2001-03-06 21:43:46 +00:00
|
|
|
|
/*-
|
2008-04-10 13:05:05 +00:00
|
|
|
|
* Copyright (c) 1998 - 2008 S<EFBFBD>ren Schmidt <sos@FreeBSD.org>
|
2001-03-06 21:43:46 +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.
|
|
|
|
|
*
|
|
|
|
|
* 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.
|
|
|
|
|
*/
|
|
|
|
|
|
2003-08-24 17:55:58 +00:00
|
|
|
|
#include <sys/cdefs.h>
|
|
|
|
|
__FBSDID("$FreeBSD$");
|
|
|
|
|
|
2002-04-05 13:13:56 +00:00
|
|
|
|
#include "opt_ata.h"
|
2001-03-06 21:43:46 +00:00
|
|
|
|
#include <sys/param.h>
|
|
|
|
|
#include <sys/systm.h>
|
|
|
|
|
#include <sys/kernel.h>
|
|
|
|
|
#include <sys/module.h>
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
#include <sys/ata.h>
|
2001-03-06 21:43:46 +00:00
|
|
|
|
#include <sys/bus.h>
|
2006-02-09 20:53:32 +00:00
|
|
|
|
#include <sys/conf.h>
|
2001-03-06 21:43:46 +00:00
|
|
|
|
#include <sys/malloc.h>
|
2004-01-11 22:08:34 +00:00
|
|
|
|
#include <sys/sema.h>
|
2003-08-24 09:22:26 +00:00
|
|
|
|
#include <sys/taskqueue.h>
|
2004-01-14 21:26:35 +00:00
|
|
|
|
#include <vm/uma.h>
|
2001-03-06 21:43:46 +00:00
|
|
|
|
#include <machine/stdarg.h>
|
|
|
|
|
#include <machine/resource.h>
|
|
|
|
|
#include <machine/bus.h>
|
|
|
|
|
#include <sys/rman.h>
|
2003-08-22 05:54:52 +00:00
|
|
|
|
#include <dev/pci/pcivar.h>
|
|
|
|
|
#include <dev/pci/pcireg.h>
|
2001-03-06 21:43:46 +00:00
|
|
|
|
#include <dev/ata/ata-all.h>
|
2003-02-20 20:02:32 +00:00
|
|
|
|
#include <dev/ata/ata-pci.h>
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
#include <ata_if.h>
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
2003-02-20 20:02:32 +00:00
|
|
|
|
/* local vars */
|
2005-10-31 15:41:29 +00:00
|
|
|
|
static MALLOC_DEFINE(M_ATAPCI, "ata_pci", "ATA driver PCI");
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
2001-09-25 17:10:39 +00:00
|
|
|
|
/* misc defines */
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
#define IOMASK 0xfffffffc
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
/*
|
|
|
|
|
* generic PCI ATA device probe
|
|
|
|
|
*/
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
int
|
2003-02-20 20:02:32 +00:00
|
|
|
|
ata_pci_probe(device_t dev)
|
2001-03-06 21:43:46 +00:00
|
|
|
|
{
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(dev);
|
|
|
|
|
char buffer[64];
|
|
|
|
|
|
|
|
|
|
/* is this a storage class device ? */
|
2001-03-06 21:43:46 +00:00
|
|
|
|
if (pci_get_class(dev) != PCIC_STORAGE)
|
2009-04-27 21:34:15 +00:00
|
|
|
|
return (ENXIO);
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
/* is this an IDE/ATA type device ? */
|
|
|
|
|
if (pci_get_subclass(dev) != PCIS_STORAGE_IDE)
|
2009-04-27 21:34:15 +00:00
|
|
|
|
return (ENXIO);
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
|
|
|
|
|
sprintf(buffer, "%s ATA controller", ata_pcivendor2str(dev));
|
|
|
|
|
device_set_desc_copy(dev, buffer);
|
|
|
|
|
ctlr->chipinit = ata_generic_chipinit;
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
/* we are a low priority handler */
|
2009-04-27 21:34:15 +00:00
|
|
|
|
return (BUS_PROBE_GENERIC);
|
2001-03-06 21:43:46 +00:00
|
|
|
|
}
|
|
|
|
|
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
int
|
2001-03-06 21:43:46 +00:00
|
|
|
|
ata_pci_attach(device_t dev)
|
|
|
|
|
{
|
2003-03-29 13:37:09 +00:00
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(dev);
|
2009-02-14 21:54:44 +00:00
|
|
|
|
device_t child;
|
2004-03-15 12:03:48 +00:00
|
|
|
|
u_int32_t cmd;
|
2004-04-27 12:54:59 +00:00
|
|
|
|
int unit;
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
2003-03-29 13:37:09 +00:00
|
|
|
|
/* do chipset specific setups only needed once */
|
2008-12-16 16:04:40 +00:00
|
|
|
|
ctlr->legacy = ata_legacy(dev);
|
|
|
|
|
if (ctlr->legacy || pci_read_config(dev, PCIR_BAR(2), 4) & IOMASK)
|
2003-03-29 13:37:09 +00:00
|
|
|
|
ctlr->channels = 2;
|
|
|
|
|
else
|
|
|
|
|
ctlr->channels = 1;
|
2009-02-16 19:10:07 +00:00
|
|
|
|
ctlr->ichannels = -1;
|
2009-02-18 22:17:48 +00:00
|
|
|
|
ctlr->ch_attach = ata_pci_ch_attach;
|
2009-02-19 00:32:55 +00:00
|
|
|
|
ctlr->ch_detach = ata_pci_ch_detach;
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
ctlr->dev = dev;
|
2003-03-29 13:37:09 +00:00
|
|
|
|
|
2004-03-15 12:03:48 +00:00
|
|
|
|
/* if needed try to enable busmastering */
|
|
|
|
|
cmd = pci_read_config(dev, PCIR_COMMAND, 2);
|
2002-04-05 13:13:56 +00:00
|
|
|
|
if (!(cmd & PCIM_CMD_BUSMASTEREN)) {
|
|
|
|
|
pci_write_config(dev, PCIR_COMMAND, cmd | PCIM_CMD_BUSMASTEREN, 2);
|
|
|
|
|
cmd = pci_read_config(dev, PCIR_COMMAND, 2);
|
|
|
|
|
}
|
2003-03-29 13:37:09 +00:00
|
|
|
|
|
2004-03-15 12:03:48 +00:00
|
|
|
|
/* if busmastering mode "stuck" use it */
|
|
|
|
|
if ((cmd & PCIM_CMD_BUSMASTEREN) == PCIM_CMD_BUSMASTEREN) {
|
|
|
|
|
ctlr->r_type1 = SYS_RES_IOPORT;
|
|
|
|
|
ctlr->r_rid1 = ATA_BMADDR_RID;
|
2004-03-17 17:50:55 +00:00
|
|
|
|
ctlr->r_res1 = bus_alloc_resource_any(dev, ctlr->r_type1, &ctlr->r_rid1,
|
|
|
|
|
RF_ACTIVE);
|
2001-03-06 21:43:46 +00:00
|
|
|
|
}
|
|
|
|
|
|
2006-01-18 09:14:55 +00:00
|
|
|
|
if (ctlr->chipinit(dev))
|
|
|
|
|
return ENXIO;
|
2003-06-07 15:19:16 +00:00
|
|
|
|
|
2003-03-29 13:37:09 +00:00
|
|
|
|
/* attach all channels on this controller */
|
2004-04-30 18:49:03 +00:00
|
|
|
|
for (unit = 0; unit < ctlr->channels; unit++) {
|
2009-02-16 19:10:07 +00:00
|
|
|
|
if ((ctlr->ichannels & (1 << unit)) == 0)
|
|
|
|
|
continue;
|
2009-02-14 21:54:44 +00:00
|
|
|
|
child = device_add_child(dev, "ata",
|
|
|
|
|
((unit == 0 || unit == 1) && ctlr->legacy) ?
