volatile char *) iscp, sizeof *iscp); /* ignore cast-qual */
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
scp->ie_iscp_ptr = (caddr_t) (uintptr_t)
((volatile char *) iscp - (volatile char *) (uintptr_t) realbase);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
iscp->ie_busy = 1;
iscp->ie_scb_offset = MK_16(realbase, scb);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
(*sc->ie_reset_586) (sc);
(*sc->ie_chan_attn) (sc);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
DELAY(100);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
if (iscp->ie_busy) {
return (0);
}
sc->iomem = (caddr_t) (uintptr_t) realbase;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
sc->iscp = iscp;
sc->scb = scb;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
/*
* Acknowledge any interrupts we may have caused...
*/
ie_ack(sc, IE_ST_WHENCE);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
return (1);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Divine the memory size of ie board UNIT.
* Better hope there's nothing important hiding just below the ie card...
*/
1997-06-27 19:36:27 +00:00
static void
find_ie_mem_size(struct ie_softc *sc)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
1997-06-27 19:36:27 +00:00
unsigned size;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
sc->iosize = 0;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
for (size = 65536; size >= 8192; size -= 8192) {
if (check_ie_present(sc)) {
1997-06-27 19:36:27 +00:00
return;
}
}
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
return;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
void
el_reset_586(struct ie_softc *sc)
{
outb(PORT(sc) + IE507_CTRL, EL_CTRL_RESET);
1997-06-27 19:36:27 +00:00
DELAY(100);
outb(PORT(sc) + IE507_CTRL, EL_CTRL_NORMAL);
1997-06-27 19:36:27 +00:00
DELAY(100);
}
void
sl_reset_586(struct ie_softc *sc)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
outb(PORT(sc) + IEATT_RESET, 0);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
void
ee16_reset_586(struct ie_softc *sc)
{
outb(PORT(sc) + IEE16_ECTRL, IEE16_RESET_586);
DELAY(100);
outb(PORT(sc) + IEE16_ECTRL, 0);
DELAY(100);
}
void
el_chan_attn(struct ie_softc *sc)
{
outb(PORT(sc) + IE507_ATTN, 1);
}
void
sl_chan_attn(struct ie_softc *sc)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
outb(PORT(sc) + IEATT_ATTN, 0);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
void
ee16_chan_attn(struct ie_softc *sc)
{
outb(PORT(sc) + IEE16_ATTN, 0);
}
u_short
1997-06-27 19:36:27 +00:00
ee16_read_eeprom(struct ie_softc *sc, int location)
{
1997-06-27 19:36:27 +00:00
int ectrl, edata;
ectrl = inb(sc->port + IEE16_ECTRL);
ectrl &= IEE16_ECTRL_MASK;
ectrl |= IEE16_ECTRL_EECS;
outb(sc->port + IEE16_ECTRL, ectrl);
ee16_eeprom_outbits(sc, IEE16_EEPROM_READ, IEE16_EEPROM_OPSIZE1);
ee16_eeprom_outbits(sc, location, IEE16_EEPROM_ADDR_SIZE);
edata = ee16_eeprom_inbits(sc);
ectrl = inb(sc->port + IEE16_ECTRL);
ectrl &= ~(IEE16_RESET_ASIC | IEE16_ECTRL_EEDI | IEE16_ECTRL_EECS);
outb(sc->port + IEE16_ECTRL, ectrl);
ee16_eeprom_clock(sc, 1);
ee16_eeprom_clock(sc, 0);
return edata;
}
static void
1997-06-27 19:36:27 +00:00
ee16_eeprom_outbits(struct ie_softc *sc, int edata, int count)
{
1997-06-27 19:36:27 +00:00
int ectrl, i;
ectrl = inb(sc->port + IEE16_ECTRL);
ectrl &= ~IEE16_RESET_ASIC;
for (i = count - 1; i >= 0; i--) {
ectrl &= ~IEE16_ECTRL_EEDI;
if (edata & (1 << i)) {
ectrl |= IEE16_ECTRL_EEDI;
}
outb(sc->port + IEE16_ECTRL, ectrl);
DELAY(1); /* eeprom data must be setup for 0.4 uSec */
ee16_eeprom_clock(sc, 1);
ee16_eeprom_clock(sc, 0);
}
ectrl &= ~IEE16_ECTRL_EEDI;
outb(sc->port + IEE16_ECTRL, ectrl);
