d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
bcf77694d1
freebsd-dev/sys/dev/pccard/pccard.c

1134 lines
30 KiB
C
Raw Normal View History

o Merge in changes to the NetBSD sources: pcmciavar.h 1.9->1.12 1.12, enami, minor coding nits 1.11, augustss, (pcmcia_devinfo NRFB) 1.10, cgd, add generic lookup routines pcmcia.c 1.14->1.23 1.23, drochner, (probe code printing, NRFB) 1.22, augustss, KNF 1.21, uch, (hpcmips tweaks NRFB) 1.20, chopps, remove bogus debug 1.19, enami, minor coding nits 1.18, augustss, (pcmcia_devinfo NRFB) 1.17, nathanw, LP64 printf fixes 1.16, cgd, add generic lookup routines 1.15, aymeric, printf fixes NRFB == not relevant to freebsd o Expand the pccard matching routines to include the ability to match against the CIS strings since our current driver database is based on that. o Add lots more ivars to get the information necessary to snag these values.
2000-09-22 01:15:26 +00:00
/* $NetBSD: pcmcia.c,v 1.23 2000/07/28 19:17:02 drochner 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
/* $FreeBSD$ */
/*
* 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>
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
#include "power_if.h"
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"
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
#define PCCARDDEBUG
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 PCCARDDEBUG
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
int pccard_debug = 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
#define DPRINTF(arg) if (pccard_debug) printf arg
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
#define DEVPRINTF(arg) if (pccard_debug) device_printf arg
o Fix memory leak in ivars o Change name of bus o Change the panic on resource allocation failure to just a message. We'll work out why this fails later in the pcic/pccbb code merge.
2000-10-19 22:36:41 +00:00
#define PRVERBOSE(arg) printf arg
#define DEVPRVERBOSE(arg) device_printf 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
#else
#define DPRINTF(arg)
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
#define DEVPRINTF(arg)
o Fix memory leak in ivars o Change name of bus o Change the panic on resource allocation failure to just a message. We'll work out why this fails later in the pcic/pccbb code merge.
2000-10-19 22:36:41 +00:00
#define PRVERBOSE(arg) if (bootverbose) printf arg
#define DEVPRVERBOSE(arg) if (bootverbose) device_printf 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
#ifdef PCCARDVERBOSE
int pccard_verbose = 1;
#else
int pccard_verbose = 0;
#endif
Tonight's cleanups. o Implement memory and I/O activation/deactivation. irq not handled. o switch pcic_chip functions around to use more convenient types. o kill __P and most of the old K&R prototypes just to be mean. o minor other nits
1999-12-07 06:44:38 +00:00
int pccard_print(void *, const char *);
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
pccard_ccr_read(pf, ccr)
struct pccard_function *pf;
int ccr;
{
return (bus_space_read_1(pf->pf_ccrt, pf->pf_ccrh,
pf->pf_ccr_offset + ccr));
}
void
pccard_ccr_write(pf, ccr, val)
struct pccard_function *pf;
int ccr;
int val;
{
if ((pf->ccr_mask) & (1 << (ccr / 2))) {
bus_space_write_1(pf->pf_ccrt, pf->pf_ccrh,
pf->pf_ccr_offset + ccr, val);
}
}
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
static int
pccard_attach_card(device_t 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
{
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_function *pf;
Change the interface to pccard_function_init. The interface is such that it looks for an acceptible one. Once it finds it, it should set the resources for the device. I say "should" because I've not written that. Also set an ivar for the child of pccard. Minor fix to the attach message printed, we lose the slot number, which I'll have to restore later. Adjust the pccard ivar so that we can save the function that corresponds to this driver so we can enable and disable it more easily. Save a pointer to the function so we know what we're dealing with. There should be some way for the driver to specify which cfg it wants to activate. For now the pccard_function_init function just picks one, but we'll have to revisit this going forward. I'm not doing it now because I'd need some way to activate the card many times and I'm not sure that is desirable or even safe with some cards.
2000-09-16 06:52:20 +00:00
struct pccard_ivar *ivar;
Matching commits to pccard for last pcic changes. We now at least to probe/attach. This is a checkpoint.
2000-06-18 05:28:59 +00:00
device_t child;
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 attached;
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
sc->intr_handler_count = 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
/*
* this is here so that when socket_enable calls gettype, trt happens
*/
STAILQ_INIT(&sc->card.pf_head);
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
DEVPRINTF((dev, "chip_socket_enable\n"));
POWER_ENABLE_SOCKET(device_get_parent(dev), 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
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
DEVPRINTF((dev, "read_cis\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
pccard_read_cis(sc);
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
DEVPRINTF((dev, "check_cis_quirks\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
pccard_check_cis_quirks(dev);
/*
* bail now if the card has no functions, or if there was an error in
* the cis.
*/
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
if (sc->card.error) {
device_printf (dev, "CARD ERROR!\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);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
}
if (STAILQ_EMPTY(&sc->card.pf_head)) {
device_printf (dev, "Card has no functions!\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);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +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
Matching commits to pccard for last pcic changes. We now at least to probe/attach. This is a checkpoint.
2000-06-18 05:28:59 +00:00
if (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
pccard_print_cis(dev);
attached = 0;
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
DEVPRINTF((dev, "functions scanning\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
STAILQ_FOREACH(pf, &sc->card.pf_head, pf_list) {
if (STAILQ_EMPTY(&pf->cfe_head))
continue;
pf->sc = sc;
pf->cfe = NULL;
Save a pointer to the dev, if any, that each function attach. Remove some bogus printfs.
2000-08-22 04:25:57 +00:00
pf->dev = NULL;
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
}
Minor tweaks to have the card be properly enabled via the CCR. Minor debug changes Minor power management pessimizations
2000-08-26 00:07:40 +00:00
#if 0
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
DEVPRINTF((dev, "chip_socket_disable\n"));
POWER_DISABLE_SOCKET(device_get_parent(dev), dev);
Minor tweaks to have the card be properly enabled via the CCR. Minor debug changes Minor power management pessimizations
2000-08-26 00:07:40 +00:00
#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
STAILQ_FOREACH(pf, &sc->card.pf_head, pf_list) {
if (STAILQ_EMPTY(&pf->cfe_head))
continue;
Matching commits to pccard for last pcic changes. We now at least to probe/attach. This is a checkpoint.
2000-06-18 05:28:59 +00:00
/*
* In NetBSD, the drivers are responsible for activating
* each function of a card. I think that in FreeBSD we
* want to activate them enough for the usual bus_*_resource
* routines will do the right thing. This many mean a
* departure from the current NetBSD model.
*
* This could get really ugly for multifunction cards. But
* it might also just fall out of the FreeBSD resource model.
*
*/
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
ivar = malloc(sizeof(struct pccard_ivar), M_DEVBUF,
M_WAITOK | M_ZERO);
Matching commits to pccard for last pcic changes. We now at least to probe/attach. This is a checkpoint.
2000-06-18 05:28:59 +00:00
child = device_add_child(dev, NULL, -1);
Change the interface to pccard_function_init. The interface is such that it looks for an acceptible one. Once it finds it, it should set the resources for the device. I say "should" because I've not written that. Also set an ivar for the child of pccard. Minor fix to the attach message printed, we lose the slot number, which I'll have to restore later. Adjust the pccard ivar so that we can save the function that corresponds to this driver so we can enable and disable it more easily. Save a pointer to the function so we know what we're dealing with. There should be some way for the driver to specify which cfg it wants to activate. For now the pccard_function_init function just picks one, but we'll have to revisit this going forward. I'm not doing it now because I'd need some way to activate the card many times and I'm not sure that is desirable or even safe with some cards.
2000-09-16 06:52:20 +00:00
device_set_ivars(child, ivar);
o Record the pccard_function in the ivar. o Use a macro function to get the ivar for the child rather than the casting directly. This should be a little safer and easier to read.
2000-10-09 03:40:15 +00:00
ivar->fcn = pf;
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
pf->dev = child;
Lots of little fixes. We no longer panic on card insert: o initialize ivars with bzero. o remove interrupt function pointer. netbsd needs it, but we don't. o add lots of comments about bogus things that I've been kludging to try to make the simple cases work. o add new ivar accessor for cis4 to match cis3. likely neither will be needed, but it doesn't hurt to have it.
2000-10-15 17:21:50 +00:00
/*
* XXX We might want to move the next two lines into
* XXX the pccard interface layer. For the moment, this
* XXX is OK, but some drivers want to pick the config
* XXX entry to use as well as some address tweaks (mostly
* XXX due to bugs in decode logic that makes some
* XXX addresses illegal or broken).
*/
Change the interface to pccard_function_init. The interface is such that it looks for an acceptible one. Once it finds it, it should set the resources for the device. I say "should" because I've not written that. Also set an ivar for the child of pccard. Minor fix to the attach message printed, we lose the slot number, which I'll have to restore later. Adjust the pccard ivar so that we can save the function that corresponds to this driver so we can enable and disable it more easily. Save a pointer to the function so we know what we're dealing with. There should be some way for the driver to specify which cfg it wants to activate. For now the pccard_function_init function just picks one, but we'll have to revisit this going forward. I'm not doing it now because I'd need some way to activate the card many times and I'm not sure that is desirable or even safe with some cards.
2000-09-16 06:52:20 +00:00
pccard_function_init(pf);
o Fix memory leak in ivars o Change name of bus o Change the panic on resource allocation failure to just a message. We'll work out why this fails later in the pcic/pccbb code merge.
