d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
04738e99b5
freebsd-dev/sys/dev/ata/ata-dma.c

1809 lines
59 KiB
C
Raw Normal View History

Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
/*-
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
* Copyright (c) 1998,1999,2000,2001,2002 S<EFBFBD>ren Schmidt <sos@FreeBSD.org>
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer,
* without modification, immediately at the beginning of the file.
* 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. 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.
*
$Id$ -> $FreeBSD$
1999-08-28 01:08:13 +00:00
* $FreeBSD$
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
*/
#include <sys/param.h>
#include <sys/systm.h>
Provide the interface to atacontrol and associated logic. see atacontrol(8) for more. Also the ATA_ENABLE_ATAPI_DMA, ATA_ENABLE_WC and ATA_ENABLE_TAGS options are gone, use the tuneables listed in ata.4 instead from the loader (this makes it possible to switch off DMA before the driver has to touch the devices on broken hardware).
2001-03-15 15:36:25 +00:00
#include <sys/ata.h>
Move the new byte order function prototypes from <sys/param.h> to <sys/endian.h>. This puts us in line with NetBSD and OpenBSD.
2002-04-26 22:48:23 +00:00
#include <sys/endian.h>
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
#include <sys/malloc.h>
Update to use the new-bus framework. No functional changes. Mostly done by Doug Rabson, minor fixes by me.
1999-04-18 20:48:15 +00:00
#include <sys/bus.h>
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
#include <pci/pcivar.h>
* Completely rewrite the alpha busspace to hide the implementation from the drivers. * Remove legacy inx/outx support from chipset and replace with macros which call busspace. * Rework pci config accesses to route through the pcib device instead of calling a MD function directly. With these changes it is possible to cleanly support machines which have more than one independantly numbered PCI busses. As a bonus, the new busspace implementation should be measurably faster than the old one.
2000-08-28 21:48:13 +00:00
#include <machine/bus.h>
Introduce busspace instead of the good old in/out instructions. Not pretty but it works (I hope)...
2001-02-06 16:44:25 +00:00
#include <sys/rman.h>
Remove unneeded includes Pointed out by: phk
2000-09-20 07:00:24 +00:00
#include <dev/ata/ata-all.h>
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
/* prototypes */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
static void ata_dmacreate(struct ata_device *, int, int);
static void ata_dmasetupd_cb(void *, bus_dma_segment_t *, int, int);
static void ata_dmasetupc_cb(void *, bus_dma_segment_t *, int, int);
static void cyrix_timing(struct ata_device *, int, int);
static void promise_timing(struct ata_device *, int, int);
static void hpt_timing(struct ata_device *, int, int);
static int hpt_cable80(struct ata_device *);
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
/* misc defines */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
#define ATAPI_DEVICE(atadev) \
((atadev->unit == ATA_MASTER && \
atadev->channel->devices & ATA_ATAPI_MASTER) || \
(atadev->unit == ATA_SLAVE && \
atadev->channel->devices & ATA_ATAPI_SLAVE))
#define MAXSEGSZ PAGE_SIZE
#define MAXTABSZ PAGE_SIZE
#define MAXCTLDMASZ (2 * (MAXTABSZ + MAXPHYS))
struct ata_dc_cb_args {
bus_addr_t maddr;
int error;
};
static void
ata_dmasetupc_cb(void *xsc, bus_dma_segment_t *segs, int nsegs, int error)
{
struct ata_dc_cb_args *cba = (struct ata_dc_cb_args *)xsc;
if (!(cba->error = error))
cba->maddr = segs[0].ds_addr;
}
Bring the 'new-bus' to the i386. This extensively changes the way the i386 platform boots, it is no longer ISA-centric, and is fully dynamic. Most old drivers compile and run without modification via 'compatability shims' to enable a smoother transition. eisa, isapnp and pccard* are not yet using the new resource manager. Once fully converted, all drivers will be loadable, including PCI and ISA. (Some other changes appear to have snuck in, including a port of Soren's ATA driver to the Alpha. Soren, back this out if you need to.) This is a checkpoint of work-in-progress, but is quite functional. The bulk of the work was done over the last few years by Doug Rabson and Garrett Wollman. Approved by: core
1999-04-16 21:22:55 +00:00
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
int
ata_dmaalloc(struct ata_device *atadev)
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
{
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
struct ata_channel *ch;
struct ata_dc_cb_args ccba;
struct ata_dmastate *ds;
int error;
ch = atadev->channel;
ds = &atadev->dmastate;
if (!ds->cdmatag) {
if ((error = bus_dma_tag_create(ch->dmatag, 1, PAGE_SIZE,
BUS_SPACE_MAXADDR_32BIT,
BUS_SPACE_MAXADDR, NULL, NULL,
MAXTABSZ, 1, MAXTABSZ,
BUS_DMA_ALLOCNOW, &ds->cdmatag)))
return error;
}
if (!ds->ddmatag) {
if ((error = bus_dma_tag_create(ch->dmatag, ch->alignment + 1, 0,
BUS_SPACE_MAXADDR_32BIT,
BUS_SPACE_MAXADDR, NULL, NULL,
MAXPHYS, ATA_DMA_ENTRIES, MAXSEGSZ,
BUS_DMA_ALLOCNOW, &ds->ddmatag)))
return error;
}
if (!ds->mdmatab) {
if ((error = bus_dmamem_alloc(ds->cdmatag, (void **)&ds->dmatab, 0,
&ds->cdmamap)))
return error;
if ((error = bus_dmamap_load(ds->cdmatag, ds->cdmamap, ds->dmatab,
MAXTABSZ, ata_dmasetupc_cb, &ccba,
0)) != 0 || ccba.error != 0) {
bus_dmamem_free(ds->cdmatag, ds->dmatab, ds->cdmamap);
return error;
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ds->mdmatab = ccba.maddr;
}
if (!ds->ddmamap) {
if ((error = bus_dmamap_create(ds->ddmatag, 0, &ds->ddmamap)) != 0)
return error;
}
return 0;
}
void
ata_dmafree(struct ata_device *atadev)
{
struct ata_dmastate *ds;
ds = &atadev->dmastate;
if (ds->mdmatab) {
bus_dmamap_unload(ds->cdmatag, ds->cdmamap);
bus_dmamem_free(ds->cdmatag, ds->dmatab, ds->cdmamap);
ds->mdmatab = 0;
ds->cdmamap = NULL;
ds->dmatab = NULL;
}
if (ds->ddmamap) {
bus_dmamap_destroy(ds->ddmatag, ds->ddmamap);
ds->ddmamap = NULL;
}
if (ds->cdmatag) {
bus_dma_tag_destroy(ds->cdmatag);
ds->cdmatag = NULL;
}
if (ds->ddmatag) {
bus_dma_tag_destroy(ds->ddmatag);
ds->ddmatag = NULL;
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
}
}
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
void
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmafreetags(struct ata_channel *ch)
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
{
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (ch->dmatag) {
bus_dma_tag_destroy(ch->dmatag);
ch->dmatag = NULL;
}
}
static void
ata_dmacreate(struct ata_device *atadev, int apiomode, int mode)
{
atadev->mode = mode;
if (!atadev->channel->dmatag) {
if (bus_dma_tag_create(NULL, 1, 0,
BUS_SPACE_MAXADDR_32BIT, BUS_SPACE_MAXADDR,
NULL, NULL, MAXCTLDMASZ, ATA_DMA_ENTRIES,
BUS_SPACE_MAXSIZE_32BIT, 0,
&atadev->channel->dmatag)) {
ata_prtdev(atadev, "DMA tag allocation failed, disabling DMA\n");
ata_dmainit(atadev, apiomode, -1, -1);
}
}
}
void
ata_dmainit(struct ata_device *atadev, int apiomode, int wdmamode, int udmamode)
{
device_t parent = device_get_parent(atadev->channel->dev);
int chiptype = atadev->channel->chiptype;
int chiprev = pci_get_revid(parent);
int channel = atadev->channel->unit;
int device = ATA_DEV(atadev->unit);
int devno = (channel << 1) + device;
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
int error;
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
/* set our most pessimistic default mode */
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
atadev->mode = ATA_PIO;
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (!atadev->channel->r_bmio)
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
/* if simplex controller, only allow DMA on primary channel */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (channel == 1) {
ATA_OUTB(atadev->channel->r_bmio, ATA_BMSTAT_PORT,
ATA_INB(atadev->channel->r_bmio, ATA_BMSTAT_PORT) &
Introduce busspace instead of the good old in/out instructions. Not pretty but it works (I hope)...
2001-02-06 16:44:25 +00:00
(ATA_BMSTAT_DMA_MASTER | ATA_BMSTAT_DMA_SLAVE));
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (ATA_INB(atadev->channel->r_bmio, ATA_BMSTAT_PORT) &
ATA_BMSTAT_DMA_SIMPLEX) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "simplex device, DMA on primary only\n");
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
}
}
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
/* DMA engine address alignment is usually 1 word (2 bytes) */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
atadev->channel->alignment = 0x1;
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
Add experimental support for the Promise ATA133 controller. This might need additional code for hotswapping use, but first I need to get my hands on actual hw... Based on docs kindly provided by Promise Inc which we thank for the close cooperation with the FreeBSD project!.
2001-10-23 08:01:47 +00:00
#if 1
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (udmamode > 2 && !atadev->param->hwres_cblid) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev,"DMA limited to UDMA33, non-ATA66 cable or device\n");
Fix a nasty bug in ata_intr, parens are a good thing if used right. Make the error recovery code a little more obvious. Inform the user if UDMA66 mode couldn't be selected due to a non ATA66 compliant 80pin cable. Minor cosmetics.
2000-05-23 19:05:56 +00:00
udmamode = 2;
}
Add experimental support for the Promise ATA133 controller. This might need additional code for hotswapping use, but first I need to get my hands on actual hw... Based on docs kindly provided by Promise Inc which we thank for the close cooperation with the FreeBSD project!.
2001-10-23 08:01:47 +00:00
#endif
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
switch (chiptype) {
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
Add PCI ID for the ICH4 ATA100 controller. Sponsored by: The Weather Channel
2002-07-19 22:14:54 +00:00
case 0x24cb8086: /* Intel ICH4 */
Add a bunch of new PCI ID's for Intel ATA chips.
2001-10-23 09:57:49 +00:00
case 0x248a8086: /* Intel ICH3 mobile */
case 0x248b8086: /* Intel ICH3 */
Add support the the Intel ICH2 mobile Tidy chip name printing a bit.
2001-04-16 21:22:34 +00:00
case 0x244a8086: /* Intel ICH2 mobile */
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
case 0x244b8086: /* Intel ICH2 */
if (udmamode >= 5) {
int32_t mask48, new48;
int16_t word54;
word54 = pci_read_config(parent, 0x54, 2);
if (word54 & (0x10 << devno)) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_UDMA5, ATA_C_F_SETXFER,ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting UDMA5 on Intel chip\n",
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
(error) ? "failed" : "success");
if (!error) {
mask48 = (1 << devno) + (3 << (16 + (devno << 2)));
new48 = (1 << devno) + (1 << (16 + (devno << 2)));
pci_write_config(parent, 0x48,
(pci_read_config(parent, 0x48, 4) &
~mask48) | new48, 4);
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
pci_write_config(parent, 0x54, word54 | (0x1000<<devno), 2);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA5);
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
return;
}
}
}
Fix the ICH2 ATA100 support, and proberly downgrade the chip setup when going to slower DMA modes on all Intel chips.
2000-08-10 07:34:49 +00:00
/* make sure eventual ATA100 mode from the BIOS is disabled */
pci_write_config(parent, 0x54,
pci_read_config(parent, 0x54, 2) & ~(0x1000<<devno),2);
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
/* FALLTHROUGH */
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
case 0x24118086: /* Intel ICH */
case 0x76018086: /* Intel ICH */
Add UDMA66 support to the Intel ICH chip. Submitted by: sakichan@lares.dti.ne.jp
2000-02-22 20:37:01 +00:00
if (udmamode >= 4) {
int32_t mask48, new48;
int16_t word54;
word54 = pci_read_config(parent, 0x54, 2);
if (word54 & (0x10 << devno)) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_UDMA4, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting UDMA4 on Intel chip\n",
Add support the the Intel ICH2 mobile Tidy chip name printing a bit.
2001-04-16 21:22:34 +00:00
(error) ? "failed" : "success");
Add UDMA66 support to the Intel ICH chip. Submitted by: sakichan@lares.dti.ne.jp
2000-02-22 20:37:01 +00:00
if (!error) {
mask48 = (1 << devno) + (3 << (16 + (devno << 2)));
new48 = (1 << devno) + (2 << (16 + (devno << 2)));
pci_write_config(parent, 0x48,
(pci_read_config(parent, 0x48, 4) &
~mask48) | new48, 4);
pci_write_config(parent, 0x54, word54 | (1 << devno), 2);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA4);
Add UDMA66 support to the Intel ICH chip. Submitted by: sakichan@lares.dti.ne.jp
2000-02-22 20:37:01 +00:00
return;
}
}
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
}
Fix the ICH2 ATA100 support, and proberly downgrade the chip setup when going to slower DMA modes on all Intel chips.
2000-08-10 07:34:49 +00:00
/* make sure eventual ATA66 mode from the BIOS is disabled */
pci_write_config(parent, 0x54,
pci_read_config(parent, 0x54, 2) & ~(1 << devno), 2);
Add UDMA66 support to the Intel ICH chip. Submitted by: sakichan@lares.dti.ne.jp
2000-02-22 20:37:01 +00:00
/* FALLTHROUGH */
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
case 0x71118086: /* Intel PIIX4 */
Add a bunch of new PCI ID's for Intel ATA chips.
2001-10-23 09:57:49 +00:00
case 0x84CA8086: /* Intel PIIX4 */
Add support for the Intel 810 chipset (ICH type of chips) Prober support for the VIA 82C686, I finally got the right datasheet. Get rid of atapi_wait, merge it into ata_wait. Avoid a couple of races by using asleep instead of tsleep. Always use 16bit transfers on ISA systems. Clear up the atapi_read/write functions.
2000-01-17 02:04:19 +00:00
case 0x71998086: /* Intel PIIX4e */
case 0x24218086: /* Intel ICH0 */
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
if (udmamode >= 2) {
Ten'th update to the new ATA/ATAPI driver: It been awhile since the last major update, as a benefit there are some cool things in this one (and new bugs probably :) )... The ATA driver has grown "real" timeout support for all devices. This means that it should be possible to get in contact with (especially) lost ATAPI devices. It also means that the ATA driver is now usable on notebooks as it will DTRT on resume. An experimental hack at utilizing the Promise66's at UDMA66 is in there, but I cant test it. If someone feels like sending me one, give me a ping. The ATAPI DMA enableling scheme has been changed, also better DMA support for the Aladdin chipset has been implemented for ATAPI devices. Note that the Aladdin apparently only can do DMA reads on ATAPI devices, and the Promise cant do ATAPI DMA at all. I have seen problems on some ATAPI devices that should be able to run in DMA mode, so if you encounter problems with hanging atapi devices during the probe, or during access, disable DMA in atapi-all.c, and let me know. It might be nessesary to do this via a "white list" for known good devices... The ATAPI CDROM driver can now use eject/close without hanging and the bug that caused reading beyond the end of a CD has been fixed. Media change is also handled proberly. DVD drives are identified and are usable as CDROM devices at least, I dont have the HW to test this further, see above :). The ATAPI tape driver has gotten some support for using the DSC method for not blocking the IDE channel during read/write when the device has full buffers. It knows about the OnStream DI-30 device, support is not completed yet, but it can function as a primitive backup medium, without filemarks, and without bad media handeling. This is because the OnStream device doesn't handle this (like everybody else) in HW. It also now supports getting/setting the record position on devices that supports it. Some rather major cleanups and rearrangements as well (cvs -b diff is your freind). I'm closing in on declaring this for beta code, most of the infrastruture is in place by now. As usual USE AT YOUR OWN RISK!!, this is still alpha level code. This driver can hose your disk real bad if anything goes wrong, but now you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-09-21 19:50:40 +00:00
int32_t mask48, new48;
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Unbreak ATAPI on the Aladdin chipset, only DMA access worked. Try to use a 32bit mask on the IO addresses, this fixes the alpha and hopefully doesn't break on any i386 machines. Try to enable both read & write cache on disks, they should be as default, but better be sure..
