powerpc.git
16 years ago[MIPS] IRQ Affinity Support for SMTC on Malta Platform
Kevin D. Kissell [Fri, 3 Aug 2007 17:38:03 +0000 (19:38 +0200)]
[MIPS] IRQ Affinity Support for SMTC on Malta Platform

Signed-off-by: Kevin D. Kissell <kevink@mips.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
16 years agoide: use only ->set_pio_mode method for programming PIO modes (take 2)
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:02 +0000 (23:54 +0200)]
ide: use only ->set_pio_mode method for programming PIO modes (take 2)

Use ->set_pio_mode method to program PIO modes in ide_set_xfer_rate()
(the only place which used ->speedproc to program PIO modes) and remove
handling of PIO modes from all ->speedproc implementations.

v2:
* Fix pmac_ide_tune_chipset() comment.

There should be no functionality changes caused by this patch.

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agosis5513: don't change UDMA settings when programming PIO
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:01 +0000 (23:54 +0200)]
sis5513: don't change UDMA settings when programming PIO

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoit8213/piix/slc90e66: don't change DMA settings when programming PIO
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:01 +0000 (23:54 +0200)]
it8213/piix/slc90e66: don't change DMA settings when programming PIO

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoalim15x3: PIO mode setup fixes
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:01 +0000 (23:54 +0200)]
alim15x3: PIO mode setup fixes

* Return failure in ->speedproc method for unsupported/invalid transfer
  modes passed from user-space (fixes theoretical OOPS in ali_tune_pio()
  and/or setting random PIO timings on host controller + disabling UDMA
  bit + setting unsupported/invalid transfer mode on the device).

* Don't disable UDMA bit when programming PIO modes.

* Add FIXME about DMA timings never being set.

* Bump driver version.

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agosiimage: fix ->set_pio_mode method to select PIO data transfer
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:01 +0000 (23:54 +0200)]
siimage: fix ->set_pio_mode method to select PIO data transfer

* Remember to select PIO data transfer (with IORDY monitored) in sil_tune_pio()
  (->set_pio_mode method) so the controller is always programmed correctly for
  PIO transfers (this is important if DMA is not going to be used).

* Don't set DMA/UDMA timings for PIO modes in siimage_tune_chipset().

* Bump driver version.

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agocs5520: don't enable VDMA in ->speedproc
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:01 +0000 (23:54 +0200)]
cs5520: don't enable VDMA in ->speedproc

There is no need to do it in cs5520_tune_chipset() (->speedproc method)
since it is correctly handled by cs5520_dma_on() (->ide_dma_on method).

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agosc1200: remove redundant warning message from sc1200_tune_chipset()
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:01 +0000 (23:54 +0200)]
sc1200: remove redundant warning message from sc1200_tune_chipset()

ide_config_drive_speed() called by sc1200_set_xfer_mode() already gives
a meaningful warning message in case of failure.

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoide-pmac: PIO mode setup fixes (take 3)
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:01 +0000 (23:54 +0200)]
ide-pmac: PIO mode setup fixes (take 3)

* Add { 0, 0 } entry to {kauai,shasta}_pio_timings[] so kauai_lookup_timing()
  always returns a valid PIO timing (fixes PIO timing not being set for devices
  with minimum PIO cycle <= 120ns).

* Add setting transfer mode on the device to pmac_ide_set_pio_mode().

* Fix pmac_ide_set_pio() to always program chipset for given PIO timing instead
  of only when the device we want to program PIO timing for is the currently
  selected one.

* Now that pmac_ide_set_pio() is fixed there is no need to set transfer mode
  on the device and program chipset for PIO in pmac_ide_tune_chipset()
  (returning 0 == success is not entirely correct but is OK for now since
  the upper layers are only checking ->speedproc return value for DMA modes).

v2:
* s/speed/XFER_PIO_0 + pio/ in pmac_ide_set_pio_mode() so it actually compiles.

* Removal of kauai_lookup_timing() return value checking went to separate patch.

v3:
* Use XFER_PIO_0 not XFER_PIO. (Noticed by Ben)

Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoicside: fix ->speedproc to return on unsupported modes (take 5)
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:00 +0000 (23:54 +0200)]
icside: fix ->speedproc to return on unsupported modes (take 5)

* All other implementations of ->speedproc return zero on success
  and non-zero on failure.  Currently it doesn't matter for icside host
  driver and isn't a bug per se since:

  - ide_set_xfer_rate() return value is ignored by all IDE core users

  - icside doesn't (yet!) use ide_tune_dma() in icside_dma_check()

  but sooner or later we will need to fix anyway - so lets do it now.

* icside_set_speed() happily accepts unsupported transfer modes which
  results in drive->drive_data being set to the maximum value (480)
  and drive->current_speed being set to the unsupported transfer mode.

