d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
538ccfb68e
freebsd-skq/sys/dev/pccard/pccard_cis.c

1298 lines
34 KiB
C
Raw Normal View History

Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
/* $NetBSD: pcmcia_cis.c,v 1.17 2000/02/10 09:01:52 chopps Exp $ */
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
/* $FreeBSD$ */
Start each of the license/copyright comments with /*-, minor shuffle of lines
2005-01-06 01:43:34 +00:00
/*-
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
* Copyright (c) 1997 Marc Horowitz. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by Marc Horowitz.
* 4. The name of the author may not be used to endorse or promote products
* derived from this software without specific prior written permission.
*
* 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.
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/malloc.h>
#include <sys/module.h>
#include <sys/kernel.h>
#include <sys/queue.h>
#include <sys/types.h>
#include <sys/bus.h>
#include <machine/bus.h>
#include <sys/rman.h>
#include <machine/resource.h>
#include <dev/pccard/pccardreg.h>
#include <dev/pccard/pccardvar.h>
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
#include <dev/pccard/pccard_cis.h>
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
checkpoint latest pccard/pcic hacking: o Eliminate cross calls between the devices. Instead move to using the newbus messaging system. Added three new card calls: attach_card, detach_card, get_type. o Eliminate interrupt routine in pccard we never use. o Move from deactivate to detach for removing cards. o Start mapping CIS memory, but it is broken and causes panics. At least it is closer to working than before. o Eliminate struct device everywhere. It was bogus. o Initialize softc for pccard device so we have valid pointers to ourselves. o Implement routine to find the pcic ivar for a child device of the pccard so we can use it to talk to the pcic hardware. o Lots of minor tiding up. This version now panics when we try to read the CIS. The next batch of work to make this work is what was outlined in my posting to mobile about resource allocation and such.
2000-04-13 06:42:58 +00:00
#include "card_if.h"
Make hw.pccard.debug and hw.pccard.cis_debug tunable/sysctl. Setting to 1 will enable more verbose debugging output from the pccard system.
2002-03-07 08:03:53 +00:00
extern int pccard_cis_debug;
checkpoint latest pccard/pcic hacking: o Eliminate cross calls between the devices. Instead move to using the newbus messaging system. Added three new card calls: attach_card, detach_card, get_type. o Eliminate interrupt routine in pccard we never use. o Move from deactivate to detach for removing cards. o Start mapping CIS memory, but it is broken and causes panics. At least it is closer to working than before. o Eliminate struct device everywhere. It was bogus. o Initialize softc for pccard device so we have valid pointers to ourselves. o Implement routine to find the pcic ivar for a child device of the pccard so we can use it to talk to the pcic hardware. o Lots of minor tiding up. This version now panics when we try to read the CIS. The next batch of work to make this work is what was outlined in my posting to mobile about resource allocation and such.
2000-04-13 06:42:58 +00:00
#define PCCARDCISDEBUG
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
#ifdef PCCARDCISDEBUG
When debugging CIS, only print 10 CISTPL_NULLs. Chances are good they are all bogus, and the cards that don't decode things quite right often have hundreds of them. This will fix starvation of small dmesg buffers and allow better debugging to happen. I thought about adding an override, but there is such a thing as too many knobs. :-)
2003-08-20 05:44:55 +00:00
#define DPRINTF(arg) do { if (pccard_cis_debug) printf arg; } while (0)
#define DEVPRINTF(arg) do { if (pccard_cis_debug) device_printf arg; } while (0)
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
#else
#define DPRINTF(arg)
Eliminate pccard_chip_* tonight. o ifdef out pccardchip.h (almost all of it, there are dangling bits o Add rid/res members to pccard_function o remove pct/pch from pccard_softc o map memory properly in scan_cis (almost, see XXX for more work) o manage ccr. o remove bogus comment I added about touching the ccr being a layering violation for pccard. It is properly done at that level. o More function prototyping
2000-01-10 06:58:17 +00:00
#define DEVPRINTF(arg)
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
#endif
Improve reading of CIS cards: (1) Align to 64k for the CIS. Some cards don't like it when we aren't aligned to a 64k boundary. I can't find anything in the standard that requires this, but I have 1/2 dozen cards that won't work at all unless I enable this. (2) Sleep 1s before scanning the CIS. This may be a nop, but has little harm. (3) The CIS can be up to 4k in some weird, odd-ball edge cases. Since we have limiters for when that's not the case, it does no harm to increase it to 4k. #1 was submitted, in a different form, by Carlos Velasco.
2004-04-12 20:56:34 +00:00
#define PCCARD_CIS_SIZE 4096
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
struct cis_state {
int count;
int gotmfc;
struct pccard_config_entry temp_cfe;
struct pccard_config_entry *default_cfe;
struct pccard_card *card;
struct pccard_function *pf;
};
Upon relection, we shouldn't allow the tuple structs to be modified by the functor, so make it a const pointer, and chase down the resulting const-poisoning. Approved by: re (scottl)
2005-07-01 15:52:50 +00:00
static int pccard_parse_cis_tuple(const struct pccard_tuple *, void *);
static int decode_funce(const struct pccard_tuple *, struct pccard_function *);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
void
Eliminate pccard_chip_* tonight. o ifdef out pccardchip.h (almost all of it, there are dangling bits o Add rid/res members to pccard_function o remove pct/pch from pccard_softc o map memory properly in scan_cis (almost, see XXX for more work) o manage ccr. o remove bogus comment I added about touching the ccr being a layering violation for pccard. It is properly done at that level. o More function prototyping
2000-01-10 06:58:17 +00:00
pccard_read_cis(struct pccard_softc *sc)
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
{
struct cis_state state;
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
bzero(&state, sizeof state);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
state.card = &sc->card;
state.card->error = 0;
state.card->cis1_major = -1;
state.card->cis1_minor = -1;
state.card->cis1_info[0] = NULL;
state.card->cis1_info[1] = NULL;
state.card->cis1_info[2] = NULL;
state.card->cis1_info[3] = NULL;
Sync to 1.16 pccarddevs to get new PCMCIA_ symbols
2001-11-11 20:15:47 +00:00
state.card->manufacturer = PCMCIA_VENDOR_INVALID;
state.card->product = PCMCIA_PRODUCT_INVALID;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
STAILQ_INIT(&state.card->pf_head);
state.pf = NULL;
Improve reading of CIS cards: (1) Align to 64k for the CIS. Some cards don't like it when we aren't aligned to a 64k boundary. I can't find anything in the standard that requires this, but I have 1/2 dozen cards that won't work at all unless I enable this. (2) Sleep 1s before scanning the CIS. This may be a nop, but has little harm. (3) The CIS can be up to 4k in some weird, odd-ball edge cases. Since we have limiters for when that's not the case, it does no harm to increase it to 4k. #1 was submitted, in a different form, by Carlos Velasco.
2004-04-12 20:56:34 +00:00
tsleep(&state, 0, "pccard", hz);
Fix pcic_detach_socket to get right pcic_handle. Pass sc->dev rather than a bogusly cast pccard_softc *sc. This allows us to insert and remove cards w/o panicing the kernel. However, the cis isn't mapped in, so the pccard_scan_cis function fails.
2000-04-04 04:12:43 +00:00
if (pccard_scan_cis(sc->dev, pccard_parse_cis_tuple,
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
&state) == -1)
state.card->error++;
}
int
Add a much-requested feature: The ability for pccard attachments to scan the CIS for interesting tuples. 95% of what can be obtained from the CIS is harvested by the pccard layer and presented to the user in standard function calls. However, there are special needs at times where the standard stuff doesn't suffice. This is for those special cases. CARD_SCAN_CIS(device_get_parent(dev), function, argp) scans the CIS of the card, passing each tuple to function with the tuple and argp as its arguments. Returning 0 continues the scan, while returning 1 terminates the scan. The value of the last invocation of function is returned from this function. int (*pccard_scan_t)(struct pccard_tuple *tuple, void *argp) function called for each tuple. Elements of the CIS tuple can be read with pccard_tuple_read_{1,2,3,4,n}(). You are reading the actual tuple memory each time, in case your card has registers in the CIS. # I suppose these things should be documented in pccard(4) or something like # that. # I plan on unifying cardbus CIS support in a similar way. Approved by: re (scottl)
2005-07-01 03:40:28 +00:00
pccard_scan_cis(device_t dev, pccard_scan_t fct, void *arg)
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
{
Eliminate pccard_chip_* tonight. o ifdef out pccardchip.h (almost all of it, there are dangling bits o Add rid/res members to pccard_function o remove pct/pch from pccard_softc o map memory properly in scan_cis (almost, see XXX for more work) o manage ccr. o remove bogus comment I added about touching the ccr being a layering violation for pccard. It is properly done at that level. o More function prototyping
2000-01-10 06:58:17 +00:00
struct resource *res;
int rid;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
struct pccard_tuple tuple;
int longlink_present;
int longlink_common;
Tag longling_addr as maybe using a bad type, I'm not sure.