1999-11-08 21:36:00 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
Ten'th update to the new ATA/ATAPI driver: It been awhile since the last major update, as a benefit there are some cool things in this one (and new bugs probably :) )... The ATA driver has grown "real" timeout support for all devices. This means that it should be possible to get in contact with (especially) lost ATAPI devices. It also means that the ATA driver is now usable on notebooks as it will DTRT on resume. An experimental hack at utilizing the Promise66's at UDMA66 is in there, but I cant test it. If someone feels like sending me one, give me a ping. The ATAPI DMA enableling scheme has been changed, also better DMA support for the Aladdin chipset has been implemented for ATAPI devices. Note that the Aladdin apparently only can do DMA reads on ATAPI devices, and the Promise cant do ATAPI DMA at all. I have seen problems on some ATAPI devices that should be able to run in DMA mode, so if you encounter problems with hanging atapi devices during the probe, or during access, disable DMA in atapi-all.c, and let me know. It might be nessesary to do this via a "white list" for known good devices... The ATAPI CDROM driver can now use eject/close without hanging and the bug that caused reading beyond the end of a CD has been fixed. Media change is also handled proberly. DVD drives are identified and are usable as CDROM devices at least, I dont have the HW to test this further, see above :). The ATAPI tape driver has gotten some support for using the DSC method for not blocking the IDE channel during read/write when the device has full buffers. It knows about the OnStream DI-30 device, support is not completed yet, but it can function as a primitive backup medium, without filemarks, and without bad media handeling. This is because the OnStream device doesn't handle this (like everybody else) in HW. It also now supports getting/setting the record position on devices that supports it. Some rather major cleanups and rearrangements as well (cvs -b diff is your freind). I'm closing in on declaring this for beta code, most of the infrastruture is in place by now. As usual USE AT YOUR OWN RISK!!, this is still alpha level code. This driver can hose your disk real bad if anything goes wrong, but now you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-09-21 19:50:40 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA2 on Intel chip\n",
Add support the the Intel ICH2 mobile Tidy chip name printing a bit.
2001-04-16 21:22:34 +00:00
(error) ? "failed" : "success");
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (!error) {
mask48 = (1 << devno) + (3 << (16 + (devno << 2)));
new48 = (1 << devno) + (2 << (16 + (devno << 2)));
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
pci_write_config(parent, 0x48,
(pci_read_config(parent, 0x48, 4) &
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
~mask48) | new48, 4);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
}
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
}
Fix the ICH2 ATA100 support, and proberly downgrade the chip setup when going to slower DMA modes on all Intel chips.
2000-08-10 07:34:49 +00:00
/* make sure eventual ATA33 mode from the BIOS is disabled */
pci_write_config(parent, 0x48,
pci_read_config(parent, 0x48, 4) & ~(1 << devno), 4);
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
/* FALLTHROUGH */
case 0x70108086: /* Intel PIIX3 */
if (wdmamode >= 2 && apiomode >= 4) {
int32_t mask40, new40, mask44, new44;
/* if SITRE not set doit for both channels */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (!((pci_read_config(parent, 0x40, 4) >> (channel<<8)) & 0x4000)){
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
new40 = pci_read_config(parent, 0x40, 4);
new44 = pci_read_config(parent, 0x44, 4);
Ten'th update to the new ATA/ATAPI driver: It been awhile since the last major update, as a benefit there are some cool things in this one (and new bugs probably :) )... The ATA driver has grown "real" timeout support for all devices. This means that it should be possible to get in contact with (especially) lost ATAPI devices. It also means that the ATA driver is now usable on notebooks as it will DTRT on resume. An experimental hack at utilizing the Promise66's at UDMA66 is in there, but I cant test it. If someone feels like sending me one, give me a ping. The ATAPI DMA enableling scheme has been changed, also better DMA support for the Aladdin chipset has been implemented for ATAPI devices. Note that the Aladdin apparently only can do DMA reads on ATAPI devices, and the Promise cant do ATAPI DMA at all. I have seen problems on some ATAPI devices that should be able to run in DMA mode, so if you encounter problems with hanging atapi devices during the probe, or during access, disable DMA in atapi-all.c, and let me know. It might be nessesary to do this via a "white list" for known good devices... The ATAPI CDROM driver can now use eject/close without hanging and the bug that caused reading beyond the end of a CD has been fixed. Media change is also handled proberly. DVD drives are identified and are usable as CDROM devices at least, I dont have the HW to test this further, see above :). The ATAPI tape driver has gotten some support for using the DSC method for not blocking the IDE channel during read/write when the device has full buffers. It knows about the OnStream DI-30 device, support is not completed yet, but it can function as a primitive backup medium, without filemarks, and without bad media handeling. This is because the OnStream device doesn't handle this (like everybody else) in HW. It also now supports getting/setting the record position on devices that supports it. Some rather major cleanups and rearrangements as well (cvs -b diff is your freind). I'm closing in on declaring this for beta code, most of the infrastruture is in place by now. As usual USE AT YOUR OWN RISK!!, this is still alpha level code. This driver can hose your disk real bad if anything goes wrong, but now you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-09-21 19:50:40 +00:00
if (!(new40 & 0x00004000)) {
new44 &= ~0x0000000f;
new44 |= ((new40&0x00003000)>>10)|((new40&0x00000300)>>8);
}
if (!(new40 & 0x40000000)) {
new44 &= ~0x000000f0;
new44 |= ((new40&0x30000000)>>22)|((new40&0x03000000)>>20);
}
new40 |= 0x40004000;
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
pci_write_config(parent, 0x40, new40, 4);
pci_write_config(parent, 0x44, new44, 4);
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
}
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Unbreak ATAPI on the Aladdin chipset, only DMA access worked. Try to use a 32bit mask on the IO addresses, this fixes the alpha and hopefully doesn't break on any i386 machines. Try to enable both read & write cache on disks, they should be as default, but better be sure..
1999-11-08 21:36:00 +00:00
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
Ten'th update to the new ATA/ATAPI driver: It been awhile since the last major update, as a benefit there are some cool things in this one (and new bugs probably :) )... The ATA driver has grown "real" timeout support for all devices. This means that it should be possible to get in contact with (especially) lost ATAPI devices. It also means that the ATA driver is now usable on notebooks as it will DTRT on resume. An experimental hack at utilizing the Promise66's at UDMA66 is in there, but I cant test it. If someone feels like sending me one, give me a ping. The ATAPI DMA enableling scheme has been changed, also better DMA support for the Aladdin chipset has been implemented for ATAPI devices. Note that the Aladdin apparently only can do DMA reads on ATAPI devices, and the Promise cant do ATAPI DMA at all. I have seen problems on some ATAPI devices that should be able to run in DMA mode, so if you encounter problems with hanging atapi devices during the probe, or during access, disable DMA in atapi-all.c, and let me know. It might be nessesary to do this via a "white list" for known good devices... The ATAPI CDROM driver can now use eject/close without hanging and the bug that caused reading beyond the end of a CD has been fixed. Media change is also handled proberly. DVD drives are identified and are usable as CDROM devices at least, I dont have the HW to test this further, see above :). The ATAPI tape driver has gotten some support for using the DSC method for not blocking the IDE channel during read/write when the device has full buffers. It knows about the OnStream DI-30 device, support is not completed yet, but it can function as a primitive backup medium, without filemarks, and without bad media handeling. This is because the OnStream device doesn't handle this (like everybody else) in HW. It also now supports getting/setting the record position on devices that supports it. Some rather major cleanups and rearrangements as well (cvs -b diff is your freind). I'm closing in on declaring this for beta code, most of the infrastruture is in place by now. As usual USE AT YOUR OWN RISK!!, this is still alpha level code. This driver can hose your disk real bad if anything goes wrong, but now you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-09-21 19:50:40 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting WDMA2 on Intel chip\n",
Add support the the Intel ICH2 mobile Tidy chip name printing a bit.
2001-04-16 21:22:34 +00:00
(error) ? "failed" : "success");
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (!error) {
if (device == ATA_MASTER) {
mask40 = 0x0000330f;
new40 = 0x00002307;
mask44 = 0;
new44 = 0;
}
else {
mask40 = 0x000000f0;
new40 = 0x00000070;
mask44 = 0x0000000f;
new44 = 0x0000000b;
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (channel) {
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
mask40 <<= 16;
new40 <<= 16;
mask44 <<= 4;
new44 <<= 4;
}
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
pci_write_config(parent, 0x40,
(pci_read_config(parent, 0x40, 4) & ~mask40)|
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
new40, 4);
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
pci_write_config(parent, 0x44,
(pci_read_config(parent, 0x44, 4) & ~mask44)|
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
new44, 4);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Ten'th update to the new ATA/ATAPI driver: It been awhile since the last major update, as a benefit there are some cool things in this one (and new bugs probably :) )... The ATA driver has grown "real" timeout support for all devices. This means that it should be possible to get in contact with (especially) lost ATAPI devices. It also means that the ATA driver is now usable on notebooks as it will DTRT on resume. An experimental hack at utilizing the Promise66's at UDMA66 is in there, but I cant test it. If someone feels like sending me one, give me a ping. The ATAPI DMA enableling scheme has been changed, also better DMA support for the Aladdin chipset has been implemented for ATAPI devices. Note that the Aladdin apparently only can do DMA reads on ATAPI devices, and the Promise cant do ATAPI DMA at all. I have seen problems on some ATAPI devices that should be able to run in DMA mode, so if you encounter problems with hanging atapi devices during the probe, or during access, disable DMA in atapi-all.c, and let me know. It might be nessesary to do this via a "white list" for known good devices... The ATAPI CDROM driver can now use eject/close without hanging and the bug that caused reading beyond the end of a CD has been fixed. Media change is also handled proberly. DVD drives are identified and are usable as CDROM devices at least, I dont have the HW to test this further, see above :). The ATAPI tape driver has gotten some support for using the DSC method for not blocking the IDE channel during read/write when the device has full buffers. It knows about the OnStream DI-30 device, support is not completed yet, but it can function as a primitive backup medium, without filemarks, and without bad media handeling. This is because the OnStream device doesn't handle this (like everybody else) in HW. It also now supports getting/setting the record position on devices that supports it. Some rather major cleanups and rearrangements as well (cvs -b diff is your freind). I'm closing in on declaring this for beta code, most of the infrastruture is in place by now. As usual USE AT YOUR OWN RISK!!, this is still alpha level code. This driver can hose your disk real bad if anything goes wrong, but now you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-09-21 19:50:40 +00:00
}
}
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
/* we could set PIO mode timings, but we assume the BIOS did that */
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
break;
case 0x12308086: /* Intel PIIX */
Add DMA support for the SiS 5591 and old PIIX chipsets. Add a missing DELAY(1) in ata_wait. Change the info from ad_version, so the ATA version from the disk can be used to quantify the DAM modes valid for this drive, ie be more selective with turning DMA on on older disks that should not support it.. Fix the probe for BIOS enabled DMA in the generic case, master/slave was reversed in the test. Check the return for ata_command in all cases, and print warnings if it fails. Call ata_dmainit with all dmamodes off when falling back to PIO mode, that should take care of both the Promise & HPT366 controllers not being able to handle the fallback... Cleanup the printf's in the drivers, use the prober device name (if possible) instead of ataN-master/slave.
1999-12-13 10:19:39 +00:00
if (wdmamode >= 2 && apiomode >= 4) {
int32_t word40;
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
word40 = pci_read_config(parent, 0x40, 4);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
word40 >>= channel * 16;
Add DMA support for the SiS 5591 and old PIIX chipsets. Add a missing DELAY(1) in ata_wait. Change the info from ad_version, so the ATA version from the disk can be used to quantify the DAM modes valid for this drive, ie be more selective with turning DMA on on older disks that should not support it.. Fix the probe for BIOS enabled DMA in the generic case, master/slave was reversed in the test. Check the return for ata_command in all cases, and print warnings if it fails. Call ata_dmainit with all dmamodes off when falling back to PIO mode, that should take care of both the Promise & HPT366 controllers not being able to handle the fallback... Cleanup the printf's in the drivers, use the prober device name (if possible) instead of ataN-master/slave.
1999-12-13 10:19:39 +00:00
/* Check for timing config usable for DMA on controller */
if (!((word40 & 0x3300) == 0x2300 &&
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
((word40 >> (device ? 4 : 0)) & 1) == 1))
Add DMA support for the SiS 5591 and old PIIX chipsets. Add a missing DELAY(1) in ata_wait. Change the info from ad_version, so the ATA version from the disk can be used to quantify the DAM modes valid for this drive, ie be more selective with turning DMA on on older disks that should not support it.. Fix the probe for BIOS enabled DMA in the generic case, master/slave was reversed in the test. Check the return for ata_command in all cases, and print warnings if it fails. Call ata_dmainit with all dmamodes off when falling back to PIO mode, that should take care of both the Promise & HPT366 controllers not being able to handle the fallback... Cleanup the printf's in the drivers, use the prober device name (if possible) instead of ataN-master/slave.
1999-12-13 10:19:39 +00:00
break;
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add DMA support for the SiS 5591 and old PIIX chipsets. Add a missing DELAY(1) in ata_wait. Change the info from ad_version, so the ATA version from the disk can be used to quantify the DAM modes valid for this drive, ie be more selective with turning DMA on on older disks that should not support it.. Fix the probe for BIOS enabled DMA in the generic case, master/slave was reversed in the test. Check the return for ata_command in all cases, and print warnings if it fails. Call ata_dmainit with all dmamodes off when falling back to PIO mode, that should take care of both the Promise & HPT366 controllers not being able to handle the fallback... Cleanup the printf's in the drivers, use the prober device name (if possible) instead of ataN-master/slave.
1999-12-13 10:19:39 +00:00
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting WDMA2 on Intel chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success");
Add DMA support for the SiS 5591 and old PIIX chipsets. Add a missing DELAY(1) in ata_wait. Change the info from ad_version, so the ATA version from the disk can be used to quantify the DAM modes valid for this drive, ie be more selective with turning DMA on on older disks that should not support it.. Fix the probe for BIOS enabled DMA in the generic case, master/slave was reversed in the test. Check the return for ata_command in all cases, and print warnings if it fails. Call ata_dmainit with all dmamodes off when falling back to PIO mode, that should take care of both the Promise & HPT366 controllers not being able to handle the fallback... Cleanup the printf's in the drivers, use the prober device name (if possible) instead of ataN-master/slave.
1999-12-13 10:19:39 +00:00
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add DMA support for the SiS 5591 and old PIIX chipsets. Add a missing DELAY(1) in ata_wait. Change the info from ad_version, so the ATA version from the disk can be used to quantify the DAM modes valid for this drive, ie be more selective with turning DMA on on older disks that should not support it.. Fix the probe for BIOS enabled DMA in the generic case, master/slave was reversed in the test. Check the return for ata_command in all cases, and print warnings if it fails. Call ata_dmainit with all dmamodes off when falling back to PIO mode, that should take care of both the Promise & HPT366 controllers not being able to handle the fallback... Cleanup the printf's in the drivers, use the prober device name (if possible) instead of ataN-master/slave.
1999-12-13 10:19:39 +00:00
}
}
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
break;
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
case 0x522910b9: /* AcerLabs Aladdin IV/V */
Add ATA66 and ATA100 mode support for Acer chipsets. MFC candidate :)
2001-04-06 19:18:35 +00:00
/* the older Aladdin doesn't support ATAPI DMA on both master & slave */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (chiprev < 0xc2 &&
atadev->channel->devices & ATA_ATAPI_MASTER &&
atadev->channel->devices & ATA_ATAPI_SLAVE) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "two atapi devices on this channel, no DMA\n");
Unbreak ATAPI on the Aladdin chipset, only DMA access worked. Try to use a 32bit mask on the IO addresses, this fixes the alpha and hopefully doesn't break on any i386 machines. Try to enable both read & write cache on disks, they should be as default, but better be sure..
1999-11-08 21:36:00 +00:00
break;
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
pci_write_config(parent, 0x58 + (channel << 2), 0x00310001, 4);
if (udmamode >= 5 && chiprev >= 0xc4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add ATA66 and ATA100 mode support for Acer chipsets. MFC candidate :)
2001-04-06 19:18:35 +00:00
ATA_UDMA5, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA5 on Acer chip\n",
Add ATA66 and ATA100 mode support for Acer chipsets. MFC candidate :)
2001-04-06 19:18:35 +00:00
(error) ? "failed" : "success");
if (!error) {
int32_t word54 = pci_read_config(parent, 0x54, 4);
pci_write_config(parent, 0x4b,
pci_read_config(parent, 0x4b, 1) | 0x01, 1);
word54 &= ~(0x000f000f << (devno << 2));
word54 |= (0x000f0005 << (devno << 2));
pci_write_config(parent, 0x54, word54, 4);
pci_write_config(parent, 0x53,
pci_read_config(parent, 0x53, 1) | 0x03, 1);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA5);
Add ATA66 and ATA100 mode support for Acer chipsets. MFC candidate :)
2001-04-06 19:18:35 +00:00
return;
}
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (udmamode >= 4 && chiprev >= 0xc2) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add ATA66 and ATA100 mode support for Acer chipsets. MFC candidate :)
2001-04-06 19:18:35 +00:00
ATA_UDMA4, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA4 on Acer chip\n",
Add ATA66 and ATA100 mode support for Acer chipsets. MFC candidate :)
2001-04-06 19:18:35 +00:00
(error) ? "failed" : "success");
if (!error) {
int32_t word54 = pci_read_config(parent, 0x54, 4);
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
Add ATA66 and ATA100 mode support for Acer chipsets. MFC candidate :)
2001-04-06 19:18:35 +00:00
pci_write_config(parent, 0x4b,
pci_read_config(parent, 0x4b, 1) | 0x01, 1);
word54 &= ~(0x000f000f << (devno << 2));
word54 |= (0x00080005 << (devno << 2));
pci_write_config(parent, 0x54, word54, 4);
pci_write_config(parent, 0x53,
pci_read_config(parent, 0x53, 1) | 0x03, 1);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA4);
Add ATA66 and ATA100 mode support for Acer chipsets. MFC candidate :)
2001-04-06 19:18:35 +00:00
return;
}
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (udmamode >= 2 && chiprev >= 0x20) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Unbreak ATAPI on the Aladdin chipset, only DMA access worked. Try to use a 32bit mask on the IO addresses, this fixes the alpha and hopefully doesn't break on any i386 machines. Try to enable both read & write cache on disks, they should be as default, but better be sure..