  Fix it.

v2:
* The initial version of the patch was broken because it didn't take into
  the account (the different from usual) return values of icside_set_speed()
  (Noticed by Russell).

v3:
* Remove no longer needed initialization/checking of cycle_time
  (Noticed by Sergei).

* No need to set drive->drive_data if DMA is not going to be used
  (Noticed by Sergei).

* Remove incorrect setting of drive->current_speed
  (Noticed by Sergei).

* Move ide_config_drive_speed() at the end of icside_set_speed().

v4:
* If DMA mode is not found in icside_dma_check() then just return "-1" and
  don't call icside_set_speed() (v3 got it wrong and "1" was returned instead).

v5:
* Return "-1"/"0" in icside_set_speed() instead of icside_dma_check() return
  value (just like it was before this patch).

Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agosgiioc4: use ide_tune_dma()
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:00 +0000 (23:54 +0200)]
sgiioc4: use ide_tune_dma()

* Add DRV_NAME define and use it instead of sgiioc4_chipset.name.

* Remove no longer needed sgiioc4_chipset.

* Remove needless clearing of ->atapi_dma from ide_dma_sgiioc4().

* Fix ide_dma_sgiioc4() to return success/failure.  Check return value in
  sgiioc4_ide_setup_pci_device() and set hwif->autodma accordingly.  Also add
  missing setting of drive->autodma.

* Add sgiioc4_speedproc() (implementation of ->speedproc method).

* Fix ->mwdma_mask (MWDMA2 mask is 0x04 not 0x02) and remove incorrect
  ->swdma_mask.  Also remove needless initialization of ->ultra_mask.

* Use ide_tune_dma() in sgiioc4_ide_dma_check(), this fixes following bugs:
  - DMA capability bit not being checked on the device
  - DMA blacklist not being checked
  - MWDMA2 mode support by device not being checked

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Jeremy Higdon <jeremy@sgi.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoamd74xx/via82cxxx: use ide_tune_dma()
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:00 +0000 (23:54 +0200)]
amd74xx/via82cxxx: use ide_tune_dma()

* Use ide_tune_dma() in amd74xx/via82cxxx driver, this fixes following bugs:
  - DMA capability bit not being checked on the device
  - DMA blacklist not being checked
  - DMA mode being programmed even if drive->autodma == 0
    (thus possibly destroying PIO timings)

* Bump driver version.

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoide: add ide_set{_max}_pio() (take 4)
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:00 +0000 (23:54 +0200)]
ide: add ide_set{_max}_pio() (take 4)

* Add IDE_HFLAG_ABUSE_{PREFETCH,FAST_DEVSEL,DMA_MODES} flags
  and set them in ht6560, cmd640, cmd64x and sc1200 host drivers.

* Add set_pio_mode_abuse() for checking if host driver has a non-standard
  ->tuneproc() implementation and use it in do_special().

* Add ide_set_pio() for setting PIO mode (it uses hwif->pio_mask to find
  the maximum PIO mode supported by the host), also add ide_set_max_pio()
  wrapper for ide_set_pio() to use for auto-tuning.  Convert users of
  ->tuneproc to use ide_set{_max}_pio() where possible.  This leaves only
  do_special(), set_using_pio(), ide_hwif_restore() and ide_set_pio() as
  a direct users of ->tuneproc.

* Remove no longer needed ide_get_best_pio_mode() calls and printk-s
  reporting PIO mode selected from ->tuneproc implementations.

* Rename ->tuneproc hook to ->set_pio_mode and make 'pio' argument const.

* Remove stale comment from ide_config_drive_speed().

v2:
* Fix "ata_" prefix (Noticed by Jeff).

v3:
* Minor cleanups/fixups per Sergei's suggestions.

v4:
* Fix compile problem in drivers/ide/pci/cmd640.c
  (Noticed by Andrew Morton).

* Improve some ->set_pio_mode comments.

Reviewed-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoide: Kconfig face-lift
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:54:00 +0000 (23:54 +0200)]
ide: Kconfig face-lift

* Convert config options for PCI host drivers to select BLK_DEV_IDEPCI.

* Same for BLK_DEV_IDEDMA_PCI and PCI DMA host drivers, also make this
  config option select BLK_DEV_IDEPCI.

* Remove no longer needed help for BLK_DEV_IDEPCI and BLK_DEV_IDEDMA_PCI.

* Remove redundant IDE_CHIPSETS config option.

As a result of this patch users have now three less config options
to worry about and can just select host drivers that they need
(IDE config menu is "flatter").

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoide: move ide_rate_filter() calls to the upper layer (take 2)
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:53:59 +0000 (23:53 +0200)]
ide: move ide_rate_filter() calls to the upper layer (take 2)

* Move ide_rate_filter() calls from host drivers to IDE core.