2000-10-19 22:36:41 +00:00
if (pccard_function_enable(pf) == 0 &&
device_probe_and_attach(child) == 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
attached++;
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
DEVPRINTF((sc->dev, "function %d CCR at %d "
Lots of little fixes. We no longer panic on card insert: o initialize ivars with bzero. o remove interrupt function pointer. netbsd needs it, but we don't. o add lots of comments about bogus things that I've been kludging to try to make the simple cases work. o add new ivar accessor for cis4 to match cis3. likely neither will be needed, but it doesn't hurt to have it.
2000-10-15 17:21:50 +00:00
"offset %x: %x %x %x %x, %x %x %x %x, %x\n",
pf->number, pf->pf_ccr_window, pf->pf_ccr_offset,
pccard_ccr_read(pf, 0x00),
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_ccr_read(pf, 0x02), pccard_ccr_read(pf, 0x04),
pccard_ccr_read(pf, 0x06), pccard_ccr_read(pf, 0x0A),
pccard_ccr_read(pf, 0x0C), pccard_ccr_read(pf, 0x0E),
pccard_ccr_read(pf, 0x10), pccard_ccr_read(pf, 0x12)));
Lots of little fixes. We no longer panic on card insert: o initialize ivars with bzero. o remove interrupt function pointer. netbsd needs it, but we don't. o add lots of comments about bogus things that I've been kludging to try to make the simple cases work. o add new ivar accessor for cis4 to match cis3. likely neither will be needed, but it doesn't hurt to have it.
2000-10-15 17:21:50 +00:00
} else {
device_delete_child(dev, child);
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
}
}
Matching commits to pccard for last pcic changes. We now at least to probe/attach. This is a checkpoint.
2000-06-18 05:28:59 +00:00
return 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
}
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
static int
pccard_detach_card(device_t dev, int flags)
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
{
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_function *pf;
/*
* We are running on either the PCCARD socket's event thread
* or in user context detaching a device by user request.
*/
STAILQ_FOREACH(pf, &sc->card.pf_head, pf_list) {
if (STAILQ_FIRST(&pf->cfe_head) == NULL)
continue;
Matching commits to pccard for last pcic changes. We now at least to probe/attach. This is a checkpoint.
2000-06-18 05:28:59 +00:00
pccard_function_disable(pf);
This is a bandaid for a problem that is not entirely fixed yet. The pccard_function_init() call creates a bunch of inactive resources that are persistant and configured on demand. When the child driver "allocates" a resource it is connected up to one of these. When the child releases the resource, we should not delete our copy, just deactivate it again. Otherwise there is nothing to recreate it again after several probe functions have run and done an alloc/release cycle. INVARIANTS shows 0xdeadc0de without this. More work is needed to do a sweep though the pccard_function_disable() call to actually delete the resources for real. Right now, we leak memory on eject (at best), so Dont Do That(TM) yet. This affects 16 bit pccards on a cardbus bridge only. This will be fixed soon, but for now it gets the cards working. Reviewed by: imp
2001-01-07 21:32:05 +00:00
/*
* XXX must also actually delete resources created by
* pccard_function_init()
*/
Save a pointer to the dev, if any, that each function attach. Remove some bogus printfs.
2000-08-22 04:25:57 +00:00
if (pf->dev)
device_delete_child(dev, pf->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
}
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
return 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
}
o Merge in changes to the NetBSD sources: pcmciavar.h 1.9->1.12 1.12, enami, minor coding nits 1.11, augustss, (pcmcia_devinfo NRFB) 1.10, cgd, add generic lookup routines pcmcia.c 1.14->1.23 1.23, drochner, (probe code printing, NRFB) 1.22, augustss, KNF 1.21, uch, (hpcmips tweaks NRFB) 1.20, chopps, remove bogus debug 1.19, enami, minor coding nits 1.18, augustss, (pcmcia_devinfo NRFB) 1.17, nathanw, LP64 printf fixes 1.16, cgd, add generic lookup routines 1.15, aymeric, printf fixes NRFB == not relevant to freebsd o Expand the pccard matching routines to include the ability to match against the CIS strings since our current driver database is based on that. o Add lots more ivars to get the information necessary to snag these values.
2000-09-22 01:15:26 +00:00
const struct pccard_product *
pccard_product_lookup(device_t dev, const struct pccard_product *tab,
size_t ent_size, pccard_product_match_fn matchfn)
{
const struct pccard_product *ent;
int matches;
u_int32_t fcn;
u_int32_t vendor;
u_int32_t prod;
char *vendorstr;
char *prodstr;
#ifdef DIAGNOSTIC
if (sizeof *ent > ent_size)
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
panic("pccard_product_lookup: bogus ent_size %ld",
o Merge in changes to the NetBSD sources: pcmciavar.h 1.9->1.12 1.12, enami, minor coding nits 1.11, augustss, (pcmcia_devinfo NRFB) 1.10, cgd, add generic lookup routines pcmcia.c 1.14->1.23 1.23, drochner, (probe code printing, NRFB) 1.22, augustss, KNF 1.21, uch, (hpcmips tweaks NRFB) 1.20, chopps, remove bogus debug 1.19, enami, minor coding nits 1.18, augustss, (pcmcia_devinfo NRFB) 1.17, nathanw, LP64 printf fixes 1.16, cgd, add generic lookup routines 1.15, aymeric, printf fixes NRFB == not relevant to freebsd o Expand the pccard matching routines to include the ability to match against the CIS strings since our current driver database is based on that. o Add lots more ivars to get the information necessary to snag these values.
2000-09-22 01:15:26 +00:00
(long) ent_size);
#endif
if (pccard_get_vendor(dev, &vendor))
return (NULL);
if (pccard_get_product(dev, &prod))
return (NULL);
if (pccard_get_function_number(dev, &fcn))
return (NULL);
if (pccard_get_vendor_str(dev, &vendorstr))
return (NULL);
if (pccard_get_product_str(dev, &prodstr))
return (NULL);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
for (ent = tab; ent->pp_name != NULL;
o Merge in changes to the NetBSD sources: pcmciavar.h 1.9->1.12 1.12, enami, minor coding nits 1.11, augustss, (pcmcia_devinfo NRFB) 1.10, cgd, add generic lookup routines pcmcia.c 1.14->1.23 1.23, drochner, (probe code printing, NRFB) 1.22, augustss, KNF 1.21, uch, (hpcmips tweaks NRFB) 1.20, chopps, remove bogus debug 1.19, enami, minor coding nits 1.18, augustss, (pcmcia_devinfo NRFB) 1.17, nathanw, LP64 printf fixes 1.16, cgd, add generic lookup routines 1.15, aymeric, printf fixes NRFB == not relevant to freebsd o Expand the pccard matching routines to include the ability to match against the CIS strings since our current driver database is based on that. o Add lots more ivars to get the information necessary to snag these values.
2000-09-22 01:15:26 +00:00
ent = (const struct pccard_product *)
((const char *) ent + ent_size)) {
matches = 1;
if (matches && ent->pp_vendor != PCCARD_VENDOR_ANY &&
vendor != ent->pp_vendor)
matches = 0;
if (matches && ent->pp_product != PCCARD_PRODUCT_ANY &&
prod != ent->pp_product)
matches = 0;
if (matches && fcn != ent->pp_expfunc)
matches = 0;
Convert the vendor/product cis strings to an array of CIS entries. This allows us to use the PCCARD_CIS #defines generated to distinguish cards that don't have a Manufacturer and OEM.
2001-01-21 08:09:00 +00:00
if (matches && ent->pp_cis[0] &&
strcmp(ent->pp_cis[0], vendorstr) != 0)
o Merge in changes to the NetBSD sources: pcmciavar.h 1.9->1.12 1.12, enami, minor coding nits 1.11, augustss, (pcmcia_devinfo NRFB) 1.10, cgd, add generic lookup routines pcmcia.c 1.14->1.23 1.23, drochner, (probe code printing, NRFB) 1.22, augustss, KNF 1.21, uch, (hpcmips tweaks NRFB) 1.20, chopps, remove bogus debug 1.19, enami, minor coding nits 1.18, augustss, (pcmcia_devinfo NRFB) 1.17, nathanw, LP64 printf fixes 1.16, cgd, add generic lookup routines 1.15, aymeric, printf fixes NRFB == not relevant to freebsd o Expand the pccard matching routines to include the ability to match against the CIS strings since our current driver database is based on that. o Add lots more ivars to get the information necessary to snag these values.
2000-09-22 01:15:26 +00:00
matches = 0;
Convert the vendor/product cis strings to an array of CIS entries. This allows us to use the PCCARD_CIS #defines generated to distinguish cards that don't have a Manufacturer and OEM.
2001-01-21 08:09:00 +00:00
if (matches && ent->pp_cis[1] &&
strcmp(ent->pp_cis[1], prodstr) != 0)
o Merge in changes to the NetBSD sources: pcmciavar.h 1.9->1.12 1.12, enami, minor coding nits 1.11, augustss, (pcmcia_devinfo NRFB) 1.10, cgd, add generic lookup routines pcmcia.c 1.14->1.23 1.23, drochner, (probe code printing, NRFB) 1.22, augustss, KNF 1.21, uch, (hpcmips tweaks NRFB) 1.20, chopps, remove bogus debug 1.19, enami, minor coding nits 1.18, augustss, (pcmcia_devinfo NRFB) 1.17, nathanw, LP64 printf fixes 1.16, cgd, add generic lookup routines 1.15, aymeric, printf fixes NRFB == not relevant to freebsd o Expand the pccard matching routines to include the ability to match against the CIS strings since our current driver database is based on that. o Add lots more ivars to get the information necessary to snag these values.