1999-11-08 21:36:00 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
Ten'th update to the new ATA/ATAPI driver: It been awhile since the last major update, as a benefit there are some cool things in this one (and new bugs probably :) )... The ATA driver has grown "real" timeout support for all devices. This means that it should be possible to get in contact with (especially) lost ATAPI devices. It also means that the ATA driver is now usable on notebooks as it will DTRT on resume. An experimental hack at utilizing the Promise66's at UDMA66 is in there, but I cant test it. If someone feels like sending me one, give me a ping. The ATAPI DMA enableling scheme has been changed, also better DMA support for the Aladdin chipset has been implemented for ATAPI devices. Note that the Aladdin apparently only can do DMA reads on ATAPI devices, and the Promise cant do ATAPI DMA at all. I have seen problems on some ATAPI devices that should be able to run in DMA mode, so if you encounter problems with hanging atapi devices during the probe, or during access, disable DMA in atapi-all.c, and let me know. It might be nessesary to do this via a "white list" for known good devices... The ATAPI CDROM driver can now use eject/close without hanging and the bug that caused reading beyond the end of a CD has been fixed. Media change is also handled proberly. DVD drives are identified and are usable as CDROM devices at least, I dont have the HW to test this further, see above :). The ATAPI tape driver has gotten some support for using the DSC method for not blocking the IDE channel during read/write when the device has full buffers. It knows about the OnStream DI-30 device, support is not completed yet, but it can function as a primitive backup medium, without filemarks, and without bad media handeling. This is because the OnStream device doesn't handle this (like everybody else) in HW. It also now supports getting/setting the record position on devices that supports it. Some rather major cleanups and rearrangements as well (cvs -b diff is your freind). I'm closing in on declaring this for beta code, most of the infrastruture is in place by now. As usual USE AT YOUR OWN RISK!!, this is still alpha level code. This driver can hose your disk real bad if anything goes wrong, but now you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-09-21 19:50:40 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA2 on Acer chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success");
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (!error) {
Add ATA66 and ATA100 mode support for Acer chipsets. MFC candidate :)
2001-04-06 19:18:35 +00:00
int32_t word54 = pci_read_config(parent, 0x54, 4);
Fix a nasty bug in ata_intr, parens are a good thing if used right. Make the error recovery code a little more obvious. Inform the user if UDMA66 mode couldn't be selected due to a non ATA66 compliant 80pin cable. Minor cosmetics.
2000-05-23 19:05:56 +00:00
word54 &= ~(0x000f000f << (devno << 2));
word54 |= (0x000a0005 << (devno << 2));
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
pci_write_config(parent, 0x54, word54, 4);
pci_write_config(parent, 0x53,
pci_read_config(parent, 0x53, 1) | 0x03, 1);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
atadev->channel->flags |= ATA_ATAPI_DMA_RO;
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
}
Ten'th update to the new ATA/ATAPI driver: It been awhile since the last major update, as a benefit there are some cool things in this one (and new bugs probably :) )... The ATA driver has grown "real" timeout support for all devices. This means that it should be possible to get in contact with (especially) lost ATAPI devices. It also means that the ATA driver is now usable on notebooks as it will DTRT on resume. An experimental hack at utilizing the Promise66's at UDMA66 is in there, but I cant test it. If someone feels like sending me one, give me a ping. The ATAPI DMA enableling scheme has been changed, also better DMA support for the Aladdin chipset has been implemented for ATAPI devices. Note that the Aladdin apparently only can do DMA reads on ATAPI devices, and the Promise cant do ATAPI DMA at all. I have seen problems on some ATAPI devices that should be able to run in DMA mode, so if you encounter problems with hanging atapi devices during the probe, or during access, disable DMA in atapi-all.c, and let me know. It might be nessesary to do this via a "white list" for known good devices... The ATAPI CDROM driver can now use eject/close without hanging and the bug that caused reading beyond the end of a CD has been fixed. Media change is also handled proberly. DVD drives are identified and are usable as CDROM devices at least, I dont have the HW to test this further, see above :). The ATAPI tape driver has gotten some support for using the DSC method for not blocking the IDE channel during read/write when the device has full buffers. It knows about the OnStream DI-30 device, support is not completed yet, but it can function as a primitive backup medium, without filemarks, and without bad media handeling. This is because the OnStream device doesn't handle this (like everybody else) in HW. It also now supports getting/setting the record position on devices that supports it. Some rather major cleanups and rearrangements as well (cvs -b diff is your freind). I'm closing in on declaring this for beta code, most of the infrastruture is in place by now. As usual USE AT YOUR OWN RISK!!, this is still alpha level code. This driver can hose your disk real bad if anything goes wrong, but now you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-09-21 19:50:40 +00:00
}
Proberly back down DMA modes on the Acer Aladdin.
2000-12-19 10:37:03 +00:00
/* make sure eventual UDMA mode from the BIOS is disabled */
Try better this time at a patch that properly backspeeds on the Acer chips.
2001-01-07 17:00:09 +00:00
pci_write_config(parent, 0x56, pci_read_config(parent, 0x56, 2) &
~(0x0008 << (devno << 2)), 2);
Proberly back down DMA modes on the Acer Aladdin.
2000-12-19 10:37:03 +00:00
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (wdmamode >= 2 && apiomode >= 4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Unbreak ATAPI on the Aladdin chipset, only DMA access worked. Try to use a 32bit mask on the IO addresses, this fixes the alpha and hopefully doesn't break on any i386 machines. Try to enable both read & write cache on disks, they should be as default, but better be sure..
1999-11-08 21:36:00 +00:00
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting WDMA2 on Acer chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success");
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (!error) {
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
pci_write_config(parent, 0x53,
pci_read_config(parent, 0x53, 1) | 0x03, 1);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
atadev->channel->flags |= ATA_ATAPI_DMA_RO;
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
}
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
}
Fix the CD driver so that the last blocks can be read even if a blocksize != 2k is used. Update the timeout code to try fallback to PIO if problems arise in DMA mode.
2000-03-05 16:52:26 +00:00
pci_write_config(parent, 0x53,
(pci_read_config(parent, 0x53, 1) & ~0x01) | 0x02, 1);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_PIO0 + apiomode,
ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting PIO%d on Acer chip\n",
(error) ? "failed" : "success",
(apiomode >= 0) ? apiomode : 0);
if (!error) {
int32_t word54 = pci_read_config(parent, 0x54, 4);
int32_t timing;
switch(ATA_PIO0 + apiomode) {
case ATA_PIO0: timing = 0x006d0003;
case ATA_PIO1: timing = 0x00580002;
case ATA_PIO2: timing = 0x00440001;
case ATA_PIO3: timing = 0x00330001;
case ATA_PIO4: timing = 0x00310001;
default: timing = 0x006d0003;
}
pci_write_config(parent, 0x58 + (channel << 2), timing, 4);
word54 &= ~(0x000f000f << (devno << 2));
word54 |= (0x00000004 << (devno << 2));
pci_write_config(parent, 0x54, word54, 4);
atadev->mode = ATA_PIO0 + apiomode;
return;
}
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
break;
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
case 0x01bc10de: /* nVIDIA nForce */
Add support for the nVidia nForce2 ATA part. Fix support for the nForce1 as well, registers are offset 0x10 against the AMD/VIA parts.
2003-01-08 16:51:41 +00:00
case 0x006510de: /* nVIDIA nForce2 */
Add AMD 768 support.
2002-03-24 12:44:23 +00:00
case 0x74411022: /* AMD 768 */
Add support for the AMD 766 southbridge incl ATA100 support Fix ATA66 mode for the AMD756, the timing was way to slow
2001-05-14 18:38:22 +00:00
case 0x74111022: /* AMD 766 */
Add support for ATA66 on newer revisions of the VIA 82C596
2000-04-09 15:08:27 +00:00
case 0x74091022: /* AMD 756 */
Add support for the VIA 8235. Submitted by: Jason Dambrosio <jason@wiz.cx>
2002-09-18 09:39:37 +00:00
case 0x05711106: /* VIA 82C571, 82C586, 82C596, 82C686, 8231,8233,8235 */
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
{
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
int via_modes[5][7] = {
{ 0x00, 0x00, 0xc0, 0x00, 0x00, 0x00, 0x00 }, /* VIA ATA33 */
{ 0x00, 0x00, 0xea, 0x00, 0xe8, 0x00, 0x00 }, /* VIA ATA66 */
{ 0x00, 0x00, 0xf4, 0x00, 0xf1, 0xf0, 0x00 }, /* VIA ATA100 */
{ 0x00, 0x00, 0xf6, 0x00, 0xf2, 0xf1, 0xf0 }, /* VIA ATA133 */
Add support for the nVidia nForce2 ATA part. Fix support for the nForce1 as well, registers are offset 0x10 against the AMD/VIA parts.
2003-01-08 16:51:41 +00:00
{ 0x00, 0x00, 0xc0, 0x00, 0xc5, 0xc6, 0xc7 }}; /* AMD/nVIDIA */
int reg = 0x53 - devno;
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
int *reg_val = NULL;
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
char *chip = "VIA";
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
Add support for the VIA 8235. Submitted by: Jason Dambrosio <jason@wiz.cx>
2002-09-18 09:39:37 +00:00
if (ata_find_dev(parent, 0x31471106, 0) || /* 8233a */
ata_find_dev(parent, 0x31771106, 0)) { /* 8235 */
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
udmamode = imin(udmamode, 6);
reg_val = via_modes[3];
Add first cut support for the ATA100 capable VIA82c686b.
2001-01-17 09:22:13 +00:00
}
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
else if (ata_find_dev(parent, 0x06861106, 0x40) || /* 82C686b */
ata_find_dev(parent, 0x82311106, 0) || /* 8231 */
ata_find_dev(parent, 0x30741106, 0) || /* 8233 */
ata_find_dev(parent, 0x31091106, 0)) { /* 8233c */
udmamode = imin(udmamode, 5);
reg_val = via_modes[2];
}
else if (ata_find_dev(parent, 0x06861106, 0x10) || /* 82C686a */
ata_find_dev(parent, 0x05961106, 0x12)) { /* 82C596b */
udmamode = imin(udmamode, 4);
reg_val = via_modes[1];
}
Correct the Northbridge test on the new ATA133 VIA's Misc cosmetics now I'm there.
2002-04-02 16:45:06 +00:00
else if (ata_find_dev(parent, 0x06861106, 0)) { /* 82C686 */
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
udmamode = imin(udmamode, 2);
reg_val = via_modes[1];
}
else if (ata_find_dev(parent, 0x05961106, 0) || /* 82C596a */
ata_find_dev(parent, 0x05861106, 0x03)) { /* 82C586b */
udmamode = imin(udmamode, 2);
reg_val = via_modes[0];
}
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
else if (chiptype == 0x74411022 || /* AMD 768 */
chiptype == 0x74111022) { /* AMD 766 */
udmamode = imin(udmamode, 5);
reg_val = via_modes[4];
chip = "AMD";
}
else if (chiptype == 0x74091022) { /* AMD 756 */
udmamode = imin(udmamode, 4);
reg_val = via_modes[4];
chip = "AMD";
}
Add support for the nVidia nForce2 ATA part. Fix support for the nForce1 as well, registers are offset 0x10 against the AMD/VIA parts.
2003-01-08 16:51:41 +00:00
else if (chiptype == 0x006510de) { /* nForce2 */
udmamode = imin(udmamode, 6);
reg += 0x10;
reg_val = via_modes[4];
chip = "nVidia";
}
else if (chiptype == 0x01bc10de) { /* nForce */
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
udmamode = imin(udmamode, 5);
Add support for the nVidia nForce2 ATA part. Fix support for the nForce1 as well, registers are offset 0x10 against the AMD/VIA parts.
2003-01-08 16:51:41 +00:00
reg += 0x10;
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
reg_val = via_modes[4];
Add support for the nVidia nForce2 ATA part. Fix support for the nForce1 as well, registers are offset 0x10 against the AMD/VIA parts.
2003-01-08 16:51:41 +00:00
chip = "nVidia";
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
}
else
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
udmamode = 0;
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
Add support for the nVidia nForce2 ATA part. Fix support for the nForce1 as well, registers are offset 0x10 against the AMD/VIA parts.
2003-01-08 16:51:41 +00:00
if (udmamode || wdmamode)
pci_write_config(parent, reg - 0x08, 0x20, 1);
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
if (udmamode >= 6) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
ATA_UDMA6, ATA_C_F_SETXFER, ATA_WAIT_READY);
Add first cut support for the ATA100 capable VIA82c686b.
2001-01-17 09:22:13 +00:00
if (bootverbose)
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
ata_prtdev(atadev, "%s setting UDMA6 on %s chip\n",
(error) ? "failed" : "success", chip);
Add first cut support for the ATA100 capable VIA82c686b.
2001-01-17 09:22:13 +00:00
if (!error) {
Add support for the nVidia nForce2 ATA part. Fix support for the nForce1 as well, registers are offset 0x10 against the AMD/VIA parts.
2003-01-08 16:51:41 +00:00
pci_write_config(parent, reg, reg_val[6], 1);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA6);
Add first cut support for the ATA100 capable VIA82c686b.
2001-01-17 09:22:13 +00:00
return;
}
}
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
if (udmamode >= 5) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
ATA_UDMA5, ATA_C_F_SETXFER, ATA_WAIT_READY);
Add first cut support for the ATA100 capable VIA82c686b.
2001-01-17 09:22:13 +00:00
if (bootverbose)
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
ata_prtdev(atadev, "%s setting UDMA5 on %s chip\n",
(error) ? "failed" : "success", chip);
Add first cut support for the ATA100 capable VIA82c686b.
2001-01-17 09:22:13 +00:00
if (!error) {
Add support for the nVidia nForce2 ATA part. Fix support for the nForce1 as well, registers are offset 0x10 against the AMD/VIA parts.
2003-01-08 16:51:41 +00:00
pci_write_config(parent, reg, reg_val[5], 1);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA5);
Add first cut support for the ATA100 capable VIA82c686b.
2001-01-17 09:22:13 +00:00
return;
}
}
Rearrange the probe a bit, hopefully this will help to eleminate some of the fake devices sometimes seen on single device ATA channels. Proberly fail on failures in ata-disk.c, retry instead of hang. Cleanup the VIA probe/init code a bit. Fix a couble of missing free's in atapi-cd.c in the changer code.
2000-08-16 07:09:04 +00:00
if (udmamode >= 4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Rearrange the probe a bit, hopefully this will help to eleminate some of the fake devices sometimes seen on single device ATA channels. Proberly fail on failures in ata-disk.c, retry instead of hang. Cleanup the VIA probe/init code a bit. Fix a couble of missing free's in atapi-cd.c in the changer code.
2000-08-16 07:09:04 +00:00
ATA_UDMA4, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
ata_prtdev(atadev, "%s setting UDMA4 on %s chip\n",
(error) ? "failed" : "success", chip);
Rearrange the probe a bit, hopefully this will help to eleminate some of the fake devices sometimes seen on single device ATA channels. Proberly fail on failures in ata-disk.c, retry instead of hang. Cleanup the VIA probe/init code a bit. Fix a couble of missing free's in atapi-cd.c in the changer code.
2000-08-16 07:09:04 +00:00
if (!error) {
Add support for the nVidia nForce2 ATA part. Fix support for the nForce1 as well, registers are offset 0x10 against the AMD/VIA parts.
2003-01-08 16:51:41 +00:00
pci_write_config(parent, reg, reg_val[4], 1);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA4);
Rearrange the probe a bit, hopefully this will help to eleminate some of the fake devices sometimes seen on single device ATA channels. Proberly fail on failures in ata-disk.c, retry instead of hang. Cleanup the VIA probe/init code a bit. Fix a couble of missing free's in atapi-cd.c in the changer code.
2000-08-16 07:09:04 +00:00
return;
}
Add support for the Intel 810 chipset (ICH type of chips) Prober support for the VIA 82C686, I finally got the right datasheet. Get rid of atapi_wait, merge it into ata_wait. Avoid a couple of races by using asleep instead of tsleep. Always use 16bit transfers on ISA systems. Clear up the atapi_read/write functions.
2000-01-17 02:04:19 +00:00
}
Rearrange the probe a bit, hopefully this will help to eleminate some of the fake devices sometimes seen on single device ATA channels. Proberly fail on failures in ata-disk.c, retry instead of hang. Cleanup the VIA probe/init code a bit. Fix a couble of missing free's in atapi-cd.c in the changer code.
2000-08-16 07:09:04 +00:00
if (udmamode >= 2) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Rearrange the probe a bit, hopefully this will help to eleminate some of the fake devices sometimes seen on single device ATA channels. Proberly fail on failures in ata-disk.c, retry instead of hang. Cleanup the VIA probe/init code a bit. Fix a couble of missing free's in atapi-cd.c in the changer code.
2000-08-16 07:09:04 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
ata_prtdev(atadev, "%s setting UDMA2 on %s chip\n",
(error) ? "failed" : "success", chip);
Rearrange the probe a bit, hopefully this will help to eleminate some of the fake devices sometimes seen on single device ATA channels. Proberly fail on failures in ata-disk.c, retry instead of hang. Cleanup the VIA probe/init code a bit. Fix a couble of missing free's in atapi-cd.c in the changer code.
