d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
a494f83f3f
freebsd-skq/sys/dev/pdq/pdq_ifsubr.c

776 lines
19 KiB
C
Raw Normal View History

- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
/* $NetBSD: pdq_ifsubr.c,v 1.38 2001/12/21 23:21:47 matt Exp $ */
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
/*-
* Copyright (c) 1995, 1996 Matt Thomas <matt@3am-software.com>
* 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. The name of the author may not be used to endorse or promote products
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
* derived from this software without specific prior written permission
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
*
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
* $NetBSD: pdq_ifsubr.c,v 1.12 1997/06/05 01:56:35 thomas Exp$
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
*/
Use __FBSDID(). Also some minor style cleanups.
2003-08-24 17:55:58 +00:00
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
/*
* DEC PDQ FDDI Controller; code for BSD derived operating systems
*
* This module provide bus independent BSD specific O/S functions.
* (ie. it provides an ifnet interface to the rest of the system)
*/
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#define PDQ_OSSUPPORT
Make INET a proper option. This will not make any of object files that LINT create change; there might be differences with INET disabled, but hardly anything compiled before without INET anyway. Now the 'obvious' things will give a proper error if compiled without inet - ipx_ip, ipfw, tcp_debug. The only thing that _should_ work (but can't be made to compile reasonably easily) is sppp :-( This commit move struct arpcom from <netinet/if_ether.h> to <net/if_arp.h>.
1998-01-08 23:42:31 +00:00
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
#include <sys/param.h>
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/lock.h>
#include <sys/mutex.h>
#include <sys/malloc.h>
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
#include <sys/socket.h>
Don't include <sys/ioctl.h> in the kernel. Stage 2: include <sys/sockio.h> instead of <sys/ioctl.h> in network files.
1997-03-24 11:33:46 +00:00
#include <sys/sockio.h>
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#include <sys/module.h>
#include <sys/bus.h>
#include <machine/bus.h>
#include <machine/resource.h>
#include <sys/rman.h>
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
#include <net/if.h>
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#include <net/if_arp.h>
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
#include <net/if_dl.h>
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#include <net/if_media.h>
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
#include <net/if_types.h>
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#include <net/fddi.h>
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
#include <net/bpf.h>
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#include <dev/pdq/pdq_freebsd.h>
This mega-merge brings Matt Thomas' 960801 FDDI driver (almost) up to -current. Thanks goes to Ulrike Nitzsche <ulrike@ifw-dresden.de> for giving me a chance to test this. Only the PCI driver is tested though. One final patch will follow in a separate commit. This is so that everything up to here can be dragged into 2.2, if we decide so. Reviewed by: joerg Submitted by: Matt Thomas <matt@3am-software.com>
1997-01-17 23:54:45 +00:00
#include <dev/pdq/pdqreg.h>
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
devclass_t pdq_devclass;
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
static void pdq_watchdog(void *);
static void
pdq_ifstop(pdq_softc_t *sc)
{
PDQ_IFNET(sc)->if_drv_flags &= ~(IFF_DRV_RUNNING | IFF_DRV_OACTIVE);
sc->sc_pdq->pdq_flags &= ~PDQ_RUNNING;
pdq_stop(sc->sc_pdq);
callout_stop(&sc->watchdog);
}
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
static void
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
pdq_ifinit_locked(pdq_softc_t *sc)
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
{
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
PDQ_LOCK_ASSERT(sc);
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
if (PDQ_IFNET(sc)->if_flags & IFF_UP) {
Propagate rename of IFF_OACTIVE and IFF_RUNNING to IFF_DRV_OACTIVE and IFF_DRV_RUNNING, as well as the move from ifnet.if_flags to ifnet.if_drv_flags. Device drivers are now responsible for synchronizing access to these flags, as they are in if_drv_flags. This helps prevent races between the network stack and device driver in maintaining the interface flags field. Many __FreeBSD__ and __FreeBSD_version checks maintained and continued; some less so. Reviewed by: pjd, bz MFC after: 7 days
2005-08-09 10:20:02 +00:00
PDQ_IFNET(sc)->if_drv_flags |= IFF_DRV_RUNNING;
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
if (PDQ_IFNET(sc)->if_flags & IFF_PROMISC) {
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
sc->sc_pdq->pdq_flags |= PDQ_PROMISC;
} else {
sc->sc_pdq->pdq_flags &= ~PDQ_PROMISC;
}
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
if (PDQ_IFNET(sc)->if_flags & IFF_LINK1) {
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
sc->sc_pdq->pdq_flags |= PDQ_PASS_SMT;
} else {
sc->sc_pdq->pdq_flags &= ~PDQ_PASS_SMT;
}
sc->sc_pdq->pdq_flags |= PDQ_RUNNING;
pdq_run(sc->sc_pdq);
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
callout_reset(&sc->watchdog, hz, pdq_watchdog, sc);
} else
pdq_ifstop(sc);
}
static void
pdq_ifinit(void *arg)
{
pdq_softc_t *sc;
sc = arg;
PDQ_LOCK(sc);
pdq_ifinit_locked(sc);
PDQ_UNLOCK(sc);
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
}
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
static void
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
pdq_watchdog(void *arg)
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
{
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
pdq_softc_t *sc;
struct ifnet *ifp;
sc = arg;
PDQ_LOCK_ASSERT(sc);
callout_reset(&sc->watchdog, hz, pdq_watchdog, sc);
if (sc->timer == 0 || --sc->timer > 0)
return;
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
/*
* No progress was made on the transmit queue for PDQ_OS_TX_TRANSMIT
* seconds. Remove all queued packets.