2003-03-18 02:38:33 +00:00
u_long longlink_addr; /* Type suspect */
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
int mfc_count;
int mfc_index;
When debugging CIS, only print 10 CISTPL_NULLs. Chances are good they are all bogus, and the cards that don't decode things quite right often have hundreds of them. This will fix starvation of small dmesg buffers and allow better debugging to happen. I thought about adding an override, but there is such a thing as too many knobs. :-)
2003-08-20 05:44:55 +00:00
#ifdef PCCARDCISDEBUG
int cis_none_cnt = 10; /* Only report 10 CIS_NONEs */
#endif
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
struct {
int common;
u_long addr;
} mfc[256 / 5];
int ret;
ret = 0;
/* allocate some memory */
Improve reading of CIS cards: (1) Align to 64k for the CIS. Some cards don't like it when we aren't aligned to a 64k boundary. I can't find anything in the standard that requires this, but I have 1/2 dozen cards that won't work at all unless I enable this. (2) Sleep 1s before scanning the CIS. This may be a nop, but has little harm. (3) The CIS can be up to 4k in some weird, odd-ball edge cases. Since we have limiters for when that's not the case, it does no harm to increase it to 4k. #1 was submitted, in a different form, by Carlos Velasco.
2004-04-12 20:56:34 +00:00
/*
* Some reports from the field suggest that a 64k memory boundary
* helps card CIS being able to be read. Try it here and see what
* the results actually are. I'm not sure I understand why this
* would make cards work better, but it is easy enough to test.
*/
Eliminate pccard_chip_* tonight. o ifdef out pccardchip.h (almost all of it, there are dangling bits o Add rid/res members to pccard_function o remove pct/pch from pccard_softc o map memory properly in scan_cis (almost, see XXX for more work) o manage ccr. o remove bogus comment I added about touching the ccr being a layering violation for pccard. It is properly done at that level. o More function prototyping
2000-01-10 06:58:17 +00:00
rid = 0;
Improve reading of CIS cards: (1) Align to 64k for the CIS. Some cards don't like it when we aren't aligned to a 64k boundary. I can't find anything in the standard that requires this, but I have 1/2 dozen cards that won't work at all unless I enable this. (2) Sleep 1s before scanning the CIS. This may be a nop, but has little harm. (3) The CIS can be up to 4k in some weird, odd-ball edge cases. Since we have limiters for when that's not the case, it does no harm to increase it to 4k. #1 was submitted, in a different form, by Carlos Velasco.
2004-04-12 20:56:34 +00:00
res = bus_alloc_resource(dev, SYS_RES_MEMORY, &rid, 0, ~0,
PCCARD_CIS_SIZE, RF_ACTIVE | rman_make_alignment_flags(64*1024));
Eliminate pccard_chip_* tonight. o ifdef out pccardchip.h (almost all of it, there are dangling bits o Add rid/res members to pccard_function o remove pct/pch from pccard_softc o map memory properly in scan_cis (almost, see XXX for more work) o manage ccr. o remove bogus comment I added about touching the ccr being a layering violation for pccard. It is properly done at that level. o More function prototyping
2000-01-10 06:58:17 +00:00
if (res == NULL) {
device_printf(dev, "can't alloc memory to read attributes\n");
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
return -1;
}
checkpoint latest pccard/pcic hacking: o Eliminate cross calls between the devices. Instead move to using the newbus messaging system. Added three new card calls: attach_card, detach_card, get_type. o Eliminate interrupt routine in pccard we never use. o Move from deactivate to detach for removing cards. o Start mapping CIS memory, but it is broken and causes panics. At least it is closer to working than before. o Eliminate struct device everywhere. It was bogus. o Initialize softc for pccard device so we have valid pointers to ourselves. o Implement routine to find the pcic ivar for a child device of the pccard so we can use it to talk to the pcic hardware. o Lots of minor tiding up. This version now panics when we try to read the CIS. The next batch of work to make this work is what was outlined in my posting to mobile about resource allocation and such.
2000-04-13 06:42:58 +00:00
CARD_SET_RES_FLAGS(device_get_parent(dev), dev, SYS_RES_MEMORY,
rid, PCCARD_A_MEM_ATTR);
Eliminate pccard_chip_* tonight. o ifdef out pccardchip.h (almost all of it, there are dangling bits o Add rid/res members to pccard_function o remove pct/pch from pccard_softc o map memory properly in scan_cis (almost, see XXX for more work) o manage ccr. o remove bogus comment I added about touching the ccr being a layering violation for pccard. It is properly done at that level. o More function prototyping
2000-01-10 06:58:17 +00:00
tuple.memt = rman_get_bustag(res);
tuple.memh = rman_get_bushandle(res);
OK. Next step: we read in CIS. I've done this by having requests to allocate memory propigate up the tree. We'll see how well this works and reevaluate if it isn't working well. Also initialize ptr in the tuple. As well as minor reorg of memory allocation. Likely need to do similar things for I/O when the time comes. I've move all defines from pccardchip.h into pccardvar.h and eliminated pccardchip.h.
2000-04-19 08:31:21 +00:00
tuple.ptr = 0;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
Put a band-aide(tm) on the CIS panic problem. This is a similar fix to what is in NetBSD. I have a few cards that tickles this bug, and this just keeps us from panicing. It doesn't actually fix the problem (that will happen once I figure out why some cards hate the address their CIS is mapped to high memory).
2003-08-18 03:07:09 +00:00
DPRINTF(("cis mem map 0x%x (resource: 0x%lx)\n",
(unsigned int) tuple.memh, rman_get_start(res)));
Eliminate pccard_chip_* tonight. o ifdef out pccardchip.h (almost all of it, there are dangling bits o Add rid/res members to pccard_function o remove pct/pch from pccard_softc o map memory properly in scan_cis (almost, see XXX for more work) o manage ccr. o remove bogus comment I added about touching the ccr being a layering violation for pccard. It is properly done at that level. o More function prototyping
2000-01-10 06:58:17 +00:00
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
tuple.mult = 2;
longlink_present = 1;
longlink_common = 1;
longlink_addr = 0;
mfc_count = 0;
mfc_index = 0;
Eliminate pccard_chip_* tonight. o ifdef out pccardchip.h (almost all of it, there are dangling bits o Add rid/res members to pccard_function o remove pct/pch from pccard_softc o map memory properly in scan_cis (almost, see XXX for more work) o manage ccr. o remove bogus comment I added about touching the ccr being a layering violation for pccard. It is properly done at that level. o More function prototyping
2000-01-10 06:58:17 +00:00
DEVPRINTF((dev, "CIS tuple chain:\n"));
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
while (1) {
while (1) {
Put a band-aide(tm) on the CIS panic problem. This is a similar fix to what is in NetBSD. I have a few cards that tickles this bug, and this just keeps us from panicing. It doesn't actually fix the problem (that will happen once I figure out why some cards hate the address their CIS is mapped to high memory).
2003-08-18 03:07:09 +00:00
/*
* Perform boundary check for insane cards.
* If CIS is too long, simulate CIS end.
* (This check may not be sufficient for
* malicious cards.)
*/
if (tuple.mult * tuple.ptr >= PCCARD_CIS_SIZE - 1
- 32 /* ad hoc value */ ) {
printf("CIS is too long -- truncating\n");
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
tuple.code = CISTPL_END;
Put a band-aide(tm) on the CIS panic problem. This is a similar fix to what is in NetBSD. I have a few cards that tickles this bug, and this just keeps us from panicing. It doesn't actually fix the problem (that will happen once I figure out why some cards hate the address their CIS is mapped to high memory).