2000-08-16 07:09:04 +00:00
if (!error) {
Add support for the nVidia nForce2 ATA part. Fix support for the nForce1 as well, registers are offset 0x10 against the AMD/VIA parts.
2003-01-08 16:51:41 +00:00
pci_write_config(parent, reg, reg_val[2], 1);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
Rearrange the probe a bit, hopefully this will help to eleminate some of the fake devices sometimes seen on single device ATA channels. Proberly fail on failures in ata-disk.c, retry instead of hang. Cleanup the VIA probe/init code a bit. Fix a couble of missing free's in atapi-cd.c in the changer code.
2000-08-16 07:09:04 +00:00
return;
}
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
}
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
if (wdmamode >= 2 && apiomode >= 4) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting WDMA2 on %s chip\n",
(error) ? "failed" : "success", chip);
if (!error) {
Add support for the nVidia nForce2 ATA part. Fix support for the nForce1 as well, registers are offset 0x10 against the AMD/VIA parts.
2003-01-08 16:51:41 +00:00
pci_write_config(parent, reg, 0x0b, 1);
pci_write_config(parent, reg - 0x08, 0x31, 1);
Add support for the nVIDIA nForce ATA controller. Collapse the VIA/AMD/nVIDIA support code into one, they are created more or less equal anyway..
2002-04-16 08:30:51 +00:00
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
return;
}
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
}
}
/* we could set PIO mode timings, but we assume the BIOS did that */
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
break;
Add DMA support for the SiS 5591 and old PIIX chipsets. Add a missing DELAY(1) in ata_wait. Change the info from ad_version, so the ATA version from the disk can be used to quantify the DAM modes valid for this drive, ie be more selective with turning DMA on on older disks that should not support it.. Fix the probe for BIOS enabled DMA in the generic case, master/slave was reversed in the test. Check the return for ata_command in all cases, and print warnings if it fails. Call ata_dmainit with all dmamodes off when falling back to PIO mode, that should take care of both the Promise & HPT366 controllers not being able to handle the fallback... Cleanup the printf's in the drivers, use the prober device name (if possible) instead of ataN-master/slave.
1999-12-13 10:19:39 +00:00
case 0x55131039: /* SiS 5591 */
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
if (ata_find_dev(parent, 0x06301039, 0x30) || /* SiS 630 */
Correct the Northbridge test on the new ATA133 VIA's Misc cosmetics now I'm there.
2002-04-02 16:45:06 +00:00
ata_find_dev(parent, 0x06331039, 0) || /* SiS 633 */
ata_find_dev(parent, 0x06351039, 0) || /* SiS 635 */
ata_find_dev(parent, 0x06401039, 0) || /* SiS 640 */
ata_find_dev(parent, 0x06451039, 0) || /* SiS 645 */
ata_find_dev(parent, 0x06501039, 0) || /* SiS 650 */
ata_find_dev(parent, 0x07301039, 0) || /* SiS 730 */
ata_find_dev(parent, 0x07331039, 0) || /* SiS 733 */
ata_find_dev(parent, 0x07351039, 0) || /* SiS 735 */
ata_find_dev(parent, 0x07401039, 0) || /* SiS 740 */
ata_find_dev(parent, 0x07451039, 0) || /* SiS 745 */
ata_find_dev(parent, 0x07501039, 0)) { /* SiS 750 */
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
int8_t reg = 0x40 + (devno << 1);
Adjust the timings for the SiS chips a bit, also add the SiS 645. Cosmetics on the Acer chips (print right modes)
2001-12-14 21:28:49 +00:00
int16_t val = pci_read_config(parent, reg, 2) & 0x0fff;
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
if (udmamode >= 5) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
ATA_UDMA5, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA5 on SiS chip\n",
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
(error) ? "failed" : "success");
if (!error) {
Adjust the timings for the SiS chips a bit, also add the SiS 645. Cosmetics on the Acer chips (print right modes)
2001-12-14 21:28:49 +00:00
pci_write_config(parent, reg, val | 0x8000, 2);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA5);
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
return;
}
}
if (udmamode >= 4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
ATA_UDMA4, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA4 on SiS chip\n",
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
(error) ? "failed" : "success");
if (!error) {
Adjust the timings for the SiS chips a bit, also add the SiS 645. Cosmetics on the Acer chips (print right modes)
2001-12-14 21:28:49 +00:00
pci_write_config(parent, reg, val | 0x9000, 2);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA4);
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
return;
}
}
if (udmamode >= 2) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA2 on SiS chip\n",
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
(error) ? "failed" : "success");
if (!error) {
Adjust the timings for the SiS chips a bit, also add the SiS 645. Cosmetics on the Acer chips (print right modes)
2001-12-14 21:28:49 +00:00
pci_write_config(parent, reg, val | 0xb000, 2);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
return;
}
}
} else if (ata_find_dev(parent, 0x05301039, 0) || /* SiS 530 */
ata_find_dev(parent, 0x05401039, 0) || /* SiS 540 */
ata_find_dev(parent, 0x06201039, 0) || /* SiS 620 */
ata_find_dev(parent, 0x06301039, 0)) { /* SiS 630 */
int8_t reg = 0x40 + (devno << 1);
int16_t val = pci_read_config(parent, reg, 2) & 0x0fff;
if (udmamode >= 4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
ATA_UDMA4, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA4 on SiS chip\n",
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
(error) ? "failed" : "success");
if (!error) {
pci_write_config(parent, reg, val | 0x9000, 2);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA4);
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
return;
}
}
if (udmamode >= 2) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA2 on SiS chip\n",
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
(error) ? "failed" : "success");
if (!error) {
Adjust the timings for the SiS chips a bit, also add the SiS 645. Cosmetics on the Acer chips (print right modes)
2001-12-14 21:28:49 +00:00
pci_write_config(parent, reg, val | 0xa000, 2);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
Initial support for the newer SiS chipsets, based on docs we finally got from SiS. This should also close PR 32421 which has patches which seem to set the timing registers wrongly according to SiS...
2001-12-02 10:48:52 +00:00
return;
}
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
} else if (udmamode >= 2 && chiprev > 0xc1) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for even more SiS chipsets. Misc cosmetics.
2001-12-25 14:44:26 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA2 on SiS chip\n",
Add support for even more SiS chipsets. Misc cosmetics.
2001-12-25 14:44:26 +00:00
(error) ? "failed" : "success");
if (!error) {
pci_write_config(parent, 0x40 + (devno << 1), 0xa301, 2);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
Add support for even more SiS chipsets. Misc cosmetics.
2001-12-25 14:44:26 +00:00
return;
Add DMA support for the SiS 5591 and old PIIX chipsets. Add a missing DELAY(1) in ata_wait. Change the info from ad_version, so the ATA version from the disk can be used to quantify the DAM modes valid for this drive, ie be more selective with turning DMA on on older disks that should not support it.. Fix the probe for BIOS enabled DMA in the generic case, master/slave was reversed in the test. Check the return for ata_command in all cases, and print warnings if it fails. Call ata_dmainit with all dmamodes off when falling back to PIO mode, that should take care of both the Promise & HPT366 controllers not being able to handle the fallback... Cleanup the printf's in the drivers, use the prober device name (if possible) instead of ataN-master/slave.
1999-12-13 10:19:39 +00:00
}
}
if (wdmamode >=2 && apiomode >= 4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add DMA support for the SiS 5591 and old PIIX chipsets. Add a missing DELAY(1) in ata_wait. Change the info from ad_version, so the ATA version from the disk can be used to quantify the DAM modes valid for this drive, ie be more selective with turning DMA on on older disks that should not support it.. Fix the probe for BIOS enabled DMA in the generic case, master/slave was reversed in the test. Check the return for ata_command in all cases, and print warnings if it fails. Call ata_dmainit with all dmamodes off when falling back to PIO mode, that should take care of both the Promise & HPT366 controllers not being able to handle the fallback... Cleanup the printf's in the drivers, use the prober device name (if possible) instead of ataN-master/slave.
1999-12-13 10:19:39 +00:00
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting WDMA2 on SiS chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success");
Add DMA support for the SiS 5591 and old PIIX chipsets. Add a missing DELAY(1) in ata_wait. Change the info from ad_version, so the ATA version from the disk can be used to quantify the DAM modes valid for this drive, ie be more selective with turning DMA on on older disks that should not support it.. Fix the probe for BIOS enabled DMA in the generic case, master/slave was reversed in the test. Check the return for ata_command in all cases, and print warnings if it fails. Call ata_dmainit with all dmamodes off when falling back to PIO mode, that should take care of both the Promise & HPT366 controllers not being able to handle the fallback... Cleanup the printf's in the drivers, use the prober device name (if possible) instead of ataN-master/slave.
1999-12-13 10:19:39 +00:00
if (!error) {
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
pci_write_config(parent, 0x40 + (devno << 1), 0x0301, 2);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
return;
}
}
/* we could set PIO mode timings, but we assume the BIOS did that */
break;
Add preliminary mostly untested support for the Silicon Image Sil680 chip.
2002-09-12 15:25:59 +00:00
case 0x06801095: /* Sil 0680 ATA133 controller */
{
u_int8_t ureg = 0xac + (device * 0x02) + (channel * 0x10);
u_int8_t uval = pci_read_config(parent, ureg, 1);
u_int8_t mreg = channel ? 0x84 : 0x80;
u_int8_t mask = device ? 0x30 : 0x03;
u_int8_t mode = pci_read_config(parent, mreg, 1);
/* enable UDMA mode */
pci_write_config(parent, mreg,
(mode & ~mask) | (device ? 0x30 : 0x03), 1);
if (udmamode >= 6) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_UDMA6, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting UDMA6 on Sil chip\n",
(error) ? "failed" : "success");
if (!error) {
pci_write_config(parent, ureg, (uval & 0x3f) | 0x01, 1);
ata_dmacreate(atadev, apiomode, ATA_UDMA6);
return;
}
}
if (udmamode >= 5) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_UDMA5, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting UDMA5 on Sil chip\n",
(error) ? "failed" : "success");
if (!error) {
pci_write_config(parent, ureg, (uval & 0x3f) | 0x02, 1);
ata_dmacreate(atadev, apiomode, ATA_UDMA5);
return;
}
}
if (udmamode >= 4) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_UDMA4, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting UDMA4 on Sil chip\n",
(error) ? "failed" : "success");
if (!error) {
pci_write_config(parent, ureg, (uval & 0x3f) | 0x03, 1);
ata_dmacreate(atadev, apiomode, ATA_UDMA4);
return;
}
}
if (udmamode >= 2) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting UDMA2 on Sil chip\n",
(error) ? "failed" : "success");
if (!error) {
pci_write_config(parent, ureg, (uval & 0x3f) | 0x07, 1);
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
return;
}
}
/* disable UDMA mode and enable WDMA mode */
pci_write_config(parent, mreg,
(mode & ~mask) | (device ? 0x20 : 0x02), 1);
if (wdmamode >= 2 && apiomode >= 4) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting WDMA2 on Sil chip\n",
(error) ? "failed" : "success");
if (!error) {
pci_write_config(parent, ureg - 0x4, 0x10c1, 2);
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
return;
}
}
/* restore PIO mode */
pci_write_config(parent, mreg, mode, 1);
}
/* we could set PIO mode timings, but we assume the BIOS did that */
break;
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
case 0x06491095: /* CMD 649 ATA100 controller */
if (udmamode >= 5) {
u_int8_t umode;
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
ATA_UDMA5, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA5 on CMD chip\n",
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
(error) ? "failed" : "success");
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
umode = pci_read_config(parent, channel ? 0x7b : 0x73, 1);
umode &= ~(device ? 0xca : 0x35);
umode |= (device ? 0x0a : 0x05);
pci_write_config(parent, channel ? 0x7b : 0x73, umode, 1);
ata_dmacreate(atadev, apiomode, ATA_UDMA5);
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
return;
}
}
/* FALLTHROUGH */
case 0x06481095: /* CMD 648 ATA66 controller */
if (udmamode >= 4) {
u_int8_t umode;
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
ATA_UDMA4, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA4 on CMD chip\n",
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
(error) ? "failed" : "success");
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
umode = pci_read_config(parent, channel ? 0x7b : 0x73, 1);
umode &= ~(device ? 0xca : 0x35);
umode |= (device ? 0x4a : 0x15);
pci_write_config(parent, channel ? 0x7b : 0x73, umode, 1);
ata_dmacreate(atadev, apiomode, ATA_UDMA4);
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
return;
}
}
if (udmamode >= 2) {
u_int8_t umode;
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA2 on CMD chip\n",
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
(error) ? "failed" : "success");
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
umode = pci_read_config(parent, channel ? 0x7b : 0x73, 1);
umode &= ~(device ? 0xca : 0x35);
umode |= (device ? 0x42 : 0x11);
pci_write_config(parent, channel ? 0x7b : 0x73, umode, 1);
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
return;
}
}
/* make sure eventual UDMA mode from the BIOS is disabled */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
pci_write_config(parent, channel ? 0x7b : 0x73,
pci_read_config(parent, channel ? 0x7b : 0x73, 1) &
Fix the breakage of tagged queueing that the busdma integration introduced. Since its now only possible to have one DMA control block at a time, we move the setup to dmastart instead.
2002-04-18 19:11:45 +00:00
~(device ? 0xca : 0x53), 1);
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
/* FALLTHROUGH */
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
case 0x06461095: /* CMD 646 ATA controller */
if (wdmamode >= 2 && apiomode >= 4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting WDMA2 on CMD chip\n",
Update the ata driver to take more advantage of newbus, this was needed to make attach/detach of devices work, which is needed for the PCCARD support. (PCCARD support is still not working though, more to come on that) Support the CMD646 chip which is used on many alphas, sadly only in WDMA2 mode, as the silicon is broken beyond belief for UDMA modes. Lots of cosmetic fixes here and there. Sorry for the size of this megapatchfromhell but it was not possible otherwise... newbus patches based on work from: dfr (Doug Rabson)
2000-02-18 20:57:33 +00:00
error ? "failed" : "success");
if (!error) {
int32_t offset = (devno < 3) ? (devno << 1) : 7;
pci_write_config(parent, 0x54 + offset, 0x3f, 1);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add DMA support for the SiS 5591 and old PIIX chipsets. Add a missing DELAY(1) in ata_wait. Change the info from ad_version, so the ATA version from the disk can be used to quantify the DAM modes valid for this drive, ie be more selective with turning DMA on on older disks that should not support it.. Fix the probe for BIOS enabled DMA in the generic case, master/slave was reversed in the test. Check the return for ata_command in all cases, and print warnings if it fails. Call ata_dmainit with all dmamodes off when falling back to PIO mode, that should take care of both the Promise & HPT366 controllers not being able to handle the fallback... Cleanup the printf's in the drivers, use the prober device name (if possible) instead of ataN-master/slave.
1999-12-13 10:19:39 +00:00
}
}
/* we could set PIO mode timings, but we assume the BIOS did that */
break;
Add (partial) support for the Cypress ATA controller. Call intr_teardown on detach. Always add non masterdevice from unit 2 upwards. Update to the pccard code, at least some cards are now working, more testing to follow.
2000-02-25 09:48:23 +00:00
case 0xc6931080: /* Cypress 82c693 ATA controller */
if (wdmamode >= 2 && apiomode >= 4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add (partial) support for the Cypress ATA controller. Call intr_teardown on detach. Always add non masterdevice from unit 2 upwards. Update to the pccard code, at least some cards are now working, more testing to follow.
2000-02-25 09:48:23 +00:00
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting WDMA2 on Cypress chip\n",
Add (partial) support for the Cypress ATA controller. Call intr_teardown on detach. Always add non masterdevice from unit 2 upwards. Update to the pccard code, at least some cards are now working, more testing to follow.
2000-02-25 09:48:23 +00:00
error ? "failed" : "success");
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
pci_write_config(atadev->channel->dev,
channel ? 0x4e:0x4c, 0x2020, 2);
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
Add (partial) support for the Cypress ATA controller. Call intr_teardown on detach. Always add non masterdevice from unit 2 upwards. Update to the pccard code, at least some cards are now working, more testing to follow.
2000-02-25 09:48:23 +00:00
return;
}
}
/* we could set PIO mode timings, but we assume the BIOS did that */
break;
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
case 0x01021078: /* Cyrix 5530 ATA33 controller */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
atadev->channel->alignment = 0xf;
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
if (udmamode >= 2) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA2 on Cyrix chip\n",
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
(error) ? "failed" : "success");
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
cyrix_timing(atadev, devno, ATA_UDMA2);
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
return;
}
}
if (wdmamode >= 2 && apiomode >= 4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting WDMA2 on Cyrix chip\n",
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
(error) ? "failed" : "success");
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
cyrix_timing(atadev, devno, ATA_WDMA2);
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
return;
}
}
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Update with latest ATA/ATAPI ver 6 rev 2 items.
2001-10-06 11:07:04 +00:00
ATA_PIO0 + apiomode, ATA_C_F_SETXFER,
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting %s on Cyrix chip\n",
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
(error) ? "failed" : "success",
Update with latest ATA/ATAPI ver 6 rev 2 items.