*/
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
ifp = PDQ_IFNET(sc);
Propagate rename of IFF_OACTIVE and IFF_RUNNING to IFF_DRV_OACTIVE and IFF_DRV_RUNNING, as well as the move from ifnet.if_flags to ifnet.if_drv_flags. Device drivers are now responsible for synchronizing access to these flags, as they are in if_drv_flags. This helps prevent races between the network stack and device driver in maintaining the interface flags field. Many __FreeBSD__ and __FreeBSD_version checks maintained and continued; some less so. Reviewed by: pjd, bz MFC after: 7 days
2005-08-09 10:20:02 +00:00
ifp->if_drv_flags &= ~IFF_DRV_OACTIVE;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
for (;;) {
struct mbuf *m;
IFQ_DEQUEUE(&ifp->if_snd, m);
if (m == NULL)
return;
PDQ_OS_DATABUF_FREE(PDQ_OS_IFP_TO_SOFTC(ifp)->sc_pdq, m);
}
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
}
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
static void
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
pdq_ifstart_locked(struct ifnet *ifp)
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
{
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
pdq_softc_t * const sc = PDQ_OS_IFP_TO_SOFTC(ifp);
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
struct mbuf *m;
int tx = 0;
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
PDQ_LOCK_ASSERT(sc);
Propagate rename of IFF_OACTIVE and IFF_RUNNING to IFF_DRV_OACTIVE and IFF_DRV_RUNNING, as well as the move from ifnet.if_flags to ifnet.if_drv_flags. Device drivers are now responsible for synchronizing access to these flags, as they are in if_drv_flags. This helps prevent races between the network stack and device driver in maintaining the interface flags field. Many __FreeBSD__ and __FreeBSD_version checks maintained and continued; some less so. Reviewed by: pjd, bz MFC after: 7 days
2005-08-09 10:20:02 +00:00
if ((ifp->if_drv_flags & IFF_DRV_RUNNING) == 0)
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
return;
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
if (sc->timer == 0)
sc->timer = PDQ_OS_TX_TIMEOUT;
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
if ((sc->sc_pdq->pdq_flags & PDQ_TXOK) == 0) {
Propagate rename of IFF_OACTIVE and IFF_RUNNING to IFF_DRV_OACTIVE and IFF_DRV_RUNNING, as well as the move from ifnet.if_flags to ifnet.if_drv_flags. Device drivers are now responsible for synchronizing access to these flags, as they are in if_drv_flags. This helps prevent races between the network stack and device driver in maintaining the interface flags field. Many __FreeBSD__ and __FreeBSD_version checks maintained and continued; some less so. Reviewed by: pjd, bz MFC after: 7 days
2005-08-09 10:20:02 +00:00
PDQ_IFNET(sc)->if_drv_flags |= IFF_DRV_OACTIVE;
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
return;
}
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
sc->sc_flags |= PDQIF_DOWNCALL;
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
for (;; tx = 1) {
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
IF_DEQUEUE(&ifp->if_snd, m);
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
if (m == NULL)
break;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#if defined(PDQ_BUS_DMA) && !defined(PDQ_BUS_DMA_NOTX)
if ((m->m_flags & M_HASTXDMAMAP) == 0) {
bus_dmamap_t map;
if (PDQ_OS_HDR_OFFSET != PDQ_RX_FC_OFFSET) {
m->m_data[0] = PDQ_FDDI_PH0;
m->m_data[1] = PDQ_FDDI_PH1;
m->m_data[2] = PDQ_FDDI_PH2;
}
if (!bus_dmamap_create(sc->sc_dmatag, m->m_pkthdr.len, 255,
m->m_pkthdr.len, 0, BUS_DMA_NOWAIT, &map)) {
if (!bus_dmamap_load_mbuf(sc->sc_dmatag, map, m,
BUS_DMA_WRITE|BUS_DMA_NOWAIT)) {
bus_dmamap_sync(sc->sc_dmatag, map, 0, m->m_pkthdr.len,
BUS_DMASYNC_PREWRITE);
M_SETCTX(m, map);
m->m_flags |= M_HASTXDMAMAP;
}
}
if ((m->m_flags & M_HASTXDMAMAP) == 0)
break;
}
#else
if (PDQ_OS_HDR_OFFSET != PDQ_RX_FC_OFFSET) {
m->m_data[0] = PDQ_FDDI_PH0;
m->m_data[1] = PDQ_FDDI_PH1;
m->m_data[2] = PDQ_FDDI_PH2;
}
#endif
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
if (pdq_queue_transmit_data(sc->sc_pdq, m) == PDQ_FALSE)
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
break;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
}
if (m != NULL) {
Propagate rename of IFF_OACTIVE and IFF_RUNNING to IFF_DRV_OACTIVE and IFF_DRV_RUNNING, as well as the move from ifnet.if_flags to ifnet.if_drv_flags. Device drivers are now responsible for synchronizing access to these flags, as they are in if_drv_flags. This helps prevent races between the network stack and device driver in maintaining the interface flags field. Many __FreeBSD__ and __FreeBSD_version checks maintained and continued; some less so. Reviewed by: pjd, bz MFC after: 7 days
2005-08-09 10:20:02 +00:00
ifp->if_drv_flags |= IFF_DRV_OACTIVE;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
IF_PREPEND(&ifp->if_snd, m);
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
}
if (tx)
PDQ_DO_TYPE2_PRODUCER(sc->sc_pdq);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
sc->sc_flags &= ~PDQIF_DOWNCALL;
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
}
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
static void
pdq_ifstart(struct ifnet *ifp)
{
pdq_softc_t * const sc = PDQ_OS_IFP_TO_SOFTC(ifp);
PDQ_LOCK(sc);
pdq_ifstart_locked(ifp);
PDQ_UNLOCK(sc);
}
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
void
pdq_os_receive_pdu(
pdq_t *pdq,