DELAY(1); /* eeprom data must be held for 0.4 uSec */
}
static int
1997-06-27 19:36:27 +00:00
ee16_eeprom_inbits(struct ie_softc *sc)
{
1997-06-27 19:36:27 +00:00
int ectrl, edata, i;
ectrl = inb(sc->port + IEE16_ECTRL);
ectrl &= ~IEE16_RESET_ASIC;
for (edata = 0, i = 0; i < 16; i++) {
edata = edata << 1;
ee16_eeprom_clock(sc, 1);
ectrl = inb(sc->port + IEE16_ECTRL);
if (ectrl & IEE16_ECTRL_EEDO) {
edata |= 1;
}
ee16_eeprom_clock(sc, 0);
}
return (edata);
}
static void
1997-06-27 19:36:27 +00:00
ee16_eeprom_clock(struct ie_softc *sc, int state)
{
1997-06-27 19:36:27 +00:00
int ectrl;
ectrl = inb(sc->port + IEE16_ECTRL);
ectrl &= ~(IEE16_RESET_ASIC | IEE16_ECTRL_EESK);
if (state) {
ectrl |= IEE16_ECTRL_EESK;
}
outb(sc->port + IEE16_ECTRL, ectrl);
DELAY(9); /* EESK must be stable for 8.38 uSec */
}
1998-04-15 17:47:40 +00:00
static __inline void
1997-06-27 19:36:27 +00:00
ee16_interrupt_enable(struct ie_softc *sc)
{
DELAY(100);
outb(sc->port + IEE16_IRQ, sc->irq_encoded | IEE16_IRQ_ENABLE);
DELAY(100);
}
void
sl_read_ether(struct ie_softc *sc, unsigned char *addr)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
1997-06-27 19:36:27 +00:00
int i;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
for (i = 0; i < 6; i++)
addr[i] = inb(PORT(sc) + i);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
static void
iereset(struct ie_softc *sc)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
struct ifnet *ifp = sc->ifp;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
if_printf(ifp, "reset\n");
ie_stop(sc);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
/*
* Stop i82586 dead in its tracks.
*/
if (command_and_wait(sc, IE_RU_ABORT | IE_CU_ABORT, 0, 0))
if_printf(ifp, "abort commands timed out\n");
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
if (command_and_wait(sc, IE_RU_DISABLE | IE_CU_STOP, 0, 0))
if_printf(ifp, "disable commands timed out\n");
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#ifdef notdef
if (!check_ie_present(sc))
1997-06-27 19:36:27 +00:00
panic("ie disappeared!");
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
#endif
if (ifp->if_flags & IFF_UP)
ieinit_locked(sc);
1995-05-30 08:16:23 +00:00
1997-06-27 19:36:27 +00:00
return;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Send a command to the controller and wait for it to either
* complete or be accepted, depending on the command. If the
* command pointer is null, then pretend that the command is
* not an action command. If the command pointer is not null,
* and the command is an action command, wait for
* ((volatile struct ie_cmd_common *)pcmd)->ie_cmd_status & MASK
* to become true.
*/
1997-06-27 19:36:27 +00:00
static int
command_and_wait(struct ie_softc *sc, int cmd, volatile void *pcmd, int mask)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
1997-06-27 19:36:27 +00:00
volatile struct ie_cmd_common *cc = pcmd;
int i;
1997-06-27 19:36:27 +00:00
sc->scb->ie_command = (u_short) cmd;
1997-06-27 19:36:27 +00:00
if (IE_ACTION_COMMAND(cmd) && pcmd) {
(*sc->ie_chan_attn) (sc);
1997-06-27 19:36:27 +00:00
/*
* Now spin-lock waiting for status. This is not a very
* nice thing to do, but I haven't figured out how, or
* indeed if, we can put the process waiting for action to
* sleep. (We may be getting called through some other
* timeout running in the kernel.)
*
* According to the packet driver, the minimum timeout
* should be .369 seconds, which we round up to .37.
1997-06-27 19:36:27 +00:00
*/
for (i = 0; i < 370; i++) {
if (cc->ie_cmd_status & mask)
return (0);
DELAY(1000);
1997-06-27 19:36:27 +00:00
}
return (1);
1997-06-27 19:36:27 +00:00
} else {
/*
* Otherwise, just wait for the command to be accepted.