2001-10-06 11:07:04 +00:00
ata_mode2str(ATA_PIO0 + apiomode));
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
cyrix_timing(atadev, devno, ATA_PIO0 + apiomode);
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
atadev->mode = ATA_PIO0 + apiomode;
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
return;
Add support for the ServerWorks CSB6. The support for the 3'rd channel is only experimental.
2003-01-19 13:03:20 +00:00
case 0x02131166: /* ServerWorks CSB6 ATA 100 controller (chan 0+1) */
Add support for the ServerWorks CSB5 chips
2002-03-18 12:13:13 +00:00
case 0x02121166: /* ServerWorks CSB5 ATA66/100 controller */
Add support for the ServerWorks CSB6. The support for the 3'rd channel is only experimental.
2003-01-19 13:03:20 +00:00
if (udmamode >= 5 && (chiptype == 0x02131166 ||
(chiptype == 0x02121166 && chiprev >= 0x92))) {
Add support for the ServerWorks CSB5 chips
2002-03-18 12:13:13 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_UDMA5, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting UDMA5 on ServerWorks chip\n",
(error) ? "failed" : "success");
if (!error) {
u_int16_t reg56;
pci_write_config(parent, 0x54,
pci_read_config(parent, 0x54, 1) |
(0x01 << devno), 1);
reg56 = pci_read_config(parent, 0x56, 2);
reg56 &= ~(0xf << (devno * 4));
reg56 |= (0x5 << (devno * 4));
pci_write_config(parent, 0x56, reg56, 2);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA5);
Add support for the ServerWorks CSB5 chips
2002-03-18 12:13:13 +00:00
return;
}
}
Add support for the ServerWorks CSB6. The support for the 3'rd channel is only experimental.
2003-01-19 13:03:20 +00:00
/* FALLTHROUGH */
case 0x02171166: /* ServerWorks CSB6 ATA 66 controller (chan 2) */
Add support for the ServerWorks CSB5 chips
2002-03-18 12:13:13 +00:00
if (udmamode >= 4) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_UDMA4, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting UDMA4 on ServerWorks chip\n",
(error) ? "failed" : "success");
if (!error) {
u_int16_t reg56;
pci_write_config(parent, 0x54,
pci_read_config(parent, 0x54, 1) |
(0x01 << devno), 1);
reg56 = pci_read_config(parent, 0x56, 2);
reg56 &= ~(0xf << (devno * 4));
reg56 |= (0x4 << (devno * 4));
pci_write_config(parent, 0x56, reg56, 2);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA4);
Add support for the ServerWorks CSB5 chips
2002-03-18 12:13:13 +00:00
return;
}
}
/* FALLTHROUGH */
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
case 0x02111166: /* ServerWorks ROSB4 ATA33 controller */
if (udmamode >= 2) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA2 on ServerWorks chip\n",
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
(error) ? "failed" : "success");
if (!error) {
Proberly back down DMA modes on the Acer Aladdin.
2000-12-19 10:37:03 +00:00
u_int16_t reg56;
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
pci_write_config(parent, 0x54,
pci_read_config(parent, 0x54, 1) |
(0x01 << devno), 1);
reg56 = pci_read_config(parent, 0x56, 2);
reg56 &= ~(0xf << (devno * 4));
reg56 |= (0x2 << (devno * 4));
pci_write_config(parent, 0x56, reg56, 2);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
return;
}
}
if (wdmamode >= 2 && apiomode >= 4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting WDMA2 on ServerWorks chip\n",
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
(error) ? "failed" : "success");
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
int offset = devno ^ 0x01;
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
int word44 = pci_read_config(parent, 0x44, 4);
pci_write_config(parent, 0x54,
pci_read_config(parent, 0x54, 1) &
~(0x01 << devno), 1);
word44 &= ~(0xff << (offset << 8));
word44 |= (0x20 << (offset << 8));
pci_write_config(parent, 0x44, 0x20, 4);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
Add support for ServerWorks ROSB4 ATA33 chipset. Add support for CMD 648 ATA66 & CMD 649 ATA100 chipsets. Fix the "resource already allocated" panic with the CMD and other braindead controllers. Add options ATA_ENABLE_TAGS, without this option tagged queuing will not be attempted.
2000-10-03 13:12:36 +00:00
return;
}
}
/* we could set PIO mode timings, but we assume the BIOS did that */
break;
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
case 0x4d69105a: /* Promise TX2 ATA133 controllers */
Add yet another chip ID for a Promise TX2 chip.
2002-04-11 11:04:23 +00:00
case 0x6269105a: /* Promise TX2 ATA133 controllers */
Fix the 48bit access support for the older Promise 66/100 controllers, the first attempt was wrong and could cause r/w timeouts. Add yet another Promise PCI id.
2003-01-19 11:47:32 +00:00
case 0x1275105a: /* Promise TX2 ATA133 controllers */
case 0x5275105a: /* Promise TX2 ATA133 controllers */
Add yet another Promise PCI id.
2002-10-01 15:21:09 +00:00
case 0x7275105a: /* Promise TX2 ATA133 controllers */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ATA_OUTB(atadev->channel->r_bmio, ATA_BMDEVSPEC_0, 0x0b);
if (udmamode >= 6 &&
!(ATA_INB(atadev->channel->r_bmio, ATA_BMDEVSPEC_1) & 0x04)) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
ATA_UDMA6, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA6 on Promise chip\n",
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
(error) ? "failed" : "success");
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA6);
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
return;
}
}
/* FALLTHROUGH */
Add support for yet another Promise ATA100 variant Minor fix to the VIA setup code.
2001-06-08 09:51:33 +00:00
case 0x4d68105a: /* Promise TX2 ATA100 controllers */
Add support for even more SiS chipsets. Misc cosmetics.
2001-12-25 14:44:26 +00:00
case 0x6268105a: /* Promise TX2 ATA100 controllers */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ATA_OUTB(atadev->channel->r_bmio, ATA_BMDEVSPEC_0, 0x0b);
if (udmamode >= 5 &&
!(ATA_INB(atadev->channel->r_bmio, ATA_BMDEVSPEC_1) & 0x04)) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
ATA_UDMA5, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA5 on Promise chip\n",
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
(error) ? "failed" : "success");
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA5);
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
return;
}
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ATA_OUTB(atadev->channel->r_bmio, ATA_BMDEVSPEC_0, 0x0b);
if (udmamode >= 4 &&
!(ATA_INB(atadev->channel->r_bmio, ATA_BMDEVSPEC_1) & 0x04)) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
ATA_UDMA4, ATA_C_F_SETXFER, ATA_WAIT_READY);
Proberly support the Promise TX2/4. Programming info kindly provided by Daniela Engert <dani@ngrt.de>
2001-10-03 08:27:29 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA4 on Promise chip\n",
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
(error) ? "failed" : "success");
Proberly support the Promise TX2/4. Programming info kindly provided by Daniela Engert <dani@ngrt.de>
2001-10-03 08:27:29 +00:00
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA4);
Proberly support the Promise TX2/4. Programming info kindly provided by Daniela Engert <dani@ngrt.de>
2001-10-03 08:27:29 +00:00
return;
}
}
if (udmamode >= 2) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Proberly support the Promise TX2/4. Programming info kindly provided by Daniela Engert <dani@ngrt.de>
2001-10-03 08:27:29 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA on Promise chip\n",
(error) ? "failed" : "success");
Proberly support the Promise TX2/4. Programming info kindly provided by Daniela Engert <dani@ngrt.de>
2001-10-03 08:27:29 +00:00
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
Proberly support the Promise TX2/4. Programming info kindly provided by Daniela Engert <dani@ngrt.de>
2001-10-03 08:27:29 +00:00
return;
}
}
if (wdmamode >= 2 && apiomode >= 4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Proberly support the Promise TX2/4. Programming info kindly provided by Daniela Engert <dani@ngrt.de>
2001-10-03 08:27:29 +00:00
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting WDMA2 on Promise chip\n",
(error) ? "failed" : "success");
Proberly support the Promise TX2/4. Programming info kindly provided by Daniela Engert <dani@ngrt.de>
2001-10-03 08:27:29 +00:00
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
Proberly support the Promise TX2/4. Programming info kindly provided by Daniela Engert <dani@ngrt.de>
2001-10-03 08:27:29 +00:00
return;
}
}
break;
case 0x0d30105a: /* Promise OEM ATA100 controllers */
Add yet another (older) Promise chip
2002-06-19 12:26:20 +00:00
case 0x4d30105a: /* Promise Ultra/FastTrak 100 controllers */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (!ATAPI_DEVICE(atadev) && udmamode >= 5 &&
!(pci_read_config(parent, 0x50, 2) & (channel ? 1<<11 : 1<<10))) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
ATA_UDMA5, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA5 on Promise chip\n",
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
(error) ? "failed" : "success");
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
promise_timing(atadev, devno, ATA_UDMA5);
ata_dmacreate(atadev, apiomode, ATA_UDMA5);
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
return;
}
}
Finally commit some of the minor things I've collected over the last month(s): Add tagged queueing support for new IBM drives. Add support for Yet Another Promise ATA 100 chip. Flush disk cache on close. Dont flush the disk cache on BIO_ORDERED anymore. Cleanup the tests for DMA on ATAPI devices. Allow to share ALL irq's even the std irg 14 & 15. Fix calculation bug in end of media code on CD's. Add REZERO on opening a CDR/CDRW. Cleanup ataioctl a bit.
2001-08-21 11:35:47 +00:00
/* FALLTHROUGH */
Add yet another (older) Promise chip
2002-06-19 12:26:20 +00:00
case 0x0d38105a: /* Promise FastTrak 66 controllers */
Finally commit some of the minor things I've collected over the last month(s): Add tagged queueing support for new IBM drives. Add support for Yet Another Promise ATA 100 chip. Flush disk cache on close. Dont flush the disk cache on BIO_ORDERED anymore. Cleanup the tests for DMA on ATAPI devices. Allow to share ALL irq's even the std irg 14 & 15. Fix calculation bug in end of media code on CD's. Add REZERO on opening a CDR/CDRW. Cleanup ataioctl a bit.
2001-08-21 11:35:47 +00:00
case 0x4d38105a: /* Promise Ultra/FastTrak 66 controllers */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (!ATAPI_DEVICE(atadev) && udmamode >= 4 &&
!(pci_read_config(parent, 0x50, 2) & (channel ? 1<<11 : 1<<10))) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Unbreak ATAPI on the Aladdin chipset, only DMA access worked. Try to use a 32bit mask on the IO addresses, this fixes the alpha and hopefully doesn't break on any i386 machines. Try to enable both read & write cache on disks, they should be as default, but better be sure..
1999-11-08 21:36:00 +00:00
ATA_UDMA4, ATA_C_F_SETXFER, ATA_WAIT_READY);
Fix IRQ allocation bug on controllers using a shared interrupt. Fix a bug which could cause panics in ad/atapi-interrupt. Add support for UDMA66 on Promise Ultra/Fasttrak controllers. Get rid of ATA_IGNORE_INTR, and introduce ATA_WAIT_INTR instead. Add a delay in the dump routine in ata-disk.c, some controllers seem to need this. Also dont use the timeout watchdog when dumping. Disable DMA on ATAPI devices as default, add option ATA_ENABLE_ATAPI_DMA for those that has HW that works. Add support for some not-up-to-spec ATAPI devices that returns data together with completition status on data moving cmd's.
1999-11-06 16:50:21 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA4 on Promise chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success");
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
promise_timing(atadev, devno, ATA_UDMA4);
ata_dmacreate(atadev, apiomode, ATA_UDMA4);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
}
Fix IRQ allocation bug on controllers using a shared interrupt. Fix a bug which could cause panics in ad/atapi-interrupt. Add support for UDMA66 on Promise Ultra/Fasttrak controllers. Get rid of ATA_IGNORE_INTR, and introduce ATA_WAIT_INTR instead. Add a delay in the dump routine in ata-disk.c, some controllers seem to need this. Also dont use the timeout watchdog when dumping. Disable DMA on ATAPI devices as default, add option ATA_ENABLE_ATAPI_DMA for those that has HW that works. Add support for some not-up-to-spec ATAPI devices that returns data together with completition status on data moving cmd's.
1999-11-06 16:50:21 +00:00
}
Finally commit some of the minor things I've collected over the last month(s): Add tagged queueing support for new IBM drives. Add support for Yet Another Promise ATA 100 chip. Flush disk cache on close. Dont flush the disk cache on BIO_ORDERED anymore. Cleanup the tests for DMA on ATAPI devices. Allow to share ALL irq's even the std irg 14 & 15. Fix calculation bug in end of media code on CD's. Add REZERO on opening a CDR/CDRW. Cleanup ataioctl a bit.
2001-08-21 11:35:47 +00:00
/* FALLTHROUGH */
case 0x4d33105a: /* Promise Ultra/FastTrak 33 controllers */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (!ATAPI_DEVICE(atadev) && udmamode >= 2) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Unbreak ATAPI on the Aladdin chipset, only DMA access worked. Try to use a 32bit mask on the IO addresses, this fixes the alpha and hopefully doesn't break on any i386 machines. Try to enable both read & write cache on disks, they should be as default, but better be sure..
1999-11-08 21:36:00 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
Ten'th update to the new ATA/ATAPI driver: It been awhile since the last major update, as a benefit there are some cool things in this one (and new bugs probably :) )... The ATA driver has grown "real" timeout support for all devices. This means that it should be possible to get in contact with (especially) lost ATAPI devices. It also means that the ATA driver is now usable on notebooks as it will DTRT on resume. An experimental hack at utilizing the Promise66's at UDMA66 is in there, but I cant test it. If someone feels like sending me one, give me a ping. The ATAPI DMA enableling scheme has been changed, also better DMA support for the Aladdin chipset has been implemented for ATAPI devices. Note that the Aladdin apparently only can do DMA reads on ATAPI devices, and the Promise cant do ATAPI DMA at all. I have seen problems on some ATAPI devices that should be able to run in DMA mode, so if you encounter problems with hanging atapi devices during the probe, or during access, disable DMA in atapi-all.c, and let me know. It might be nessesary to do this via a "white list" for known good devices... The ATAPI CDROM driver can now use eject/close without hanging and the bug that caused reading beyond the end of a CD has been fixed. Media change is also handled proberly. DVD drives are identified and are usable as CDROM devices at least, I dont have the HW to test this further, see above :). The ATAPI tape driver has gotten some support for using the DSC method for not blocking the IDE channel during read/write when the device has full buffers. It knows about the OnStream DI-30 device, support is not completed yet, but it can function as a primitive backup medium, without filemarks, and without bad media handeling. This is because the OnStream device doesn't handle this (like everybody else) in HW. It also now supports getting/setting the record position on devices that supports it. Some rather major cleanups and rearrangements as well (cvs -b diff is your freind). I'm closing in on declaring this for beta code, most of the infrastruture is in place by now. As usual USE AT YOUR OWN RISK!!, this is still alpha level code. This driver can hose your disk real bad if anything goes wrong, but now you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-09-21 19:50:40 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA2 on Promise chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success");
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
promise_timing(atadev, devno, ATA_UDMA2);
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
}
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (!ATAPI_DEVICE(atadev) && wdmamode >= 2 && apiomode >= 4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Unbreak ATAPI on the Aladdin chipset, only DMA access worked. Try to use a 32bit mask on the IO addresses, this fixes the alpha and hopefully doesn't break on any i386 machines. Try to enable both read & write cache on disks, they should be as default, but better be sure..
1999-11-08 21:36:00 +00:00
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
Ten'th update to the new ATA/ATAPI driver: It been awhile since the last major update, as a benefit there are some cool things in this one (and new bugs probably :) )... The ATA driver has grown "real" timeout support for all devices. This means that it should be possible to get in contact with (especially) lost ATAPI devices. It also means that the ATA driver is now usable on notebooks as it will DTRT on resume. An experimental hack at utilizing the Promise66's at UDMA66 is in there, but I cant test it. If someone feels like sending me one, give me a ping. The ATAPI DMA enableling scheme has been changed, also better DMA support for the Aladdin chipset has been implemented for ATAPI devices. Note that the Aladdin apparently only can do DMA reads on ATAPI devices, and the Promise cant do ATAPI DMA at all. I have seen problems on some ATAPI devices that should be able to run in DMA mode, so if you encounter problems with hanging atapi devices during the probe, or during access, disable DMA in atapi-all.c, and let me know. It might be nessesary to do this via a "white list" for known good devices... The ATAPI CDROM driver can now use eject/close without hanging and the bug that caused reading beyond the end of a CD has been fixed. Media change is also handled proberly. DVD drives are identified and are usable as CDROM devices at least, I dont have the HW to test this further, see above :). The ATAPI tape driver has gotten some support for using the DSC method for not blocking the IDE channel during read/write when the device has full buffers. It knows about the OnStream DI-30 device, support is not completed yet, but it can function as a primitive backup medium, without filemarks, and without bad media handeling. This is because the OnStream device doesn't handle this (like everybody else) in HW. It also now supports getting/setting the record position on devices that supports it. Some rather major cleanups and rearrangements as well (cvs -b diff is your freind). I'm closing in on declaring this for beta code, most of the infrastruture is in place by now. As usual USE AT YOUR OWN RISK!!, this is still alpha level code. This driver can hose your disk real bad if anything goes wrong, but now you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-09-21 19:50:40 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting WDMA2 on Promise chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success");
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
promise_timing(atadev, devno, ATA_WDMA2);
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
}
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
}
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Update with latest ATA/ATAPI ver 6 rev 2 items.