2000-09-22 01:15:26 +00:00
matches = 0;
Convert the vendor/product cis strings to an array of CIS entries. This allows us to use the PCCARD_CIS #defines generated to distinguish cards that don't have a Manufacturer and OEM.
2001-01-21 08:09:00 +00:00
/* XXX need to match cis[2] and cis[3] also XXX */
o Merge in changes to the NetBSD sources: pcmciavar.h 1.9->1.12 1.12, enami, minor coding nits 1.11, augustss, (pcmcia_devinfo NRFB) 1.10, cgd, add generic lookup routines pcmcia.c 1.14->1.23 1.23, drochner, (probe code printing, NRFB) 1.22, augustss, KNF 1.21, uch, (hpcmips tweaks NRFB) 1.20, chopps, remove bogus debug 1.19, enami, minor coding nits 1.18, augustss, (pcmcia_devinfo NRFB) 1.17, nathanw, LP64 printf fixes 1.16, cgd, add generic lookup routines 1.15, aymeric, printf fixes NRFB == not relevant to freebsd o Expand the pccard matching routines to include the ability to match against the CIS strings since our current driver database is based on that. o Add lots more ivars to get the information necessary to snag these values.
2000-09-22 01:15:26 +00:00
if (matchfn != NULL)
matches = (*matchfn)(dev, ent, matches);
if (matches)
return (ent);
}
return (NULL);
}
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
static int
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
pccard_card_gettype(device_t dev, int *type)
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
{
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_function *pf;
/*
* set the iftype to memory if this card has no functions (not yet
* probed), or only one function, and that is not initialized yet or
* that is memory.
*/
pf = STAILQ_FIRST(&sc->card.pf_head);
if (pf == NULL ||
(STAILQ_NEXT(pf, pf_list) == NULL &&
(pf->cfe == NULL || pf->cfe->iftype == PCCARD_IFTYPE_MEMORY)))
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
*type = PCCARD_IFTYPE_MEMORY;
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
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
*type = PCCARD_IFTYPE_IO;
return 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
}
/*
* Initialize a PCCARD function. May be called as long as the function is
* disabled.
*/
void
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
pccard_function_init(struct pccard_function *pf)
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
{
Change the interface to pccard_function_init. The interface is such that it looks for an acceptible one. Once it finds it, it should set the resources for the device. I say "should" because I've not written that. Also set an ivar for the child of pccard. Minor fix to the attach message printed, we lose the slot number, which I'll have to restore later. Adjust the pccard ivar so that we can save the function that corresponds to this driver so we can enable and disable it more easily. Save a pointer to the function so we know what we're dealing with. There should be some way for the driver to specify which cfg it wants to activate. For now the pccard_function_init function just picks one, but we'll have to revisit this going forward. I'm not doing it now because I'd need some way to activate the card many times and I'm not sure that is desirable or even safe with some cards.
2000-09-16 06:52:20 +00:00
struct pccard_config_entry *cfe;
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
int i;
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
struct pccard_ivar *devi = PCCARD_IVAR(pf->dev);
struct resource_list *rl = &devi->resources;
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
struct resource_list_entry *rle;
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
struct resource *r = 0;
device_t bus;
o Remove lots of kludges with iospace and irq. o Make sure that I/O space is aligned properly based on the length of the resource we're allocating. o Add bootverbose message when we fail.
2000-10-22 04:46:56 +00:00
int start;
int end;
Change the interface to pccard_function_init. The interface is such that it looks for an acceptible one. Once it finds it, it should set the resources for the device. I say "should" because I've not written that. Also set an ivar for the child of pccard. Minor fix to the attach message printed, we lose the slot number, which I'll have to restore later. Adjust the pccard ivar so that we can save the function that corresponds to this driver so we can enable and disable it more easily. Save a pointer to the function so we know what we're dealing with. There should be some way for the driver to specify which cfg it wants to activate. For now the pccard_function_init function just picks one, but we'll have to revisit this going forward. I'm not doing it now because I'd need some way to activate the card many times and I'm not sure that is desirable or even safe with some cards.
2000-09-16 06:52:20 +00:00
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
if (pf->pf_flags & PFF_ENABLED) {
printf("pccard_function_init: function is enabled");
return;
}
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
bus = device_get_parent(pf->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
/* Remember which configuration entry we are using. */
Lots of little fixes. We no longer panic on card insert: o initialize ivars with bzero. o remove interrupt function pointer. netbsd needs it, but we don't. o add lots of comments about bogus things that I've been kludging to try to make the simple cases work. o add new ivar accessor for cis4 to match cis3. likely neither will be needed, but it doesn't hurt to have it.
2000-10-15 17:21:50 +00:00
for (cfe = STAILQ_FIRST(&pf->cfe_head); cfe != NULL;
cfe = STAILQ_NEXT(cfe, cfe_list)) {
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
for (i = 0; i < cfe->num_iospace; i++)
cfe->iores[i] = NULL;
cfe->irqres = NULL;
for (i = 0; i < cfe->num_iospace; i++) {
o Remove lots of kludges with iospace and irq. o Make sure that I/O space is aligned properly based on the length of the resource we're allocating. o Add bootverbose message when we fail.
2000-10-22 04:46:56 +00:00
start = cfe->iospace[i].start;
if (start)
end = start + cfe->iospace[i].length - 1;
else
end = ~0;
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
cfe->iorid[i] = i;
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
r = cfe->iores[i] = bus_alloc_resource(bus,
o Remove lots of kludges with iospace and irq. o Make sure that I/O space is aligned properly based on the length of the resource we're allocating. o Add bootverbose message when we fail.
2000-10-22 04:46:56 +00:00
SYS_RES_IOPORT, &cfe->iorid[i], start, end,
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
cfe->iospace[i].length,
o Remove lots of kludges with iospace and irq. o Make sure that I/O space is aligned properly based on the length of the resource we're allocating. o Add bootverbose message when we fail.
2000-10-22 04:46:56 +00:00
rman_make_alignment_flags(cfe->iospace[i].length));
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
if (cfe->iores[i] == 0)
goto not_this_one;
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
resource_list_add(rl, SYS_RES_IOPORT, cfe->iorid[i],
rman_get_start(r), rman_get_end(r),
cfe->iospace[i].length);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
{
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
rle = resource_list_find(rl, SYS_RES_IOPORT,
cfe->iorid[i]);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
rle->res = r;
}
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
}
if (cfe->num_memspace > 0) {
goto not_this_one;
}
if (cfe->irqmask) {
cfe->irqrid = 0;
Get the interrupt correct.
2000-10-20 20:27:22 +00:00
r = cfe->irqres = bus_alloc_resource(bus, SYS_RES_IRQ,
o Remove lots of kludges with iospace and irq. o Make sure that I/O space is aligned properly based on the length of the resource we're allocating. o Add bootverbose message when we fail.
2000-10-22 04:46:56 +00:00
&cfe->irqrid, 0, ~0, 1, 0);
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
if (cfe->irqres == 0)
goto not_this_one;
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
resource_list_add(rl, SYS_RES_IRQ, cfe->irqrid,
rman_get_start(r), rman_get_end(r), 1);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
{
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
rle = resource_list_find(rl, SYS_RES_IRQ,
cfe->irqrid);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
rle->res = r;
}
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
}
/* XXX Don't know how to deal with maxtwins */
/* If we get to here, we've allocated all we need */
Lots of little fixes. We no longer panic on card insert: o initialize ivars with bzero. o remove interrupt function pointer. netbsd needs it, but we don't. o add lots of comments about bogus things that I've been kludging to try to make the simple cases work. o add new ivar accessor for cis4 to match cis3. likely neither will be needed, but it doesn't hurt to have it.
2000-10-15 17:21:50 +00:00
pf->cfe = cfe;
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
break;
not_this_one:;
o Remove lots of kludges with iospace and irq. o Make sure that I/O space is aligned properly based on the length of the resource we're allocating. o Add bootverbose message when we fail.
2000-10-22 04:46:56 +00:00
DEVPRVERBOSE((bus, "Allocation failed for cfe %d\n",
cfe->number));
o Fix memory leak in ivars o Change name of bus o Change the panic on resource allocation failure to just a message. We'll work out why this fails later in the pcic/pccbb code merge.
2000-10-19 22:36:41 +00:00
/*
* Release resources that we partially allocated
* from this config entry.
*/
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
for (i = 0; i < cfe->num_iospace; i++) {
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
resource_list_delete(rl, SYS_RES_IOPORT, i);
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
if (cfe->iores[i])
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
bus_release_resource(bus, SYS_RES_IOPORT,
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
cfe->iorid[i], cfe->iores[i]);
cfe->iores[i] = NULL;
}
if (cfe->irqmask && cfe->irqres) {
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
resource_list_delete(rl, SYS_RES_IRQ, cfe->irqrid);
bus_release_resource(bus, SYS_RES_IRQ,
First cut at allocating the resources from the CIS after the probe, but before the attach. Things aren't completely working, but this is a good checkpoint. Also, initialize the dev member of the function as soon as we add it to the parent.
2000-10-16 07:51:12 +00:00
cfe->irqrid, cfe->irqres);
cfe->irqres = NULL;
}
Lots of little fixes. We no longer panic on card insert: o initialize ivars with bzero. o remove interrupt function pointer. netbsd needs it, but we don't. o add lots of comments about bogus things that I've been kludging to try to make the simple cases work. o add new ivar accessor for cis4 to match cis3. likely neither will be needed, but it doesn't hurt to have it.