*/
(*sc->ie_chan_attn) (sc);
1997-06-27 19:36:27 +00:00
while (sc->scb->ie_command); /* spin lock */
1997-06-27 19:36:27 +00:00
return (0);
}
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Run the time-domain reflectometer...
*/
1997-06-27 19:36:27 +00:00
static void
run_tdr(struct ie_softc *sc, volatile struct ie_tdr_cmd *cmd)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
1997-06-27 19:36:27 +00:00
int result;
cmd->com.ie_cmd_status = 0;
cmd->com.ie_cmd_cmd = IE_CMD_TDR | IE_CMD_LAST;
cmd->com.ie_cmd_link = 0xffff;
cmd->ie_tdr_time = 0;
sc->scb->ie_command_list = MK_16(MEM(sc), cmd);
1997-06-27 19:36:27 +00:00
cmd->ie_tdr_time = 0;
if (command_and_wait(sc, IE_CU_START, cmd, IE_STAT_COMPL))
1997-06-27 19:36:27 +00:00
result = 0x2000;
else
result = cmd->ie_tdr_time;
ie_ack(sc, IE_ST_WHENCE);
1997-06-27 19:36:27 +00:00
if (result & IE_TDR_SUCCESS)
return;
if (result & IE_TDR_XCVR) {
if_printf(sc->ifp, "transceiver problem\n");
1997-06-27 19:36:27 +00:00
} else if (result & IE_TDR_OPEN) {
if_printf(sc->ifp, "TDR detected an open %d clocks away\n",
1997-06-27 19:36:27 +00:00
result & IE_TDR_TIME);
} else if (result & IE_TDR_SHORT) {
if_printf(sc->ifp, "TDR detected a short %d clocks away\n",
1997-06-27 19:36:27 +00:00
result & IE_TDR_TIME);
} else {
if_printf(sc->ifp, "TDR returned unknown status %x\n", result);
1997-06-27 19:36:27 +00:00
}
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
1997-06-27 19:36:27 +00:00
static void
start_receiver(struct ie_softc *sc)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
sc->scb->ie_recv_list = MK_16(MEM(sc), sc->rframes[0]);
command_and_wait(sc, IE_RU_START, 0, 0);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
ie_ack(sc, IE_ST_WHENCE);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Here is a helper routine for iernr() and ieinit(). This sets up
* the RFA.
*/
static v_caddr_t
setup_rfa(struct ie_softc *sc, v_caddr_t ptr)
1997-06-27 19:36:27 +00:00
{
volatile struct ie_recv_frame_desc *rfd = (volatile void *)ptr;
1997-06-27 19:36:27 +00:00
volatile struct ie_recv_buf_desc *rbd;
int i;
/* First lay them out */
for (i = 0; i < sc->nframes; i++) {
sc->rframes[i] = rfd;
bzero((volatile char *) rfd, sizeof *rfd); /* ignore cast-qual */
1997-06-27 19:36:27 +00:00
rfd++;
}
ptr = Alignvol(rfd); /* ignore cast-qual */
1997-06-27 19:36:27 +00:00
/* Now link them together */
for (i = 0; i < sc->nframes; i++) {
sc->rframes[i]->ie_fd_next =
MK_16(MEM(sc), sc->rframes[(i + 1) % sc->nframes]);
1997-06-27 19:36:27 +00:00
}
/* Finally, set the EOL bit on the last one. */
sc->rframes[sc->nframes - 1]->ie_fd_last |= IE_FD_LAST;
1997-06-27 19:36:27 +00:00
/*
* Now lay out some buffers for the incoming frames. Note that we
* set aside a bit of slop in each buffer, to make sure that we have
* enough space to hold a single frame in every buffer.
*/
rbd = (volatile void *) ptr;
1997-06-27 19:36:27 +00:00
for (i = 0; i < sc->nrxbufs; i++) {
sc->rbuffs[i] = rbd;
bzero((volatile char *)rbd, sizeof *rbd);
ptr = Alignvol(ptr + sizeof *rbd);
1997-06-27 19:36:27 +00:00
rbd->ie_rbd_length = IE_RBUF_SIZE;
rbd->ie_rbd_buffer = MK_24(MEM(sc), ptr);
sc->cbuffs[i] = (volatile void *) ptr;
1997-06-27 19:36:27 +00:00
ptr += IE_RBUF_SIZE;
rbd = (volatile void *) ptr;
1997-06-27 19:36:27 +00:00
}
/* Now link them together */
for (i = 0; i < sc->nrxbufs; i++) {
sc->rbuffs[i]->ie_rbd_next =
MK_16(MEM(sc), sc->rbuffs[(i + 1) % sc->nrxbufs]);
1997-06-27 19:36:27 +00:00
}
/* Tag EOF on the last one */
sc->rbuffs[sc->nrxbufs - 1]->ie_rbd_length |= IE_RBD_LAST;
1997-06-27 19:36:27 +00:00
/*
* We use the head and tail pointers on receive to keep track of the
* order in which RFDs and RBDs are used.