|
|
|
|
|
unit : devclass_find_free_unit(ata_devclass, 2));
|
|
|
|
|
if (child == NULL)
|
|
|
|
|
device_printf(dev, "failed to add ata child device\n");
|
|
|
|
|
else
|
|
|
|
|
device_set_ivars(child, (void *)(intptr_t)unit);
|
2004-04-30 18:49:03 +00:00
|
|
|
|
}
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
bus_generic_attach(dev);
|
|
|
|
|
return 0;
|
2004-04-13 09:44:20 +00:00
|
|
|
|
}
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
int
|
2004-03-15 12:03:48 +00:00
|
|
|
|
ata_pci_detach(device_t dev)
|
|
|
|
|
{
|
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(dev);
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
device_t *children;
|
|
|
|
|
int nchildren, i;
|
2004-03-15 12:03:48 +00:00
|
|
|
|
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
/* detach & delete all children */
|
|
|
|
|
if (!device_get_children(dev, &children, &nchildren)) {
|
|
|
|
|
for (i = 0; i < nchildren; i++)
|
|
|
|
|
device_delete_child(dev, children[i]);
|
|
|
|
|
free(children, M_TEMP);
|
2004-03-15 12:03:48 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (ctlr->r_irq) {
|
|
|
|
|
bus_teardown_intr(dev, ctlr->r_irq, ctlr->handle);
|
2009-02-15 20:37:55 +00:00
|
|
|
|
bus_release_resource(dev, SYS_RES_IRQ, ctlr->r_irq_rid, ctlr->r_irq);
|
|
|
|
|
if (ctlr->r_irq_rid != ATA_IRQ_RID)
|
|
|
|
|
pci_release_msi(dev);
|
2004-03-15 12:03:48 +00:00
|
|
|
|
}
|
|
|
|
|
if (ctlr->r_res2)
|
|
|
|
|
bus_release_resource(dev, ctlr->r_type2, ctlr->r_rid2, ctlr->r_res2);
|
|
|
|
|
if (ctlr->r_res1)
|
|
|
|
|
bus_release_resource(dev, ctlr->r_type1, ctlr->r_rid1, ctlr->r_res1);
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
2008-09-18 12:12:34 +00:00
|
|
|
|
int
|
|
|
|
|
ata_pci_suspend(device_t dev)
|
|
|
|
|
{
|
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(dev);
|
|
|
|
|
int error = 0;
|
|
|
|
|
|
|
|
|
|
bus_generic_suspend(dev);
|
|
|
|
|
if (ctlr->suspend)
|
|
|
|
|
error = ctlr->suspend(dev);
|
|
|
|
|
return error;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
ata_pci_resume(device_t dev)
|
|
|
|
|
{
|
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(dev);
|
|
|
|
|
int error = 0;
|
|
|
|
|
|
|
|
|
|
if (ctlr->resume)
|
|
|
|
|
error = ctlr->resume(dev);
|
|
|
|
|
bus_generic_resume(dev);
|
|
|
|
|
return error;
|
|
|
|
|
}
|
|
|
|
|
|
Change the way in which AHCI+PATA combined controllers, such as JMicron
and Marvell handled. Instead of trying to attach two different drivers to
single device, wrapping each call, make one of them (atajmicron, atamarvell)
attach do device solely, but create child device for AHCI driver,
passing it all required resources. It is quite easy, as none of
resources are shared, except IRQ.
As result, it:
- makes drivers operation more independent and straitforward,
- allows to use new ahci(4) driver with such devices, adding support for
new features, such as PMP and NCQ, same time keeping legacy PATA support,
- will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
|
|
|
|
int
|
|
|
|
|
ata_pci_read_ivar(device_t dev, device_t child, int which, uintptr_t *result)
|
|
|
|
|
{
|
|
|
|
|
|
|
|
|
|
return (BUS_READ_IVAR(device_get_parent(dev), dev, which, result));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
ata_pci_write_ivar(device_t dev, device_t child, int which, uintptr_t value)
|
|
|
|
|
{
|
|
|
|
|
|
|
|
|
|
return (BUS_WRITE_IVAR(device_get_parent(dev), dev, which, value));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
uint32_t
|
|
|
|
|
ata_pci_read_config(device_t dev, device_t child, int reg, int width)
|
|
|
|
|
{
|
|
|
|
|
|
|
|
|
|
return (pci_read_config(dev, reg, width));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
ata_pci_write_config(device_t dev, device_t child, int reg,
|
|
|
|
|
uint32_t val, int width)
|
|
|
|
|
{
|
|
|
|
|
|
|
|
|
|
pci_write_config(dev, reg, val, width);
|
|
|
|
|
}
|
|
|
|
|
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
struct resource *
|
2001-03-06 21:43:46 +00:00
|
|
|
|
ata_pci_alloc_resource(device_t dev, device_t child, int type, int *rid,
|
|
|
|
|
u_long start, u_long end, u_long count, u_int flags)
|
|
|
|
|
{
|
Change the way in which AHCI+PATA combined controllers, such as JMicron
and Marvell handled. Instead of trying to attach two different drivers to
single device, wrapping each call, make one of them (atajmicron, atamarvell)
attach do device solely, but create child device for AHCI driver,
passing it all required resources. It is quite easy, as none of
resources are shared, except IRQ.
As result, it:
- makes drivers operation more independent and straitforward,
- allows to use new ahci(4) driver with such devices, adding support for
new features, such as PMP and NCQ, same time keeping legacy PATA support,
- will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
|
|
|
|
struct ata_pci_controller *controller = device_get_softc(dev);
|
|
|
|
|
struct resource *res = NULL;
|
|
|
|
|
|
|
|
|
|
if (device_get_devclass(child) == ata_devclass) {
|
|
|
|
|
int unit = ((struct ata_channel *)device_get_softc(child))->unit;
|
|
|
|
|
int myrid;
|
|
|
|
|
|
|
|
|
|
if (type == SYS_RES_IOPORT) {
|
|
|
|
|
switch (*rid) {
|
|
|
|
|
case ATA_IOADDR_RID:
|
|
|
|
|
if (controller->legacy) {
|
|
|
|
|
start = (unit ? ATA_SECONDARY : ATA_PRIMARY);
|
|
|
|
|
count = ATA_IOSIZE;
|
|
|
|
|
end = start + count - 1;
|
|
|
|
|
}
|
|
|
|
|
myrid = PCIR_BAR(0) + (unit << 3);
|
|
|
|
|
res = BUS_ALLOC_RESOURCE(device_get_parent(dev), dev,
|
|
|
|
|
SYS_RES_IOPORT, &myrid,
|
|
|
|
|
start, end, count, flags);
|
|
|
|
|
break;
|
|
|
|
|
case ATA_CTLADDR_RID:
|
|
|
|
|
if (controller->legacy) {
|
|
|
|
|
start = (unit ? ATA_SECONDARY : ATA_PRIMARY) +
|
|
|
|
|
ATA_CTLOFFSET;
|
|
|
|
|
count = ATA_CTLIOSIZE;
|
|
|
|
|
end = start + count - 1;
|
|
|
|
|
}
|
|
|
|
|
myrid = PCIR_BAR(1) + (unit << 3);
|
|
|
|
|
res = BUS_ALLOC_RESOURCE(device_get_parent(dev), dev,
|
|
|
|
|
SYS_RES_IOPORT, &myrid,
|
|
|
|
|
start, end, count, flags);
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
if (type == SYS_RES_IRQ && *rid == ATA_IRQ_RID) {
|
|
|
|
|
if (controller->legacy) {
|
|
|
|
|
int irq = (unit == 0 ? 14 : 15);
|
2003-02-20 20:02:32 +00:00
|
|
|
|
|
Change the way in which AHCI+PATA combined controllers, such as JMicron
and Marvell handled. Instead of trying to attach two different drivers to
single device, wrapping each call, make one of them (atajmicron, atamarvell)
attach do device solely, but create child device for AHCI driver,
passing it all required resources. It is quite easy, as none of
resources are shared, except IRQ.
As result, it:
- makes drivers operation more independent and straitforward,
- allows to use new ahci(4) driver with such devices, adding support for
new features, such as PMP and NCQ, same time keeping legacy PATA support,
- will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
|
|
|
|
res = BUS_ALLOC_RESOURCE(device_get_parent(dev), child,
|
|
|
|
|
SYS_RES_IRQ, rid, irq, irq, 1, flags);
|
|
|
|
|
} else
|
|
|
|
|
res = controller->r_irq;
|
|
|
|
|
}
|
|
|
|
|
} else {
|
|
|
|
|
if (type == SYS_RES_IRQ) {
|
|
|
|
|
if (*rid != ATA_IRQ_RID)
|
|
|
|
|
return (NULL);
|
|
|
|
|
res = controller->r_irq;
|
|
|
|
|
} else {
|
|
|
|
|
res = BUS_ALLOC_RESOURCE(device_get_parent(dev), dev,
|
|
|
|
|
type, rid, start, end, count, flags);
|
|
|
|
|
}
|
2001-03-06 21:43:46 +00:00
|
|
|
|
}
|
Change the way in which AHCI+PATA combined controllers, such as JMicron
and Marvell handled. Instead of trying to attach two different drivers to
single device, wrapping each call, make one of them (atajmicron, atamarvell)
attach do device solely, but create child device for AHCI driver,
passing it all required resources. It is quite easy, as none of
resources are shared, except IRQ.