2003-08-18 03:07:09 +00:00
} else {
/* get the tuple code */
tuple.code = pccard_cis_read_1(&tuple, tuple.ptr);
}
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
/* two special-case tuples */
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
if (tuple.code == CISTPL_NULL) {
When debugging CIS, only print 10 CISTPL_NULLs. Chances are good they are all bogus, and the cards that don't decode things quite right often have hundreds of them. This will fix starvation of small dmesg buffers and allow better debugging to happen. I thought about adding an override, but there is such a thing as too many knobs. :-)
2003-08-20 05:44:55 +00:00
#ifdef PCCARDCISDEBUG
if (cis_none_cnt > 0)
DPRINTF(("CISTPL_NONE\n 00\n"));
else if (cis_none_cnt == 0)
DPRINTF(("TOO MANY CIS_NONE\n"));
cis_none_cnt--;
#endif
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
tuple.ptr++;
continue;
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
} else if (tuple.code == CISTPL_END) {
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
DPRINTF(("CISTPL_END\n ff\n"));
/* Call the function for the END tuple, since
the CIS semantics depend on it */
Add a much-requested feature: The ability for pccard attachments to scan the CIS for interesting tuples. 95% of what can be obtained from the CIS is harvested by the pccard layer and presented to the user in standard function calls. However, there are special needs at times where the standard stuff doesn't suffice. This is for those special cases. CARD_SCAN_CIS(device_get_parent(dev), function, argp) scans the CIS of the card, passing each tuple to function with the tuple and argp as its arguments. Returning 0 continues the scan, while returning 1 terminates the scan. The value of the last invocation of function is returned from this function. int (*pccard_scan_t)(struct pccard_tuple *tuple, void *argp) function called for each tuple. Elements of the CIS tuple can be read with pccard_tuple_read_{1,2,3,4,n}(). You are reading the actual tuple memory each time, in case your card has registers in the CIS. # I suppose these things should be documented in pccard(4) or something like # that. # I plan on unifying cardbus CIS support in a similar way. Approved by: re (scottl)
2005-07-01 03:40:28 +00:00
if ((*fct)(&tuple, arg)) {
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
ret = 1;
goto done;
}
tuple.ptr++;
break;
}
/* now all the normal tuples */
tuple.length = pccard_cis_read_1(&tuple, tuple.ptr + 1);
switch (tuple.code) {
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_LONGLINK_A:
case CISTPL_LONGLINK_C:
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (tuple.length < 4) {
DPRINTF(("CISTPL_LONGLINK_%s too "
"short %d\n",
longlink_common ? "C" : "A",
tuple.length));
break;
}
longlink_present = 1;
longlink_common = (tuple.code ==
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
CISTPL_LONGLINK_C) ? 1 : 0;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
longlink_addr = pccard_tuple_read_4(&tuple, 0);
DPRINTF(("CISTPL_LONGLINK_%s %lx\n",
longlink_common ? "C" : "A",
longlink_addr));
break;
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_NO_LINK:
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
longlink_present = 0;
DPRINTF(("CISTPL_NO_LINK\n"));
break;
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_CHECKSUM:
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (tuple.length < 5) {
DPRINTF(("CISTPL_CHECKSUM too "
"short %d\n", tuple.length));
break;
} {
int16_t offset;
u_long addr, length;
u_int cksum, sum;
int i;
Do the cast away of unsignedness in a way that is more commprehensible.
2002-02-19 05:01:43 +00:00
offset = (uint16_t)
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
pccard_tuple_read_2(&tuple, 0);
length = pccard_tuple_read_2(&tuple, 2);
cksum = pccard_tuple_read_1(&tuple, 4);
addr = tuple.ptr + offset;
DPRINTF(("CISTPL_CHECKSUM addr=%lx "
"len=%lx cksum=%x",
addr, length, cksum));
/*
* XXX do more work to deal with
* distant regions
*/
if ((addr >= PCCARD_CIS_SIZE) ||
((addr + length) >=
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
PCCARD_CIS_SIZE)) {
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
DPRINTF((" skipped, "
"too distant\n"));
break;
}
sum = 0;
for (i = 0; i < length; i++)
sum +=
bus_space_read_1(tuple.memt,
tuple.memh,
addr + tuple.mult * i);
if (cksum != (sum & 0xff)) {
DPRINTF((" failed sum=%x\n",
sum));
Eliminate pccard_chip_* tonight. o ifdef out pccardchip.h (almost all of it, there are dangling bits o Add rid/res members to pccard_function o remove pct/pch from pccard_softc o map memory properly in scan_cis (almost, see XXX for more work) o manage ccr. o remove bogus comment I added about touching the ccr being a layering violation for pccard. It is properly done at that level. o More function prototyping
2000-01-10 06:58:17 +00:00
device_printf(dev,
"CIS checksum failed\n");
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
#if 0
/*
* XXX Some working cards have
* XXX bad checksums!!
*/
ret = -1;
#endif
} else {
DPRINTF((" ok\n"));
}
}
break;
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_LONGLINK_MFC:
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (tuple.length < 1) {
DPRINTF(("CISTPL_LONGLINK_MFC too "
"short %d\n", tuple.length));
break;
}
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
if (((tuple.length - 1) % 5) != 0) {
DPRINTF(("CISTPL_LONGLINK_MFC bogus "
"length %d\n", tuple.length));
break;
}
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
/*
* this is kind of ad hoc, as I don't have
* any real documentation
*/
{
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
int i, tmp_count;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
/*
* put count into tmp var so that
* if we have to bail (because it's
* a bogus count) it won't be
* remembered for later use.
*/
tmp_count =
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
pccard_tuple_read_1(&tuple, 0);
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
DPRINTF(("CISTPL_LONGLINK_MFC %d",
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
tmp_count));
/*
* make _sure_ it's the right size;
* if too short, it may be a weird
* (unknown/undefined) format
*/
if (tuple.length != (tmp_count*5 + 1)) {
DPRINTF((" bogus length %d\n",
tuple.length));
break;
}
/*
* sanity check for a programming
* error which is difficult to find
* when debugging.
*/
if (tmp_count >
howmany(sizeof mfc, sizeof mfc[0]))
panic("CISTPL_LONGLINK_MFC mfc "
"count would blow stack");
mfc_count = tmp_count;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
for (i = 0; i < mfc_count; i++) {
mfc[i].common =
(pccard_tuple_read_1(&tuple,
1 + 5 * i) ==
PCCARD_MFC_MEM_COMMON) ?
1 : 0;
mfc[i].addr =
pccard_tuple_read_4(&tuple,
1 + 5 * i + 1);
DPRINTF((" %s:%lx",
mfc[i].common ? "common" :
"attr", mfc[i].addr));
}
DPRINTF(("\n"));
}
/*
* for LONGLINK_MFC, fall through to the
* function. This tuple has structural and
* semantic content.
*/
default:
{
Add a much-requested feature: The ability for pccard attachments to scan the CIS for interesting tuples. 95% of what can be obtained from the CIS is harvested by the pccard layer and presented to the user in standard function calls. However, there are special needs at times where the standard stuff doesn't suffice. This is for those special cases. CARD_SCAN_CIS(device_get_parent(dev), function, argp) scans the CIS of the card, passing each tuple to function with the tuple and argp as its arguments. Returning 0 continues the scan, while returning 1 terminates the scan. The value of the last invocation of function is returned from this function. int (*pccard_scan_t)(struct pccard_tuple *tuple, void *argp) function called for each tuple. Elements of the CIS tuple can be read with pccard_tuple_read_{1,2,3,4,n}(). You are reading the actual tuple memory each time, in case your card has registers in the CIS. # I suppose these things should be documented in pccard(4) or something like # that. # I plan on unifying cardbus CIS support in a similar way. Approved by: re (scottl)
2005-07-01 03:40:28 +00:00
if ((*fct)(&tuple, arg)) {
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
ret = 1;
goto done;
}
}
break;
} /* switch */
#ifdef PCCARDCISDEBUG
/* print the tuple */
{
int i;
DPRINTF((" %02x %02x", tuple.code,
tuple.length));
for (i = 0; i < tuple.length; i++) {
DPRINTF((" %02x",
pccard_tuple_read_1(&tuple, i)));
if ((i % 16) == 13)
DPRINTF(("\n"));
}
implement MFC links properly (and I think long links too). This make the sprint wireless card try to attach. Sadly, the pci code at the bridge keeps this from happening. Bug w/o PR: jhb :-)
2001-12-04 13:48:16 +00:00
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if ((i % 16) != 14)
DPRINTF(("\n"));
}
#endif
/* skip to the next tuple */
tuple.ptr += 2 + tuple.length;
}
/*
* the chain is done. Clean up and move onto the next one,
* if any. The loop is here in the case that there is an MFC
* card with no longlink (which defaults to existing, == 0).
* In general, this means that if one pointer fails, it will
* try the next one, instead of just bailing.
*/
while (1) {
if (longlink_present) {
implement MFC links properly (and I think long links too). This make the sprint wireless card try to attach. Sadly, the pci code at the bridge keeps this from happening. Bug w/o PR: jhb :-)
2001-12-04 13:48:16 +00:00
CARD_SET_RES_FLAGS(device_get_parent(dev), dev,
SYS_RES_MEMORY, rid, longlink_common ?
Improve support of MFC cards (Multi-function cards). This commit allows us to properly parse cards with attribute memory based CIS that before wouldn't parse correctly, sometimes with a panic. This allows me to get my 3C562 modem/ethernet card to fail to attach due to problems in the ep and sio drivers rather than due to problems in the CIS parsing code :-). We weren't setting the address to jump to for the function entries. This caused us to only work when the addional entries were after the first ones. On the 3C562/3C563 card this was not the case. We were also mapping Attribute memory when common memory was asked for in the target of the LONGLINK_{A,C} or LONGLINK_MFC. My IBM Home And Away Modem/LAN card still fails for reasons unknown.