2001-10-06 11:07:04 +00:00
ATA_PIO0 + apiomode,
Add support for VIA 82C596 controller Better shared irq handeling for Promise & HPT366 controllers Setup prober PIO mode timings on Promise & HPT366 controllers Update Copyright headers to be Y2K compliant :)
2000-01-03 10:26:59 +00:00
ATA_C_F_SETXFER, ATA_WAIT_READY);
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting PIO%d on Promise chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success",
(apiomode >= 0) ? apiomode : 0);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
promise_timing(atadev, devno, ATA_PIO0 + apiomode);
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
atadev->mode = ATA_PIO0 + apiomode;
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
Add experimental support for the HighPoint HPT372 ATA133 controller. Based on docs kindly provided by HighPoint which we thank for the close cooperation with the FreeBSD project!.
2001-10-23 09:25:57 +00:00
case 0x00041103: /* HighPoint HPT366/368/370/372 controllers */
Add support for the Highpoint HPT372 based cards (rocketraid 133). HW Sponsored by: Mike Tancsa
2002-02-18 11:57:56 +00:00
case 0x00051103: /* HighPoint HPT372 controllers */
Add support for the HighPoint HPT374 4 channel ATA chip. Sponsored by: Isilon Systems.
2002-02-11 15:48:04 +00:00
case 0x00081103: /* HighPoint HPT374 controllers */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (!ATAPI_DEVICE(atadev) && udmamode >= 6 && hpt_cable80(atadev) &&
((chiptype == 0x00041103 && chiprev >= 0x05) ||
(chiptype == 0x00051103 && chiprev >= 0x01) ||
(chiptype == 0x00081103 && chiprev >= 0x07))) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add experimental support for the HighPoint HPT372 ATA133 controller. Based on docs kindly provided by HighPoint which we thank for the close cooperation with the FreeBSD project!.
2001-10-23 09:25:57 +00:00
ATA_UDMA6, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA6 on HighPoint chip\n",
Add experimental support for the HighPoint HPT372 ATA133 controller. Based on docs kindly provided by HighPoint which we thank for the close cooperation with the FreeBSD project!.
2001-10-23 09:25:57 +00:00
(error) ? "failed" : "success");
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
hpt_timing(atadev, devno, ATA_UDMA6);
ata_dmacreate(atadev, apiomode, ATA_UDMA6);
Add experimental support for the HighPoint HPT372 ATA133 controller. Based on docs kindly provided by HighPoint which we thank for the close cooperation with the FreeBSD project!.
2001-10-23 09:25:57 +00:00
return;
}
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (!ATAPI_DEVICE(atadev) && udmamode >= 5 && hpt_cable80(atadev) &&
((chiptype == 0x00041103 && chiprev >= 0x03) ||
(chiptype == 0x00051103 && chiprev >= 0x01) ||
(chiptype == 0x00081103 && chiprev >= 0x07))) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
ATA_UDMA5, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA5 on HighPoint chip\n",
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
(error) ? "failed" : "success");
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
hpt_timing(atadev, devno, ATA_UDMA5);
ata_dmacreate(atadev, apiomode, ATA_UDMA5);
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
return;
}
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (!ATAPI_DEVICE(atadev) && udmamode >= 4 && hpt_cable80(atadev)) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Unbreak ATAPI on the Aladdin chipset, only DMA access worked. Try to use a 32bit mask on the IO addresses, this fixes the alpha and hopefully doesn't break on any i386 machines. Try to enable both read & write cache on disks, they should be as default, but better be sure..
1999-11-08 21:36:00 +00:00
ATA_UDMA4, ATA_C_F_SETXFER, ATA_WAIT_READY);
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA4 on HighPoint chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success");
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
hpt_timing(atadev, devno, ATA_UDMA4);
ata_dmacreate(atadev, apiomode, ATA_UDMA4);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
}
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (!ATAPI_DEVICE(atadev) && udmamode >= 2) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Unbreak ATAPI on the Aladdin chipset, only DMA access worked. Try to use a 32bit mask on the IO addresses, this fixes the alpha and hopefully doesn't break on any i386 machines. Try to enable both read & write cache on disks, they should be as default, but better be sure..
1999-11-08 21:36:00 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
Ten'th update to the new ATA/ATAPI driver: It been awhile since the last major update, as a benefit there are some cool things in this one (and new bugs probably :) )... The ATA driver has grown "real" timeout support for all devices. This means that it should be possible to get in contact with (especially) lost ATAPI devices. It also means that the ATA driver is now usable on notebooks as it will DTRT on resume. An experimental hack at utilizing the Promise66's at UDMA66 is in there, but I cant test it. If someone feels like sending me one, give me a ping. The ATAPI DMA enableling scheme has been changed, also better DMA support for the Aladdin chipset has been implemented for ATAPI devices. Note that the Aladdin apparently only can do DMA reads on ATAPI devices, and the Promise cant do ATAPI DMA at all. I have seen problems on some ATAPI devices that should be able to run in DMA mode, so if you encounter problems with hanging atapi devices during the probe, or during access, disable DMA in atapi-all.c, and let me know. It might be nessesary to do this via a "white list" for known good devices... The ATAPI CDROM driver can now use eject/close without hanging and the bug that caused reading beyond the end of a CD has been fixed. Media change is also handled proberly. DVD drives are identified and are usable as CDROM devices at least, I dont have the HW to test this further, see above :). The ATAPI tape driver has gotten some support for using the DSC method for not blocking the IDE channel during read/write when the device has full buffers. It knows about the OnStream DI-30 device, support is not completed yet, but it can function as a primitive backup medium, without filemarks, and without bad media handeling. This is because the OnStream device doesn't handle this (like everybody else) in HW. It also now supports getting/setting the record position on devices that supports it. Some rather major cleanups and rearrangements as well (cvs -b diff is your freind). I'm closing in on declaring this for beta code, most of the infrastruture is in place by now. As usual USE AT YOUR OWN RISK!!, this is still alpha level code. This driver can hose your disk real bad if anything goes wrong, but now you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-09-21 19:50:40 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting UDMA2 on HighPoint chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success");
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
hpt_timing(atadev, devno, ATA_UDMA2);
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
}
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (!ATAPI_DEVICE(atadev) && wdmamode >= 2 && apiomode >= 4) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Unbreak ATAPI on the Aladdin chipset, only DMA access worked. Try to use a 32bit mask on the IO addresses, this fixes the alpha and hopefully doesn't break on any i386 machines. Try to enable both read & write cache on disks, they should be as default, but better be sure..
1999-11-08 21:36:00 +00:00
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
Ten'th update to the new ATA/ATAPI driver: It been awhile since the last major update, as a benefit there are some cool things in this one (and new bugs probably :) )... The ATA driver has grown "real" timeout support for all devices. This means that it should be possible to get in contact with (especially) lost ATAPI devices. It also means that the ATA driver is now usable on notebooks as it will DTRT on resume. An experimental hack at utilizing the Promise66's at UDMA66 is in there, but I cant test it. If someone feels like sending me one, give me a ping. The ATAPI DMA enableling scheme has been changed, also better DMA support for the Aladdin chipset has been implemented for ATAPI devices. Note that the Aladdin apparently only can do DMA reads on ATAPI devices, and the Promise cant do ATAPI DMA at all. I have seen problems on some ATAPI devices that should be able to run in DMA mode, so if you encounter problems with hanging atapi devices during the probe, or during access, disable DMA in atapi-all.c, and let me know. It might be nessesary to do this via a "white list" for known good devices... The ATAPI CDROM driver can now use eject/close without hanging and the bug that caused reading beyond the end of a CD has been fixed. Media change is also handled proberly. DVD drives are identified and are usable as CDROM devices at least, I dont have the HW to test this further, see above :). The ATAPI tape driver has gotten some support for using the DSC method for not blocking the IDE channel during read/write when the device has full buffers. It knows about the OnStream DI-30 device, support is not completed yet, but it can function as a primitive backup medium, without filemarks, and without bad media handeling. This is because the OnStream device doesn't handle this (like everybody else) in HW. It also now supports getting/setting the record position on devices that supports it. Some rather major cleanups and rearrangements as well (cvs -b diff is your freind). I'm closing in on declaring this for beta code, most of the infrastruture is in place by now. As usual USE AT YOUR OWN RISK!!, this is still alpha level code. This driver can hose your disk real bad if anything goes wrong, but now you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-09-21 19:50:40 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting WDMA2 on HighPoint chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success");
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
hpt_timing(atadev, devno, ATA_WDMA2);
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
}
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
}
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Update with latest ATA/ATAPI ver 6 rev 2 items.
2001-10-06 11:07:04 +00:00
ATA_PIO0 + apiomode,
Add support for VIA 82C596 controller Better shared irq handeling for Promise & HPT366 controllers Setup prober PIO mode timings on Promise & HPT366 controllers Update Copyright headers to be Y2K compliant :)
2000-01-03 10:26:59 +00:00
ATA_C_F_SETXFER, ATA_WAIT_READY);
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting PIO%d on HighPoint chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success",
(apiomode >= 0) ? apiomode : 0);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
hpt_timing(atadev, devno, ATA_PIO0 + apiomode);
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
atadev->mode = ATA_PIO0 + apiomode;
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
Add support for the PC98 platform to the ATA driver. This mostly consists of functionality to serialize accesses to the two ATA channels (which can also be used to "fix" certain PCI based controllers). Add support for Acard controllers. Enable the ATA driver in PC98 GENERIC, and add device hints. Update man page with latest support. The PC98 core team has kindly provided me with a PC98 machine that made this all possible, thanks to all that contributed to that effort, without that this would probably newer have been possible.. Approved by: re@
2002-12-03 20:20:44 +00:00
case 0x00091191: /* Acard ATP865R controller */
case 0x00081191: /* Acard ATP865 controller */
if (ATAPI_DEVICE(atadev))
break;
if (udmamode >= 6) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_UDMA6, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting up UDMA6 mode on Acard chip\n",
(error) ? "failed" : "success");
if (!error) {
u_int16_t reg44 = pci_read_config(parent, 0x44, 2);
reg44 &= ~(0x000f << (devno << 2));
reg44 |= (0x0007 << (devno << 2));
pci_write_config(parent, 0x44, reg44, 2);
pci_write_config(parent, 0x4a, 0xa6, 1);
pci_write_config(parent, 0x40 + devno, 0x031, 1);
ata_dmacreate(atadev, apiomode, ATA_UDMA6);
return;
}
}
if (udmamode >= 5) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_UDMA5, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting up UDMA5 mode on Acard chip\n",
(error) ? "failed" : "success");
if (!error) {
u_int16_t reg44 = pci_read_config(parent, 0x44, 2);
reg44 &= ~(0x000f << (devno << 2));
reg44 |= (0x0006 << (devno << 2));
pci_write_config(parent, 0x44, reg44, 2);
pci_write_config(parent, 0x4a, 0xa6, 1);
pci_write_config(parent, 0x40 + devno, 0x031, 1);
ata_dmacreate(atadev, apiomode, ATA_UDMA5);
return;
}
}
/* FALLTHROUGH */
case 0x00071191: /* Acard ATP860R controller */
case 0x00061191: /* Acard ATP860 controller */
if (ATAPI_DEVICE(atadev))
break;
if (udmamode >= 4) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_UDMA4, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting up UDMA4 mode on Acard chip\n",
(error) ? "failed" : "success");
if (!error) {
u_int16_t reg44 = pci_read_config(parent, 0x44, 2);
reg44 &= ~(0x000f << (devno << 2));
reg44 |= (0x0005 << (devno << 2));
pci_write_config(parent, 0x44, reg44, 2);
pci_write_config(parent, 0x4a, 0xa6, 1);
pci_write_config(parent, 0x40 + devno, 0x031, 1);
ata_dmacreate(atadev, apiomode, ATA_UDMA4);
return;
}
}
if (udmamode >= 2) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Fix typo PR: 45375
2003-01-09 13:54:07 +00:00
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
Add support for the PC98 platform to the ATA driver. This mostly consists of functionality to serialize accesses to the two ATA channels (which can also be used to "fix" certain PCI based controllers). Add support for Acard controllers. Enable the ATA driver in PC98 GENERIC, and add device hints. Update man page with latest support. The PC98 core team has kindly provided me with a PC98 machine that made this all possible, thanks to all that contributed to that effort, without that this would probably newer have been possible.. Approved by: re@
2002-12-03 20:20:44 +00:00
if (bootverbose)
ata_prtdev(atadev, "%s setting up UDMA2 mode on Acard chip\n",
(error) ? "failed" : "success");
if (!error) {
u_int16_t reg44 = pci_read_config(parent, 0x44, 2);
reg44 &= ~(0x000f << (devno << 2));
reg44 |= (0x0003 << (devno << 2));
pci_write_config(parent, 0x44, reg44, 2);
pci_write_config(parent, 0x4a, 0xa6, 1);
pci_write_config(parent, 0x40 + devno, 0x031, 1);
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
return;
}
}
if (wdmamode >= 2 && apiomode >= 4) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting up WDMA2 mode on Acard chip\n",
(error) ? "failed" : "success");
if (!error) {
u_int16_t reg44 = pci_read_config(parent, 0x44, 2);
reg44 &= ~(0x000f << (devno << 2));
pci_write_config(parent, 0x44, reg44, 2);
pci_write_config(parent, 0x4a, 0xa6, 1);
pci_write_config(parent, 0x40 + devno, 0x031, 1);
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
return;
}
}
/* we could set PIO mode timings, but we assume the BIOS did that */
break;
case 0x00051191: /* Acard ATP850UF controller */
if (ATAPI_DEVICE(atadev))
break;
if (udmamode >= 2) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_UDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting up UDMA2 mode on Acard chip\n",
(error) ? "failed" : "success");
if (!error) {
u_int8_t reg54 = pci_read_config(parent, 0x54, 1);
reg54 |= (0x03 << (devno << 1));
pci_write_config(parent, 0x54, reg54, 1);
pci_write_config(parent, 0x4a, 0xa6, 1);
pci_write_config(parent, 0x40 + (devno << 1), 0x0301, 2);
ata_dmacreate(atadev, apiomode, ATA_UDMA2);
return;
}
}
if (wdmamode >= 2 && apiomode >= 4) {
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
if (bootverbose)
ata_prtdev(atadev, "%s setting up WDMA2 mode on Acard chip\n",
(error) ? "failed" : "success");
if (!error) {
u_int8_t reg54 = pci_read_config(parent, 0x54, 1);
reg54 &= ~(0x03 << (devno << 1));
pci_write_config(parent, 0x54, reg54, 1);
pci_write_config(parent, 0x4a, 0xa6, 1);
pci_write_config(parent, 0x40 + (devno << 1), 0x0301, 2);
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
return;
}
}
/* we could set PIO mode timings, but we assume the BIOS did that */
break;
Add support for the Cenatek Rocket Drive.
2002-02-12 16:59:28 +00:00
case 0x000116ca: /* Cenatek Rocket Drive controller */
if (wdmamode >= 0 &&
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
(ATA_INB(atadev->channel->r_bmio, ATA_BMSTAT_PORT) &
(device ? ATA_BMSTAT_DMA_SLAVE : ATA_BMSTAT_DMA_MASTER)))
ata_dmacreate(atadev, apiomode, ATA_DMA);
Add support for the Cenatek Rocket Drive.
2002-02-12 16:59:28 +00:00
else
atadev->mode = ATA_PIO;
return;
Support quad & max speeds in wormcontrol. A bit more general cleanup.
1999-09-22 12:04:51 +00:00
default: /* unknown controller chip */
/* better not try generic DMA on ATAPI devices it almost never works */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (ATAPI_DEVICE(atadev))
Support quad & max speeds in wormcontrol. A bit more general cleanup.
1999-09-22 12:04:51 +00:00
break;
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
/* if controller says its setup for DMA take the easy way out */
/* the downside is we dont know what DMA mode we are in */
Add support for the Cenatek Rocket Drive.
2002-02-12 16:59:28 +00:00
if ((udmamode >= 0 || wdmamode >= 2) &&
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
(ATA_INB(atadev->channel->r_bmio, ATA_BMSTAT_PORT) &
(device ? ATA_BMSTAT_DMA_SLAVE : ATA_BMSTAT_DMA_MASTER))) {
ata_dmacreate(atadev, apiomode, ATA_DMA);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
}
Support quad & max speeds in wormcontrol. A bit more general cleanup.
1999-09-22 12:04:51 +00:00
/* well, we have no support for this, but try anyways */
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if ((wdmamode >= 2 && apiomode >= 4) && atadev->channel->r_bmio) {
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0,
Unbreak ATAPI on the Aladdin chipset, only DMA access worked. Try to use a 32bit mask on the IO addresses, this fixes the alpha and hopefully doesn't break on any i386 machines. Try to enable both read & write cache on disks, they should be as default, but better be sure..