As result, it:
- makes drivers operation more independent and straitforward,
- allows to use new ahci(4) driver with such devices, adding support for
new features, such as PMP and NCQ, same time keeping legacy PATA support,
- will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
|
|
|
|
return (res);
|
2001-03-06 21:43:46 +00:00
|
|
|
|
}
|
|
|
|
|
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
int
|
2001-03-06 21:43:46 +00:00
|
|
|
|
ata_pci_release_resource(device_t dev, device_t child, int type, int rid,
|
|
|
|
|
struct resource *r)
|
|
|
|
|
{
|
|
|
|
|
|
Change the way in which AHCI+PATA combined controllers, such as JMicron
and Marvell handled. Instead of trying to attach two different drivers to
single device, wrapping each call, make one of them (atajmicron, atamarvell)
attach do device solely, but create child device for AHCI driver,
passing it all required resources. It is quite easy, as none of
resources are shared, except IRQ.
As result, it:
- makes drivers operation more independent and straitforward,
- allows to use new ahci(4) driver with such devices, adding support for
new features, such as PMP and NCQ, same time keeping legacy PATA support,
- will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
|
|
|
|
if (device_get_devclass(child) == ata_devclass) {
|
|
|
|
|
struct ata_pci_controller *controller = device_get_softc(dev);
|
|
|
|
|
int unit = ((struct ata_channel *)device_get_softc(child))->unit;
|
|
|
|
|
|
|
|
|
|
if (type == SYS_RES_IOPORT) {
|
|
|
|
|
switch (rid) {
|
|
|
|
|
case ATA_IOADDR_RID:
|
|
|
|
|
return BUS_RELEASE_RESOURCE(device_get_parent(dev), dev,
|
|
|
|
|
SYS_RES_IOPORT,
|
|
|
|
|
PCIR_BAR(0) + (unit << 3), r);
|
|
|
|
|
case ATA_CTLADDR_RID:
|
|
|
|
|
return BUS_RELEASE_RESOURCE(device_get_parent(dev), dev,
|
|
|
|
|
SYS_RES_IOPORT,
|
|
|
|
|
PCIR_BAR(1) + (unit << 3), r);
|
|
|
|
|
default:
|
|
|
|
|
return ENOENT;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
if (type == SYS_RES_IRQ) {
|
|
|
|
|
if (rid != ATA_IRQ_RID)
|
|
|
|
|
return ENOENT;
|
|
|
|
|
if (controller->legacy) {
|
|
|
|
|
return BUS_RELEASE_RESOURCE(device_get_parent(dev), child,
|
|
|
|
|
SYS_RES_IRQ, rid, r);
|
|
|
|
|
} else
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
} else {
|
|
|
|
|
if (type == SYS_RES_IRQ) {
|
|
|
|
|
if (rid != ATA_IRQ_RID)
|
|
|
|
|
return (ENOENT);
|
|
|
|
|
return (0);
|
|
|
|
|
} else {
|
|
|
|
|
return (BUS_RELEASE_RESOURCE(device_get_parent(dev), child,
|
|
|
|
|
type, rid, r));
|
|
|
|
|
}
|
2004-06-15 11:02:09 +00:00
|
|
|
|
}
|
Change the way in which AHCI+PATA combined controllers, such as JMicron
and Marvell handled. Instead of trying to attach two different drivers to
single device, wrapping each call, make one of them (atajmicron, atamarvell)
attach do device solely, but create child device for AHCI driver,
passing it all required resources. It is quite easy, as none of
resources are shared, except IRQ.
As result, it:
- makes drivers operation more independent and straitforward,
- allows to use new ahci(4) driver with such devices, adding support for
new features, such as PMP and NCQ, same time keeping legacy PATA support,
- will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
|
|
|
|
return (EINVAL);
|
2001-03-06 21:43:46 +00:00
|
|
|
|
}
|
|
|
|
|
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
int
|
2001-03-06 21:43:46 +00:00
|
|
|
|
ata_pci_setup_intr(device_t dev, device_t child, struct resource *irq,
|
2007-02-23 12:19:07 +00:00
|
|
|
|
int flags, driver_filter_t *filter, driver_intr_t *function,
|
|
|
|
|
void *argument, void **cookiep)
|
2001-03-06 21:43:46 +00:00
|
|
|
|
{
|
2003-02-25 14:46:30 +00:00
|
|
|
|
struct ata_pci_controller *controller = device_get_softc(dev);
|
2003-02-20 20:02:32 +00:00
|
|
|
|
|
Change the way in which AHCI+PATA combined controllers, such as JMicron
and Marvell handled. Instead of trying to attach two different drivers to
single device, wrapping each call, make one of them (atajmicron, atamarvell)
attach do device solely, but create child device for AHCI driver,
passing it all required resources. It is quite easy, as none of
resources are shared, except IRQ.
As result, it:
- makes drivers operation more independent and straitforward,
- allows to use new ahci(4) driver with such devices, adding support for
new features, such as PMP and NCQ, same time keeping legacy PATA support,
- will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
|
|
|
|
if (controller->legacy) {
|
|
|
|
|
return BUS_SETUP_INTR(device_get_parent(dev), child, irq,
|
|
|
|
|
flags, filter, function, argument, cookiep);
|
|
|
|
|
} else {
|
|
|
|
|
struct ata_pci_controller *controller = device_get_softc(dev);
|
|
|
|
|
int unit;
|
|
|
|
|
|
|
|
|
|
if (filter != NULL) {
|
|
|
|
|
printf("ata-pci.c: we cannot use a filter here\n");
|
|
|
|
|
return (EINVAL);
|
|
|
|
|
}
|
|
|
|
|
if (device_get_devclass(child) == ata_devclass)
|
|
|
|
|
unit = ((struct ata_channel *)device_get_softc(child))->unit;
|
|
|
|
|
else
|
|
|
|
|
unit = ATA_PCI_MAX_CH - 1;
|
|
|
|
|
controller->interrupt[unit].function = function;
|
|
|
|
|
controller->interrupt[unit].argument = argument;
|
|
|
|
|
*cookiep = controller;
|
|
|
|
|
return 0;
|
2007-02-23 12:19:07 +00:00
|
|
|
|
}
|
2001-03-06 21:43:46 +00:00
|
|
|
|
}
|
|
|
|
|
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
int
|
2001-03-06 21:43:46 +00:00
|
|
|
|
ata_pci_teardown_intr(device_t dev, device_t child, struct resource *irq,
|
|
|
|
|
void *cookie)
|
|
|
|
|
{
|
2003-02-25 14:46:30 +00:00
|
|
|
|
struct ata_pci_controller *controller = device_get_softc(dev);
|
|
|
|
|
|
Change the way in which AHCI+PATA combined controllers, such as JMicron
and Marvell handled. Instead of trying to attach two different drivers to
single device, wrapping each call, make one of them (atajmicron, atamarvell)
attach do device solely, but create child device for AHCI driver,
passing it all required resources. It is quite easy, as none of
resources are shared, except IRQ.