2002-03-29 08:05:39 +00:00
PCCARD_A_MEM_COM : PCCARD_A_MEM_ATTR);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
DPRINTF(("cis mem map %x\n",
(unsigned int) tuple.memh));
tuple.mult = longlink_common ? 1 : 2;
Improve support of MFC cards (Multi-function cards). This commit allows us to properly parse cards with attribute memory based CIS that before wouldn't parse correctly, sometimes with a panic. This allows me to get my 3C562 modem/ethernet card to fail to attach due to problems in the ep and sio drivers rather than due to problems in the CIS parsing code :-). We weren't setting the address to jump to for the function entries. This caused us to only work when the addional entries were after the first ones. On the 3C562/3C563 card this was not the case. We were also mapping Attribute memory when common memory was asked for in the target of the LONGLINK_{A,C} or LONGLINK_MFC. My IBM Home And Away Modem/LAN card still fails for reasons unknown.
2002-03-29 08:05:39 +00:00
tuple.ptr = longlink_addr;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
longlink_present = 0;
longlink_common = 1;
longlink_addr = 0;
} else if (mfc_count && (mfc_index < mfc_count)) {
implement MFC links properly (and I think long links too). This make the sprint wireless card try to attach. Sadly, the pci code at the bridge keeps this from happening. Bug w/o PR: jhb :-)
2001-12-04 13:48:16 +00:00
CARD_SET_RES_FLAGS(device_get_parent(dev), dev,
SYS_RES_MEMORY, rid, mfc[mfc_index].common
Improve support of MFC cards (Multi-function cards). This commit allows us to properly parse cards with attribute memory based CIS that before wouldn't parse correctly, sometimes with a panic. This allows me to get my 3C562 modem/ethernet card to fail to attach due to problems in the ep and sio drivers rather than due to problems in the CIS parsing code :-). We weren't setting the address to jump to for the function entries. This caused us to only work when the addional entries were after the first ones. On the 3C562/3C563 card this was not the case. We were also mapping Attribute memory when common memory was asked for in the target of the LONGLINK_{A,C} or LONGLINK_MFC. My IBM Home And Away Modem/LAN card still fails for reasons unknown.
2002-03-29 08:05:39 +00:00
? PCCARD_A_MEM_COM : PCCARD_A_MEM_ATTR);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
DPRINTF(("cis mem map %x\n",
(unsigned int) tuple.memh));
/* set parse state, and point at the next one */
tuple.mult = mfc[mfc_index].common ? 1 : 2;
Improve support of MFC cards (Multi-function cards). This commit allows us to properly parse cards with attribute memory based CIS that before wouldn't parse correctly, sometimes with a panic. This allows me to get my 3C562 modem/ethernet card to fail to attach due to problems in the ep and sio drivers rather than due to problems in the CIS parsing code :-). We weren't setting the address to jump to for the function entries. This caused us to only work when the addional entries were after the first ones. On the 3C562/3C563 card this was not the case. We were also mapping Attribute memory when common memory was asked for in the target of the LONGLINK_{A,C} or LONGLINK_MFC. My IBM Home And Away Modem/LAN card still fails for reasons unknown.
2002-03-29 08:05:39 +00:00
tuple.ptr = mfc[mfc_index].addr;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
mfc_index++;
} else {
goto done;
}
/* make sure that the link is valid */
tuple.code = pccard_cis_read_1(&tuple, tuple.ptr);
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
if (tuple.code != CISTPL_LINKTARGET) {
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
DPRINTF(("CISTPL_LINKTARGET expected, "
"code %02x observed\n", tuple.code));
continue;
}
tuple.length = pccard_cis_read_1(&tuple, tuple.ptr + 1);
if (tuple.length < 3) {
DPRINTF(("CISTPL_LINKTARGET too short %d\n",
tuple.length));
continue;
}
if ((pccard_tuple_read_1(&tuple, 0) != 'C') ||
(pccard_tuple_read_1(&tuple, 1) != 'I') ||
(pccard_tuple_read_1(&tuple, 2) != 'S')) {
DPRINTF(("CISTPL_LINKTARGET magic "
"%02x%02x%02x incorrect\n",
pccard_tuple_read_1(&tuple, 0),
pccard_tuple_read_1(&tuple, 1),
pccard_tuple_read_1(&tuple, 2)));
continue;
}
tuple.ptr += 2 + tuple.length;
break;
}
}
done:
Eliminate pccard_chip_* tonight. o ifdef out pccardchip.h (almost all of it, there are dangling bits o Add rid/res members to pccard_function o remove pct/pch from pccard_softc o map memory properly in scan_cis (almost, see XXX for more work) o manage ccr. o remove bogus comment I added about touching the ccr being a layering violation for pccard. It is properly done at that level. o More function prototyping
2000-01-10 06:58:17 +00:00
bus_release_resource(dev, SYS_RES_MEMORY, rid, res);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
return (ret);
}
/* XXX this is incredibly verbose. Not sure what trt is */
void
pccard_print_cis(device_t dev)
{
o Move to using PCCARD_SOFTC(dev) o fill in the size of the actual softc, rather than 1 in data structure o minor debugging improvements.
2000-08-19 19:22:04 +00:00
struct pccard_softc *sc = PCCARD_SOFTC(dev);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
struct pccard_card *card = &sc->card;
struct pccard_function *pf;
struct pccard_config_entry *cfe;
int i;
device_printf(dev, "CIS version ");
if (card->cis1_major == 4) {
if (card->cis1_minor == 0)
printf("PCCARD 1.0\n");
else if (card->cis1_minor == 1)
printf("PCCARD 2.0 or 2.1\n");
} else if (card->cis1_major >= 5)
printf("PC Card Standard %d.%d\n", card->cis1_major, card->cis1_minor);
else
printf("unknown (major=%d, minor=%d)\n",
card->cis1_major, card->cis1_minor);
device_printf(dev, "CIS info: ");
for (i = 0; i < 4; i++) {
if (card->cis1_info[i] == NULL)
break;
if (i)
printf(", ");
printf("%s", card->cis1_info[i]);
}
printf("\n");
device_printf(dev, "Manufacturer code 0x%x, product 0x%x\n",
card->manufacturer, card->product);
STAILQ_FOREACH(pf, &card->pf_head, pf_list) {
device_printf(dev, "function %d: ", pf->number);
switch (pf->function) {
case PCCARD_FUNCTION_UNSPEC:
printf("unspecified");
break;
case PCCARD_FUNCTION_MULTIFUNCTION:
printf("multi-function");
break;
case PCCARD_FUNCTION_MEMORY:
printf("memory");
break;
case PCCARD_FUNCTION_SERIAL:
printf("serial port");
break;
case PCCARD_FUNCTION_PARALLEL:
printf("parallel port");
break;
case PCCARD_FUNCTION_DISK:
printf("fixed disk");
break;
case PCCARD_FUNCTION_VIDEO:
printf("video adapter");
break;
case PCCARD_FUNCTION_NETWORK:
printf("network adapter");
break;
case PCCARD_FUNCTION_AIMS:
printf("auto incrementing mass storage");
break;
case PCCARD_FUNCTION_SCSI:
printf("SCSI bridge");
break;
case PCCARD_FUNCTION_SECURITY:
printf("Security services");
break;
case PCCARD_FUNCTION_INSTRUMENT:
printf("Instrument");
break;
default:
printf("unknown (%d)", pf->function);
break;
}
Fix printf() format errors. Reviewed by: imp
2002-11-14 14:02:32 +00:00
printf(", ccr addr %x mask %x\n", pf->ccr_base, pf->ccr_mask);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
STAILQ_FOREACH(cfe, &pf->cfe_head, cfe_list) {
device_printf(dev, "function %d, config table entry "
"%d: ", pf->number, cfe->number);
switch (cfe->iftype) {
case PCCARD_IFTYPE_MEMORY:
printf("memory card");
break;
case PCCARD_IFTYPE_IO:
printf("I/O card");
break;
default:
printf("card type unknown");
break;
}
printf("; irq mask %x", cfe->irqmask);
if (cfe->num_iospace) {
printf("; iomask %lx, iospace", cfe->iomask);
Update pccard code to latest NetBSD code. This is the last merge before newbusification hits full steam ahead. All: Adjust NetBSD labels to reflect new base versions. dev/pcic/i82365.c: 1.24 Interface change for kernel threads 1.25 Massive unification for cardbus dev/pcic/i82365var.h 1.8 Massive unification for cardbus dev/pcic/i82365_isasubr.c 1.3 Massive unification for cardbus dev/pccard/pccard_cis.c 1.11 Massive unification for cardbus (better device printing, better memspace calcs) dev/pccard/pccard_cis_quirks.c 1.4,1.5 Lotsa 3com devices dev/pccard/pccardchip.h 1.4 Massive unification for cardbus dev/pccard/pccarddevs 1.33..1.59 Lots of devices