2000-10-15 17:21:50 +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
}
/* Enable a PCCARD function */
int
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_function_enable(struct pccard_function *pf)
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_function *tmp;
int reg;
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_t dev = pf->sc->dev;
Undo the interface change to CARD_GET_MEMORY_OFFSET. It wasn't tested by even a compile of the OLDCARD code, was unapproved by me the keeper of OLDCARD and broke OLDCARD and the ray driver. Adjust new code to cope with the older interface. If the interface changes in the future, it ***MUST*** be cleared by me so that the OLDCARD impacts taken into account. It code in card_if.m is used jointly by both OLDCARD and NEWCARD.
2001-01-07 08:08:54 +00:00
o Fix memory leak in ivars o Change name of bus o Change the panic on resource allocation failure to just a message. We'll work out why this fails later in the pcic/pccbb code merge.
2000-10-19 22:36:41 +00:00
if (pf->cfe == NULL) {
DEVPRVERBOSE((dev, "No config entry could be allocated.\n"));
return ENOMEM;
}
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
/*
* Increase the reference count on the socket, enabling power, if
* necessary.
*/
if (pf->sc->sc_enabled_count++ == 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
POWER_ENABLE_SOCKET(device_get_parent(dev), dev);
DEVPRINTF((dev, "++enabled_count = %d\n", pf->sc->sc_enabled_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
if (pf->pf_flags & PFF_ENABLED) {
/*
* Don't do anything if we're already enabled.
*/
return (0);
}
/*
* it's possible for different functions' CCRs to be in the same
* underlying page. Check for that.
*/
STAILQ_FOREACH(tmp, &pf->sc->card.pf_head, pf_list) {
if ((tmp->pf_flags & PFF_ENABLED) &&
(pf->ccr_base >= (tmp->ccr_base - tmp->pf_ccr_offset)) &&
((pf->ccr_base + PCCARD_CCR_SIZE) <=
(tmp->ccr_base - tmp->pf_ccr_offset +
tmp->pf_ccr_realsize))) {
pf->pf_ccrt = tmp->pf_ccrt;
pf->pf_ccrh = tmp->pf_ccrh;
pf->pf_ccr_realsize = tmp->pf_ccr_realsize;
/*
* pf->pf_ccr_offset = (tmp->pf_ccr_offset -
* tmp->ccr_base) + pf->ccr_base;
*/
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
/* pf->pf_ccr_offset =
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
(tmp->pf_ccr_offset + pf->ccr_base) -
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
tmp->ccr_base; */
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
pf->pf_ccr_window = tmp->pf_ccr_window;
break;
}
}
if (tmp == NULL) {
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
pf->ccr_rid = 0;
pf->ccr_res = bus_alloc_resource(dev, SYS_RES_MEMORY,
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
&pf->ccr_rid, 0, ~0, 1 << 10, RF_ACTIVE);
if (!pf->ccr_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
goto bad;
Undo the interface change to CARD_GET_MEMORY_OFFSET. It wasn't tested by even a compile of the OLDCARD code, was unapproved by me the keeper of OLDCARD and broke OLDCARD and the ray driver. Adjust new code to cope with the older interface. If the interface changes in the future, it ***MUST*** be cleared by me so that the OLDCARD impacts taken into account. It code in card_if.m is used jointly by both OLDCARD and NEWCARD.
2001-01-07 08:08:54 +00:00
DEVPRINTF((dev, "ccr_res == %lx-%lx, base=%lx\n",
rman_get_start(pf->ccr_res), rman_get_end(pf->ccr_res),
pf->ccr_base));
Matching commits to pccard for last pcic changes. We now at least to probe/attach. This is a checkpoint.
2000-06-18 05:28:59 +00:00
CARD_SET_RES_FLAGS(device_get_parent(dev), dev, SYS_RES_MEMORY,
pf->ccr_rid, PCCARD_A_MEM_ATTR);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
CARD_SET_MEMORY_OFFSET(device_get_parent(dev), dev,
o Now that I've had time to test the new interface, reintegrate it back in. o Fix OLDCARD to use the new interface. o Rename the offsetp argument to deltap to more closely reflect what it is returning (it returns the delta from the requested value to the actual value). o Remove duplicate $FreeBSD$ in pccbb.c o Allow deltap to be NULL. o Convert new isa pcic driver and add XXX comments that this function isn't actually implemented there (which means that NEWCARD pccard stuff won't work there until it is). o Revert attempts to make old inferface work in NEWCARD. Subitted by: peter (Parts of the new version code)
2001-01-07 16:31:09 +00:00
pf->ccr_rid, pf->ccr_base, &pf->pf_ccr_offset);
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
pf->pf_ccrt = rman_get_bustag(pf->ccr_res);
pf->pf_ccrh = rman_get_bushandle(pf->ccr_res);
pf->pf_ccr_realsize = 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
}
reg = (pf->cfe->number & PCCARD_CCR_OPTION_CFINDEX);
reg |= PCCARD_CCR_OPTION_LEVIREQ;
if (pccard_mfc(pf->sc)) {
reg |= (PCCARD_CCR_OPTION_FUNC_ENABLE |
PCCARD_CCR_OPTION_ADDR_DECODE);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
/*
Lots of little fixes. We no longer panic on card insert: o initialize ivars with bzero. o remove interrupt function pointer. netbsd needs it, but we don't. o add lots of comments about bogus things that I've been kludging to try to make the simple cases work. o add new ivar accessor for cis4 to match cis3. likely neither will be needed, but it doesn't hurt to have it.
2000-10-15 17:21:50 +00:00
* XXX Need to enable PCCARD_CCR_OPTION_IRQ_ENABLE if
* XXX we have an interrupt handler, but we don't know that
* XXX at this point.
*/
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
/* reg |= PCCARD_CCR_OPTION_IREQ_ENABLE;*/
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_ccr_write(pf, PCCARD_CCR_OPTION, reg);
reg = 0;
if ((pf->cfe->flags & PCCARD_CFE_IO16) == 0)
reg |= PCCARD_CCR_STATUS_IOIS8;
if (pf->cfe->flags & PCCARD_CFE_AUDIO)
reg |= PCCARD_CCR_STATUS_AUDIO;
pccard_ccr_write(pf, PCCARD_CCR_STATUS, reg);
pccard_ccr_write(pf, PCCARD_CCR_SOCKETCOPY, 0);
if (pccard_mfc(pf->sc)) {
long tmp, iosize;
tmp = pf->pf_mfc_iomax - pf->pf_mfc_iobase;
/* round up to nearest (2^n)-1 */
for (iosize = 1; iosize < tmp; iosize <<= 1)
;
iosize--;
pccard_ccr_write(pf, PCCARD_CCR_IOBASE0,
pf->pf_mfc_iobase & 0xff);
pccard_ccr_write(pf, PCCARD_CCR_IOBASE1,
(pf->pf_mfc_iobase >> 8) & 0xff);
pccard_ccr_write(pf, PCCARD_CCR_IOBASE2, 0);
pccard_ccr_write(pf, PCCARD_CCR_IOBASE3, 0);
pccard_ccr_write(pf, PCCARD_CCR_IOSIZE, iosize);
}
#ifdef PCCARDDEBUG
if (pccard_debug) {
STAILQ_FOREACH(tmp, &pf->sc->card.pf_head, pf_list) {
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
device_printf(tmp->sc->dev,
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
"function %d CCR at %d offset %x: "
"%x %x %x %x, %x %x %x %x, %x\n",
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
tmp->number, tmp->pf_ccr_window,
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
tmp->pf_ccr_offset,
pccard_ccr_read(tmp, 0x00),
pccard_ccr_read(tmp, 0x02),
pccard_ccr_read(tmp, 0x04),
pccard_ccr_read(tmp, 0x06),
pccard_ccr_read(tmp, 0x0A),
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
pccard_ccr_read(tmp, 0x0C),
Checkpoint of today's changes. We now get to the point where the pccard layer is trying to access the now unexistant chip functions. o Added DEVPRINTF which is like DPRINTF only calls device_printf. o Made it possible to define PCICDEBUG o Remove ph_parent and use the softc pointer sc instead in pcic_handle. o Remove all references to dv_xname o Add some debug messages. o enable MI attach/detach calling for pccard. o convert pcic_chip_socket_{en,dis}able to pcic_{dis,en}able_socket and connect them to the power_{enable,disbale}_socket. o Remove pccard pointer from pcic_softc. o GC some unused pccard functions. o Convert pccard_chip_socket* to POWER_ENABLE_SOCKET o kill pccard_attach_args. o power_if.m updates. More to come.
2000-01-06 07:30:28 +00:00
pccard_ccr_read(tmp, 0x0E),
pccard_ccr_read(tmp, 0x10),
pccard_ccr_read(tmp, 0x12));
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
pf->pf_flags |= PFF_ENABLED;
return (0);
bad:
/*
* Decrement the reference count, and power down the socket, if
* necessary.
*/
if (--pf->sc->sc_enabled_count == 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
POWER_DISABLE_SOCKET(device_get_parent(dev), dev);
Minor tweaks to have the card be properly enabled via the CCR. Minor debug changes Minor power management pessimizations
2000-08-26 00:07:40 +00:00
DEVPRINTF((dev, "bad --enabled_count = %d\n", pf->sc->sc_enabled_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
return (1);
}
/* Disable PCCARD function. */
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_function_disable(struct pccard_function *pf)
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_function *tmp;
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_t dev = pf->sc->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
if (pf->cfe == NULL)
Matching commits to pccard for last pcic changes. We now at least to probe/attach. This is a checkpoint.