*/
sc->rfhead = 0;
sc->rftail = sc->nframes - 1;
sc->rbhead = 0;
sc->rbtail = sc->nrxbufs - 1;
1997-06-27 19:36:27 +00:00
sc->scb->ie_recv_list = MK_16(MEM(sc), sc->rframes[0]);
sc->rframes[0]->ie_fd_buf_desc = MK_16(MEM(sc), sc->rbuffs[0]);
1997-06-27 19:36:27 +00:00
ptr = Alignvol(ptr);
1997-06-27 19:36:27 +00:00
return (ptr);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* Run the multicast setup command.
*/
1997-06-27 19:36:27 +00:00
static int
mc_setup(struct ie_softc *sc)
1997-06-27 19:36:27 +00:00
{
volatile struct ie_mcast_cmd *cmd = (volatile void *)sc->xmit_cbuffs[0];
1997-06-27 19:36:27 +00:00
cmd->com.ie_cmd_status = 0;
cmd->com.ie_cmd_cmd = IE_CMD_MCAST | IE_CMD_LAST;
cmd->com.ie_cmd_link = 0xffff;
/* ignore cast-qual */
bcopy((v_caddr_t) sc->mcast_addrs, (v_caddr_t) cmd->ie_mcast_addrs,
sc->mcast_count * sizeof *sc->mcast_addrs);
1997-06-27 19:36:27 +00:00
cmd->ie_mcast_bytes = sc->mcast_count * 6; /* grrr... */
1997-06-27 19:36:27 +00:00
sc->scb->ie_command_list = MK_16(MEM(sc), cmd);
if (command_and_wait(sc, IE_CU_START, cmd, IE_STAT_COMPL)
1997-06-27 19:36:27 +00:00
|| !(cmd->com.ie_cmd_status & IE_STAT_OK)) {
if_printf(sc->ifp, "multicast address setup command failed\n");
1997-06-27 19:36:27 +00:00
return (0);
}
return (1);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
/*
* This routine takes the environment generated by check_ie_present()
* and adds to it all the other structures we need to operate the adapter.
* This includes executing the CONFIGURE, IA-SETUP, and MC-SETUP commands,
* starting the receiver unit, and clearing interrupts.
*/
static void
ieinit(xsc)
void *xsc;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
struct ie_softc *sc = xsc;
IE_LOCK(sc);
ieinit_locked(sc);
IE_UNLOCK(sc);
}
static void
ieinit_locked(struct ie_softc *sc)
{
struct ifnet *ifp = sc->ifp;
volatile struct ie_sys_ctl_block *scb = sc->scb;
caddr_t ptr;
1997-06-27 19:36:27 +00:00
int i;
ptr = Alignvol((volatile char *) scb + sizeof *scb);
1997-06-27 19:36:27 +00:00
/*
* Send the configure command first.
*/
{
volatile struct ie_config_cmd *cmd = (volatile void *) ptr;
1997-06-27 19:36:27 +00:00
ie_setup_config(cmd, sc->promisc,
sc->hard_type == IE_STARLAN10);
1997-06-27 19:36:27 +00:00
cmd->com.ie_cmd_status = 0;
cmd->com.ie_cmd_cmd = IE_CMD_CONFIG | IE_CMD_LAST;
cmd->com.ie_cmd_link = 0xffff;
scb->ie_command_list = MK_16(MEM(sc), cmd);
1997-06-27 19:36:27 +00:00
if (command_and_wait(sc, IE_CU_START, cmd, IE_STAT_COMPL)
1997-06-27 19:36:27 +00:00
|| !(cmd->com.ie_cmd_status & IE_STAT_OK)) {
if_printf(ifp, "configure command failed\n");
1997-06-27 19:36:27 +00:00
return;
}
}
/*
* Now send the Individual Address Setup command.