struct mbuf *m,
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
size_t pktlen,
int drop)
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
{
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
pdq_softc_t *sc = pdq->pdq_os_ctx;
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
struct ifnet *ifp = PDQ_IFNET(sc);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
struct fddi_header *fh;
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
network interface driver changes: o don't strip the Ethernet header from inbound packets; pass packets up the stack intact (required significant changes to some drivers) o reference common definitions in net/ethernet.h (e.g. ETHER_ALIGN) o track ether_ifattach/ether_ifdetach API changes o track bpf changes (use BPF_TAP and BPF_MTAP) o track vlan changes (ifnet capabilities, revised processing scheme, etc.) o use if_input to pass packets "up" o call ether_ioctl for default handling of ioctls Reviewed by: many Approved by: re
2002-11-14 23:54:55 +00:00
ifp->if_ipackets++;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#if defined(PDQ_BUS_DMA)
{
/*
* Even though the first mbuf start at the first fddi header octet,
* the dmamap starts PDQ_OS_HDR_OFFSET octets earlier. Any additional
* mbufs will start normally.
*/
int offset = PDQ_OS_HDR_OFFSET;
struct mbuf *m0;
for (m0 = m; m0 != NULL; m0 = m0->m_next, offset = 0) {
pdq_os_databuf_sync(sc, m0, offset, m0->m_len, BUS_DMASYNC_POSTREAD);
bus_dmamap_unload(sc->sc_dmatag, M_GETCTX(m0, bus_dmamap_t));
bus_dmamap_destroy(sc->sc_dmatag, M_GETCTX(m0, bus_dmamap_t));
m0->m_flags &= ~M_HASRXDMAMAP;
M_SETCTX(m0, NULL);
}
}
#endif
m->m_pkthdr.len = pktlen;
fh = mtod(m, struct fddi_header *);
if (drop || (fh->fddi_fc & (FDDIFC_L|FDDIFC_F)) != FDDIFC_LLC_ASYNC) {
network interface driver changes: o don't strip the Ethernet header from inbound packets; pass packets up the stack intact (required significant changes to some drivers) o reference common definitions in net/ethernet.h (e.g. ETHER_ALIGN) o track ether_ifattach/ether_ifdetach API changes o track bpf changes (use BPF_TAP and BPF_MTAP) o track vlan changes (ifnet capabilities, revised processing scheme, etc.) o use if_input to pass packets "up" o call ether_ioctl for default handling of ioctls Reviewed by: many Approved by: re
2002-11-14 23:54:55 +00:00
ifp->if_iqdrops++;
ifp->if_ierrors++;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
PDQ_OS_DATABUF_FREE(pdq, m);
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
return;
}
network interface driver changes: o don't strip the Ethernet header from inbound packets; pass packets up the stack intact (required significant changes to some drivers) o reference common definitions in net/ethernet.h (e.g. ETHER_ALIGN) o track ether_ifattach/ether_ifdetach API changes o track bpf changes (use BPF_TAP and BPF_MTAP) o track vlan changes (ifnet capabilities, revised processing scheme, etc.) o use if_input to pass packets "up" o call ether_ioctl for default handling of ioctls Reviewed by: many Approved by: re
2002-11-14 23:54:55 +00:00
m->m_pkthdr.rcvif = ifp;
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
PDQ_UNLOCK(sc);
network interface driver changes: o don't strip the Ethernet header from inbound packets; pass packets up the stack intact (required significant changes to some drivers) o reference common definitions in net/ethernet.h (e.g. ETHER_ALIGN) o track ether_ifattach/ether_ifdetach API changes o track bpf changes (use BPF_TAP and BPF_MTAP) o track vlan changes (ifnet capabilities, revised processing scheme, etc.) o use if_input to pass packets "up" o call ether_ioctl for default handling of ioctls Reviewed by: many Approved by: re
2002-11-14 23:54:55 +00:00
(*ifp->if_input)(ifp, m);
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
PDQ_LOCK(sc);
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
}
void
pdq_os_restart_transmitter(
pdq_t *pdq)
{
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
pdq_softc_t *sc = pdq->pdq_os_ctx;
Propagate rename of IFF_OACTIVE and IFF_RUNNING to IFF_DRV_OACTIVE and IFF_DRV_RUNNING, as well as the move from ifnet.if_flags to ifnet.if_drv_flags. Device drivers are now responsible for synchronizing access to these flags, as they are in if_drv_flags. This helps prevent races between the network stack and device driver in maintaining the interface flags field. Many __FreeBSD__ and __FreeBSD_version checks maintained and continued; some less so. Reviewed by: pjd, bz MFC after: 7 days
2005-08-09 10:20:02 +00:00
PDQ_IFNET(sc)->if_drv_flags &= ~IFF_DRV_OACTIVE;
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
if (IFQ_IS_EMPTY(&PDQ_IFNET(sc)->if_snd) == 0) {
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
sc->timer = PDQ_OS_TX_TIMEOUT;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
if ((sc->sc_flags & PDQIF_DOWNCALL) == 0)
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
pdq_ifstart_locked(PDQ_IFNET(sc));
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
} else {
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
sc->timer = 0;
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
}
}
void
pdq_os_transmit_done(
pdq_t *pdq,
struct mbuf *m)
{
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
pdq_softc_t *sc = pdq->pdq_os_ctx;
Fix preprocessor code to check for a symbol being defined prior to checking for non-zero.