As result, it:
- makes drivers operation more independent and straitforward,
- allows to use new ahci(4) driver with such devices, adding support for
new features, such as PMP and NCQ, same time keeping legacy PATA support,
- will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
|
|
|
|
if (controller->legacy) {
|
|
|
|
|
return BUS_TEARDOWN_INTR(device_get_parent(dev), child, irq, cookie);
|
|
|
|
|
} else {
|
|
|
|
|
struct ata_pci_controller *controller = device_get_softc(dev);
|
|
|
|
|
int unit;
|
|
|
|
|
|
|
|
|
|
if (device_get_devclass(child) == ata_devclass)
|
|
|
|
|
unit = ((struct ata_channel *)device_get_softc(child))->unit;
|
|
|
|
|
else
|
|
|
|
|
unit = ATA_PCI_MAX_CH - 1;
|
|
|
|
|
controller->interrupt[unit].function = NULL;
|
|
|
|
|
controller->interrupt[unit].argument = NULL;
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
2001-03-06 21:43:46 +00:00
|
|
|
|
}
|
2003-03-29 13:37:09 +00:00
|
|
|
|
|
2009-12-06 00:10:13 +00:00
|
|
|
|
int
|
|
|
|
|
ata_generic_setmode(device_t dev, int target, int mode)
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
{
|
|
|
|
|
|
2009-12-06 00:10:13 +00:00
|
|
|
|
return (min(mode, ATA_UDMA2));
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
}
|
|
|
|
|
|
2009-12-06 00:10:13 +00:00
|
|
|
|
int
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
ata_generic_chipinit(device_t dev)
|
|
|
|
|
{
|
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(dev);
|
|
|
|
|
|
|
|
|
|
if (ata_setup_interrupt(dev, ata_generic_intr))
|
|
|
|
|
return ENXIO;
|
|
|
|
|
ctlr->setmode = ata_generic_setmode;
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
2005-04-08 09:37:47 +00:00
|
|
|
|
int
|
2009-02-18 22:17:48 +00:00
|
|
|
|
ata_pci_ch_attach(device_t dev)
|
2003-03-29 13:37:09 +00:00
|
|
|
|
{
|
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(device_get_parent(dev));
|
2005-04-05 14:51:43 +00:00
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
2005-04-06 10:22:56 +00:00
|
|
|
|
struct resource *io = NULL, *ctlio = NULL;
|
2003-03-29 13:37:09 +00:00
|
|
|
|
int i, rid;
|
|
|
|
|
|
|
|
|
|
rid = ATA_IOADDR_RID;
|
2005-04-25 07:57:04 +00:00
|
|
|
|
if (!(io = bus_alloc_resource_any(dev, SYS_RES_IOPORT, &rid, RF_ACTIVE)))
|
2003-03-29 13:37:09 +00:00
|
|
|
|
return ENXIO;
|
|
|
|
|
|
2005-04-06 10:22:56 +00:00
|
|
|
|
rid = ATA_CTLADDR_RID;
|
2005-04-25 07:57:04 +00:00
|
|
|
|
if (!(ctlio = bus_alloc_resource_any(dev, SYS_RES_IOPORT, &rid,RF_ACTIVE))){
|
2003-03-29 13:37:09 +00:00
|
|
|
|
bus_release_resource(dev, SYS_RES_IOPORT, ATA_IOADDR_RID, io);
|
|
|
|
|
return ENXIO;
|
|
|
|
|
}
|
|
|
|
|
|
2009-02-18 22:17:48 +00:00
|
|
|
|
ata_pci_dmainit(dev);
|
|
|
|
|
|
2005-04-06 10:22:56 +00:00
|
|
|
|
for (i = ATA_DATA; i <= ATA_COMMAND; i ++) {
|
2003-03-29 13:37:09 +00:00
|
|
|
|
ch->r_io[i].res = io;
|
|
|
|
|
ch->r_io[i].offset = i;
|
|
|
|
|
}
|
2005-04-06 10:22:56 +00:00
|
|
|
|
ch->r_io[ATA_CONTROL].res = ctlio;
|
2008-12-16 16:04:40 +00:00
|
|
|
|
ch->r_io[ATA_CONTROL].offset = ctlr->legacy ? 0 : 2;
|
2003-04-07 14:12:12 +00:00
|
|
|
|
ch->r_io[ATA_IDX_ADDR].res = io;
|
2005-04-30 16:22:07 +00:00
|
|
|
|
ata_default_registers(dev);
|
2004-03-15 12:03:48 +00:00
|
|
|
|
if (ctlr->r_res1) {
|
2003-03-29 13:37:09 +00:00
|
|
|
|
for (i = ATA_BMCMD_PORT; i <= ATA_BMDTP_PORT; i++) {
|
2004-03-15 12:03:48 +00:00
|
|
|
|
ch->r_io[i].res = ctlr->r_res1;
|
2005-04-06 10:22:56 +00:00
|
|
|
|
ch->r_io[i].offset = (i - ATA_BMCMD_PORT) + (ch->unit*ATA_BMIOSIZE);
|
2003-03-29 13:37:09 +00:00
|
|
|
|
}
|
|
|
|
|
}
|
2004-04-13 09:44:20 +00:00
|
|
|
|
|
2006-01-18 09:14:55 +00:00
|
|
|
|
ata_pci_hw(dev);
|
2003-03-29 13:37:09 +00:00
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
2009-02-19 00:32:55 +00:00
|
|
|
|
int
|
|
|
|
|
ata_pci_ch_detach(device_t dev)
|
|
|
|
|
{
|
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
|
|
|
|
|
|
|
|
|
ata_pci_dmafini(dev);
|
|
|
|
|
|
|
|
|
|
bus_release_resource(dev, SYS_RES_IOPORT, ATA_CTLADDR_RID,
|
|
|
|
|
ch->r_io[ATA_CONTROL].res);
|
|
|
|
|
bus_release_resource(dev, SYS_RES_IOPORT, ATA_IOADDR_RID,
|
|
|
|
|
ch->r_io[ATA_IDX_ADDR].res);
|
|
|
|
|
|
|
|
|
|
return (0);
|
|
|
|
|
}
|
|
|
|
|
|
2008-04-12 17:21:22 +00:00
|
|
|
|
int
|
2006-01-18 09:14:55 +00:00
|
|
|
|
ata_pci_status(device_t dev)
|
|
|
|
|
{
|
2008-12-16 16:04:40 +00:00
|
|
|
|
struct ata_pci_controller *controller =
|
|
|
|
|
device_get_softc(device_get_parent(dev));
|
2006-01-18 09:14:55 +00:00
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
|
|
|
|
|
2008-12-16 16:04:40 +00:00
|
|
|
|
if ((dumping || !controller->legacy) &&
|
2008-04-10 13:05:05 +00:00
|
|
|
|
((ch->flags & ATA_ALWAYS_DMASTAT) ||
|
|
|
|
|
(ch->dma.flags & ATA_DMA_ACTIVE))) {
|
2006-01-18 09:14:55 +00:00
|
|
|
|
int bmstat = ATA_IDX_INB(ch, ATA_BMSTAT_PORT) & ATA_BMSTAT_MASK;
|
|
|
|
|
|
2009-12-05 13:40:51 +00:00
|
|
|
|
if ((bmstat & ATA_BMSTAT_INTERRUPT) == 0)
|
2006-01-18 09:14:55 +00:00
|
|
|
|
return 0;
|
|
|
|
|
ATA_IDX_OUTB(ch, ATA_BMSTAT_PORT, bmstat & ~ATA_BMSTAT_ERROR);
|
|
|
|
|
DELAY(1);
|
|
|
|
|
}
|
|
|
|
|
if (ATA_IDX_INB(ch, ATA_ALTSTAT) & ATA_S_BUSY) {
|
2006-01-18 13:10:17 +00:00
|
|
|
|
DELAY(100);
|
|
|
|
|
if (ATA_IDX_INB(ch, ATA_ALTSTAT) & ATA_S_BUSY)
|
|
|
|
|
return 0;
|
2006-01-18 09:14:55 +00:00
|
|
|
|
}
|
|
|
|
|
return 1;
|
|
|
|
|
}
|
|
|
|
|
|
2008-04-10 13:05:05 +00:00
|
|
|
|
void
|
|
|
|
|
ata_pci_hw(device_t dev)
|
|
|
|
|
{
|
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
|
|
|
|
|
|
|
|
|
ata_generic_hw(dev);
|
|
|
|
|
ch->hw.status = ata_pci_status;
|
|
|
|
|
}
|
|
|
|
|
|
2003-03-29 13:37:09 +00:00
|
|
|
|
static int
|
2008-04-10 13:05:05 +00:00
|
|
|
|
ata_pci_dmastart(struct ata_request *request)