2000-06-18 05:28:59 +00:00
panic("pccard_function_disable: function not initialized");
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 ((pf->pf_flags & PFF_ENABLED) == 0) {
/*
* Don't do anything if we're already disabled.
*/
return;
}
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
if (pf->intr_handler != NULL) {
pf->intr_handler = NULL;
pf->intr_handler_arg = NULL;
pf->intr_handler_cookie = NULL;
pccard_ccr_write(pf, PCCARD_CCR_OPTION,
pccard_ccr_read(pf, PCCARD_CCR_OPTION) &
~PCCARD_CCR_OPTION_IREQ_ENABLE);
if (pf->sc->intr_handler_count == 1) {
struct pccard_ivar *ivar = PCCARD_IVAR(pf->dev);
struct resource_list_entry *rle = NULL;
pf->sc->intr_handler_count--;
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
rle = resource_list_find(&ivar->resources, SYS_RES_IRQ,
0);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
if (rle == NULL)
panic("No IRQ for pccard?");
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
bus_teardown_intr(dev, rle->res,
&pf->sc->intr_handler_count);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +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
/*
* it's possible for different functions' CCRs to be in the same
* underlying page. Check for that. Note we mark us as disabled
* first to avoid matching ourself.
*/
pf->pf_flags &= ~PFF_ENABLED;
STAILQ_FOREACH(tmp, &pf->sc->card.pf_head, pf_list) {
if ((tmp->pf_flags & PFF_ENABLED) &&
(pf->ccr_base >= (tmp->ccr_base - tmp->pf_ccr_offset)) &&
((pf->ccr_base + PCCARD_CCR_SIZE) <=
(tmp->ccr_base - tmp->pf_ccr_offset + tmp->pf_ccr_realsize)))
break;
}
/* Not used by anyone else; unmap the CCR. */
if (tmp == NULL) {
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
bus_release_resource(dev, SYS_RES_MEMORY, pf->ccr_rid,
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
pf->ccr_res);
pf->ccr_res = NULL;
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
}
/*
* Decrement the reference count, and power down the socket, if
* necessary.
*/
if (--pf->sc->sc_enabled_count == 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
POWER_DISABLE_SOCKET(device_get_parent(dev), dev);
DEVPRINTF((dev, "--enabled_count = %d\n", pf->sc->sc_enabled_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
}
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 0
/* XXX These functions are needed, but not like this XXX */
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
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_io_map(struct pccard_function *pf, int width, bus_addr_t offset,
bus_size_t size, struct pccard_io_handle *pcihp, int *windowp)
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 reg;
Lots of little fixes. We no longer panic on card insert: o initialize ivars with bzero. o remove interrupt function pointer. netbsd needs it, but we don't. o add lots of comments about bogus things that I've been kludging to try to make the simple cases work. o add new ivar accessor for cis4 to match cis3. likely neither will be needed, but it doesn't hurt to have it.
2000-10-15 17:21:50 +00:00
if (pccard_chip_io_map(pf->sc->pct, pf->sc->pch, width, offset, size,
pcihp, windowp))
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);
/*
* XXX in the multifunction multi-iospace-per-function case, this
* needs to cooperate with io_alloc to make sure that the spaces
* don't overlap, and that the ccr's are set correctly
*/
if (pccard_mfc(pf->sc)) {
long tmp, iosize;
if (pf->pf_mfc_iomax == 0) {
pf->pf_mfc_iobase = pcihp->addr + offset;
pf->pf_mfc_iomax = pf->pf_mfc_iobase + size;
} else {
/* this makes the assumption that nothing overlaps */
if (pf->pf_mfc_iobase > pcihp->addr + offset)
pf->pf_mfc_iobase = pcihp->addr + offset;
if (pf->pf_mfc_iomax < pcihp->addr + offset + size)
pf->pf_mfc_iomax = pcihp->addr + offset + size;
}
tmp = pf->pf_mfc_iomax - pf->pf_mfc_iobase;
/* round up to nearest (2^n)-1 */
for (iosize = 1; iosize >= tmp; iosize <<= 1)
;
iosize--;
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
pccard_ccr_write(pf, PCCARD_CCR_IOBASE0,
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
pf->pf_mfc_iobase & 0xff);
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_ccr_write(pf, PCCARD_CCR_IOBASE1,
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
(pf->pf_mfc_iobase >> 8) & 0xff);
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_ccr_write(pf, PCCARD_CCR_IOBASE2, 0);
pccard_ccr_write(pf, PCCARD_CCR_IOBASE3, 0);
pccard_ccr_write(pf, PCCARD_CCR_IOSIZE, iosize);
reg = pccard_ccr_read(pf, PCCARD_CCR_OPTION);
reg |= PCCARD_CCR_OPTION_ADDR_DECODE;
pccard_ccr_write(pf, PCCARD_CCR_OPTION, reg);
}
return (0);
}
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_io_unmap(struct pccard_function *pf, int window)
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_chip_io_unmap(pf->sc->pct, pf->sc->pch, window);
/* XXX Anything for multi-function cards? */
}
#endif
Implement indirection in the pccard probe/attach. This should make it possible to have different probe/attach semantics between the two systems and yet still use the same driver for both. Compatibility methods for OLDCARD drivers. We use these routines to make it possible to call the OLDCARD driver's probe routine in the context that it expects. For OLDCARD these are implemented as pass throughs to the device_{probe,attach} routines. For NEWCARD they are implemented such such that probe becomes strictly a matching routine and attach does both the old probe and old attach. compat devices should use the following: /* Device interface */ DEVMETHOD(device_probe), pccard_compat_probe), DEVMETHOD(device_attach), pccard_compat_attach), /* Card interface */ DEVMETHOD(card_compat_match, foo_match), /* newly written */ DEVMETHOD(card_compat_probe, foo_probe), /* old probe */ DEVMETHOD(card_compat_attach, foo_attach), /* old attach */ This will allow a single driver binary image to be used for both OLDCARD and NEWCARD. Drivers wishing to not retain OLDCARD compatibility needn't do this. ep driver minorly updated. sn driver updated more than minorly. Add module dependencies to allow module to load. Also change name to if_sn. Add some debugging code. attempt to fix the cannot allocate memory problem I'd been seeing. Minor formatting nits.
2000-09-19 04:39:20 +00:00
/*
* simulate the old "probe" routine. In the new world order, the driver
* needs to grab devices while in the old they were assigned to the device by
* the pccardd process. These symbols are exported to the upper layers.
*/
int
pccard_compat_probe(device_t dev)
{
return (CARD_COMPAT_MATCH(dev));
}
int
pccard_compat_attach(device_t dev)
{
int err;
err = CARD_COMPAT_PROBE(dev);
if (err == 0)
err = CARD_COMPAT_ATTACH(dev);
return (err);
}
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
#define PCCARD_NPORT 2
#define PCCARD_NMEM 5
#define PCCARD_NIRQ 1
#define PCCARD_NDRQ 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
static int
pccard_add_children(device_t dev, int busno)
{
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
/* Call parent to scan for any current children */
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 0;
}
static int
pccard_probe(device_t dev)
{
o Fix memory leak in ivars o Change name of bus o Change the panic on resource allocation failure to just a message. We'll work out why this fails later in the pcic/pccbb code merge.
2000-10-19 22:36:41 +00:00
device_set_desc(dev, "16-bit PCCard bus");
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 pccard_add_children(dev, device_get_unit(dev));
}
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
static int
pccard_attach(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);
Matching commits to pccard for last pcic changes. We now at least to probe/attach. This is a checkpoint.
2000-06-18 05:28:59 +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
sc->dev = dev;
Matching commits to pccard for last pcic changes. We now at least to probe/attach. This is a checkpoint.
2000-06-18 05:28:59 +00:00
sc->sc_enabled_count = 0;
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
return bus_generic_attach(dev);
}
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
static void
pccard_print_resources(struct resource_list *rl, const char *name, int type,
int count, const char *format)
{
struct resource_list_entry *rle;
int printed;
int i;
printed = 0;
for (i = 0; i < count; i++) {
rle = resource_list_find(rl, type, i);
if (rle) {
if (printed == 0)
printf(" %s ", name);
else if (printed > 0)
printf(",");
printed++;
printf(format, rle->start);
if (rle->count > 1) {
printf("-");
printf(format, rle->start + rle->count - 1);
}
} else if (i > 3) {
/* check the first few regardless */
break;
}
}
}
static int
pccard_print_child(device_t dev, device_t child)
{
o Record the pccard_function in the ivar. o Use a macro function to get the ivar for the child rather than the casting directly. This should be a little safer and easier to read.
2000-10-09 03:40:15 +00:00
struct pccard_ivar *devi = PCCARD_IVAR(child);
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
struct resource_list *rl = &devi->resources;
int retval = 0;
retval += bus_print_child_header(dev, child);
retval += printf(" at");
if (devi) {
pccard_print_resources(rl, "port", SYS_RES_IOPORT,
PCCARD_NPORT, "%#lx");
pccard_print_resources(rl, "iomem", SYS_RES_MEMORY,
PCCARD_NMEM, "%#lx");
pccard_print_resources(rl, "irq", SYS_RES_IRQ, PCCARD_NIRQ,
"%ld");
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
pccard_print_resources(rl, "drq", SYS_RES_DRQ, PCCARD_NDRQ,
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
"%ld");
o Remove a boatload of debugging printfs. o Report function number and config index on probe line o Activate the resources (I hope) when RF_ACTIVE is set on those resources I'm allocating on behalf of my children. o Always enable interrupts on multifunction cards in the multifunction register.