2007-06-08 01:49:04 +00:00
#if defined(NBPFILTER) && NBPFILTER > 0
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
if (PQD_IFNET(sc)->if_bpf != NULL)
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
PDQ_BPF_MTAP(sc, m);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#endif
PDQ_OS_DATABUF_FREE(pdq, m);
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
PDQ_IFNET(sc)->if_opackets++;
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
}
void
pdq_os_addr_fill(
pdq_t *pdq,
pdq_lanaddr_t *addr,
size_t num_addrs)
{
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
pdq_softc_t *sc = pdq->pdq_os_ctx;
struct ifnet *ifp;
Catch up with the moving target: merge the change from rev 1.19 of the now dead sys/pci/if_pdq.c which has been committed about by the same time i made my tests with Matt's code. LINT should compile now again. Well, that's a clear case where ``CVS writer locks'' would certainly (not) have helped. :-]
1997-01-18 13:03:21 +00:00
struct ifmultiaddr *ifma;
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
ifp = sc->ifp;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
/*
* ADDR_FILTER_SET is always issued before FILTER_SET so
* we can play with PDQ_ALLMULTI and not worry about
* queueing a FILTER_SET ourselves.
*/
pdq->pdq_flags &= ~PDQ_ALLMULTI;
#if defined(IFF_ALLMULTI)
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
PDQ_IFNET(sc)->if_flags &= ~IFF_ALLMULTI;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#endif
Use if_maddr_rlock()/if_maddr_runlock() rather than IF_ADDR_LOCK()/ IF_ADDR_UNLOCK() across network device drivers when accessing the per-interface multicast address list, if_multiaddrs. This will allow us to change the locking strategy without affecting our driver programming interface or binary interface. For two wireless drivers, remove unnecessary locking, since they don't actually access the multicast address list. Approved by: re (kib) MFC after: 6 weeks
2009-06-26 11:45:06 +00:00
if_maddr_rlock(PDQ_IFNET(sc));
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
for (ifma = TAILQ_FIRST(&PDQ_IFNET(sc)->if_multiaddrs); ifma && num_addrs > 0;
Convert if_multiaddrs from LIST to TAILQ so that it can be traversed backwards in the three drivers which want to do that. Reviewed by: mikeh
2001-02-06 10:12:15 +00:00
ifma = TAILQ_NEXT(ifma, ifma_link)) {
Catch up with the moving target: merge the change from rev 1.19 of the now dead sys/pci/if_pdq.c which has been committed about by the same time i made my tests with Matt's code. LINT should compile now again. Well, that's a clear case where ``CVS writer locks'' would certainly (not) have helped. :-]
1997-01-18 13:03:21 +00:00
char *mcaddr;
if (ifma->ifma_addr->sa_family != AF_LINK)
continue;
mcaddr = LLADDR((struct sockaddr_dl *)ifma->ifma_addr);
((u_short *) addr->lanaddr_bytes)[0] = ((u_short *) mcaddr)[0];
((u_short *) addr->lanaddr_bytes)[1] = ((u_short *) mcaddr)[1];
((u_short *) addr->lanaddr_bytes)[2] = ((u_short *) mcaddr)[2];
addr++;
num_addrs--;
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
}
Use if_maddr_rlock()/if_maddr_runlock() rather than IF_ADDR_LOCK()/ IF_ADDR_UNLOCK() across network device drivers when accessing the per-interface multicast address list, if_multiaddrs. This will allow us to change the locking strategy without affecting our driver programming interface or binary interface. For two wireless drivers, remove unnecessary locking, since they don't actually access the multicast address list. Approved by: re (kib) MFC after: 6 weeks
2009-06-26 11:45:06 +00:00
if_maddr_runlock(PDQ_IFNET(sc));
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
/*
* If not all the address fit into the CAM, turn on all-multicast mode.