1999-11-28 05:49:27 +00:00
for (i = 0; i < cfe->num_iospace; i++) {
printf(" %lx", cfe->iospace[i].start);
if (cfe->iospace[i].length)
printf("-%lx",
cfe->iospace[i].start +
cfe->iospace[i].length - 1);
}
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
}
if (cfe->num_memspace) {
printf("; memspace");
Update pccard code to latest NetBSD code. This is the last merge before newbusification hits full steam ahead. All: Adjust NetBSD labels to reflect new base versions. dev/pcic/i82365.c: 1.24 Interface change for kernel threads 1.25 Massive unification for cardbus dev/pcic/i82365var.h 1.8 Massive unification for cardbus dev/pcic/i82365_isasubr.c 1.3 Massive unification for cardbus dev/pccard/pccard_cis.c 1.11 Massive unification for cardbus (better device printing, better memspace calcs) dev/pccard/pccard_cis_quirks.c 1.4,1.5 Lotsa 3com devices dev/pccard/pccardchip.h 1.4 Massive unification for cardbus dev/pccard/pccarddevs 1.33..1.59 Lots of devices
1999-11-28 05:49:27 +00:00
for (i = 0; i < cfe->num_memspace; i++) {
printf(" %lx",
cfe->memspace[i].cardaddr);
if (cfe->memspace[i].length)
printf("-%lx",
cfe->memspace[i].cardaddr +
cfe->memspace[i].length - 1);
if (cfe->memspace[i].hostaddr)
printf("@%lx",
cfe->memspace[i].hostaddr);
}
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
}
if (cfe->maxtwins)
printf("; maxtwins %d", cfe->maxtwins);
printf(";");
if (cfe->flags & PCCARD_CFE_MWAIT_REQUIRED)
printf(" mwait_required");
if (cfe->flags & PCCARD_CFE_RDYBSY_ACTIVE)
printf(" rdybsy_active");
if (cfe->flags & PCCARD_CFE_WP_ACTIVE)
printf(" wp_active");
if (cfe->flags & PCCARD_CFE_BVD_ACTIVE)
printf(" bvd_active");
if (cfe->flags & PCCARD_CFE_IO8)
printf(" io8");
if (cfe->flags & PCCARD_CFE_IO16)
printf(" io16");
if (cfe->flags & PCCARD_CFE_IRQSHARE)
printf(" irqshare");
if (cfe->flags & PCCARD_CFE_IRQPULSE)
printf(" irqpulse");
if (cfe->flags & PCCARD_CFE_IRQLEVEL)
printf(" irqlevel");
if (cfe->flags & PCCARD_CFE_POWERDOWN)
printf(" powerdown");
if (cfe->flags & PCCARD_CFE_READONLY)
printf(" readonly");
if (cfe->flags & PCCARD_CFE_AUDIO)
printf(" audio");
printf("\n");
}
}
if (card->error)
device_printf(dev, "%d errors found while parsing CIS\n",
card->error);
}
Add a much-requested feature: The ability for pccard attachments to scan the CIS for interesting tuples. 95% of what can be obtained from the CIS is harvested by the pccard layer and presented to the user in standard function calls. However, there are special needs at times where the standard stuff doesn't suffice. This is for those special cases. CARD_SCAN_CIS(device_get_parent(dev), function, argp) scans the CIS of the card, passing each tuple to function with the tuple and argp as its arguments. Returning 0 continues the scan, while returning 1 terminates the scan. The value of the last invocation of function is returned from this function. int (*pccard_scan_t)(struct pccard_tuple *tuple, void *argp) function called for each tuple. Elements of the CIS tuple can be read with pccard_tuple_read_{1,2,3,4,n}(). You are reading the actual tuple memory each time, in case your card has registers in the CIS. # I suppose these things should be documented in pccard(4) or something like # that. # I plan on unifying cardbus CIS support in a similar way. Approved by: re (scottl)
2005-07-01 03:40:28 +00:00
static int
Upon relection, we shouldn't allow the tuple structs to be modified by the functor, so make it a const pointer, and chase down the resulting const-poisoning. Approved by: re (scottl)
2005-07-01 15:52:50 +00:00
pccard_parse_cis_tuple(const struct pccard_tuple *tuple, void *arg)
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
{
/* most of these are educated guesses */
static struct pccard_config_entry init_cfe = {
-1, PCCARD_CFE_RDYBSY_ACTIVE | PCCARD_CFE_WP_ACTIVE |
PCCARD_CFE_BVD_ACTIVE, PCCARD_IFTYPE_MEMORY,
};
struct cis_state *state = arg;
switch (tuple->code) {
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_END:
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
/* if we've seen a LONGLINK_MFC, and this is the first
* END after it, reset the function list.
*
* XXX This might also be the right place to start a
* new function, but that assumes that a function
* definition never crosses any longlink, and I'm not
* sure about that. This is probably safe for MFC
* cards, but what we have now isn't broken, so I'd
* rather not change it.
*/
if (state->gotmfc == 1) {
struct pccard_function *pf, *pfnext;
for (pf = STAILQ_FIRST(&state->card->pf_head);
pf != NULL; pf = pfnext) {
pfnext = STAILQ_NEXT(pf, pf_list);
free(pf, M_DEVBUF);
}
STAILQ_INIT(&state->card->pf_head);
state->count = 0;
state->gotmfc = 2;
state->pf = NULL;
}
break;
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_LONGLINK_MFC:
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
/*
* this tuple's structure was dealt with in scan_cis. here,
* record the fact that the MFC tuple was seen, so that
* functions declared before the MFC link can be cleaned
* up.
*/
state->gotmfc = 1;
break;
#ifdef PCCARDCISDEBUG
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_DEVICE:
case CISTPL_DEVICE_A:
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
{
u_int reg, dtype, dspeed;
reg = pccard_tuple_read_1(tuple, 0);
dtype = reg & PCCARD_DTYPE_MASK;
dspeed = reg & PCCARD_DSPEED_MASK;
DPRINTF(("CISTPL_DEVICE%s type=",
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
(tuple->code == CISTPL_DEVICE) ? "" : "_A"));
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
switch (dtype) {
case PCCARD_DTYPE_NULL:
DPRINTF(("null"));
break;
case PCCARD_DTYPE_ROM:
DPRINTF(("rom"));
break;
case PCCARD_DTYPE_OTPROM:
DPRINTF(("otprom"));
break;
case PCCARD_DTYPE_EPROM:
DPRINTF(("eprom"));
break;
case PCCARD_DTYPE_EEPROM:
DPRINTF(("eeprom"));
break;
case PCCARD_DTYPE_FLASH:
DPRINTF(("flash"));
break;
case PCCARD_DTYPE_SRAM:
DPRINTF(("sram"));
break;
case PCCARD_DTYPE_DRAM:
DPRINTF(("dram"));
break;
case PCCARD_DTYPE_FUNCSPEC:
DPRINTF(("funcspec"));
break;
case PCCARD_DTYPE_EXTEND:
DPRINTF(("extend"));
break;
default:
DPRINTF(("reserved"));
break;
}
DPRINTF((" speed="));
switch (dspeed) {
case PCCARD_DSPEED_NULL:
DPRINTF(("null"));
break;
case PCCARD_DSPEED_250NS:
DPRINTF(("250ns"));
break;
case PCCARD_DSPEED_200NS:
DPRINTF(("200ns"));
break;
case PCCARD_DSPEED_150NS:
DPRINTF(("150ns"));
break;
case PCCARD_DSPEED_100NS:
DPRINTF(("100ns"));
break;
case PCCARD_DSPEED_EXT:
DPRINTF(("ext"));
break;
default:
DPRINTF(("reserved"));
break;
}
}
DPRINTF(("\n"));
break;
#endif
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_VERS_1:
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (tuple->length < 6) {
DPRINTF(("CISTPL_VERS_1 too short %d\n",
tuple->length));
break;
} {
int start, i, ch, count;
state->card->cis1_major = pccard_tuple_read_1(tuple, 0);
state->card->cis1_minor = pccard_tuple_read_1(tuple, 1);
for (count = 0, start = 0, i = 0;
(count < 4) && ((i + 4) < 256); i++) {
ch = pccard_tuple_read_1(tuple, 2 + i);
if (ch == 0xff)
break;
state->card->cis1_info_buf[i] = ch;
if (ch == 0) {
state->card->cis1_info[count] =
state->card->cis1_info_buf + start;
start = i + 1;
count++;
}
}
DPRINTF(("CISTPL_VERS_1\n"));
}
break;
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_MANFID:
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (tuple->length < 4) {
DPRINTF(("CISTPL_MANFID too short %d\n",
tuple->length));
break;
}
state->card->manufacturer = pccard_tuple_read_2(tuple, 0);
state->card->product = pccard_tuple_read_2(tuple, 2);
Add u_int16 prodext value in CISTPL_MANF_ID. This gets a fifth byte when manufacturer id tuple length is 5. This change is for xe driver. This is a dirty hack. But there is no better idea. Reviewd by: imp
2002-02-20 14:30:46 +00:00
/*
Better comment for the product ID thing.