1999-11-08 21:36:00 +00:00
ATA_WDMA2, ATA_C_F_SETXFER, ATA_WAIT_READY);
Ten'th update to the new ATA/ATAPI driver: It been awhile since the last major update, as a benefit there are some cool things in this one (and new bugs probably :) )... The ATA driver has grown "real" timeout support for all devices. This means that it should be possible to get in contact with (especially) lost ATAPI devices. It also means that the ATA driver is now usable on notebooks as it will DTRT on resume. An experimental hack at utilizing the Promise66's at UDMA66 is in there, but I cant test it. If someone feels like sending me one, give me a ping. The ATAPI DMA enableling scheme has been changed, also better DMA support for the Aladdin chipset has been implemented for ATAPI devices. Note that the Aladdin apparently only can do DMA reads on ATAPI devices, and the Promise cant do ATAPI DMA at all. I have seen problems on some ATAPI devices that should be able to run in DMA mode, so if you encounter problems with hanging atapi devices during the probe, or during access, disable DMA in atapi-all.c, and let me know. It might be nessesary to do this via a "white list" for known good devices... The ATAPI CDROM driver can now use eject/close without hanging and the bug that caused reading beyond the end of a CD has been fixed. Media change is also handled proberly. DVD drives are identified and are usable as CDROM devices at least, I dont have the HW to test this further, see above :). The ATAPI tape driver has gotten some support for using the DSC method for not blocking the IDE channel during read/write when the device has full buffers. It knows about the OnStream DI-30 device, support is not completed yet, but it can function as a primitive backup medium, without filemarks, and without bad media handeling. This is because the OnStream device doesn't handle this (like everybody else) in HW. It also now supports getting/setting the record position on devices that supports it. Some rather major cleanups and rearrangements as well (cvs -b diff is your freind). I'm closing in on declaring this for beta code, most of the infrastruture is in place by now. As usual USE AT YOUR OWN RISK!!, this is still alpha level code. This driver can hose your disk real bad if anything goes wrong, but now you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-09-21 19:50:40 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting WDMA2 on generic chip\n",
General cleanup. Dont be so verbose in the probe, only ONE line printed now, to get more info boot verbose. Centralise most printf's in ata-all & ata-dma to use the ata_printf function, it saves alot of codelines. Repeat the identify command if drive fails the first. Protect the timeout functions with splbio. Dont update the transfer details before we are sure the transfer succeded, this way they are proberly retried on errors. Move the handling of next_writeable to userland. Use the READ_CD command to read CD's. That enables us to read _anything_ via the normal read/write interface. This kindof obsoletes the READAUDIO ioctl, but we keep that for now.
2000-01-24 20:45:24 +00:00
(error) ? "failed" : "success");
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
if (!error) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmacreate(atadev, apiomode, ATA_WDMA2);
Cleanup the ata_dmainit function a bit. Also allow BIOS setup DMA on unknown controllers.
2000-01-28 13:35:43 +00:00
return;
Add DMA support for the VIA 82C586 & 82C686 chips, also rearrange to fall back to slower speeds if the faster ones fails to probe. Log and retry request on UDMA CRC errors. Fix a couple of warnings.
1999-11-24 21:40:05 +00:00
}
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
}
}
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
error = ata_command(atadev, ATA_C_SETFEATURES, 0, ATA_PIO0 + apiomode,
Add support for the Cenatek Rocket Drive.
2002-02-12 16:59:28 +00:00
ATA_C_F_SETXFER, ATA_WAIT_READY);
Add support for VIA 82C596 controller Better shared irq handeling for Promise & HPT366 controllers Setup prober PIO mode timings on Promise & HPT366 controllers Update Copyright headers to be Y2K compliant :)
2000-01-03 10:26:59 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "%s setting PIO%d on generic chip\n",
fix support for shared IRQ's. fix support for multiple HPT & Promise controllers. support mixed 33/66 devices on the Promise 66 controllers. fix the refcount stuff in the atapi drivers. misc cleanups.
2000-02-04 10:20:22 +00:00
(error) ? "failed" : "success", apiomode < 0 ? 0 : apiomode);
if (!error)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
atadev->mode = ATA_PIO0 + apiomode;
Fix the CD driver so that the last blocks can be read even if a blocksize != 2k is used. Update the timeout code to try fallback to PIO if problems arise in DMA mode.
2000-03-05 16:52:26 +00:00
else {
fix support for shared IRQ's. fix support for multiple HPT & Promise controllers. support mixed 33/66 devices on the Promise 66 controllers. fix the refcount stuff in the atapi drivers. misc cleanups.
2000-02-04 10:20:22 +00:00
if (bootverbose)
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_prtdev(atadev, "using PIO mode set by BIOS\n");
atadev->mode = ATA_PIO;
Fix the CD driver so that the last blocks can be read even if a blocksize != 2k is used. Update the timeout code to try fallback to PIO if problems arise in DMA mode.
2000-03-05 16:52:26 +00:00
}
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
struct ata_dmasetup_data_cb_args {
struct ata_dmaentry *dmatab;
int error;
};
static void
ata_dmasetupd_cb(void *xsc, bus_dma_segment_t *segs, int nsegs, int error)
{
struct ata_dmasetup_data_cb_args *cba =
(struct ata_dmasetup_data_cb_args *)xsc;
bus_size_t cnt;
u_int32_t lastcount;
int i, j;
cba->error = error;
if (error != 0)
return;
lastcount = j = 0;
for (i = 0; i < nsegs; i++) {
/*
* A maximum segment size was specified for bus_dma_tag_create, but
* some busdma code does not seem to honor this, so fix up if needed.
*/
for (cnt = 0; cnt < segs[i].ds_len; cnt += MAXSEGSZ, j++) {
cba->dmatab[j].base = htole32(segs[i].ds_addr + cnt);
lastcount = ulmin(segs[i].ds_len - cnt, MAXSEGSZ) & 0xffff;
cba->dmatab[j].count = htole32(lastcount);
}
}
cba->dmatab[j - 1].count = htole32(lastcount | ATA_DMA_EOT);
}
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
int
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmasetup(struct ata_device *atadev, caddr_t data, int32_t count)
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
{
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
struct ata_channel *ch = atadev->channel;
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
if (((uintptr_t)data & ch->alignment) || (count & ch->alignment)) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_prtdev(atadev, "non aligned DMA transfer attempted\n");
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
return -1;
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
}
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
if (!count) {
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_prtdev(atadev, "zero length DMA transfer attempted\n");
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
return -1;
}
return 0;
}
Fix the breakage of tagged queueing that the busdma integration introduced. Since its now only possible to have one DMA control block at a time, we move the setup to dmastart instead.
2002-04-18 19:11:45 +00:00
int
ata_dmastart(struct ata_device *atadev, caddr_t data, int32_t count, int dir)
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
{
Fix the breakage of tagged queueing that the busdma integration introduced. Since its now only possible to have one DMA control block at a time, we move the setup to dmastart instead.
2002-04-18 19:11:45 +00:00
struct ata_channel *ch = atadev->channel;
struct ata_dmastate *ds = &atadev->dmastate;
struct ata_dmasetup_data_cb_args cba;
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
Fix the breakage of tagged queueing that the busdma integration introduced. Since its now only possible to have one DMA control block at a time, we move the setup to dmastart instead.
2002-04-18 19:11:45 +00:00
if (ds->flags & ATA_DS_ACTIVE)
panic("ata_dmasetup: transfer active on this device!");
Add code that works around the problem that the older Promise controllers (ultra/fasttrak-66/100) fails on 48bit accesses.
2003-01-08 10:03:31 +00:00
switch(ch->chiptype) {
case 0x0d38105a: /* Promise Fasttrak 66 */
case 0x4d38105a: /* Promise Ultra/Fasttrak 66 */
case 0x0d30105a: /* Promise OEM ATA 100 */
case 0x4d30105a: /* Promise Ultra/Fasttrak 100 */
Fix the 48bit access support for the older Promise 66/100 controllers, the first attempt was wrong and could cause r/w timeouts. Add yet another Promise PCI id.
2003-01-19 11:47:32 +00:00
if (ch->flags & ATA_48BIT_ACTIVE) {
Small change to the previous commit, zero out the 48BIT flag in ata_command instead of in dmadone.
2003-01-19 20:18:07 +00:00
ATA_OUTB(ch->r_bmio, (ch->unit ? 0x09 : 0x11),
ATA_INB(ch->r_bmio, (ch->unit ? 0x09 : 0x11)) |
(ch->unit ? 0x08 : 0x02));
ATA_OUTL(ch->r_bmio, (ch->unit ? 0x1c : 0x20),
(dir ? 0x05000000 : 0x06000000) | (count >> 1));
Fix the 48bit access support for the older Promise 66/100 controllers, the first attempt was wrong and could cause r/w timeouts. Add yet another Promise PCI id.
2003-01-19 11:47:32 +00:00
}
Add code that works around the problem that the older Promise controllers (ultra/fasttrak-66/100) fails on 48bit accesses.
2003-01-08 10:03:31 +00:00
}
Fix the breakage of tagged queueing that the busdma integration introduced. Since its now only possible to have one DMA control block at a time, we move the setup to dmastart instead.
2002-04-18 19:11:45 +00:00
cba.dmatab = ds->dmatab;
bus_dmamap_sync(ds->cdmatag, ds->cdmamap, BUS_DMASYNC_PREWRITE);
if (bus_dmamap_load(ds->ddmatag, ds->ddmamap, data, count,
ata_dmasetupd_cb, &cba, 0) || cba.error)
return -1;
bus_dmamap_sync(ds->cdmatag, ds->cdmamap, BUS_DMASYNC_POSTWRITE);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
bus_dmamap_sync(ds->ddmatag, ds->ddmamap, dir ? BUS_DMASYNC_PREREAD :
BUS_DMASYNC_PREWRITE);
Fix the breakage of tagged queueing that the busdma integration introduced. Since its now only possible to have one DMA control block at a time, we move the setup to dmastart instead.
2002-04-18 19:11:45 +00:00
ch->flags |= ATA_DMA_ACTIVE;
Add code that works around the problem that the older Promise controllers (ultra/fasttrak-66/100) fails on 48bit accesses.
2003-01-08 10:03:31 +00:00
ds->flags = dir ? (ATA_DS_ACTIVE | ATA_DS_READ) : ATA_DS_ACTIVE;
Fix the breakage of tagged queueing that the busdma integration introduced. Since its now only possible to have one DMA control block at a time, we move the setup to dmastart instead.
2002-04-18 19:11:45 +00:00
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ATA_OUTL(ch->r_bmio, ATA_BMDTP_PORT, ds->mdmatab);
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ATA_OUTB(ch->r_bmio, ATA_BMCMD_PORT, dir ? ATA_BMCMD_WRITE_READ : 0);
ATA_OUTB(ch->r_bmio, ATA_BMSTAT_PORT,
Add code that works around the problem that the older Promise controllers (ultra/fasttrak-66/100) fails on 48bit accesses.
2003-01-08 10:03:31 +00:00
(ATA_INB(ch->r_bmio, ATA_BMSTAT_PORT) |
(ATA_BMSTAT_INTERRUPT | ATA_BMSTAT_ERROR)));
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ATA_OUTB(ch->r_bmio, ATA_BMCMD_PORT,
Add code that works around the problem that the older Promise controllers (ultra/fasttrak-66/100) fails on 48bit accesses.
2003-01-08 10:03:31 +00:00
ATA_INB(ch->r_bmio, ATA_BMCMD_PORT) | ATA_BMCMD_START_STOP);
Fix the breakage of tagged queueing that the busdma integration introduced. Since its now only possible to have one DMA control block at a time, we move the setup to dmastart instead.
2002-04-18 19:11:45 +00:00
return 0;
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
}
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
int
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ata_dmadone(struct ata_device *atadev)
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
{
Add code that works around the problem that the older Promise controllers (ultra/fasttrak-66/100) fails on 48bit accesses.
2003-01-08 10:03:31 +00:00
struct ata_channel *ch = atadev->channel;
struct ata_dmastate *ds = &atadev->dmastate;
Introduce busspace instead of the good old in/out instructions. Not pretty but it works (I hope)...
2001-02-06 16:44:25 +00:00
int error;
Add code that works around the problem that the older Promise controllers (ultra/fasttrak-66/100) fails on 48bit accesses.
2003-01-08 10:03:31 +00:00
switch(ch->chiptype) {
case 0x0d38105a: /* Promise Fasttrak 66 */
case 0x4d38105a: /* Promise Ultra/Fasttrak 66 */
case 0x0d30105a: /* Promise OEM ATA 100 */
case 0x4d30105a: /* Promise Ultra/Fasttrak 100 */
Fix the 48bit access support for the older Promise 66/100 controllers, the first attempt was wrong and could cause r/w timeouts. Add yet another Promise PCI id.
2003-01-19 11:47:32 +00:00
if (ch->flags & ATA_48BIT_ACTIVE) {
Small change to the previous commit, zero out the 48BIT flag in ata_command instead of in dmadone.
2003-01-19 20:18:07 +00:00
ATA_OUTB(ch->r_bmio, (ch->unit ? 0x09 : 0x11),
ATA_INB(ch->r_bmio, (ch->unit ? 0x09 : 0x11)) &
~(ch->unit ? 0x08 : 0x02));
ATA_OUTL(ch->r_bmio, (ch->unit ? 0x1c : 0x20), 0);
Fix the 48bit access support for the older Promise 66/100 controllers, the first attempt was wrong and could cause r/w timeouts. Add yet another Promise PCI id.
2003-01-19 11:47:32 +00:00
}
Add code that works around the problem that the older Promise controllers (ultra/fasttrak-66/100) fails on 48bit accesses.
2003-01-08 10:03:31 +00:00
}
Fix the 48bit access support for the older Promise 66/100 controllers, the first attempt was wrong and could cause r/w timeouts. Add yet another Promise PCI id.
2003-01-19 11:47:32 +00:00
error = ATA_INB(ch->r_bmio, ATA_BMSTAT_PORT);
ATA_OUTB(ch->r_bmio, ATA_BMCMD_PORT,
ATA_INB(ch->r_bmio, ATA_BMCMD_PORT) & ~ATA_BMCMD_START_STOP);
ATA_OUTB(ch->r_bmio, ATA_BMSTAT_PORT,ATA_BMSTAT_INTERRUPT|ATA_BMSTAT_ERROR);
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
bus_dmamap_sync(ds->ddmatag, ds->ddmamap, (ds->flags & ATA_DS_READ) != 0 ?
BUS_DMASYNC_POSTREAD : BUS_DMASYNC_POSTWRITE);
bus_dmamap_unload(ds->ddmatag, ds->ddmamap);
Fix the 48bit access support for the older Promise 66/100 controllers, the first attempt was wrong and could cause r/w timeouts. Add yet another Promise PCI id.
2003-01-19 11:47:32 +00:00
Fix the breakage of tagged queueing that the busdma integration introduced. Since its now only possible to have one DMA control block at a time, we move the setup to dmastart instead.
2002-04-18 19:11:45 +00:00
ch->flags &= ~ATA_DMA_ACTIVE;
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ds->flags = 0;
Fix the breakage of tagged queueing that the busdma integration introduced. Since its now only possible to have one DMA control block at a time, we move the setup to dmastart instead.
2002-04-18 19:11:45 +00:00
return (error & ATA_BMSTAT_MASK);
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
}
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
int
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
ata_dmastatus(struct ata_channel *ch)
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
{
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
return ATA_INB(ch->r_bmio, ATA_BMSTAT_PORT) & ATA_BMSTAT_MASK;
Fourth update to the new ATA/ATAPI driver: Well, better late than newer, but things has been hectic around here, sorry for the long delay. DMA support has been added to the ATA disk driver. This only works on Intel PIIX3/4, Acer Aladdin and Promise controllers. The promise support works without the BIOS on the board, and timing modes are set to support up to UDMA speed. This solves the problems with having more than one promise controller in the same system. There is support for "generic" DMA, that might work on other controllers, but now you have been warned :) More chipset specific code will come soon, I have to find testers with the approbiate HW, more on that when I have it ready. The system now uses its own major numbers, please run MAKEDEV with the devices you need (ad?, acd?, afd?, ast?). For now the disk driver will also attach to the old wd major so one can at least boot without this step, but be warned, this will eventually go away. The bootblocks will have to be changed before one can boot directly from an "ad" device though. Fixed problems: All known hang problems should be solved The probe code has been sligthly changed, this should solve the reports I have lying around (I hope). Hangs when accessing ata & atapi device on the same channel simultaniously. A real braino in ata_start caused this, fixed. As usual USE AT YOUR OWN RISK!!, this is still pre alpha level code. Especially the DMA support can hose your disk real bad if anything goes wrong, agaiin you have been warned :) But please tell me how it works for you! Enjoy! -Søren
1999-03-28 18:57:20 +00:00
}
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
static void
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
cyrix_timing(struct ata_device *atadev, int devno, int mode)
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
{
u_int32_t reg20 = 0x0000e132;
u_int32_t reg24 = 0x00017771;
switch (mode) {
case ATA_PIO0: reg20 = 0x0000e132; break;
case ATA_PIO1: reg20 = 0x00018121; break;
case ATA_PIO2: reg20 = 0x00024020; break;
case ATA_PIO3: reg20 = 0x00032010; break;
case ATA_PIO4: reg20 = 0x00040010; break;
case ATA_WDMA2: reg24 = 0x00002020; break;
case ATA_UDMA2: reg24 = 0x00911030; break;
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
ATA_OUTL(atadev->channel->r_bmio, (devno << 3) + 0x20, reg20);
ATA_OUTL(atadev->channel->r_bmio, (devno << 3) + 0x24, reg24);
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
}
Add support for the Promise TX4. Rearrange the support for the VIA chips, and add experimental support for ATA133 on the newest chips.