*/
{
volatile struct ie_iasetup_cmd *cmd = (volatile void *) ptr;
1997-06-27 19:36:27 +00:00
cmd->com.ie_cmd_status = 0;
cmd->com.ie_cmd_cmd = IE_CMD_IASETUP | IE_CMD_LAST;
cmd->com.ie_cmd_link = 0xffff;
bcopy((volatile char *)IF_LLADDR(ifp),
(volatile char *)&cmd->ie_address, sizeof cmd->ie_address);
scb->ie_command_list = MK_16(MEM(sc), cmd);
if (command_and_wait(sc, IE_CU_START, cmd, IE_STAT_COMPL)
1997-06-27 19:36:27 +00:00
|| !(cmd->com.ie_cmd_status & IE_STAT_OK)) {
if_printf(ifp, "individual address "
"setup command failed\n");
1997-06-27 19:36:27 +00:00
return;
}
}
/*
* Now run the time-domain reflectometer.
*/
run_tdr(sc, (volatile void *) ptr);
1997-06-27 19:36:27 +00:00
/*
* Acknowledge any interrupts we have generated thus far.
*/
ie_ack(sc, IE_ST_WHENCE);
1997-06-27 19:36:27 +00:00
/*
* Set up the RFA.
*/
ptr = setup_rfa(sc, ptr);
1997-06-27 19:36:27 +00:00
/*
* Finally, the transmit command and buffer are the last little bit
* of work.
*/
/* transmit command buffers */
for (i = 0; i < sc->ntxbufs; i++) {
sc->xmit_cmds[i] = (volatile void *) ptr;
ptr += sizeof *sc->xmit_cmds[i];
ptr = Alignvol(ptr);
sc->xmit_buffs[i] = (volatile void *)ptr;
ptr += sizeof *sc->xmit_buffs[i];
ptr = Alignvol(ptr);
1997-06-27 19:36:27 +00:00
}
/* transmit buffers */
for (i = 0; i < sc->ntxbufs - 1; i++) {
sc->xmit_cbuffs[i] = (volatile void *)ptr;
1997-06-27 19:36:27 +00:00
ptr += IE_BUF_LEN;
ptr = Alignvol(ptr);
1997-06-27 19:36:27 +00:00
}
sc->xmit_cbuffs[sc->ntxbufs - 1] = (volatile void *) ptr;
1997-06-27 19:36:27 +00:00
for (i = 1; i < sc->ntxbufs; i++) {
bzero((v_caddr_t) sc->xmit_cmds[i], sizeof *sc->xmit_cmds[i]);
bzero((v_caddr_t) sc->xmit_buffs[i], sizeof *sc->xmit_buffs[i]);
1997-06-27 19:36:27 +00:00
}
/*
* This must be coordinated with iestart() and ietint().
*/
sc->xmit_cmds[0]->ie_xmit_status = IE_STAT_COMPL;
1997-06-27 19:36:27 +00:00
/* take the ee16 out of loopback */
if (sc->hard_type == IE_EE16) {
1997-06-27 19:36:27 +00:00
u_int8_t bart_config;
bart_config = inb(PORT(sc) + IEE16_CONFIG);
1997-06-27 19:36:27 +00:00
bart_config &= ~IEE16_BART_LOOPBACK;
/* inb doesn't get bit! */
bart_config |= IEE16_BART_MCS16_TEST;
outb(PORT(sc) + IEE16_CONFIG, bart_config);
ee16_interrupt_enable(sc);
ee16_chan_attn(sc);
1997-06-27 19:36:27 +00:00
}
ifp->if_drv_flags |= IFF_DRV_RUNNING; /* tell higher levels
1997-06-27 19:36:27 +00:00
* we're here */
ifp->if_drv_flags &= ~IFF_DRV_OACTIVE;
start_receiver(sc);
1997-06-27 19:36:27 +00:00
return;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
1997-06-27 19:36:27 +00:00
static void
ie_stop(struct ie_softc *sc)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
struct ifnet *ifp = sc->ifp;
ifp->if_drv_flags &= ~(IFF_DRV_RUNNING | IFF_DRV_OACTIVE);
command_and_wait(sc, IE_RU_DISABLE, 0, 0);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
static int
ieioctl(struct ifnet *ifp, u_long command, caddr_t data)
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
{
int error = 0;
struct ie_softc *sc = ifp->if_softc;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
1997-06-27 19:36:27 +00:00
switch (command) {
case SIOCSIFFLAGS:
/*
* Note that this device doesn't have an "all multicast"
* mode, so we must turn on promiscuous mode and do the
* filtering manually.