2002-10-07 06:35:04 +00:00
* This is for xe driver. But not limited to that driver.
Add u_int16 prodext value in CISTPL_MANF_ID. This gets a fifth byte when manufacturer id tuple length is 5. This change is for xe driver. This is a dirty hack. But there is no better idea. Reviewd by: imp
2002-02-20 14:30:46 +00:00
* In PC Card Standard,
* Manufacturer ID: 2byte.
Better comment for the product ID thing.
2002-10-07 06:35:04 +00:00
* Product ID: typically 2bytes, but there's no limit on its
* size. prodext is a two byte field, so maybe we should
* also handle the '6' case. So far no cards have surfaced
* with a length of '6'.
Add u_int16 prodext value in CISTPL_MANF_ID. This gets a fifth byte when manufacturer id tuple length is 5. This change is for xe driver. This is a dirty hack. But there is no better idea. Reviewd by: imp
2002-02-20 14:30:46 +00:00
*/
The supposed OLD STYLE network MAC id tuple was really just a buggy expression in the card in question. Since that driver uses a different mechanism, retire the workaround for this bug.
2005-07-13 14:59:06 +00:00
if (tuple->length == 5 )
Add u_int16 prodext value in CISTPL_MANF_ID. This gets a fifth byte when manufacturer id tuple length is 5. This change is for xe driver. This is a dirty hack. But there is no better idea. Reviewd by: imp
2002-02-20 14:30:46 +00:00
state->card->prodext = pccard_tuple_read_1(tuple, 4);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
DPRINTF(("CISTPL_MANFID\n"));
break;
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_FUNCID:
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (tuple->length < 1) {
DPRINTF(("CISTPL_FUNCID too short %d\n",
tuple->length));
break;
}
if ((state->pf == NULL) || (state->gotmfc == 2)) {
state->pf = malloc(sizeof(*state->pf), M_DEVBUF,
Make a few more mallocs use M_ZERO. Submitted by: josh@zipperup.org Submitted by: Robert Drehmel <robd@gmx.net> Approved by: imp
2000-10-29 16:29:05 +00:00
M_NOWAIT | M_ZERO);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
state->pf->number = state->count++;
state->pf->last_config_index = -1;
STAILQ_INIT(&state->pf->cfe_head);
STAILQ_INSERT_TAIL(&state->card->pf_head, state->pf,
pf_list);
}
state->pf->function = pccard_tuple_read_1(tuple, 0);
DPRINTF(("CISTPL_FUNCID\n"));
break;
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_FUNCE:
Update cis tuple parser, add a pccarddevs entry, and improve PCCARD_IVAR_ETHADDR in pccard_read_ivar(). Change points: (1) Read Function Ext tuple. (2) Add Ratoc REX-R280 entry(fe driver). (3) Take ether address from function ext tuple. Reviewed by: imp Obtained from: NetBSD
2001-09-02 06:37:41 +00:00
if (state->pf == NULL || state->pf->function <= 0) {
DPRINTF(("CISTPL_FUNCE is not followed by "
"valid CISTPL_FUNCID\n"));
break;
}
Upon relection, we shouldn't allow the tuple structs to be modified by the functor, so make it a const pointer, and chase down the resulting const-poisoning. Approved by: re (scottl)
2005-07-01 15:52:50 +00:00
if (tuple->length >= 2)
Update cis tuple parser, add a pccarddevs entry, and improve PCCARD_IVAR_ETHADDR in pccard_read_ivar(). Change points: (1) Read Function Ext tuple. (2) Add Ratoc REX-R280 entry(fe driver). (3) Take ether address from function ext tuple. Reviewed by: imp Obtained from: NetBSD
2001-09-02 06:37:41 +00:00
decode_funce(tuple, state->pf);
DPRINTF(("CISTPL_FUNCE\n"));
break;
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_CONFIG:
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (tuple->length < 3) {
DPRINTF(("CISTPL_CONFIG too short %d\n",
tuple->length));
break;
} {
u_int reg, rasz, rmsz, rfsz;
int i;
reg = pccard_tuple_read_1(tuple, 0);
rasz = 1 + ((reg & PCCARD_TPCC_RASZ_MASK) >>
PCCARD_TPCC_RASZ_SHIFT);
rmsz = 1 + ((reg & PCCARD_TPCC_RMSZ_MASK) >>
PCCARD_TPCC_RMSZ_SHIFT);
rfsz = ((reg & PCCARD_TPCC_RFSZ_MASK) >>
PCCARD_TPCC_RFSZ_SHIFT);
if (tuple->length < (rasz + rmsz + rfsz)) {
DPRINTF(("CISTPL_CONFIG (%d,%d,%d) too "
"short %d\n", rasz, rmsz, rfsz,
tuple->length));
break;
}
if (state->pf == NULL) {
state->pf = malloc(sizeof(*state->pf),
Make a few more mallocs use M_ZERO. Submitted by: josh@zipperup.org Submitted by: Robert Drehmel <robd@gmx.net> Approved by: imp
2000-10-29 16:29:05 +00:00
M_DEVBUF, M_NOWAIT | M_ZERO);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
state->pf->number = state->count++;
state->pf->last_config_index = -1;
STAILQ_INIT(&state->pf->cfe_head);
STAILQ_INSERT_TAIL(&state->card->pf_head,
state->pf, pf_list);
state->pf->function = PCCARD_FUNCTION_UNSPEC;
}
state->pf->last_config_index =
pccard_tuple_read_1(tuple, 1);
state->pf->ccr_base = 0;
for (i = 0; i < rasz; i++)
state->pf->ccr_base |=
((pccard_tuple_read_1(tuple, 2 + i)) <<
(i * 8));
state->pf->ccr_mask = 0;
for (i = 0; i < rmsz; i++)
state->pf->ccr_mask |=
((pccard_tuple_read_1(tuple,
2 + rasz + i)) << (i * 8));
/* skip the reserved area and subtuples */
/* reset the default cfe for each cfe list */
state->temp_cfe = init_cfe;
state->default_cfe = &state->temp_cfe;
}
DPRINTF(("CISTPL_CONFIG\n"));
break;
o move the cis tuple definitions into a common file. o minor optimization of cardbus_cis processing. Remove a bunch of generic entries that are handled by generic. o no longer need the card_get_type stuff.
2003-10-07 03:33:54 +00:00
case CISTPL_CFTABLE_ENTRY:
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
{
int idx, i, j;
u_int reg, reg2;
u_int intface, def, num;
u_int power, timing, iospace, irq, memspace, misc;
struct pccard_config_entry *cfe;
idx = 0;
reg = pccard_tuple_read_1(tuple, idx);
idx++;
intface = reg & PCCARD_TPCE_INDX_INTFACE;
def = reg & PCCARD_TPCE_INDX_DEFAULT;
num = reg & PCCARD_TPCE_INDX_NUM_MASK;
/*
* this is a little messy. Some cards have only a
* cfentry with the default bit set. So, as we go
* through the list, we add new indexes to the queue,
* and keep a pointer to the last one with the
* default bit set. if we see a record with the same
* index, as the default, we stash the default and
* replace the queue entry. otherwise, we just add
* new entries to the queue, pointing the default ptr
* at them if the default bit is set. if we get to
* the end with the default pointer pointing at a
* record which hasn't had a matching index, that's
* ok; it just becomes a cfentry like any other.
*/
/*
* if the index in the cis differs from the default
* cis, create new entry in the queue and start it
* with the current default
*/
if (num != state->default_cfe->number) {
cfe = (struct pccard_config_entry *)
malloc(sizeof(*cfe), M_DEVBUF, M_NOWAIT);
deal with malloc failure Noticed by: Coverity Prevent analysis tool
2005-03-26 21:34:12 +00:00
if (cfe == NULL) {
DPRINTF(("no memory for config entry\n"));
goto abort_cfe;
}
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
*cfe = *state->default_cfe;
STAILQ_INSERT_TAIL(&state->pf->cfe_head,
cfe, cfe_list);
cfe->number = num;
/*
* if the default bit is set in the cis, then
* point the new default at whatever is being
* filled in
*/
if (def)
state->default_cfe = cfe;
} else {
/*
* the cis index matches the default index,
* fill in the default cfentry. It is
* assumed that the cfdefault index is in the
* queue. For it to be otherwise, the cis
* index would have to be -1 (initial
* condition) which is not possible, or there
* would have to be a preceding cis entry
* which had the same cis index and had the
* default bit unset. Neither condition
* should happen. If it does, this cfentry
* is lost (written into temp space), which
* is an acceptable failure mode.