|
2003-03-29 13:37:09 +00:00
|
|
|
|
{
|
2008-04-10 13:05:05 +00:00
|
|
|
|
struct ata_channel *ch = device_get_softc(request->parent);
|
|
|
|
|
|
|
|
|
|
ATA_DEBUG_RQ(request, "dmastart");
|
2005-04-30 16:22:07 +00:00
|
|
|
|
|
2003-03-29 13:37:09 +00:00
|
|
|
|
ATA_IDX_OUTB(ch, ATA_BMSTAT_PORT, (ATA_IDX_INB(ch, ATA_BMSTAT_PORT) |
|
2003-04-07 14:12:12 +00:00
|
|
|
|
(ATA_BMSTAT_INTERRUPT | ATA_BMSTAT_ERROR)));
|
2008-04-17 12:29:35 +00:00
|
|
|
|
ATA_IDX_OUTL(ch, ATA_BMDTP_PORT, request->dma->sg_bus);
|
2008-04-10 13:05:05 +00:00
|
|
|
|
ch->dma.flags |= ATA_DMA_ACTIVE;
|
2003-08-24 09:22:26 +00:00
|
|
|
|
ATA_IDX_OUTB(ch, ATA_BMCMD_PORT,
|
2004-05-10 20:23:25 +00:00
|
|
|
|
(ATA_IDX_INB(ch, ATA_BMCMD_PORT) & ~ATA_BMCMD_WRITE_READ) |
|
2008-04-10 13:05:05 +00:00
|
|
|
|
((request->flags & ATA_R_READ) ? ATA_BMCMD_WRITE_READ : 0)|
|
2003-10-21 19:20:37 +00:00
|
|
|
|
ATA_BMCMD_START_STOP);
|
2003-03-29 13:37:09 +00:00
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int
|
2008-04-10 13:05:05 +00:00
|
|
|
|
ata_pci_dmastop(struct ata_request *request)
|
2003-03-29 13:37:09 +00:00
|
|
|
|
{
|
2008-04-10 13:05:05 +00:00
|
|
|
|
struct ata_channel *ch = device_get_softc(request->parent);
|
2003-03-29 13:37:09 +00:00
|
|
|
|
int error;
|
|
|
|
|
|
2008-04-10 13:05:05 +00:00
|
|
|
|
ATA_DEBUG_RQ(request, "dmastop");
|
|
|
|
|
|
2003-03-29 13:37:09 +00:00
|
|
|
|
ATA_IDX_OUTB(ch, ATA_BMCMD_PORT,
|
2003-04-07 14:12:12 +00:00
|
|
|
|
ATA_IDX_INB(ch, ATA_BMCMD_PORT) & ~ATA_BMCMD_START_STOP);
|
2008-04-10 13:05:05 +00:00
|
|
|
|
ch->dma.flags &= ~ATA_DMA_ACTIVE;
|
2004-09-26 11:42:42 +00:00
|
|
|
|
error = ATA_IDX_INB(ch, ATA_BMSTAT_PORT) & ATA_BMSTAT_MASK;
|
2003-03-29 13:37:09 +00:00
|
|
|
|
ATA_IDX_OUTB(ch, ATA_BMSTAT_PORT, ATA_BMSTAT_INTERRUPT | ATA_BMSTAT_ERROR);
|
2003-08-24 09:22:26 +00:00
|
|
|
|
return error;
|
2003-03-29 13:37:09 +00:00
|
|
|
|
}
|
|
|
|
|
|
2005-05-03 07:55:07 +00:00
|
|
|
|
static void
|
|
|
|
|
ata_pci_dmareset(device_t dev)
|
|
|
|
|
{
|
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
2008-04-10 13:05:05 +00:00
|
|
|
|
struct ata_request *request;
|
2005-05-03 07:55:07 +00:00
|
|
|
|
|
|
|
|
|
ATA_IDX_OUTB(ch, ATA_BMCMD_PORT,
|
|
|
|
|
ATA_IDX_INB(ch, ATA_BMCMD_PORT) & ~ATA_BMCMD_START_STOP);
|
2008-04-10 13:05:05 +00:00
|
|
|
|
ch->dma.flags &= ~ATA_DMA_ACTIVE;
|
2005-05-03 07:55:07 +00:00
|
|
|
|
ATA_IDX_OUTB(ch, ATA_BMSTAT_PORT, ATA_BMSTAT_INTERRUPT | ATA_BMSTAT_ERROR);
|
2008-04-10 13:05:05 +00:00
|
|
|
|
if ((request = ch->running)) {
|
2009-10-31 13:24:14 +00:00
|
|
|
|
device_printf(dev, "DMA reset calling unload\n");
|
2008-04-10 13:05:05 +00:00
|
|
|
|
ch->dma.unload(request);
|
|
|
|
|
}
|
2005-05-03 07:55:07 +00:00
|
|
|
|
}
|
|
|
|
|
|
2006-02-16 17:09:24 +00:00
|
|
|
|
void
|
2005-04-30 16:22:07 +00:00
|
|
|
|
ata_pci_dmainit(device_t dev)
|
2003-03-29 13:37:09 +00:00
|
|
|
|
{
|
2005-04-30 16:22:07 +00:00
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
|
|
|
|
|
|
|
|
|
ata_dmainit(dev);
|
2008-04-10 13:05:05 +00:00
|
|
|
|
ch->dma.start = ata_pci_dmastart;
|
|
|
|
|
ch->dma.stop = ata_pci_dmastop;
|
|
|
|
|
ch->dma.reset = ata_pci_dmareset;
|
2003-03-29 13:37:09 +00:00
|
|
|
|
}
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
2009-02-19 00:32:55 +00:00
|
|
|
|
void
|
|
|
|
|
ata_pci_dmafini(device_t dev)
|
|
|
|
|
{
|
|
|
|
|
|
|
|
|
|
ata_dmafini(dev);
|
|
|
|
|
}
|
2007-11-18 14:44:52 +00:00
|
|
|
|
|
2001-03-06 21:43:46 +00:00
|
|
|
|
static device_method_t ata_pci_methods[] = {
|
|
|
|
|
/* device interface */
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
DEVMETHOD(device_probe, ata_pci_probe),
|
|
|
|
|
DEVMETHOD(device_attach, ata_pci_attach),
|
|
|
|
|
DEVMETHOD(device_detach, ata_pci_detach),
|
2008-09-18 12:12:34 +00:00
|
|
|
|
DEVMETHOD(device_suspend, ata_pci_suspend),
|
|
|
|
|
DEVMETHOD(device_resume, ata_pci_resume),
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
DEVMETHOD(device_shutdown, bus_generic_shutdown),
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
|
|
|
|
/* bus methods */
|
Change the way in which AHCI+PATA combined controllers, such as JMicron
and Marvell handled. Instead of trying to attach two different drivers to
single device, wrapping each call, make one of them (atajmicron, atamarvell)
attach do device solely, but create child device for AHCI driver,
passing it all required resources. It is quite easy, as none of
resources are shared, except IRQ.
As result, it:
- makes drivers operation more independent and straitforward,
- allows to use new ahci(4) driver with such devices, adding support for
new features, such as PMP and NCQ, same time keeping legacy PATA support,
- will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
|
|
|
|
DEVMETHOD(bus_read_ivar, ata_pci_read_ivar),
|
|
|
|
|
DEVMETHOD(bus_write_ivar, ata_pci_write_ivar),
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
DEVMETHOD(bus_alloc_resource, ata_pci_alloc_resource),
|
|
|
|
|
DEVMETHOD(bus_release_resource, ata_pci_release_resource),
|
|
|
|
|
DEVMETHOD(bus_activate_resource, bus_generic_activate_resource),
|
|
|
|
|
DEVMETHOD(bus_deactivate_resource, bus_generic_deactivate_resource),
|
|
|
|
|
DEVMETHOD(bus_setup_intr, ata_pci_setup_intr),
|
|
|
|
|
DEVMETHOD(bus_teardown_intr, ata_pci_teardown_intr),
|
Change the way in which AHCI+PATA combined controllers, such as JMicron
and Marvell handled. Instead of trying to attach two different drivers to
single device, wrapping each call, make one of them (atajmicron, atamarvell)
attach do device solely, but create child device for AHCI driver,
passing it all required resources. It is quite easy, as none of
resources are shared, except IRQ.