*/
if (ifma != NULL) {
pdq->pdq_flags |= PDQ_ALLMULTI;
#if defined(IFF_ALLMULTI)
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
PDQ_IFNET(sc)->if_flags |= IFF_ALLMULTI;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#endif
}
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
}
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#if defined(IFM_FDDI)
static int
pdq_ifmedia_change(
struct ifnet *ifp)
{
pdq_softc_t * const sc = PDQ_OS_IFP_TO_SOFTC(ifp);
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
PDQ_LOCK(sc);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
if (sc->sc_ifmedia.ifm_media & IFM_FDX) {
if ((sc->sc_pdq->pdq_flags & PDQ_WANT_FDX) == 0) {
sc->sc_pdq->pdq_flags |= PDQ_WANT_FDX;
if (sc->sc_pdq->pdq_flags & PDQ_RUNNING)
pdq_run(sc->sc_pdq);
}
} else if (sc->sc_pdq->pdq_flags & PDQ_WANT_FDX) {
sc->sc_pdq->pdq_flags &= ~PDQ_WANT_FDX;
if (sc->sc_pdq->pdq_flags & PDQ_RUNNING)
pdq_run(sc->sc_pdq);
}
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
PDQ_UNLOCK(sc);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
return 0;
}
static void
pdq_ifmedia_status(
struct ifnet *ifp,
struct ifmediareq *ifmr)
{
pdq_softc_t * const sc = PDQ_OS_IFP_TO_SOFTC(ifp);
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
PDQ_LOCK(sc);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
ifmr->ifm_status = IFM_AVALID;
if (sc->sc_pdq->pdq_flags & PDQ_IS_ONRING)
ifmr->ifm_status |= IFM_ACTIVE;
ifmr->ifm_active = (ifmr->ifm_current & ~IFM_FDX);
if (sc->sc_pdq->pdq_flags & PDQ_IS_FDX)
ifmr->ifm_active |= IFM_FDX;
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
PDQ_UNLOCK(sc);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
}
void
pdq_os_update_status(
pdq_t *pdq,
const void *arg)
{
pdq_softc_t * const sc = pdq->pdq_os_ctx;
const pdq_response_status_chars_get_t *rsp = arg;
int media = 0;
switch (rsp->status_chars_get.pmd_type[0]) {
case PDQ_PMD_TYPE_ANSI_MUTLI_MODE: media = IFM_FDDI_MMF; break;
case PDQ_PMD_TYPE_ANSI_SINGLE_MODE_TYPE_1: media = IFM_FDDI_SMF; break;
case PDQ_PMD_TYPE_ANSI_SIGNLE_MODE_TYPE_2: media = IFM_FDDI_SMF; break;
case PDQ_PMD_TYPE_UNSHIELDED_TWISTED_PAIR: media = IFM_FDDI_UTP; break;
default: media |= IFM_MANUAL;
}
if (rsp->status_chars_get.station_type == PDQ_STATION_TYPE_DAS)
media |= IFM_FDDI_DA;
sc->sc_ifmedia.ifm_media = media | IFM_FDDI;
}
#endif /* defined(IFM_FDDI) */
static int
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
pdq_ifioctl(
struct ifnet *ifp,
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
u_long cmd,
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
caddr_t data)
{
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
pdq_softc_t *sc = PDQ_OS_IFP_TO_SOFTC(ifp);
int error = 0;
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
switch (cmd) {
case SIOCSIFFLAGS: {
pdq_ifinit(sc);
break;
}
case SIOCADDMULTI:
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
case SIOCDELMULTI: {
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
PDQ_LOCK(sc);
Propagate rename of IFF_OACTIVE and IFF_RUNNING to IFF_DRV_OACTIVE and IFF_DRV_RUNNING, as well as the move from ifnet.if_flags to ifnet.if_drv_flags. Device drivers are now responsible for synchronizing access to these flags, as they are in if_drv_flags. This helps prevent races between the network stack and device driver in maintaining the interface flags field. Many __FreeBSD__ and __FreeBSD_version checks maintained and continued; some less so. Reviewed by: pjd, bz MFC after: 7 days
2005-08-09 10:20:02 +00:00
if (PDQ_IFNET(sc)->if_drv_flags & IFF_DRV_RUNNING) {
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
pdq_run(sc->sc_pdq);
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
error = 0;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
}
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
PDQ_UNLOCK(sc);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
break;
}
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#if defined(IFM_FDDI) && defined(SIOCSIFMEDIA)
case SIOCSIFMEDIA:
case SIOCGIFMEDIA: {
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
struct ifreq *ifr = (struct ifreq *)data;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
error = ifmedia_ioctl(ifp, ifr, &sc->sc_ifmedia, cmd);
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
break;
}
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#endif
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
default: {
Offer unhandled IOCTLS to fddi_ioctl().