*/
cfe = state->default_cfe;
/*
* if the cis entry does not have the default
* bit set, copy the default out of the way
* first.
*/
if (!def) {
state->temp_cfe = *state->default_cfe;
state->default_cfe = &state->temp_cfe;
}
}
if (intface) {
reg = pccard_tuple_read_1(tuple, idx);
idx++;
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
cfe->flags &= ~(PCCARD_CFE_MWAIT_REQUIRED
| PCCARD_CFE_RDYBSY_ACTIVE
| PCCARD_CFE_WP_ACTIVE
| PCCARD_CFE_BVD_ACTIVE);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (reg & PCCARD_TPCE_IF_MWAIT)
cfe->flags |= PCCARD_CFE_MWAIT_REQUIRED;
if (reg & PCCARD_TPCE_IF_RDYBSY)
cfe->flags |= PCCARD_CFE_RDYBSY_ACTIVE;
if (reg & PCCARD_TPCE_IF_WP)
cfe->flags |= PCCARD_CFE_WP_ACTIVE;
if (reg & PCCARD_TPCE_IF_BVD)
cfe->flags |= PCCARD_CFE_BVD_ACTIVE;
cfe->iftype = reg & PCCARD_TPCE_IF_IFTYPE;
}
reg = pccard_tuple_read_1(tuple, idx);
idx++;
power = reg & PCCARD_TPCE_FS_POWER_MASK;
timing = reg & PCCARD_TPCE_FS_TIMING;
iospace = reg & PCCARD_TPCE_FS_IOSPACE;
irq = reg & PCCARD_TPCE_FS_IRQ;
memspace = reg & PCCARD_TPCE_FS_MEMSPACE_MASK;
misc = reg & PCCARD_TPCE_FS_MISC;
if (power) {
/* skip over power, don't save */
/* for each parameter selection byte */
for (i = 0; i < power; i++) {
reg = pccard_tuple_read_1(tuple, idx);
idx++;
/* for each bit */
for (j = 0; j < 7; j++) {
/* if the bit is set */
if ((reg >> j) & 0x01) {
/* skip over bytes */
do {
reg2 = pccard_tuple_read_1(tuple, idx);
idx++;
/*
* until
* non-extensi
* on byte
*/
} while (reg2 & 0x80);
}
}
}
}
if (timing) {
/* skip over timing, don't save */
reg = pccard_tuple_read_1(tuple, idx);
idx++;
if ((reg & PCCARD_TPCE_TD_RESERVED_MASK) !=
PCCARD_TPCE_TD_RESERVED_MASK)
idx++;
if ((reg & PCCARD_TPCE_TD_RDYBSY_MASK) !=
PCCARD_TPCE_TD_RDYBSY_MASK)
idx++;
if ((reg & PCCARD_TPCE_TD_WAIT_MASK) !=
PCCARD_TPCE_TD_WAIT_MASK)
idx++;
}
if (iospace) {
if (tuple->length <= idx) {
DPRINTF(("ran out of space before TCPE_IO\n"));
goto abort_cfe;
}
reg = pccard_tuple_read_1(tuple, idx);
idx++;
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
cfe->flags &=
~(PCCARD_CFE_IO8 | PCCARD_CFE_IO16);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (reg & PCCARD_TPCE_IO_BUSWIDTH_8BIT)
cfe->flags |= PCCARD_CFE_IO8;
if (reg & PCCARD_TPCE_IO_BUSWIDTH_16BIT)
cfe->flags |= PCCARD_CFE_IO16;
cfe->iomask =
reg & PCCARD_TPCE_IO_IOADDRLINES_MASK;
if (reg & PCCARD_TPCE_IO_HASRANGE) {
reg = pccard_tuple_read_1(tuple, idx);
idx++;
cfe->num_iospace = 1 + (reg &
PCCARD_TPCE_IO_RANGE_COUNT);
if (cfe->num_iospace >
(sizeof(cfe->iospace) /
sizeof(cfe->iospace[0]))) {
DPRINTF(("too many io "
"spaces %d",
cfe->num_iospace));
state->card->error++;
break;
}
for (i = 0; i < cfe->num_iospace; i++) {
switch (reg & PCCARD_TPCE_IO_RANGE_ADDRSIZE_MASK) {
case PCCARD_TPCE_IO_RANGE_ADDRSIZE_ONE:
cfe->iospace[i].start =
pccard_tuple_read_1(tuple, idx);
idx++;
break;
case PCCARD_TPCE_IO_RANGE_ADDRSIZE_TWO:
cfe->iospace[i].start =
pccard_tuple_read_2(tuple, idx);
idx += 2;
break;
case PCCARD_TPCE_IO_RANGE_ADDRSIZE_FOUR:
cfe->iospace[i].start =
pccard_tuple_read_4(tuple, idx);
idx += 4;
break;
}
switch (reg &
PCCARD_TPCE_IO_RANGE_LENGTHSIZE_MASK) {
case PCCARD_TPCE_IO_RANGE_LENGTHSIZE_ONE:
cfe->iospace[i].length =
pccard_tuple_read_1(tuple, idx);
idx++;
break;
case PCCARD_TPCE_IO_RANGE_LENGTHSIZE_TWO:
cfe->iospace[i].length =
pccard_tuple_read_2(tuple, idx);
idx += 2;
break;
case PCCARD_TPCE_IO_RANGE_LENGTHSIZE_FOUR:
cfe->iospace[i].length =
pccard_tuple_read_4(tuple, idx);
idx += 4;
break;
}
cfe->iospace[i].length++;
}
} else {
cfe->num_iospace = 1;
cfe->iospace[0].start = 0;
cfe->iospace[0].length =
(1 << cfe->iomask);
}
}
if (irq) {
if (tuple->length <= idx) {
DPRINTF(("ran out of space before TCPE_IR\n"));
goto abort_cfe;
}
reg = pccard_tuple_read_1(tuple, idx);
idx++;
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
cfe->flags &= ~(PCCARD_CFE_IRQSHARE
| PCCARD_CFE_IRQPULSE
| PCCARD_CFE_IRQLEVEL);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (reg & PCCARD_TPCE_IR_SHARE)
cfe->flags |= PCCARD_CFE_IRQSHARE;
if (reg & PCCARD_TPCE_IR_PULSE)
cfe->flags |= PCCARD_CFE_IRQPULSE;
if (reg & PCCARD_TPCE_IR_LEVEL)
cfe->flags |= PCCARD_CFE_IRQLEVEL;
if (reg & PCCARD_TPCE_IR_HASMASK) {
/*
* it's legal to ignore the
* special-interrupt bits, so I will
*/
cfe->irqmask =
pccard_tuple_read_2(tuple, idx);
idx += 2;
} else {
cfe->irqmask =
(1 << (reg & PCCARD_TPCE_IR_IRQ));
}
}
if (memspace) {
if (tuple->length <= idx) {
DPRINTF(("ran out of space before TCPE_MS\n"));
goto abort_cfe;
}
memspace is set to some value by masking off bits. When these bits are equal to PCCARD_TPCE_FS_MEMSPACE_NONE, memspace will be zero, so testing for this case inside of the if statement results in dead code. We'd fail to set a value to zero that's already zero (since it is initialized to 0 indirectly) with this code being there. Well, except in the very rare case that we have a card that has a defualt entry that includes a memory space followed by one that has no memory space (these are extremely rare, I don't recall ever having seen one :-). Fix this by setting num_memspace to 0 in a more appropriate place. Submitted by: Coverity Prevent analysis tool
2005-02-17 21:05:04 +00:00
if (memspace == PCCARD_TPCE_FS_MEMSPACE_LENGTH) {
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
cfe->num_memspace = 1;
cfe->memspace[0].length = 256 *
pccard_tuple_read_2(tuple, idx);
idx += 2;
cfe->memspace[0].cardaddr = 0;
cfe->memspace[0].hostaddr = 0;
} else if (memspace ==
PCCARD_TPCE_FS_MEMSPACE_LENGTHADDR) {
cfe->num_memspace = 1;
cfe->memspace[0].length = 256 *
pccard_tuple_read_2(tuple, idx);
idx += 2;
cfe->memspace[0].cardaddr = 256 *
pccard_tuple_read_2(tuple, idx);
idx += 2;
cfe->memspace[0].hostaddr = cfe->memspace[0].cardaddr;
} else {
int lengthsize;
int cardaddrsize;
int hostaddrsize;
reg = pccard_tuple_read_1(tuple, idx);
idx++;
Flesh out the pccard bus_ methods with either the generic one (where it would work), or a specialized one. Most of these have been creatively stolen from pccard_nkb, which in turn stole from isa showing that generic bus_ versions of bus_{set,get,delete}_resource might be profitable. Fix a couple of minor bugs introduced in the last round of updates from NetBSD. Start on the pccard_ivar structure which will hold the resources and slot number. Add tcic as a possible attachment for pccard and rename the attachment for pcicx to pcic since the name has changed since I originally wrote this stuff. Next up: stringing together the various memory and I/O allocation/mapping primitives in i82365.c, final touches on the isa attach routine and other fun stuff in that line of attach.