2000-10-18 02:21:00 +00:00
retval += printf(" function %d config %d", devi->fcn->number,
devi->fcn->cfe->number);
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
}
retval += bus_print_child_footer(dev, child);
return (retval);
}
static int
pccard_set_resource(device_t dev, device_t child, int type, int rid,
u_long start, u_long count)
{
o Record the pccard_function in the ivar. o Use a macro function to get the ivar for the child rather than the casting directly. This should be a little safer and easier to read.
2000-10-09 03:40:15 +00:00
struct pccard_ivar *devi = PCCARD_IVAR(child);
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
struct resource_list *rl = &devi->resources;
if (type != SYS_RES_IOPORT && type != SYS_RES_MEMORY
&& type != SYS_RES_IRQ && type != SYS_RES_DRQ)
return EINVAL;
if (rid < 0)
return EINVAL;
if (type == SYS_RES_IOPORT && rid >= PCCARD_NPORT)
return EINVAL;
if (type == SYS_RES_MEMORY && rid >= PCCARD_NMEM)
return EINVAL;
if (type == SYS_RES_IRQ && rid >= PCCARD_NIRQ)
return EINVAL;
if (type == SYS_RES_DRQ && rid >= PCCARD_NDRQ)
return EINVAL;
resource_list_add(rl, type, rid, start, start + count - 1, count);
return 0;
}
static int
pccard_get_resource(device_t dev, device_t child, int type, int rid,
u_long *startp, u_long *countp)
{
o Record the pccard_function in the ivar. o Use a macro function to get the ivar for the child rather than the casting directly. This should be a little safer and easier to read.
2000-10-09 03:40:15 +00:00
struct pccard_ivar *devi = PCCARD_IVAR(child);
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
struct resource_list *rl = &devi->resources;
struct resource_list_entry *rle;
rle = resource_list_find(rl, type, rid);
if (!rle)
return ENOENT;
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
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
if (startp)
*startp = rle->start;
if (countp)
*countp = rle->count;
return 0;
}
static void
pccard_delete_resource(device_t dev, device_t child, int type, int rid)
{
o Record the pccard_function in the ivar. o Use a macro function to get the ivar for the child rather than the casting directly. This should be a little safer and easier to read.
2000-10-09 03:40:15 +00:00
struct pccard_ivar *devi = PCCARD_IVAR(child);
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
struct resource_list *rl = &devi->resources;
resource_list_delete(rl, type, rid);
}
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
static int
pccard_set_res_flags(device_t dev, device_t child, int type, int rid,
u_int32_t flags)
{
return CARD_SET_RES_FLAGS(device_get_parent(dev), child, type,
rid, flags);
}
static int
pccard_set_memory_offset(device_t dev, device_t child, int rid,
o Now that I've had time to test the new interface, reintegrate it back in. o Fix OLDCARD to use the new interface. o Rename the offsetp argument to deltap to more closely reflect what it is returning (it returns the delta from the requested value to the actual value). o Remove duplicate $FreeBSD$ in pccbb.c o Allow deltap to be NULL. o Convert new isa pcic driver and add XXX comments that this function isn't actually implemented there (which means that NEWCARD pccard stuff won't work there until it is). o Revert attempts to make old inferface work in NEWCARD. Subitted by: peter (Parts of the new version code)
2001-01-07 16:31:09 +00:00
u_int32_t offset, u_int32_t *deltap)
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +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
{
return CARD_SET_MEMORY_OFFSET(device_get_parent(dev), child, rid,
o Now that I've had time to test the new interface, reintegrate it back in. o Fix OLDCARD to use the new interface. o Rename the offsetp argument to deltap to more closely reflect what it is returning (it returns the delta from the requested value to the actual value). o Remove duplicate $FreeBSD$ in pccbb.c o Allow deltap to be NULL. o Convert new isa pcic driver and add XXX comments that this function isn't actually implemented there (which means that NEWCARD pccard stuff won't work there until it is). o Revert attempts to make old inferface work in NEWCARD. Subitted by: peter (Parts of the new version code)
2001-01-07 16:31:09 +00:00
offset, deltap);
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
}
Implement indirection in the pccard probe/attach. This should make it possible to have different probe/attach semantics between the two systems and yet still use the same driver for both. Compatibility methods for OLDCARD drivers. We use these routines to make it possible to call the OLDCARD driver's probe routine in the context that it expects. For OLDCARD these are implemented as pass throughs to the device_{probe,attach} routines. For NEWCARD they are implemented such such that probe becomes strictly a matching routine and attach does both the old probe and old attach. compat devices should use the following: /* Device interface */ DEVMETHOD(device_probe), pccard_compat_probe), DEVMETHOD(device_attach), pccard_compat_attach), /* Card interface */ DEVMETHOD(card_compat_match, foo_match), /* newly written */ DEVMETHOD(card_compat_probe, foo_probe), /* old probe */ DEVMETHOD(card_compat_attach, foo_attach), /* old attach */ This will allow a single driver binary image to be used for both OLDCARD and NEWCARD. Drivers wishing to not retain OLDCARD compatibility needn't do this. ep driver minorly updated. sn driver updated more than minorly. Add module dependencies to allow module to load. Also change name to if_sn. Add some debugging code. attempt to fix the cannot allocate memory problem I'd been seeing. Minor formatting nits.
2000-09-19 04:39:20 +00:00
static int
pccard_read_ivar(device_t bus, device_t child, int which, u_char *result)
{
o Record the pccard_function in the ivar. o Use a macro function to get the ivar for the child rather than the casting directly. This should be a little safer and easier to read.
2000-10-09 03:40:15 +00:00
struct pccard_ivar *devi = PCCARD_IVAR(child);
o Implement get_ivars so matching routines will work (this causes compat probe routines to work). o Have a null driver_added routines. We need to be careful about probing until after we know we have a card. For the moment, we do nothing (which is safe). This fixes a panic when a driver is loaded w/o a card in the slot. XXX still need to fix the resource list code. It is totally busted and XXX causes a panic in the child printing routine after the probe has XXX succeeded.
2000-10-07 05:48:51 +00:00
struct pccard_function *func = devi->fcn;
struct pccard_softc *sc = PCCARD_SOFTC(bus);
Implement indirection in the pccard probe/attach. This should make it possible to have different probe/attach semantics between the two systems and yet still use the same driver for both. Compatibility methods for OLDCARD drivers. We use these routines to make it possible to call the OLDCARD driver's probe routine in the context that it expects. For OLDCARD these are implemented as pass throughs to the device_{probe,attach} routines. For NEWCARD they are implemented such such that probe becomes strictly a matching routine and attach does both the old probe and old attach. compat devices should use the following: /* Device interface */ DEVMETHOD(device_probe), pccard_compat_probe), DEVMETHOD(device_attach), pccard_compat_attach), /* Card interface */ DEVMETHOD(card_compat_match, foo_match), /* newly written */ DEVMETHOD(card_compat_probe, foo_probe), /* old probe */ DEVMETHOD(card_compat_attach, foo_attach), /* old attach */ This will allow a single driver binary image to be used for both OLDCARD and NEWCARD. Drivers wishing to not retain OLDCARD compatibility needn't do this. ep driver minorly updated. sn driver updated more than minorly. Add module dependencies to allow module to load. Also change name to if_sn. Add some debugging code. attempt to fix the cannot allocate memory problem I'd been seeing. Minor formatting nits.
2000-09-19 04:39:20 +00:00
/* PCCARD_IVAR_ETHADDR unhandled from oldcard */
o Implement get_ivars so matching routines will work (this causes compat probe routines to work). o Have a null driver_added routines. We need to be careful about probing until after we know we have a card. For the moment, we do nothing (which is safe). This fixes a panic when a driver is loaded w/o a card in the slot. XXX still need to fix the resource list code. It is totally busted and XXX causes a panic in the child printing routine after the probe has XXX succeeded.
2000-10-07 05:48:51 +00:00
switch (which) {
default:
case PCCARD_IVAR_ETHADDR:
return (ENOENT);
break;
case PCCARD_IVAR_VENDOR:
*(u_int32_t *) result = sc->card.manufacturer;
break;
case PCCARD_IVAR_PRODUCT:
*(u_int32_t *) result = sc->card.product;
break;
case PCCARD_IVAR_FUNCTION_NUMBER:
o Record the pccard_function in the ivar. o Use a macro function to get the ivar for the child rather than the casting directly. This should be a little safer and easier to read.
2000-10-09 03:40:15 +00:00
if (!func) {
device_printf(bus, "No function number, bug!\n");
return (ENOENT);
}
*(u_int32_t *) result = func->number;
o Implement get_ivars so matching routines will work (this causes compat probe routines to work). o Have a null driver_added routines. We need to be careful about probing until after we know we have a card. For the moment, we do nothing (which is safe). This fixes a panic when a driver is loaded w/o a card in the slot. XXX still need to fix the resource list code. It is totally busted and XXX causes a panic in the child printing routine after the probe has XXX succeeded.
2000-10-07 05:48:51 +00:00
break;
case PCCARD_IVAR_VENDOR_STR:
*(char **) result = sc->card.cis1_info[0];
break;
case PCCARD_IVAR_PRODUCT_STR:
*(char **) result = sc->card.cis1_info[1];
break;
case PCCARD_IVAR_CIS3_STR:
*(char **) result = sc->card.cis1_info[2];
break;
Lots of little fixes. We no longer panic on card insert: o initialize ivars with bzero. o remove interrupt function pointer. netbsd needs it, but we don't. o add lots of comments about bogus things that I've been kludging to try to make the simple cases work. o add new ivar accessor for cis4 to match cis3. likely neither will be needed, but it doesn't hurt to have it.