*/
IE_LOCK(sc);
1997-06-27 19:36:27 +00:00
if ((ifp->if_flags & IFF_UP) == 0 &&
(ifp->if_drv_flags & IFF_DRV_RUNNING)) {
ie_stop(sc);
1997-06-27 19:36:27 +00:00
} else if ((ifp->if_flags & IFF_UP) &&
(ifp->if_drv_flags & IFF_DRV_RUNNING) == 0) {
sc->promisc =
1997-06-27 19:36:27 +00:00
ifp->if_flags & (IFF_PROMISC | IFF_ALLMULTI);
ieinit_locked(sc);
} else if (sc->promisc ^
1997-06-27 19:36:27 +00:00
(ifp->if_flags & (IFF_PROMISC | IFF_ALLMULTI))) {
sc->promisc =
1997-06-27 19:36:27 +00:00
ifp->if_flags & (IFF_PROMISC | IFF_ALLMULTI);
ieinit_locked(sc);
1997-06-27 19:36:27 +00:00
}
IE_UNLOCK(sc);
1997-06-27 19:36:27 +00:00
break;
case SIOCADDMULTI:
case SIOCDELMULTI:
/*
* Update multicast listeners
*/
/* reset multicast filtering */
IE_LOCK(sc);
ie_mc_reset(sc);
IE_UNLOCK(sc);
1997-06-27 19:36:27 +00:00
error = 0;
break;
default:
error = ether_ioctl(ifp, command, data);
break;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
1997-06-27 19:36:27 +00:00
return (error);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
1997-06-27 19:36:27 +00:00
static void
ie_mc_reset(struct ie_softc *sc)
1997-06-27 19:36:27 +00:00
{
struct ifmultiaddr *ifma;
/*
* Step through the list of addresses.
*/
sc->mcast_count = 0;
if_maddr_rlock(sc->ifp);
TAILQ_FOREACH(ifma, &sc->ifp->if_multiaddrs, ifma_link) {
1997-06-27 19:36:27 +00:00
if (ifma->ifma_addr->sa_family != AF_LINK)
continue;
/* XXX - this is broken... */
if (sc->mcast_count >= MAXMCAST) {
sc->ifp->if_flags |= IFF_ALLMULTI;
if (sc->ifp->if_flags & IFF_UP)
ieinit_locked(sc);
1997-06-27 19:36:27 +00:00
goto setflag;
}
bcopy(LLADDR((struct sockaddr_dl *) ifma->ifma_addr),
&(sc->mcast_addrs[sc->mcast_count]), 6);
sc->mcast_count++;
1997-06-27 19:36:27 +00:00
}
if_maddr_runlock(sc->ifp);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
setflag:
sc->want_mcsetup = 1;
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
#ifdef DEBUG
1998-02-09 06:11:36 +00:00
static void
1997-06-27 19:36:27 +00:00
print_rbd(volatile struct ie_recv_buf_desc * rbd)
{
printf("RBD at %p:\n"
"actual %04x, next %04x, buffer %p\n"
1997-06-27 19:36:27 +00:00
"length %04x, mbz %04x\n",
(volatile void *) rbd,
rbd->ie_rbd_actual, rbd->ie_rbd_next,
(void *) rbd->ie_rbd_buffer,
1997-06-27 19:36:27 +00:00
rbd->ie_rbd_length, rbd->mbz);
> Gcc2 generates memcpy calls when doing some structure copying in the > code. As this memcpy generation is done by the compiler it doesn't > work to have memcpy a macro. Instead I made it a static function that > calls bcopy. Okay, I tracked down the d*mn thing that is causing the memcpy to be emmitted by gcc2. In the following patch is a fix to eliminate the one place that was causing gcc to do the memcpy in function ie_readframe for the ie_recv_frame_desc rfd = *(ie->rframes[num]) structure copy. I also fixed the code so that the NS (XNS) support would compile with out error and removed the cruft #define ns_* stuff from the softc structures (I did this a while back for all the other either drivers) Removed the $Log$ and all of the old log stuff, that is all kept in the cvs tree, no need to duplicate it in the file... I'll copy it into my commit message so that it is all retained, except the revision numbers are going to get screwed up. It now compiles cleanly under my LINT kernel... > I have not yet tested this on the AT&T fiver cards yet. But, Wollman > has used it on various other AT&T cards. I'm pretty sure Terry also > uses this driver. $Log$ stuff from Garretts original file * $Log$ * Revision 1.32 1993/06/20 15:56:15 wollman * Fixed warning message. * * Revision 1.31 93/06/20 15:52:28 wollman * More compilation lint. * * Revision 1.30 93/06/20 15:48:16 wollman * Commented all the rest