2005-03-24 01:58:20 +00:00
error = fddi_ioctl(ifp, cmd, data);
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
break;
}
}
return error;
}
#ifndef IFF_NOTRAILERS
#define IFF_NOTRAILERS 0
#endif
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
int
pdq_ifattach(pdq_softc_t *sc, const pdq_uint8_t *llc, pdq_type_t type)
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
{
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
struct ifnet *ifp;
ifp = PDQ_IFNET(sc) = if_alloc(IFT_FDDI);
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
if (ifp == NULL) {
device_printf(sc->dev, "can not if_alloc()\n");
return (ENOSPC);
}
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
Change callers of mtx_init() to pass in an appropriate lock type name. In most cases NULL is passed, but in some cases such as network driver locks (which use the MTX_NETWORK_LOCK macro) and UMA zone locks, a name is used. Tested on: i386, alpha, sparc64
2002-04-04 21:03:38 +00:00
mtx_init(&sc->mtx, device_get_nameunit(sc->dev), MTX_NETWORK_LOCK,
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
MTX_DEF);
callout_init_mtx(&sc->watchdog, &sc->mtx, 0);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
if_initname(ifp, device_get_name(sc->dev), device_get_unit(sc->dev));
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
ifp->if_softc = sc;
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
ifp->if_init = pdq_ifinit;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
ifp->if_snd.ifq_maxlen = IFQ_MAXLEN;
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
ifp->if_flags = IFF_BROADCAST|IFF_SIMPLEX|IFF_NOTRAILERS|IFF_MULTICAST;
ifp->if_ioctl = pdq_ifioctl;
ifp->if_start = pdq_ifstart;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
#if defined(IFM_FDDI)
{
const int media = sc->sc_ifmedia.ifm_media;
ifmedia_init(&sc->sc_ifmedia, IFM_FDX,
pdq_ifmedia_change, pdq_ifmedia_status);
ifmedia_add(&sc->sc_ifmedia, media, 0, 0);
ifmedia_set(&sc->sc_ifmedia, media);
}
#endif
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
sc->sc_pdq = pdq_initialize(sc->mem_bst, sc->mem_bsh, ifp->if_xname, -1,
sc, type);
if (sc->sc_pdq == NULL) {
device_printf(sc->dev, "Initialization failed.\n");
return (ENXIO);
}
- Make IFP2ENADDR() a pointer to IF_LLADDR() rather than another copy of Ethernet address. - Change iso88025_ifattach() and fddi_ifattach() to accept MAC address as an argument, similar to ether_ifattach(), to make this work.
2005-11-11 07:36:14 +00:00
fddi_ifattach(ifp, llc, FDDI_BPF_SUPPORTED);
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
return (0);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
}
void
pdq_ifdetach (pdq_softc_t *sc)
{
struct ifnet *ifp;
Stop embedding struct ifnet at the top of driver softcs. Instead the struct ifnet or the layer 2 common structure it was embedded in have been replaced with a struct ifnet pointer to be filled by a call to the new function, if_alloc(). The layer 2 common structure is also allocated via if_alloc() based on the interface type. It is hung off the new struct ifnet member, if_l2com. This change removes the size of these structures from the kernel ABI and will allow us to better manage them as interfaces come and go. Other changes of note: - Struct arpcom is no longer referenced in normal interface code. Instead the Ethernet address is accessed via the IFP2ENADDR() macro. To enforce this ac_enaddr has been renamed to _ac_enaddr. - The second argument to ether_ifattach is now always the mac address from driver private storage rather than sometimes being ac_enaddr. Reviewed by: sobomax, sam
2005-06-10 16:49:24 +00:00
ifp = sc->ifp;
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
fddi_ifdetach(ifp, FDDI_BPF_SUPPORTED);
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
PDQ_LOCK(sc);
pdq_ifstop(sc);
PDQ_UNLOCK(sc);
callout_drain(&sc->watchdog);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
pdq_free(sc->dev);
return;
}
void
pdq_free (device_t dev)
{
pdq_softc_t *sc;
sc = device_get_softc(dev);
if (sc->io)
bus_release_resource(dev, sc->io_type, sc->io_rid, sc->io);
if (sc->mem)
bus_release_resource(dev, sc->mem_type, sc->mem_rid, sc->mem);
if (sc->irq_ih)
bus_teardown_intr(dev, sc->irq, sc->irq_ih);
if (sc->irq)
bus_release_resource(dev, SYS_RES_IRQ, sc->irq_rid, sc->irq);
Several fixes to these drivers. Note that these two drivers are actually just two different attachments (EISA and PCI) to a single driver. - Add real locking. Previously these drivers only acquired their lock in their interrupt handler or in the ioctl routine (but too broadly in the latter). No locking was used for the stack calling down into the driver via if_init() or if_start(), for device shutdown or detach. Also, the interrupt handler held the driver lock while calling if_input(). All this stuff should be fixed in the locking changes. - Really fix these drivers to handle if_alloc(). The front-end attachments were using if_initname() before the ifnet was allocated. Fix this by moving some of the duplicated logic from each driver into pdq_ifattach(). While here, make pdq_ifattach() return an error so that the driver just fails to attach if if_alloc() fails rather than panic'ing. Also, defer freeing the ifnet until the driver has stopped using it during detach. - Add a new private timer to drive the watchdog timer. - Pass the softc pointer to the interrupt handlers instead of the device_t so we can avoid the use of device_get_softc() and to better match what other drivers do.