2000-10-15 17:21:50 +00:00
case PCCARD_IVAR_CIS4_STR:
*(char **) result = sc->card.cis1_info[2];
break;
o Implement get_ivars so matching routines will work (this causes compat probe routines to work). o Have a null driver_added routines. We need to be careful about probing until after we know we have a card. For the moment, we do nothing (which is safe). This fixes a panic when a driver is loaded w/o a card in the slot. XXX still need to fix the resource list code. It is totally busted and XXX causes a panic in the child printing routine after the probe has XXX succeeded.
2000-10-07 05:48:51 +00:00
}
return (0);
Implement indirection in the pccard probe/attach. This should make it possible to have different probe/attach semantics between the two systems and yet still use the same driver for both. Compatibility methods for OLDCARD drivers. We use these routines to make it possible to call the OLDCARD driver's probe routine in the context that it expects. For OLDCARD these are implemented as pass throughs to the device_{probe,attach} routines. For NEWCARD they are implemented such such that probe becomes strictly a matching routine and attach does both the old probe and old attach. compat devices should use the following: /* Device interface */ DEVMETHOD(device_probe), pccard_compat_probe), DEVMETHOD(device_attach), pccard_compat_attach), /* Card interface */ DEVMETHOD(card_compat_match, foo_match), /* newly written */ DEVMETHOD(card_compat_probe, foo_probe), /* old probe */ DEVMETHOD(card_compat_attach, foo_attach), /* old attach */ This will allow a single driver binary image to be used for both OLDCARD and NEWCARD. Drivers wishing to not retain OLDCARD compatibility needn't do this. ep driver minorly updated. sn driver updated more than minorly. Add module dependencies to allow module to load. Also change name to if_sn. Add some debugging code. attempt to fix the cannot allocate memory problem I'd been seeing. Minor formatting nits.
2000-09-19 04:39:20 +00:00
}
o Implement get_ivars so matching routines will work (this causes compat probe routines to work). o Have a null driver_added routines. We need to be careful about probing until after we know we have a card. For the moment, we do nothing (which is safe). This fixes a panic when a driver is loaded w/o a card in the slot. XXX still need to fix the resource list code. It is totally busted and XXX causes a panic in the child printing routine after the probe has XXX succeeded.
2000-10-07 05:48:51 +00:00
static void
pccard_driver_added(device_t dev, driver_t *driver)
{
Lots of little fixes. We no longer panic on card insert: o initialize ivars with bzero. o remove interrupt function pointer. netbsd needs it, but we don't. o add lots of comments about bogus things that I've been kludging to try to make the simple cases work. o add new ivar accessor for cis4 to match cis3. likely neither will be needed, but it doesn't hurt to have it.
2000-10-15 17:21:50 +00:00
/*
* XXX eventually we need to attach stuff when we know we
* XXX have kids. For now we do nothing because we normally
* XXX add children ourselves. We don't want to necessarily
* XXX force a reprobe.
*/
o Implement get_ivars so matching routines will work (this causes compat probe routines to work). o Have a null driver_added routines. We need to be careful about probing until after we know we have a card. For the moment, we do nothing (which is safe). This fixes a panic when a driver is loaded w/o a card in the slot. XXX still need to fix the resource list code. It is totally busted and XXX causes a panic in the child printing routine after the probe has XXX succeeded.
2000-10-07 05:48:51 +00:00
}
Implement indirection in the pccard probe/attach. This should make it possible to have different probe/attach semantics between the two systems and yet still use the same driver for both. Compatibility methods for OLDCARD drivers. We use these routines to make it possible to call the OLDCARD driver's probe routine in the context that it expects. For OLDCARD these are implemented as pass throughs to the device_{probe,attach} routines. For NEWCARD they are implemented such such that probe becomes strictly a matching routine and attach does both the old probe and old attach. compat devices should use the following: /* Device interface */ DEVMETHOD(device_probe), pccard_compat_probe), DEVMETHOD(device_attach), pccard_compat_attach), /* Card interface */ DEVMETHOD(card_compat_match, foo_match), /* newly written */ DEVMETHOD(card_compat_probe, foo_probe), /* old probe */ DEVMETHOD(card_compat_attach, foo_attach), /* old attach */ This will allow a single driver binary image to be used for both OLDCARD and NEWCARD. Drivers wishing to not retain OLDCARD compatibility needn't do this. ep driver minorly updated. sn driver updated more than minorly. Add module dependencies to allow module to load. Also change name to if_sn. Add some debugging code. attempt to fix the cannot allocate memory problem I'd been seeing. Minor formatting nits.
2000-09-19 04:39:20 +00:00
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
static struct resource *
pccard_alloc_resource(device_t dev, device_t child, int type, int *rid,
u_long start, u_long end, u_long count, u_int flags)
{
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
struct resource_list_entry *rle = NULL;
/* XXX: This is an ugly way to fudge the resources. */
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
if (device_get_parent(child) == dev) {
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
struct pccard_ivar *devi = PCCARD_IVAR(child);
struct resource_list *rl = &devi->resources;
rle = resource_list_find(rl, type, *rid);
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
}
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
if (rle != NULL) {
if (flags & RF_ACTIVE)
bus_activate_resource(dev, type, rle->rid, rle->res);
return (rle->res);
o Remove a boatload of debugging printfs. o Report function number and config index on probe line o Activate the resources (I hope) when RF_ACTIVE is set on those resources I'm allocating on behalf of my children. o Always enable interrupts on multifunction cards in the multifunction register.
2000-10-18 02:21:00 +00:00
}
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
return (bus_generic_alloc_resource(dev, child, type, rid, start,
end, count, flags));
}
static int
pccard_release_resource(device_t dev, device_t child, int type, int rid,
struct resource *r)
{
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
struct resource_list_entry *rle = NULL;
if (device_get_parent(child) == dev) {
struct pccard_ivar *devi = PCCARD_IVAR(child);
struct resource_list *rl = &devi->resources;
rle = resource_list_find(rl, type, rid);
}
if (rle != NULL) {
This is a bandaid for a problem that is not entirely fixed yet. The pccard_function_init() call creates a bunch of inactive resources that are persistant and configured on demand. When the child driver "allocates" a resource it is connected up to one of these. When the child releases the resource, we should not delete our copy, just deactivate it again. Otherwise there is nothing to recreate it again after several probe functions have run and done an alloc/release cycle. INVARIANTS shows 0xdeadc0de without this. More work is needed to do a sweep though the pccard_function_disable() call to actually delete the resources for real. Right now, we leak memory on eject (at best), so Dont Do That(TM) yet. This affects 16 bit pccards on a cardbus bridge only. This will be fixed soon, but for now it gets the cards working. Reviewed by: imp
2001-01-07 21:32:05 +00:00
return bus_deactivate_resource(dev, type, rle->rid, rle->res);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
}
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
return bus_generic_release_resource(dev, child, type, rid, r);
}
static int
pccard_activate_resource(device_t dev, device_t child, int type, int rid,
struct resource *r)
{
o Remove a boatload of debugging printfs. o Report function number and config index on probe line o Activate the resources (I hope) when RF_ACTIVE is set on those resources I'm allocating on behalf of my children. o Always enable interrupts on multifunction cards in the multifunction register.
2000-10-18 02:21:00 +00:00
/* XXX need to write to the COR to activate this for mf cards */
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
struct resource_list_entry *rle = NULL;
if (device_get_parent(child) == dev) {
struct pccard_ivar *devi = PCCARD_IVAR(child);
struct resource_list *rl = &devi->resources;
rle = resource_list_find(rl, type, rid);
}
if (rle != NULL) {
return (bus_activate_resource(dev, type, rle->rid, rle->res));
}
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
return (bus_generic_activate_resource(dev, child, type, rid, r));
}
static int
pccard_deactivate_resource(device_t dev, device_t child, int type, int rid,
struct resource *r)
{
o Remove a boatload of debugging printfs. o Report function number and config index on probe line o Activate the resources (I hope) when RF_ACTIVE is set on those resources I'm allocating on behalf of my children. o Always enable interrupts on multifunction cards in the multifunction register.
2000-10-18 02:21:00 +00:00
/* XXX need to write to the COR to deactivate this for mf cards */
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
struct resource_list_entry *rle = NULL;
if (device_get_parent(child) == dev) {
struct pccard_ivar *devi = PCCARD_IVAR(child);
struct resource_list *rl = &devi->resources;
rle = resource_list_find(rl, type, rid);
}
if (rle != NULL) {
return (bus_deactivate_resource(dev, type, rle->rid, rle->res));
}
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
return (bus_generic_deactivate_resource(dev, child, type, rid, r));
}
o Fix memory leak in ivars o Change name of bus o Change the panic on resource allocation failure to just a message. We'll work out why this fails later in the pcic/pccbb code merge.