of the cast-qual warnings. Might be * fixed someday by changing the prototypes of bcopy(), bzero(), etc. * * Revision 1.29 93/06/20 15:40:13 wollman * Started fixing some of the cast-qual warnings (most just commented). * * Revision 1.28 93/06/12 20:39:14 wollman * Changed definition of MK_24 to silence -Wcast-qual. * * Revision 1.27 93/06/08 17:05:57 wollman * Attach routine now returns void. * * Revision 1.26 93/06/08 16:31:21 wollman * Deleted attempt to set watchdog timer, since we don't use it. * * Revision 1.25 93/04/16 22:15:31 wollman * Fixed infinite loop in iestart() which would cause the transmitter * to constantly step on the Ethernet, sending the same two packets over * and over again. This crashes the router (really!) and makes other * network users very angry. * * Revision 1.24 93/04/14 23:57:47 wollman * Increased transmit buffer size from 1504 to 1512, just in case I don't * understand how the MTU is used. * * Revision 1.23 93/04/14 23:16:19 wollman * Fix iestart() so that it only tries to execute a command if one was * queued up (oops...). * * Revision 1.22 93/04/14 23:02:16 wollman * Unknown changes. * * Revision 1.21 93/04/12 18:29:33 wollman * Added multiple transmit commands. Note that they are always queued in * pairs (there are only two to begin with), and the I flag is only set * on the second one. This should decrease interrupt activity somewhat. * * Revision 1.20 93/04/12 15:36:05 wollman * Changed setup of config block to take into account difference between * StarLAN (10BASE-T) and Ethernet: StarLAN is Manchester, Ethernet is NRZ. * * Revision 1.19 93/03/23 18:58:44 wollman * Raised count of receive buffers from 32 to 48 and included analysis of * memory usage in a comment near that definition. * * Revision 1.18 93/03/18 03:59:53 wollman * Last try to get large TCP windows to work... * * Revision 1.17 93/03/18 03:54:10 wollman * Added RNR processing as a part of receive interrupts. Hopefully this * will make everything Just Work now. * * Revision 1.16 93/03/18 03:23:24 wollman * Still having problems with RNR conditions. Basic problem: say we * set TCP window to 16k, which is a semsible value. Then smart * partners will blast 16k to us, of which we can fit 8k minus a bit * into our buffers. (Need to look at that memory map..) So, we get * an RNR interrupt. The problem is, what do we need to do in this case? * The original code doesn't work, but it doesn't hang the system either. * * Revision 1.15 93/03/18 02:37:00 wollman * Disabled watchdog timer. This gets called from a timeout somewhere, * which is running up at softclock. This is at a higher priority than * we are, so we can get interrupted in the process of receiving something * by the timer expiration, which is not a nice thing to have happen. * Question: how can we keep this functionality, but do it in sync with * our interrupts? (Ugh!) * * Revision 1.14 93/03/18 02:35:46 wollman * Fixed some more problems with receive code---it works! * * Revision 1.13 93/03/17 20:15:59 wollman * Think we fixed the serious (showstopper) bugs in 1.12. * * Revision 1.12 93/03/15 17:45:39 wollman * Changed filtering logic dramatically to avoid the unnecessary copyin * for BPF or multicast routing when the packet is not for us. (We * have to run in promiscuous mode in this case.) We still need to * deal with the problem of allocating one mbuf per receive buffer; * we should be able to get away with a single mbuf for all reasonable * incoming packets, and using an mbuf cluster will help avoid unnecessary * copying of data. * * Revision 1.11 93/01/30 20:23:15 wollman * Fiddled with reset code to hopefully make it work better when the * watchdog timer expires. Question: why