As result, it:
- makes drivers operation more independent and straitforward,
- allows to use new ahci(4) driver with such devices, adding support for
new features, such as PMP and NCQ, same time keeping legacy PATA support,
- will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
|
|
|
|
DEVMETHOD(pci_read_config, ata_pci_read_config),
|
|
|
|
|
DEVMETHOD(pci_write_config, ata_pci_write_config),
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
|
2001-03-06 21:43:46 +00:00
|
|
|
|
{ 0, 0 }
|
|
|
|
|
};
|
|
|
|
|
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
devclass_t ata_pci_devclass;
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
|
2001-03-06 21:43:46 +00:00
|
|
|
|
static driver_t ata_pci_driver = {
|
|
|
|
|
"atapci",
|
|
|
|
|
ata_pci_methods,
|
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
|
|
|
|
sizeof(struct ata_pci_controller),
|
2001-03-06 21:43:46 +00:00
|
|
|
|
};
|
|
|
|
|
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
DRIVER_MODULE(atapci, pci, ata_pci_driver, ata_pci_devclass, 0, 0);
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
MODULE_VERSION(atapci, 1);
|
|
|
|
|
MODULE_DEPEND(atapci, ata, 1, 1, 1);
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
|
|
|
|
static int
|
2005-03-31 15:05:40 +00:00
|
|
|
|
ata_pcichannel_probe(device_t dev)
|
2001-03-06 21:43:46 +00:00
|
|
|
|
{
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
char buffer[32];
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
Change the way in which AHCI+PATA combined controllers, such as JMicron
and Marvell handled. Instead of trying to attach two different drivers to
single device, wrapping each call, make one of them (atajmicron, atamarvell)
attach do device solely, but create child device for AHCI driver,
passing it all required resources. It is quite easy, as none of
resources are shared, except IRQ.
As result, it:
- makes drivers operation more independent and straitforward,
- allows to use new ahci(4) driver with such devices, adding support for
new features, such as PMP and NCQ, same time keeping legacy PATA support,
- will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
|
|
|
|
if ((intptr_t)device_get_ivars(dev) < 0)
|
|
|
|
|
return (ENXIO);
|
2009-02-14 21:54:44 +00:00
|
|
|
|
sprintf(buffer, "ATA channel %d", (int)(intptr_t)device_get_ivars(dev));
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
device_set_desc_copy(dev, buffer);
|
|
|
|
|
|
2004-08-12 08:20:36 +00:00
|
|
|
|
return ata_probe(dev);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int
|
2005-03-31 15:05:40 +00:00
|
|
|
|
ata_pcichannel_attach(device_t dev)
|
2004-08-12 08:20:36 +00:00
|
|
|
|
{
|
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(device_get_parent(dev));
|
2009-02-14 21:54:44 +00:00
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
2004-08-12 08:20:36 +00:00
|
|
|
|
int error;
|
|
|
|
|
|
2009-02-19 12:47:24 +00:00
|
|
|
|
if (ch->attached)
|
|
|
|
|
return (0);
|
|
|
|
|
ch->attached = 1;
|
2009-02-14 21:54:44 +00:00
|
|
|
|
|
|
|
|
|
ch->unit = (intptr_t)device_get_ivars(dev);
|
|
|
|
|
|
2009-04-29 21:17:18 +00:00
|
|
|
|
resource_int_value(device_get_name(dev),
|
|
|
|
|
device_get_unit(dev), "pm_level", &ch->pm_level);
|
|
|
|
|
|
2009-02-18 22:17:48 +00:00
|
|
|
|
if ((error = ctlr->ch_attach(dev)))
|
2004-06-15 11:02:09 +00:00
|
|
|
|
return error;
|
|
|
|
|
|
2008-06-11 06:44:58 +00:00
|
|
|
|
return ata_attach(dev);
|
2004-08-12 08:20:36 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int
|
2005-03-31 15:05:40 +00:00
|
|
|
|
ata_pcichannel_detach(device_t dev)
|
2004-08-12 08:20:36 +00:00
|
|
|
|
{
|
2009-02-18 22:17:48 +00:00
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(device_get_parent(dev));
|
2009-02-19 12:47:24 +00:00
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
2004-08-12 08:20:36 +00:00
|
|
|
|
int error;
|
|
|
|
|
|
2009-02-19 12:47:24 +00:00
|
|
|
|
if (!ch->attached)
|
|
|
|
|
return (0);
|
|
|
|
|
ch->attached = 0;
|
|
|
|
|
|
2004-08-12 08:20:36 +00:00
|
|
|
|
if ((error = ata_detach(dev)))
|
|
|
|
|
return error;
|
|
|
|
|
|
2009-02-18 22:17:48 +00:00
|
|
|
|
if (ctlr->ch_detach)
|
|
|
|
|
return (ctlr->ch_detach(dev));
|
2005-04-05 14:51:43 +00:00
|
|
|
|
|
2009-02-18 22:17:48 +00:00
|
|
|
|
return (0);
|
2001-03-06 21:43:46 +00:00
|
|
|
|
}
|
2009-03-09 20:48:57 +00:00
|
|
|
|
static int
|
|
|
|
|
ata_pcichannel_suspend(device_t dev)
|
|
|
|
|
{
|
2009-03-30 22:18:38 +00:00
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(device_get_parent(dev));
|
2009-03-09 20:48:57 +00:00
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
2009-03-30 22:18:38 +00:00
|
|
|
|
int error;
|
2009-03-09 20:48:57 +00:00
|
|
|
|
|
|
|
|
|
if (!ch->attached)
|
|
|
|
|
return (0);
|
|
|
|
|
|
2009-03-30 22:18:38 +00:00
|
|
|
|
if ((error = ata_suspend(dev)))
|
|
|
|
|
return (error);
|
|
|
|
|
|
|
|
|
|
if (ctlr->ch_suspend != NULL && (error = ctlr->ch_suspend(dev)))
|
|
|
|
|
return (error);
|
|
|
|
|
|
|
|
|
|
return (0);
|
2009-03-09 20:48:57 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
ata_pcichannel_resume(device_t dev)
|
|
|
|
|
{
|
2009-03-30 22:18:38 +00:00
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(device_get_parent(dev));
|
2009-03-09 20:48:57 +00:00
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
2009-03-30 22:18:38 +00:00
|
|
|
|
int error;
|
2009-03-09 20:48:57 +00:00
|
|
|
|
|
|
|
|
|
if (!ch->attached)
|
|
|
|
|
return (0);
|
|
|
|
|
|
2009-03-30 22:18:38 +00:00
|
|
|
|
if (ctlr->ch_resume != NULL && (error = ctlr->ch_resume(dev)))
|
|
|
|
|
return (error);
|
|
|
|
|
|
2009-03-09 20:48:57 +00:00
|
|
|
|
return ata_resume(dev);
|
|
|
|
|
}
|
|
|
|
|
|
2001-03-06 21:43:46 +00:00
|
|
|
|
|
2005-03-31 15:05:40 +00:00
|
|
|
|
static int
|
|
|
|
|
ata_pcichannel_locking(device_t dev, int mode)
|
|
|
|
|
{
|
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(device_get_parent(dev));
|
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
|
|
|
|
|
|
|
|
|
if (ctlr->locking)
|
2005-04-30 16:22:07 +00:00
|
|
|
|
return ctlr->locking(dev, mode);
|
2005-03-31 15:05:40 +00:00
|
|
|
|
else
|
|
|
|
|
return ch->unit;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
ata_pcichannel_reset(device_t dev)
|
|
|
|
|
{
|
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(device_get_parent(dev));
|
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
|
|
|
|
|
2005-05-03 07:55:07 +00:00
|
|
|
|
/* if DMA engine present reset it */
|
2008-04-10 13:05:05 +00:00
|
|
|
|
if (ch->dma.reset)
|
|
|
|
|
ch->dma.reset(dev);
|
2005-04-08 09:37:47 +00:00
|
|
|
|
|
|
|
|
|
/* reset the controller HW */
|
2005-03-31 15:05:40 +00:00
|
|
|
|
if (ctlr->reset)
|
2005-04-30 16:22:07 +00:00
|
|
|
|
ctlr->reset(dev);
|
2005-04-28 22:08:08 +00:00
|
|
|
|
else
|
2005-04-30 16:22:07 +00:00
|
|
|
|
ata_generic_reset(dev);
|
2005-03-31 15:05:40 +00:00
|
|
|
|
}
|
|
|
|
|
|
2009-12-06 00:10:13 +00:00
|
|
|
|
static int
|
|
|
|
|
ata_pcichannel_setmode(device_t dev, int target, int mode)
|
|
|
|
|
{
|
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(device_get_parent(dev));
|
|
|
|
|
|
|
|
|
|
if (ctlr->setmode)
|
|
|
|
|
return (ctlr->setmode(dev, target, mode));
|
|
|
|
|
else
|
|
|
|
|
return (ata_generic_setmode(dev, target, mode));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
ata_pcichannel_getrev(device_t dev, int target)
|
2005-03-31 15:05:40 +00:00
|
|
|
|
{
|
2009-12-06 00:10:13 +00:00
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(device_get_parent(dev));
|
2010-02-22 10:45:40 +00:00
|
|
|
|
struct ata_channel *ch = device_get_softc(dev);
|
|
|
|
|
|
|
|
|
|
if (ch->flags & ATA_SATA) {
|
|
|
|
|
if (ctlr->getrev)
|
|
|
|
|
return (ctlr->getrev(dev, target));
|
|
|
|
|
else
|
|
|
|
|
return (0xff);
|
|
|
|
|
} else
|
2009-12-06 00:10:13 +00:00
|
|
|
|
return (0);
|
2005-03-31 15:05:40 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static device_method_t ata_pcichannel_methods[] = {
|
2001-03-06 21:43:46 +00:00
|
|
|
|
/* device interface */
|
2005-03-31 15:05:40 +00:00
|
|
|
|
DEVMETHOD(device_probe, ata_pcichannel_probe),
|
|
|
|
|
DEVMETHOD(device_attach, ata_pcichannel_attach),
|
|
|
|
|
DEVMETHOD(device_detach, ata_pcichannel_detach),
|
This is the much rumoured ATA mkIII update that I've been working on.