2000-10-19 22:36:41 +00:00
static void
pccard_child_detached(device_t parent, device_t dev)
{
struct pccard_ivar *ivar = PCCARD_IVAR(dev);
if (parent == device_get_parent(dev))
free(ivar, M_DEVBUF);
}
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
static void
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
pccard_intr(void *arg)
{
struct pccard_softc *sc = (struct pccard_softc *) arg;
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
struct pccard_function *pf;
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
STAILQ_FOREACH(pf, &sc->card.pf_head, pf_list) {
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
if (pf->intr_handler != NULL) {
int reg = pccard_ccr_read(pf, PCCARD_CCR_STATUS);
if (reg & PCCARD_CCR_STATUS_INTR) {
pccard_ccr_write(pf, PCCARD_CCR_STATUS,
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
reg & ~PCCARD_CCR_STATUS_INTR);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
pf->intr_handler(pf->intr_handler_arg);
}
}
}
}
static int
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
pccard_setup_intr(device_t dev, device_t child, struct resource *irq,
int flags, driver_intr_t *intr, void *arg, void **cookiep)
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
{
struct pccard_ivar *ivar = PCCARD_IVAR(child);
struct pccard_function *func = ivar->fcn;
struct resource_list_entry *rle = NULL;
struct pccard_softc *sc = device_get_softc(dev);
if (func->intr_handler != NULL)
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
panic("Only one interrupt handler per function allowed\n");
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
rle = resource_list_find(&ivar->resources, SYS_RES_IRQ, 0);
if (rle == NULL || rle->res != irq)
panic("irq in setup_intr does not match allocated irq\n");
func->intr_handler = intr;
func->intr_handler_arg = arg;
For bus_setup_intr, cookiep is a pointer to storage for the parent bus to use. We need to set it here. This fixes the problem where a probe routine establishes and disestablishes the interrupt and then we get a panic in the probe routine. However, we pass the pointer to the interrupt hanlder count to the parent bus, which writes its own cookie there, so there may be some problems with that which isn't apparent at the moment. Commit made from: laptop running NEWCARD with sn driver (which works, but gets the wrong ethernet address).
2001-01-07 20:36:27 +00:00
func->intr_handler_cookie = *cookiep = func;
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
pccard_ccr_write(func, PCCARD_CCR_OPTION,
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
pccard_ccr_read(func, PCCARD_CCR_OPTION) |
PCCARD_CCR_OPTION_IREQ_ENABLE);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
if (sc->intr_handler_count++ == 0) {
rle = resource_list_find(&ivar->resources, SYS_RES_IRQ, 0);
if (rle == NULL)
panic("No IRQ for pccard?");
bus_setup_intr(dev, rle->res, INTR_TYPE_TTY/* | INTR_FAST*/,
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
pccard_intr, sc, (void*)&sc->intr_handler_count);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
}
return 0;
}
static int
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
pccard_teardown_intr(device_t dev, device_t child, struct resource *r,
void *cookie)
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
{
struct pccard_ivar *ivar = PCCARD_IVAR(child);
struct pccard_function *func = ivar->fcn;
struct pccard_softc *sc = device_get_softc(dev);
if (func->intr_handler_cookie != cookie)
panic("pccard teardown of unknown interrupt handler\n");
func->intr_handler = NULL;
func->intr_handler_arg = NULL;
func->intr_handler_cookie = NULL;
pccard_ccr_write(func, PCCARD_CCR_OPTION,
This file follows style(9), so adjust various nits in the last few commits to conform to style(9), plus one other convention that I use: o Declare variables at the start of the function, rather than in blocks when it doesn't help understanding (mine). o 80 column limit. o BSD style statement continuation, rather than "gnu" style.
2001-01-07 20:40:23 +00:00
pccard_ccr_read(func, PCCARD_CCR_OPTION) &
~PCCARD_CCR_OPTION_IREQ_ENABLE);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
if (--sc->intr_handler_count == 0) {
struct resource_list_entry *rle = NULL;
rle = resource_list_find(&ivar->resources, SYS_RES_IRQ, 0);
if (rle == NULL)
panic("No IRQ for pccard?");
bus_teardown_intr(dev, rle->res, &sc->intr_handler_count);
}
return 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
static device_method_t pccard_methods[] = {
/* Device interface */
DEVMETHOD(device_probe, pccard_probe),
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
DEVMETHOD(device_attach, pccard_attach),
Matching commits to pccard for last pcic changes. We now at least to probe/attach. This is a checkpoint.
2000-06-18 05:28:59 +00:00
DEVMETHOD(device_detach, bus_generic_detach),
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
DEVMETHOD(device_shutdown, bus_generic_shutdown),
DEVMETHOD(device_suspend, bus_generic_suspend),
DEVMETHOD(device_resume, bus_generic_resume),
/* Bus interface */
DEVMETHOD(bus_print_child, pccard_print_child),
o Implement get_ivars so matching routines will work (this causes compat probe routines to work). o Have a null driver_added routines. We need to be careful about probing until after we know we have a card. For the moment, we do nothing (which is safe). This fixes a panic when a driver is loaded w/o a card in the slot. XXX still need to fix the resource list code. It is totally busted and XXX causes a panic in the child printing routine after the probe has XXX succeeded.
2000-10-07 05:48:51 +00:00
DEVMETHOD(bus_driver_added, pccard_driver_added),
o Fix memory leak in ivars o Change name of bus o Change the panic on resource allocation failure to just a message. We'll work out why this fails later in the pcic/pccbb code merge.
2000-10-19 22:36:41 +00:00
DEVMETHOD(bus_child_detached, pccard_child_detached),
More NEWCARD fixes. We now almost properly print the probe message. o Remember the resources we allocate for the config entry. o When we get the resource, do an resource_list_add and do a resource_list_delete if we fail later in the resource list. o In the pccard bus, we allocate the resources. When a child asks for them, just return the resources that we allocated (thanks to Paul Richards and Mike Smith for the idea).
2000-10-17 06:29:21 +00:00
DEVMETHOD(bus_alloc_resource, pccard_alloc_resource),
DEVMETHOD(bus_release_resource, pccard_release_resource),
DEVMETHOD(bus_activate_resource, pccard_activate_resource),
DEVMETHOD(bus_deactivate_resource, pccard_deactivate_resource),
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
DEVMETHOD(bus_setup_intr, pccard_setup_intr),
DEVMETHOD(bus_teardown_intr, pccard_teardown_intr),
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
DEVMETHOD(bus_set_resource, pccard_set_resource),
DEVMETHOD(bus_get_resource, pccard_get_resource),
DEVMETHOD(bus_delete_resource, pccard_delete_resource),
Implement indirection in the pccard probe/attach. This should make it possible to have different probe/attach semantics between the two systems and yet still use the same driver for both. Compatibility methods for OLDCARD drivers. We use these routines to make it possible to call the OLDCARD driver's probe routine in the context that it expects. For OLDCARD these are implemented as pass throughs to the device_{probe,attach} routines. For NEWCARD they are implemented such such that probe becomes strictly a matching routine and attach does both the old probe and old attach. compat devices should use the following: /* Device interface */ DEVMETHOD(device_probe), pccard_compat_probe), DEVMETHOD(device_attach), pccard_compat_attach), /* Card interface */ DEVMETHOD(card_compat_match, foo_match), /* newly written */ DEVMETHOD(card_compat_probe, foo_probe), /* old probe */ DEVMETHOD(card_compat_attach, foo_attach), /* old attach */ This will allow a single driver binary image to be used for both OLDCARD and NEWCARD. Drivers wishing to not retain OLDCARD compatibility needn't do this. ep driver minorly updated. sn driver updated more than minorly. Add module dependencies to allow module to load. Also change name to if_sn. Add some debugging code. attempt to fix the cannot allocate memory problem I'd been seeing. Minor formatting nits.
2000-09-19 04:39:20 +00:00
DEVMETHOD(bus_read_ivar, pccard_read_ivar),
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
/* Card Interface */
DEVMETHOD(card_set_res_flags, pccard_set_res_flags),
DEVMETHOD(card_set_memory_offset, pccard_set_memory_offset),
DEVMETHOD(card_get_type, pccard_card_gettype),
DEVMETHOD(card_attach_card, pccard_attach_card),
DEVMETHOD(card_detach_card, pccard_detach_card),
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
{ 0, 0 }
};
static driver_t pccard_driver = {
"pccard",
pccard_methods,
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
sizeof(struct pccard_softc)
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
};
devclass_t pccard_devclass;
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
DRIVER_MODULE(pccard, pcic, pccard_driver, pccard_devclass, 0, 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
DRIVER_MODULE(pccard, pc98pcic, pccard_driver, pccard_devclass, 0, 0);
DRIVER_MODULE(pccard, pccbb, pccard_driver, pccard_devclass, 0, 0);
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
DRIVER_MODULE(pccard, tcic, pccard_driver, pccard_devclass, 0, 0);
Save a pointer to the dev, if any, that each function attach. Remove some bogus printfs.
2000-08-22 04:25:57 +00:00
MODULE_VERSION(pccard, 1);
* Better kld support in pccbb/cardbus - pccbb no longer needs to remember whether a card is inserted. - pccbb reissues insertion on load of cardbus/pccard modules. - got rid of unnecessary delays in power functions. - Cardbus children are no longer deleted if probe/attach fails. - non-attached child devices are reprobed at driver_added. * CARD interface to read CIS - added card_cis_read/card_cis_free interface to read arbitrary CIS data. This currently is only implemented in cardbus. * pccard begins to work - pccard can now use higher memory space (and uses it by default). - set_memory_offset interface changed. - fixed ccr access, which was broken at multiple locations. - implement an interrupt handler - pccard can now share interrupts. - resource alloc/release/activate/deactivate functions gutted: some resources are allocated by the bridge before the child device is probed or attached. Thus the resource "belongs" to the bridge, and the pccard_*_resource functions need to fudge the owner/rid. - changed some error conditions to panics to speed debugging. * Mutex fix - Giant is entered at the beginning of thread
2001-01-06 18:04:55 +00:00
/*MODULE_DEPEND(pccard, pcic, 1, 1, 1);*/
Reference in New Issue Copy Permalink