1999-11-29 06:42:55 +00:00
cfe->num_memspace = (reg &
PCCARD_TPCE_MS_COUNT) + 1;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (cfe->num_memspace >
(sizeof(cfe->memspace) /
sizeof(cfe->memspace[0]))) {
DPRINTF(("too many mem "
"spaces %d",
cfe->num_memspace));
state->card->error++;
break;
}
lengthsize =
((reg & PCCARD_TPCE_MS_LENGTH_SIZE_MASK) >>
PCCARD_TPCE_MS_LENGTH_SIZE_SHIFT);
cardaddrsize =
((reg & PCCARD_TPCE_MS_CARDADDR_SIZE_MASK) >>
PCCARD_TPCE_MS_CARDADDR_SIZE_SHIFT);
hostaddrsize =
(reg & PCCARD_TPCE_MS_HOSTADDR) ? cardaddrsize : 0;
if (lengthsize == 0) {
DPRINTF(("cfe memspace "
"lengthsize == 0"));
state->card->error++;
}
for (i = 0; i < cfe->num_memspace; i++) {
if (lengthsize) {
cfe->memspace[i].length =
256 * pccard_tuple_read_n(tuple, lengthsize,
idx);
idx += lengthsize;
} else {
cfe->memspace[i].length = 0;
}
if (cfe->memspace[i].length == 0) {
DPRINTF(("cfe->memspace[%d].length == 0",
i));
state->card->error++;
}
if (cardaddrsize) {
cfe->memspace[i].cardaddr =
256 * pccard_tuple_read_n(tuple, cardaddrsize,
idx);
idx += cardaddrsize;
} else {
cfe->memspace[i].cardaddr = 0;
}
if (hostaddrsize) {
cfe->memspace[i].hostaddr =
256 * pccard_tuple_read_n(tuple, hostaddrsize,
idx);
idx += hostaddrsize;
} else {
cfe->memspace[i].hostaddr = 0;
}
}
}
memspace is set to some value by masking off bits. When these bits are equal to PCCARD_TPCE_FS_MEMSPACE_NONE, memspace will be zero, so testing for this case inside of the if statement results in dead code. We'd fail to set a value to zero that's already zero (since it is initialized to 0 indirectly) with this code being there. Well, except in the very rare case that we have a card that has a defualt entry that includes a memory space followed by one that has no memory space (these are extremely rare, I don't recall ever having seen one :-). Fix this by setting num_memspace to 0 in a more appropriate place. Submitted by: Coverity Prevent analysis tool
2005-02-17 21:05:04 +00:00
} else
cfe->num_memspace = 0;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (misc) {
if (tuple->length <= idx) {
DPRINTF(("ran out of space before TCPE_MI\n"));
goto abort_cfe;
}
reg = pccard_tuple_read_1(tuple, idx);
idx++;
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
cfe->flags &= ~(PCCARD_CFE_POWERDOWN
| PCCARD_CFE_READONLY
| PCCARD_CFE_AUDIO);
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (reg & PCCARD_TPCE_MI_PWRDOWN)
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
cfe->flags |= PCCARD_CFE_POWERDOWN;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (reg & PCCARD_TPCE_MI_READONLY)
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
cfe->flags |= PCCARD_CFE_READONLY;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
if (reg & PCCARD_TPCE_MI_AUDIO)
Merge changes from NetBSD through version 1.17 of this file. These give us slightly better error checking than before and interpret what default bits mean better. See the NetBSD CVS tree for the authors of these changes (revs 1.10 .. 1.17).
2002-10-07 23:03:17 +00:00
cfe->flags |= PCCARD_CFE_AUDIO;
Moderately hacked pccard code from newconfig. It is somewhat in incomplete and likely has problem. The code was originally pcmcia, but I renamed it to pccard and made it compile on FreeBSD -current. I converted SIMPLEQ to STAILQ as well as a few sc->dev.xname -> device_printf changes. This is a green port of fairly mature code. I derived this work from the FreeBSD newconfig project (http://www.jp.freebsd.org/newconfig). Any problems with it are likely introduced by me. Obtained from: newconfig project
1999-10-26 06:52:31 +00:00
cfe->maxtwins = reg & PCCARD_TPCE_MI_MAXTWINS;
while (reg & PCCARD_TPCE_MI_EXT) {
reg = pccard_tuple_read_1(tuple, idx);
idx++;
}
}
/* skip all the subtuples */
}
abort_cfe:
DPRINTF(("CISTPL_CFTABLE_ENTRY\n"));
break;
default:
DPRINTF(("unhandled CISTPL %x\n", tuple->code));
break;
}
return (0);
}
Update cis tuple parser, add a pccarddevs entry, and improve PCCARD_IVAR_ETHADDR in pccard_read_ivar(). Change points: (1) Read Function Ext tuple. (2) Add Ratoc REX-R280 entry(fe driver). (3) Take ether address from function ext tuple. Reviewed by: imp Obtained from: NetBSD
2001-09-02 06:37:41 +00:00
static int
Upon relection, we shouldn't allow the tuple structs to be modified by the functor, so make it a const pointer, and chase down the resulting const-poisoning. Approved by: re (scottl)
2005-07-01 15:52:50 +00:00
decode_funce(const struct pccard_tuple *tuple, struct pccard_function *pf)
Update cis tuple parser, add a pccarddevs entry, and improve PCCARD_IVAR_ETHADDR in pccard_read_ivar(). Change points: (1) Read Function Ext tuple. (2) Add Ratoc REX-R280 entry(fe driver). (3) Take ether address from function ext tuple. Reviewed by: imp Obtained from: NetBSD
2001-09-02 06:37:41 +00:00
{
Some older PC Cards have a weird format for FUNCE tuples. They appear as type 0, rather than the usualy type 4. Assume that this format is from an old standard and go with it. The Fujitsu FMV-186A and Silicom Ethernet cards I have both have tuples with this format, and they are both pretty old cards. # if somebody knows for sure, please let me know.
2005-01-21 02:11:48 +00:00
int i;
int len;
Update cis tuple parser, add a pccarddevs entry, and improve PCCARD_IVAR_ETHADDR in pccard_read_ivar(). Change points: (1) Read Function Ext tuple. (2) Add Ratoc REX-R280 entry(fe driver). (3) Take ether address from function ext tuple. Reviewed by: imp Obtained from: NetBSD
2001-09-02 06:37:41 +00:00
int type = pccard_tuple_read_1(tuple, 0);
switch (pf->function) {
case PCCARD_FUNCTION_DISK:
if (type == PCCARD_TPLFE_TYPE_DISK_DEVICE_INTERFACE) {
pf->pf_funce_disk_interface
= pccard_tuple_read_1(tuple, 1);
}
break;
case PCCARD_FUNCTION_NETWORK:
if (type == PCCARD_TPLFE_TYPE_LAN_NID) {
Some older PC Cards have a weird format for FUNCE tuples. They appear as type 0, rather than the usualy type 4. Assume that this format is from an old standard and go with it. The Fujitsu FMV-186A and Silicom Ethernet cards I have both have tuples with this format, and they are both pretty old cards. # if somebody knows for sure, please let me know.
2005-01-21 02:11:48 +00:00
len = pccard_tuple_read_1(tuple, 1);
Update cis tuple parser, add a pccarddevs entry, and improve PCCARD_IVAR_ETHADDR in pccard_read_ivar(). Change points: (1) Read Function Ext tuple. (2) Add Ratoc REX-R280 entry(fe driver). (3) Take ether address from function ext tuple. Reviewed by: imp Obtained from: NetBSD
2001-09-02 06:37:41 +00:00
if (tuple->length < 2 + len || len > 8) {
/* tuple length not enough or nid too long */
break;
}
for (i = 0; i < len; i++) {
pf->pf_funce_lan_nid[i]
= pccard_tuple_read_1(tuple, i + 2);
}
pf->pf_funce_lan_nidlen = len;
}
break;
default:
break;
}
return 0;
}
Reference in New Issue Copy Permalink