o ATA is now fully newbus'd and split into modules.
This means that on a modern system you just load "atapci and ata"
to get the base support, and then one or more of the device
subdrivers "atadisk atapicd atapifd atapist ataraid".
All can be loaded/unloaded anytime, but for obvious reasons you
dont want to unload atadisk when you have mounted filesystems.
o The device identify part of the probe has been rewritten to fix
the problems with odd devices the old had, and to try to remove
so of the long delays some HW could provoke. Also probing is done
without the need for interrupts, making earlier probing possible.
o SATA devices can be hot inserted/removed and devices will be created/
removed in /dev accordingly.
NOTE: only supported on controllers that has this feature:
Promise and Silicon Image for now.
On other controllers the usual atacontrol detach/attach dance is
still needed.
o Support for "atomic" composite ATA requests used for RAID.
o ATA RAID support has been rewritten and and now supports these
metadata formats:
"Adaptec HostRAID"
"Highpoint V2 RocketRAID"
"Highpoint V3 RocketRAID"
"Intel MatrixRAID"
"Integrated Technology Express"
"LSILogic V2 MegaRAID"
"LSILogic V3 MegaRAID"
"Promise FastTrak"
"Silicon Image Medley"
"FreeBSD PseudoRAID"
o Update the ioctl API to match new RAID levels etc.
o Update atacontrol to know about the new RAID levels etc
NOTE: you need to recompile atacontrol with the new sys/ata.h,
make world will take care of that.
NOTE2: that rebuild is done differently from the old system as
the rebuild is now done piggybacked on read requests to the
array, so atacontrol simply starts a background "dd" to rebuild
the array.
o The reinit code has been worked over to be much more robust.
o The timeout code has been overhauled for races.
o Support of new chipsets.
o Lots of fixes for bugs found while doing the modulerization and
reviewing the old code.
Missing or changed features from current ATA:
o atapi-cd no longer has support for ATAPI changers. Todays its
much cheaper and alot faster to copy those CD images to disk
and serve them from there. Besides they dont seem to be made
anymore, maybe for that exact reason.
o ATA RAID can only read metadata from all the above metadata formats,
not write all of them (Promise and Highpoint V2 so far). This means
that arrays can be picked up from the BIOS, but they cannot be
created from FreeBSD. There is more to it than just the missing
write metadata support, those formats are not unique to a given
controller like Promise and Highpoint formats, instead they exist
for several types, and even worse, some controllers can have
different formats and its impossible to tell which one.
The outcome is that we cannot reliably create the metadata of those
formats and be sure the controller BIOS will understand it.
However write support is needed to update/fail/rebuild the arrays
properly so it sits fairly high on the TODO list.
o So far atapicam is not supported with these changes. When/if this
will change is up to the maintainer of atapi-cam so go there for
questions.
HW donated by: Webveveriet AS
HW donated by: Frode Nordahl
HW donated by: Yahoo!
HW donated by: Sentex
Patience by: Vife and my boys (and even the cats)
2005-03-30 12:03:40 +00:00
|
|
|
|
DEVMETHOD(device_shutdown, bus_generic_shutdown),
|
2009-03-09 20:48:57 +00:00
|
|
|
|
DEVMETHOD(device_suspend, ata_pcichannel_suspend),
|
|
|
|
|
DEVMETHOD(device_resume, ata_pcichannel_resume),
|
2005-03-31 15:05:40 +00:00
|
|
|
|
|
|
|
|
|
/* ATA methods */
|
|
|
|
|
DEVMETHOD(ata_setmode, ata_pcichannel_setmode),
|
2009-12-06 00:10:13 +00:00
|
|
|
|
DEVMETHOD(ata_getrev, ata_pcichannel_getrev),
|
2005-03-31 15:05:40 +00:00
|
|
|
|
DEVMETHOD(ata_locking, ata_pcichannel_locking),
|
|
|
|
|
DEVMETHOD(ata_reset, ata_pcichannel_reset),
|
|
|
|
|
|
2001-03-06 21:43:46 +00:00
|
|
|
|
{ 0, 0 }
|
|
|
|
|
};
|
|
|
|
|
|
2005-03-31 15:05:40 +00:00
|
|
|
|
driver_t ata_pcichannel_driver = {
|
2001-03-06 21:43:46 +00:00
|
|
|
|
"ata",
|
2005-03-31 15:05:40 +00:00
|
|
|
|
ata_pcichannel_methods,
|
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
|
|
|
|
sizeof(struct ata_channel),
|
2001-03-06 21:43:46 +00:00
|
|
|
|
};
|
|
|
|
|
|
2005-03-31 15:05:40 +00:00
|
|
|
|
DRIVER_MODULE(ata, atapci, ata_pcichannel_driver, ata_devclass, 0, 0);
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* misc support fucntions
|
|
|
|
|
*/
|
|
|
|
|
int
|
|
|
|
|
ata_legacy(device_t dev)
|
|
|
|
|
{
|
|
|
|
|
return (((pci_read_config(dev, PCIR_PROGIF, 1)&PCIP_STORAGE_IDE_MASTERDEV)&&
|
|
|
|
|
((pci_read_config(dev, PCIR_PROGIF, 1) &
|
|
|
|
|
(PCIP_STORAGE_IDE_MODEPRIM | PCIP_STORAGE_IDE_MODESEC)) !=
|
|
|
|
|
(PCIP_STORAGE_IDE_MODEPRIM | PCIP_STORAGE_IDE_MODESEC))) ||
|
|
|
|
|
(!pci_read_config(dev, PCIR_BAR(0), 4) &&
|
|
|
|
|
!pci_read_config(dev, PCIR_BAR(1), 4) &&
|
|
|
|
|
!pci_read_config(dev, PCIR_BAR(2), 4) &&
|
|
|
|
|
!pci_read_config(dev, PCIR_BAR(3), 4) &&
|
|
|
|
|
!pci_read_config(dev, PCIR_BAR(5), 4)));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
ata_generic_intr(void *data)
|
|
|
|
|
{
|
|
|
|
|
struct ata_pci_controller *ctlr = data;
|
|
|
|
|
struct ata_channel *ch;
|
|
|
|
|
int unit;
|
|
|
|
|
|
Change the way in which AHCI+PATA combined controllers, such as JMicron
and Marvell handled. Instead of trying to attach two different drivers to
single device, wrapping each call, make one of them (atajmicron, atamarvell)
attach do device solely, but create child device for AHCI driver,
passing it all required resources. It is quite easy, as none of
resources are shared, except IRQ.