2009-11-19 19:25:47 +00:00
if (sc->ifp)
if_free(sc->ifp);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
/*
* Destroy the mutex.
*/
if (mtx_initialized(&sc->mtx) != 0) {
mtx_destroy(&sc->mtx);
}
return;
}
#if defined(PDQ_BUS_DMA)
int
pdq_os_memalloc_contig(
pdq_t *pdq)
{
pdq_softc_t * const sc = pdq->pdq_os_ctx;
bus_dma_segment_t db_segs[1], ui_segs[1], cb_segs[1];
int db_nsegs = 0, ui_nsegs = 0;
int steps = 0;
int not_ok;
not_ok = bus_dmamem_alloc(sc->sc_dmatag,
sizeof(*pdq->pdq_dbp), sizeof(*pdq->pdq_dbp),
sizeof(*pdq->pdq_dbp), db_segs, 1, &db_nsegs,
BUS_DMA_NOWAIT);
if (!not_ok) {
steps = 1;
not_ok = bus_dmamem_map(sc->sc_dmatag, db_segs, db_nsegs,
sizeof(*pdq->pdq_dbp), (caddr_t *) &pdq->pdq_dbp,
BUS_DMA_NOWAIT);
}
if (!not_ok) {
steps = 2;
not_ok = bus_dmamap_create(sc->sc_dmatag, db_segs[0].ds_len, 1,
0x2000, 0, BUS_DMA_NOWAIT, &sc->sc_dbmap);
}
if (!not_ok) {
steps = 3;
not_ok = bus_dmamap_load(sc->sc_dmatag, sc->sc_dbmap,
pdq->pdq_dbp, sizeof(*pdq->pdq_dbp),
NULL, BUS_DMA_NOWAIT);
}
if (!not_ok) {
steps = 4;
pdq->pdq_pa_descriptor_block = sc->sc_dbmap->dm_segs[0].ds_addr;
not_ok = bus_dmamem_alloc(sc->sc_dmatag,
PDQ_OS_PAGESIZE, PDQ_OS_PAGESIZE, PDQ_OS_PAGESIZE,
ui_segs, 1, &ui_nsegs, BUS_DMA_NOWAIT);
}
if (!not_ok) {
steps = 5;
not_ok = bus_dmamem_map(sc->sc_dmatag, ui_segs, ui_nsegs,
PDQ_OS_PAGESIZE,
(caddr_t *) &pdq->pdq_unsolicited_info.ui_events,
BUS_DMA_NOWAIT);
}
if (!not_ok) {
steps = 6;
not_ok = bus_dmamap_create(sc->sc_dmatag, ui_segs[0].ds_len, 1,
PDQ_OS_PAGESIZE, 0, BUS_DMA_NOWAIT,
&sc->sc_uimap);
}
if (!not_ok) {
steps = 7;
not_ok = bus_dmamap_load(sc->sc_dmatag, sc->sc_uimap,
pdq->pdq_unsolicited_info.ui_events,
PDQ_OS_PAGESIZE, NULL, BUS_DMA_NOWAIT);
}
if (!not_ok) {
steps = 8;
pdq->pdq_unsolicited_info.ui_pa_bufstart = sc->sc_uimap->dm_segs[0].ds_addr;
cb_segs[0] = db_segs[0];
cb_segs[0].ds_addr += offsetof(pdq_descriptor_block_t, pdqdb_consumer);
cb_segs[0].ds_len = sizeof(pdq_consumer_block_t);
not_ok = bus_dmamem_map(sc->sc_dmatag, cb_segs, 1,
sizeof(*pdq->pdq_cbp), (caddr_t *) &pdq->pdq_cbp,
BUS_DMA_NOWAIT|BUS_DMA_COHERENT);
}
if (!not_ok) {
steps = 9;
not_ok = bus_dmamap_create(sc->sc_dmatag, cb_segs[0].ds_len, 1,
0x2000, 0, BUS_DMA_NOWAIT, &sc->sc_cbmap);
}
if (!not_ok) {
steps = 10;
not_ok = bus_dmamap_load(sc->sc_dmatag, sc->sc_cbmap,
(caddr_t) pdq->pdq_cbp, sizeof(*pdq->pdq_cbp),
NULL, BUS_DMA_NOWAIT);
}
if (!not_ok) {
pdq->pdq_pa_consumer_block = sc->sc_cbmap->dm_segs[0].ds_addr;
return not_ok;
}
switch (steps) {
case 11: {
bus_dmamap_unload(sc->sc_dmatag, sc->sc_cbmap);
/* FALL THROUGH */
}
case 10: {
bus_dmamap_destroy(sc->sc_dmatag, sc->sc_cbmap);
/* FALL THROUGH */
}
case 9: {
bus_dmamem_unmap(sc->sc_dmatag,
(caddr_t) pdq->pdq_cbp, sizeof(*pdq->pdq_cbp));
/* FALL THROUGH */
}
case 8: {
bus_dmamap_unload(sc->sc_dmatag, sc->sc_uimap);
/* FALL THROUGH */
}
case 7: {
bus_dmamap_destroy(sc->sc_dmatag, sc->sc_uimap);
/* FALL THROUGH */
}
case 6: {
bus_dmamem_unmap(sc->sc_dmatag,
(caddr_t) pdq->pdq_unsolicited_info.ui_events,
PDQ_OS_PAGESIZE);
/* FALL THROUGH */
}
case 5: {
bus_dmamem_free(sc->sc_dmatag, ui_segs, ui_nsegs);