does it make a difference that * we do the same thing through ieioctl() as we were doing before. The * old code, when iereset() was called, the receive unit would completely * die. But, a cycle of ifconfig ie0 down ; ifconfig ie0 up' would bring * things back to normal. So, we try to mimic that now. Grrr... * * Revision 1.10 93/01/29 17:26:50 wollman * Finally fixed multicast routing/BPF mb_map full problem!!!! * It turned out that, when we received a packet not for us, we * would just return from ie_readframe(), *WITHOUT* freeing the * mbuf that the packet was in. So, when running in promiscuous * mode (i.e., multicast routing or BPF without -p), we would have * (on our extremely busy subnet 4) a memory leak of epic proportions. * * We now hold on to a single mbuf chain allocated in this way, * and the next time a packet is read, we free it. Also, the * next time a packet is transmitted, we free it. This way, * we never lose more than 6 mbufs total (and on average more like * 3). * * Revision 1.9 93/01/23 14:22:21 wollman * Added some StarLAN Fiber modifications from L. Jonas Olsson. * * Revision 1.8 93/01/23 14:02:34 wollman * Added support for IP multicasting. * * Revision 1.7 92/11/21 17:51:59 wollman * Added StarLAN Fiber version point to my driver. * * Revision 1.6 92/11/20 18:21:45 wollman * Fixed documentation, made out-of-mbuf condition an error. * * Revision 1.5 92/11/20 18:14:02 wollman * Move change of ethertype to host byte order so that it's after * bpf_mtap. Hopefully, this will allow us to listen to incoming * IP packets. * * Revision 1.4 92/11/20 18:11:00 wollman * Added code to drop packets when we run out of mbufs rather than * panicking. This should alleviate the problems observed when using tcpdump. * * Revision 1.3 92/11/14 14:12:40 wollman * Added BPF support code. * * Revision 1.2 92/11/14 13:30:26 wollman * Added RCS keywords *
1993-10-12 06:52:34 +00:00
}
1997-06-27 19:36:27 +00:00
#endif /* DEBUG */
int
ie_alloc_resources (device_t dev)
{
struct ie_softc * sc;
int error;
error = 0;
sc = device_get_softc(dev);
sc->io_res = bus_alloc_resource_any(dev, SYS_RES_IOPORT, &sc->io_rid,
RF_ACTIVE);
if (!sc->io_res) {
device_printf(dev, "No I/O space?!\n");
error = ENOMEM;
goto bad;
}
sc->io_bt = rman_get_bustag(sc->io_res);
sc->io_bh = rman_get_bushandle(sc->io_res);
sc->mem_res = bus_alloc_resource_any(dev, SYS_RES_MEMORY, &sc->mem_rid,
RF_ACTIVE);
if (!sc->mem_res) {
device_printf(dev, "No Memory!\n");
error = ENOMEM;
goto bad;
}
sc->mem_bt = rman_get_bustag(sc->mem_res);
sc->mem_bh = rman_get_bushandle(sc->mem_res);
sc->irq_res = bus_alloc_resource_any(dev, SYS_RES_IRQ, &sc->irq_rid,
RF_ACTIVE);
if (!sc->irq_res) {
device_printf(dev, "No IRQ!\n");
error = ENOMEM;
goto bad;
}
sc->port = rman_get_start(sc->io_res); /* XXX hack */
sc->iomembot = rman_get_virtual(sc->mem_res);
sc->iosize = rman_get_size(sc->mem_res);
return (0);
bad:
return (error);
}
void
ie_release_resources (device_t dev)
{
struct ie_softc * sc;
sc = device_get_softc(dev);
if (sc->irq_ih)
bus_teardown_intr(dev, sc->irq_res, sc->irq_ih);
if (sc->rframes)
free(sc->rframes, M_DEVBUF);
if (sc->io_res)
bus_release_resource(dev, SYS_RES_IOPORT,
sc->io_rid, sc->io_res);
if (sc->irq_res)
bus_release_resource(dev, SYS_RES_IRQ,
sc->irq_rid, sc->irq_res);
if (sc->mem_res)
bus_release_resource(dev, SYS_RES_MEMORY,
sc->mem_rid, sc->mem_res);
if (sc->ifp)
if_free(sc->ifp);
return;
}
int
ie_detach (device_t dev)
{
struct ie_softc * sc;
struct ifnet * ifp;
sc = device_get_softc(dev);
ifp = sc->ifp;
IE_LOCK(sc);
if (sc->hard_type == IE_EE16)
ee16_shutdown(sc);
ie_stop(sc);
IE_UNLOCK(sc);
ether_ifdetach(ifp);
ie_release_resources(dev);
mtx_destroy(&sc->lock);
return (0);
}