As result, it:
- makes drivers operation more independent and straitforward,
- allows to use new ahci(4) driver with such devices, adding support for
new features, such as PMP and NCQ, same time keeping legacy PATA support,
- will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
|
|
|
|
for (unit = 0; unit < ATA_PCI_MAX_CH; unit++) {
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
if ((ch = ctlr->interrupt[unit].argument))
|
|
|
|
|
ctlr->interrupt[unit].function(ch);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
ata_setup_interrupt(device_t dev, void *intr_func)
|
|
|
|
|
{
|
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(dev);
|
2009-02-15 20:37:55 +00:00
|
|
|
|
int i, msi = 0;
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
|
2008-12-16 16:04:40 +00:00
|
|
|
|
if (!ctlr->legacy) {
|
2009-02-15 20:37:55 +00:00
|
|
|
|
if (resource_int_value(device_get_name(dev),
|
|
|
|
|
device_get_unit(dev), "msi", &i) == 0 && i != 0)
|
|
|
|
|
msi = 1;
|
|
|
|
|
if (msi && pci_msi_count(dev) > 0 && pci_alloc_msi(dev, &msi) == 0) {
|
|
|
|
|
ctlr->r_irq_rid = 0x1;
|
|
|
|
|
} else {
|
|
|
|
|
ctlr->r_irq_rid = ATA_IRQ_RID;
|
|
|
|
|
}
|
|
|
|
|
if (!(ctlr->r_irq = bus_alloc_resource_any(dev, SYS_RES_IRQ,
|
|
|
|
|
&ctlr->r_irq_rid, RF_SHAREABLE | RF_ACTIVE))) {
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
device_printf(dev, "unable to map interrupt\n");
|
|
|
|
|
return ENXIO;
|
|
|
|
|
}
|
|
|
|
|
if ((bus_setup_intr(dev, ctlr->r_irq, ATA_INTR_FLAGS, NULL,
|
|
|
|
|
intr_func, ctlr, &ctlr->handle))) {
|
|
|
|
|
/* SOS XXX release r_irq */
|
|
|
|
|
device_printf(dev, "unable to setup interrupt\n");
|
|
|
|
|
return ENXIO;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
ata_set_desc(device_t dev)
|
|
|
|
|
{
|
|
|
|
|
struct ata_pci_controller *ctlr = device_get_softc(dev);
|
|
|
|
|
char buffer[128];
|
|
|
|
|
|
|
|
|
|
sprintf(buffer, "%s %s %s controller",
|
|
|
|
|
ata_pcivendor2str(dev), ctlr->chip->text,
|
|
|
|
|
ata_mode2str(ctlr->chip->max_dma));
|
|
|
|
|
device_set_desc_copy(dev, buffer);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
struct ata_chip_id *
|
|
|
|
|
ata_match_chip(device_t dev, struct ata_chip_id *index)
|
|
|
|
|
{
|
2009-04-27 18:10:42 +00:00
|
|
|
|
uint32_t devid;
|
|
|
|
|
uint8_t revid;
|
|
|
|
|
|
|
|
|
|
devid = pci_get_devid(dev);
|
|
|
|
|
revid = pci_get_revid(dev);
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
while (index->chipid != 0) {
|
2009-04-27 18:10:42 +00:00
|
|
|
|
if (devid == index->chipid && revid >= index->chiprev)
|
|
|
|
|
return (index);
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
index++;
|
|
|
|
|
}
|
2009-04-27 18:10:42 +00:00
|
|
|
|
return (NULL);
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
struct ata_chip_id *
|
|
|
|
|
ata_find_chip(device_t dev, struct ata_chip_id *index, int slot)
|
|
|
|
|
{
|
2009-04-27 19:39:18 +00:00
|
|
|
|
struct ata_chip_id *idx;
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
device_t *children;
|
|
|
|
|
int nchildren, i;
|
2009-04-27 19:39:18 +00:00
|
|
|
|
uint8_t s;
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
|
|
|
|
|
if (device_get_children(device_get_parent(dev), &children, &nchildren))
|
2009-04-27 19:39:18 +00:00
|
|
|
|
return (NULL);
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
|
2009-04-27 19:39:18 +00:00
|
|
|
|
for (i = 0; i < nchildren; i++) {
|
|
|
|
|
s = pci_get_slot(children[i]);
|
|
|
|
|
if ((slot >= 0 && s == slot) || (slot < 0 && s <= -slot)) {
|
|
|
|
|
idx = ata_match_chip(children[i], index);
|
|
|
|
|
if (idx != NULL) {
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
free(children, M_TEMP);
|
2009-04-27 19:39:18 +00:00
|
|
|
|
return (idx);
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
free(children, M_TEMP);
|
2009-04-27 19:39:18 +00:00
|
|
|
|
return (NULL);
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
char *
|
|
|
|
|
ata_pcivendor2str(device_t dev)
|
|
|
|
|
{
|
|
|
|
|
switch (pci_get_vendor(dev)) {
|
|
|
|
|
case ATA_ACARD_ID: return "Acard";
|
|
|
|
|
case ATA_ACER_LABS_ID: return "AcerLabs";
|
|
|
|
|
case ATA_AMD_ID: return "AMD";
|
|
|
|
|
case ATA_ADAPTEC_ID: return "Adaptec";
|
|
|
|
|
case ATA_ATI_ID: return "ATI";
|
|
|
|
|
case ATA_CYRIX_ID: return "Cyrix";
|
|
|
|
|
case ATA_CYPRESS_ID: return "Cypress";
|
|
|
|
|
case ATA_HIGHPOINT_ID: return "HighPoint";
|
|
|
|
|
case ATA_INTEL_ID: return "Intel";
|
|
|
|
|
case ATA_ITE_ID: return "ITE";
|
|
|
|
|
case ATA_JMICRON_ID: return "JMicron";
|
|
|
|
|
case ATA_MARVELL_ID: return "Marvell";
|
2010-01-26 15:25:24 +00:00
|
|
|
|
case ATA_MARVELL2_ID: return "Marvell";
|
This is the roumored ATA modulerisation works, and it needs a little explanation.
If you just config KERNEL as usual there should be no apparent changes, you'll get all chipset support code compiled in.
However there is now a way to only compile in code for chipsets needed on a pr vendor basis. ATA now has the following "device" entries:
atacore: ATA core functionality, always needed for any ATA setup
atacard: CARDBUS support
atacbus: PC98 cbus support
ataisa: ISA bus support
atapci: PCI bus support only generic chipset support.
ataahci: AHCI support, also pulled in by some vendor modules.
ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, atamarvell, atamicron, atanational, atanetcell, atanvidia, atapromise, ataserverworks, atasiliconimage, atasis, atavia; Vendor support, ie atavia for VIA chipsets
atadisk: ATA disk driver
ataraid: ATA softraid driver
atapicd: ATAPI cd/dvd driver
atapifd: ATAPI floppy/flashdisk driver
atapist: ATAPI tape driver
atausb: ATA<>USB bridge
atapicam: ATA<>CAM bridge
This makes it possible to config a kernel with just VIA chipset support by having the following ATA lines in the kernel config file:
device atacore
device atapci
device atavia
And then you need the atadisk, atapicd etc lines in there just as usual.
If you use ATA as modules loaded at boot there is few changes except the rename of the "ata" module to "atacore", things looks just as usual.
However under atapci you now have a whole bunch of vendor specific drivers, that you can kldload individually depending on you needs. Drivers have the same names as used in the kernel config explained above.
2008-10-09 12:56:57 +00:00
|
|
|
|
case ATA_NATIONAL_ID: return "National";
|
|
|
|
|
case ATA_NETCELL_ID: return "Netcell";
|
|
|
|
|
case ATA_NVIDIA_ID: return "nVidia";
|
|
|
|
|
case ATA_PROMISE_ID: return "Promise";
|
|
|
|
|
case ATA_SERVERWORKS_ID: return "ServerWorks";
|
|
|
|
|
case ATA_SILICON_IMAGE_ID: return "SiI";
|
|
|
|
|
case ATA_SIS_ID: return "SiS";
|
|
|
|
|
case ATA_VIA_ID: return "VIA";
|
|
|
|
|
case ATA_CENATEK_ID: return "Cenatek";
|
|
|
|
|
case ATA_MICRON_ID: return "Micron";
|
|
|
|
|
default: return "Generic";
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
ata_mode2idx(int mode)
|
|
|
|
|
{
|
|
|
|
|
if ((mode & ATA_DMA_MASK) == ATA_UDMA0)
|
|
|
|
|
return (mode & ATA_MODE_MASK) + 8;
|
|
|
|
|
if ((mode & ATA_DMA_MASK) == ATA_WDMA0)
|
|
|
|
|
return (mode & ATA_MODE_MASK) + 5;
|
|
|
|
|
return (mode & ATA_MODE_MASK) - ATA_PIO0;
|
|
|
|
|
}
|
|
|
|
|
|