/* FALL THROUGH */
}
case 4: {
bus_dmamap_unload(sc->sc_dmatag, sc->sc_dbmap);
/* FALL THROUGH */
}
case 3: {
bus_dmamap_destroy(sc->sc_dmatag, sc->sc_dbmap);
/* FALL THROUGH */
}
case 2: {
bus_dmamem_unmap(sc->sc_dmatag,
(caddr_t) pdq->pdq_dbp,
sizeof(*pdq->pdq_dbp));
/* FALL THROUGH */
}
case 1: {
bus_dmamem_free(sc->sc_dmatag, db_segs, db_nsegs);
/* FALL THROUGH */
}
}
return not_ok;
}
extern void
pdq_os_descriptor_block_sync(
pdq_os_ctx_t *sc,
size_t offset,
size_t length,
int ops)
{
bus_dmamap_sync(sc->sc_dmatag, sc->sc_dbmap, offset, length, ops);
Finally import the 960801 of Matt Thomas' DEC FDDI driver. I'm importing it onto a vendor branch first, in the hope that this will make future maintenance easier. The conflicts are (hopefully) unimportant. More commits that actually bring this into the source tree will follow. Submitted by: Matt Thomas (thomas@lkg.dec.com)
1997-01-17 23:19:49 +00:00
}
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
extern void
pdq_os_consumer_block_sync(
pdq_os_ctx_t *sc,
int ops)
{
bus_dmamap_sync(sc->sc_dmatag, sc->sc_cbmap, 0, sizeof(pdq_consumer_block_t), ops);
}
extern void
pdq_os_unsolicited_event_sync(
pdq_os_ctx_t *sc,
size_t offset,
size_t length,
int ops)
{
bus_dmamap_sync(sc->sc_dmatag, sc->sc_uimap, offset, length, ops);
}
extern void
pdq_os_databuf_sync(
pdq_os_ctx_t *sc,
struct mbuf *m,
size_t offset,
size_t length,
int ops)
{
bus_dmamap_sync(sc->sc_dmatag, M_GETCTX(m, bus_dmamap_t), offset, length, ops);
}
extern void
pdq_os_databuf_free(
pdq_os_ctx_t *sc,
struct mbuf *m)
{
if (m->m_flags & (M_HASRXDMAMAP|M_HASTXDMAMAP)) {
bus_dmamap_t map = M_GETCTX(m, bus_dmamap_t);
bus_dmamap_unload(sc->sc_dmatag, map);
bus_dmamap_destroy(sc->sc_dmatag, map);
m->m_flags &= ~(M_HASRXDMAMAP|M_HASTXDMAMAP);
}
m_freem(m);
}
extern struct mbuf *
pdq_os_databuf_alloc(
pdq_os_ctx_t *sc)
{
struct mbuf *m;
bus_dmamap_t map;
Back out M_* changes, per decision of the TRB. Approved by: trb
2003-02-19 05:47:46 +00:00
MGETHDR(m, M_DONTWAIT, MT_DATA);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
if (m == NULL) {
printf("%s: can't alloc small buf\n", sc->sc_dev.dv_xname);
return NULL;
}
Back out M_* changes, per decision of the TRB. Approved by: trb
2003-02-19 05:47:46 +00:00
MCLGET(m, M_DONTWAIT);
- Merge the pdq driver (if_fpa and if_fea) from NetBSD. Among other things this gets us ifmedia support. - Update fddi_ifattach() to take an additional argument.
2002-03-29 11:22:22 +00:00
if ((m->m_flags & M_EXT) == 0) {
printf("%s: can't alloc cluster\n", sc->sc_dev.dv_xname);
m_free(m);
return NULL;
}
m->m_pkthdr.len = m->m_len = PDQ_OS_DATABUF_SIZE;
if (bus_dmamap_create(sc->sc_dmatag, PDQ_OS_DATABUF_SIZE,
1, PDQ_OS_DATABUF_SIZE, 0, BUS_DMA_NOWAIT, &map)) {
printf("%s: can't create dmamap\n", sc->sc_dev.dv_xname);
m_free(m);
return NULL;
}
if (bus_dmamap_load_mbuf(sc->sc_dmatag, map, m,
BUS_DMA_READ|BUS_DMA_NOWAIT)) {
printf("%s: can't load dmamap\n", sc->sc_dev.dv_xname);
bus_dmamap_destroy(sc->sc_dmatag, map);
m_free(m);
return NULL;
}
m->m_flags |= M_HASRXDMAMAP;
M_SETCTX(m, map);
return m;
}
#endif
Reference in New Issue Copy Permalink