2002-01-28 13:17:10 +00:00
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
static void
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
promise_timing(struct ata_device *atadev, int devno, int mode)
Add support for VIA 82C596 controller Better shared irq handeling for Promise & HPT366 controllers Setup prober PIO mode timings on Promise & HPT366 controllers Update Copyright headers to be Y2K compliant :)
2000-01-03 10:26:59 +00:00
{
fix support for shared IRQ's. fix support for multiple HPT & Promise controllers. support mixed 33/66 devices on the Promise 66 controllers. fix the refcount stuff in the atapi drivers. misc cleanups.
2000-02-04 10:20:22 +00:00
u_int32_t timing = 0;
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
/* XXX: Endianess */
fix support for shared IRQ's. fix support for multiple HPT & Promise controllers. support mixed 33/66 devices on the Promise 66 controllers. fix the refcount stuff in the atapi drivers. misc cleanups.
2000-02-04 10:20:22 +00:00
struct promise_timing {
u_int8_t pa:4;
u_int8_t prefetch:1;
u_int8_t iordy:1;
u_int8_t errdy:1;
u_int8_t syncin:1;
u_int8_t pb:5;
u_int8_t mb:3;
u_int8_t mc:4;
u_int8_t dmaw:1;
u_int8_t dmar:1;
u_int8_t iordyp:1;
u_int8_t dmarqp:1;
u_int8_t reserved:8;
} *t = (struct promise_timing*)&timing;
t->iordy = 1; t->iordyp = 1;
if (mode >= ATA_DMA) {
t->prefetch = 1; t->errdy = 1; t->syncin = 1;
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
switch (atadev->channel->chiptype) {
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
case 0x4d33105a: /* Promise Ultra/Fasttrak 33 */
fix support for shared IRQ's. fix support for multiple HPT & Promise controllers. support mixed 33/66 devices on the Promise 66 controllers. fix the refcount stuff in the atapi drivers. misc cleanups.
2000-02-04 10:20:22 +00:00
switch (mode) {
default:
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
case ATA_PIO0: t->pa = 9; t->pb = 19; t->mb = 7; t->mc = 15; break;
case ATA_PIO1: t->pa = 5; t->pb = 12; t->mb = 7; t->mc = 15; break;
case ATA_PIO2: t->pa = 3; t->pb = 8; t->mb = 7; t->mc = 15; break;
case ATA_PIO3: t->pa = 2; t->pb = 6; t->mb = 7; t->mc = 15; break;
case ATA_PIO4: t->pa = 1; t->pb = 4; t->mb = 7; t->mc = 15; break;
case ATA_WDMA2: t->pa = 3; t->pb = 7; t->mb = 3; t->mc = 3; break;
case ATA_UDMA2: t->pa = 3; t->pb = 7; t->mb = 1; t->mc = 1; break;
fix support for shared IRQ's. fix support for multiple HPT & Promise controllers. support mixed 33/66 devices on the Promise 66 controllers. fix the refcount stuff in the atapi drivers. misc cleanups.
2000-02-04 10:20:22 +00:00
}
break;
Add yet another (older) Promise chip
2002-06-19 12:26:20 +00:00
case 0x0d38105a: /* Promise Fasttrak 66 */
Add support for tagged queuing on ATA drives. There is only support for IBM's DPTA and DTLA series of drives (no other disk vendors are known to support this) on non-Promise controllers (promise controllers lockup when given the tagged queuing specific commands). It gives especially master/slave comboes about 5% better performance. Add support for the Promise ATA100 OEM chip (pdc20265) Add support for the Cyrix 5530 Change the way status is read from the drives, use the alternate status reg when possible. Better support for DEVFS, the acdXtY devices are now created when needed. Lots of little cleanups.
2000-09-19 11:08:39 +00:00
case 0x4d38105a: /* Promise Ultra/Fasttrak 66 */
case 0x0d30105a: /* Promise OEM ATA 100 */
Add yet another (older) Promise chip
2002-06-19 12:26:20 +00:00
case 0x4d30105a: /* Promise Ultra/Fasttrak 100 */
fix support for shared IRQ's. fix support for multiple HPT & Promise controllers. support mixed 33/66 devices on the Promise 66 controllers. fix the refcount stuff in the atapi drivers. misc cleanups.
2000-02-04 10:20:22 +00:00
switch (mode) {
default:
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
case ATA_PIO0: t->pa = 15; t->pb = 31; t->mb = 7; t->mc = 15; break;
case ATA_PIO1: t->pa = 10; t->pb = 24; t->mb = 7; t->mc = 15; break;
case ATA_PIO2: t->pa = 6; t->pb = 16; t->mb = 7; t->mc = 15; break;
case ATA_PIO3: t->pa = 4; t->pb = 12; t->mb = 7; t->mc = 15; break;
case ATA_PIO4: t->pa = 2; t->pb = 8; t->mb = 7; t->mc = 15; break;
case ATA_WDMA2: t->pa = 6; t->pb = 14; t->mb = 6; t->mc = 6; break;
case ATA_UDMA2: t->pa = 6; t->pb = 14; t->mb = 2; t->mc = 2; break;
case ATA_UDMA4: t->pa = 3; t->pb = 7; t->mb = 1; t->mc = 1; break;
case ATA_UDMA5: t->pa = 3; t->pb = 7; t->mb = 1; t->mc = 1; break;
fix support for shared IRQ's. fix support for multiple HPT & Promise controllers. support mixed 33/66 devices on the Promise 66 controllers. fix the refcount stuff in the atapi drivers. misc cleanups.
2000-02-04 10:20:22 +00:00
}
break;
Add support for VIA 82C596 controller Better shared irq handeling for Promise & HPT366 controllers Setup prober PIO mode timings on Promise & HPT366 controllers Update Copyright headers to be Y2K compliant :)
2000-01-03 10:26:59 +00:00
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
pci_write_config(device_get_parent(atadev->channel->dev),
0x60 + (devno << 2), timing, 4);
Add support for VIA 82C596 controller Better shared irq handeling for Promise & HPT366 controllers Setup prober PIO mode timings on Promise & HPT366 controllers Update Copyright headers to be Y2K compliant :)
2000-01-03 10:26:59 +00:00
}
static void
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
hpt_timing(struct ata_device *atadev, int devno, int mode)
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
{
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
device_t parent = device_get_parent(atadev->channel->dev);
u_int32_t chiptype = atadev->channel->chiptype;
int chiprev = pci_get_revid(parent);
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
u_int32_t timing;
Even more Highpoint RAID support. Fix the 80pin cable detection system.
2002-03-08 21:36:49 +00:00
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (chiptype == 0x00081103 && chiprev >= 0x07) {
Add support for the Highpoint HPT372 based cards (rocketraid 133). HW Sponsored by: Mike Tancsa
2002-02-18 11:57:56 +00:00
switch (mode) { /* HPT374 */
Add support for the HighPoint HPT374 4 channel ATA chip. Sponsored by: Isilon Systems.
2002-02-11 15:48:04 +00:00
case ATA_PIO0: timing = 0x0ac1f48a; break;
case ATA_PIO1: timing = 0x0ac1f465; break;
case ATA_PIO2: timing = 0x0a81f454; break;
case ATA_PIO3: timing = 0x0a81f443; break;
case ATA_PIO4: timing = 0x0a81f442; break;
Major update of the ATA RAID code, part 3: Add code to properly detach/attach disks that are part of a RAID. Mark a disk that is attached on an ATA channel belonging to a RAID as a spare disk that can be used for rebuilding failed RAID1's. Add support for rebuilding failed RAID1's. Several fixes to the detach/attach code. For replacing a disk in a failed RAID1 do the following: Find the controller channel# of the failed disk. Exec 'atacontrol detach <channel#>' to free the disk from the system. Replace the failed disk with a new one of at least the same size. If your have your disks in drawers/enclosures this can be done with the system still running. Exec 'atacontrol attach <channel#>' to add the disk to the system and mark it as a valid spare for rebuild. Exec 'atacontrol rebuild <array#>' The system will rebuild the array on the fly, the array can still be used during this, although with slower performance. Please let me know of any problems with this! Sponsored by: Advanis Inc. MFC after: 2 weeks
2002-03-03 15:36:21 +00:00
case ATA_WDMA2: timing = 0x22808242; break;
case ATA_UDMA2: timing = 0x120c8242; break;
case ATA_UDMA4: timing = 0x12ac8242; break;
case ATA_UDMA5: timing = 0x12848242; break;
case ATA_UDMA6: timing = 0x12808242; break;
Add support for the HighPoint HPT374 4 channel ATA chip. Sponsored by: Isilon Systems.
2002-02-11 15:48:04 +00:00
default: timing = 0x0d029d5e;
}
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
else if ((chiptype == 0x00041103 && chiprev >= 0x05) ||
(chiptype == 0x00051103 && chiprev >= 0x01)) {
Add support for the Highpoint HPT372 based cards (rocketraid 133). HW Sponsored by: Mike Tancsa
2002-02-18 11:57:56 +00:00
switch (mode) { /* HPT372 */
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
case ATA_PIO0: timing = 0x0d029d5e; break;
case ATA_PIO1: timing = 0x0d029d26; break;
case ATA_PIO2: timing = 0x0c829ca6; break;
case ATA_PIO3: timing = 0x0c829c84; break;
case ATA_PIO4: timing = 0x0c829c62; break;
Add experimental support for the HighPoint HPT372 ATA133 controller. Based on docs kindly provided by HighPoint which we thank for the close cooperation with the FreeBSD project!.
2001-10-23 09:25:57 +00:00
case ATA_WDMA2: timing = 0x2c829262; break;
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
case ATA_UDMA2: timing = 0x1c91dc62; break;
case ATA_UDMA4: timing = 0x1c8ddc62; break;
case ATA_UDMA5: timing = 0x1c6ddc62; break;
case ATA_UDMA6: timing = 0x1c81dc62; break;
Add experimental support for the HighPoint HPT372 ATA133 controller. Based on docs kindly provided by HighPoint which we thank for the close cooperation with the FreeBSD project!.
2001-10-23 09:25:57 +00:00
default: timing = 0x0d029d5e;
}
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
else if (chiptype == 0x00041103 && chiprev >= 0x03) {
Add support for the Highpoint HPT372 based cards (rocketraid 133). HW Sponsored by: Mike Tancsa
2002-02-18 11:57:56 +00:00
switch (mode) { /* HPT370 */
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
case ATA_PIO0: timing = 0x06914e57; break;
case ATA_PIO1: timing = 0x06914e43; break;
case ATA_PIO2: timing = 0x06514e33; break;
case ATA_PIO3: timing = 0x06514e22; break;
case ATA_PIO4: timing = 0x06514e21; break;
Major update of the ATA RAID code, part 1: Overhaul of the attach/detach code and structures, there were some nasty bugs in the old implementation. This made it possible to collapse the ATA/ATAPI device control structures into one generic structure. A note here, the kernel is NOT ready for detach of active devices, it fails all over in random places, but for inactive devices it works. However for ATA RAID this works, since the RAID abstration layer insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the physical disks. Proberly detect the RAID's from the BIOS, and mark critical RAID1 arrays as such, but continue if there is enough of the mirror left to do so. Properly fail arrays on a live system. For RAID0 that means return EIO, and for RAID1 it means continue on the still working part of the mirror if possible, else return EIO. If the state changes, log this to the console. Allow for Promise & Highpoint controllers/arrays to coexist on the same machine. It is not possible to distribute arrays over different makes of controllers though. If Promise SuperSwap enclosures are used, signal disk state on the status LED on the front. Misc fixes that I had lying around for various minor bugs. Sponsored by: Advanis Inc.
2002-02-04 19:23:40 +00:00
case ATA_WDMA2: timing = 0x26514e21; break;
case ATA_UDMA2: timing = 0x16494e31; break;
case ATA_UDMA4: timing = 0x16454e31; break;
case ATA_UDMA5: timing = 0x16454e31; break;
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
default: timing = 0x06514e57;
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
}
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
pci_write_config(parent, 0x40 + (devno << 2) , timing, 4);
}
Add support for the Highpoint HPT372 based cards (rocketraid 133). HW Sponsored by: Mike Tancsa
2002-02-18 11:57:56 +00:00
else { /* HPT36[68] */
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
switch (pci_read_config(parent, 0x41 + (devno << 2), 1)) {
case 0x85: /* 25Mhz */
switch (mode) {
Even more Highpoint RAID support. Fix the 80pin cable detection system.
2002-03-08 21:36:49 +00:00
case ATA_PIO0: timing = 0x40d08585; break;
case ATA_PIO1: timing = 0x40d08572; break;
case ATA_PIO2: timing = 0x40ca8542; break;
case ATA_PIO3: timing = 0x40ca8532; break;
case ATA_PIO4: timing = 0x40ca8521; break;
case ATA_WDMA2: timing = 0x20ca8521; break;
case ATA_UDMA2: timing = 0x10cf8521; break;
case ATA_UDMA4: timing = 0x10c98521; break;
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
default: timing = 0x01208585;
}
break;
default:
case 0xa7: /* 33MHz */
switch (mode) {
Even more Highpoint RAID support. Fix the 80pin cable detection system.
2002-03-08 21:36:49 +00:00
case ATA_PIO0: timing = 0x40d0a7aa; break;
case ATA_PIO1: timing = 0x40d0a7a3; break;
case ATA_PIO2: timing = 0x40d0a753; break;
case ATA_PIO3: timing = 0x40c8a742; break;
case ATA_PIO4: timing = 0x40c8a731; break;
case ATA_WDMA2: timing = 0x20c8a731; break;
case ATA_UDMA2: timing = 0x10caa731; break;
case ATA_UDMA4: timing = 0x10c9a731; break;
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
default: timing = 0x0120a7a7;
}
break;
case 0xd9: /* 40Mhz */
switch (mode) {
Even more Highpoint RAID support. Fix the 80pin cable detection system.
2002-03-08 21:36:49 +00:00
case ATA_PIO0: timing = 0x4018d9d9; break;
case ATA_PIO1: timing = 0x4010d9c7; break;
case ATA_PIO2: timing = 0x4010d997; break;
case ATA_PIO3: timing = 0x4010d974; break;
case ATA_PIO4: timing = 0x4008d963; break;
case ATA_WDMA2: timing = 0x2008d943; break;
case ATA_UDMA2: timing = 0x100bd943; break;
case ATA_UDMA4: timing = 0x100fd943; break;
Add experimental code for ATA100 support on: Promise Ultra100 / Fasttrak100 HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100) Intel ICH2 (Intel 815E based motherboards) So far I can read >90MB/s on the Promise and the HPT370. I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems writing is still done in ATA66 mode :( The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
default: timing = 0x0120d9d9;
}
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
}
}
Even more Highpoint RAID support. Fix the 80pin cable detection system.
2002-03-08 21:36:49 +00:00
pci_write_config(parent, 0x40 + (devno << 2) , timing, 4);
}
static int
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
hpt_cable80(struct ata_device *atadev)
Even more Highpoint RAID support. Fix the 80pin cable detection system.
2002-03-08 21:36:49 +00:00
{
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
device_t parent = device_get_parent(atadev->channel->dev);
Even more Highpoint RAID support. Fix the 80pin cable detection system.
2002-03-08 21:36:49 +00:00
u_int8_t reg, val, res;
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
if (atadev->channel->chiptype==0x00081103 && pci_get_function(parent)==1) {
reg = atadev->channel->unit ? 0x57 : 0x53;
Even more Highpoint RAID support. Fix the 80pin cable detection system.
2002-03-08 21:36:49 +00:00
val = pci_read_config(parent, reg, 1);
pci_write_config(parent, reg, val | 0x80, 1);
}
else {
reg = 0x5b;
val = pci_read_config(parent, reg, 1);
pci_write_config(parent, reg, val & 0xfe, 1);
}
Make the ATA driver compile & work on the sparc64 platform. Initial work & code by tmm. Lots of changes and rearrangements by yours truely to make busdma be a little less a PITA (but I still dont like it).
2002-04-05 13:13:56 +00:00
res = pci_read_config(parent, 0x5a, 1) & (atadev->channel->unit ? 0x1:0x2);
Even more Highpoint RAID support. Fix the 80pin cable detection system.
2002-03-08 21:36:49 +00:00
pci_write_config(parent, reg, val, 1);
return !res;
Add support for the HPT366 chip, this is used on the Abit boards and their HotRod controller and on SIIG PCI ultra DMA controller. These changes also made lots of the Promise code go away, its all much more generic this way. Get rid of atapi_immed_cmd, instead use the queue to move atapi commands from interrupt context if nessesary, the entire atapi layer has gotten an overhaul. Lots of fixes to utililize the new features in subr_disk.c etc, and get rid of the last biots of softc arrays in the drivers, the only one left is atadevices which cannot easily go away (yet). Use our own malloc names, its a lot easier to track memory usage this way. General cleanup overall.
1999-10-09 19:57:13 +00:00
}
Reference in New Issue Copy Permalink