* Make ide_rate_filter() static.

* Make 'speed' argument of ->speedproc const.

v2:
* Fix it8213_tune_chipset() comment.

There should be no functionality changes caused by this patch.

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agosis5513: add ->udma_filter method for chipset_family >= ATA_133
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:53:59 +0000 (23:53 +0200)]
sis5513: add ->udma_filter method for chipset_family >= ATA_133

* Add sis5513_ata133_udma_filter() for chipset_family >= ATA_133,
  use it to filter UDMA6 mode if ATA133 bit is disabled.

* Remove no longer needed UDMA6 limiting logic from sis5513_tune_chipset().

* Bump driver version.

There should be no functionality changes caused by this patch.

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoide: mode limiting fixes for user requested speed changes
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:53:59 +0000 (23:53 +0200)]
ide: mode limiting fixes for user requested speed changes

* Add an extra argument to ide_max_dma_mode() for passing requested transfer
  mode.  Use it as an upper limit when finding the best DMA for device/host.

* Rename ide_max_dma_mode() to ide_find_dma_mode() and at the same time add
  ide_max_dma_mode() wrapper which passes XFER_UDMA_6 as a requested mode to
  ide_find_dma_mode().  Also add inline ide_find_dma_mode() version for
  CONFIG_BLK_DEV_IDEDMA=n case.

* Pass requested transfer mode from ide_find_dma_mode() to ide_get_mode_mask()
  to avoid false warning from eighty_ninty_three().

* Use ide_find_dma_mode() to limit the user requested transfer mode in
  ide_rate_filter().  Also limit the requested mode by host max PIO mode.

Above changes make ide_rate_filter() to:

* Clip desired transfer mode down if it is invalid (values 0x0F, 0x13-0x19
  and 0x25-0x39, values > 0x46 were already clipped down, same for values
  0x25-0x39 but iff UDMA was not supported by the host).

* Clip desired transfer mode down if it is currently unsupported by IDE core
  (PIO6 and MWDMA3-4, the latter were already clipped down but iff UDMA was
  not supported by the host).

* Clip desired transfer mode down according to the host capabilities
  (UDMA modes were already clipped down but MWDMA/SWDMA/PIO weren't,
  also ->atapi_dma flag was not respected).

* Clip desired transfer mode down according to the device capabilities
  (except PIO modes for now which require mode work) - shouldn't be a
  problem since ide_set_xfer_rate() is called _after_ device has accepted
  given transfer mode.

and also result in a number of host driver specific bugfixes:

* icside
  - clip unsupported PIO5 mode down
  - fix unsupported/invalid modes being set in drive->current_speed

* ide-cris
  - clip unsupported PIO5 and SWDMA0-2 modes down
  - clip DMA modes down for ATAPI devices
  - fix BUG() on unsupported/invalid modes

* au1xxx-ide
  - clip unsupported PIO5, SWDMA0-2 and MWDMA0-2
    (if BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA=n) modes down

* aec62xx
  - clip unsupported PIO5 and SWDMA0-2 modes down
  - clip DMA modes down for ATAPI devices
  - fix 0x00 being programmed as PIO timing for unsupported/invalid modes
  - fix unsupported/invalid modes being set on the device

* alim15x3
  - clip DMA modes down for ATAPI devices (chipset revision == 0x20 only)
  - fix theoretical OOPS for 0x0F mode
  - fix unsupported/invalid modes being set on the device

* amd74xx
  - clip unsupported SWDMA0-2 (on COBRA_7401 revs <= 7) modes down
  - fix random PIO timings being set for unsupported/invalid modes
  - fix unsupported/invalid modes being set on the device

* atiixp
  - clip unsupported PIO5 and SWDMA0-2 modes down
  - fix cached MWDMA mode being cleared for unsupported/invalid modes
  - fix PIO{0,2} timings being programmed for unsupported/invalid modes
  - fix theoretical OOPS for PIO5-6 and 0x0F modes
  - fix unsupported/invalid modes being set on the device

* cmd64x
  - clip unsupported SWDMA0-2 modes down

* cs5530
  - clip unsupported PIO5 and SWDMA0-2 modes down
  - fix unsupported/invalid modes being set on the device
  - fix BUG() on unsupported/invalid modes
    (which happened if the device accepted the setting)

* cs5535
  - clip unsupported PIO5 and SWDMA0-2 modes down
  - fix unsupported/invalid modes being set on the device
  - fix theoretical OOPS for PIO5-6 and 0x0F modes

* hpt34x
  - clip DMA modes down for ATAPI devices
  - fix invalid timings being programmed for unsupported/invalid modes
  - fix unsupported/invalid modes being set on the device

* hpt366
  - clip unsupported PIO5 and SWDMA0-2 modes down
  - fix PIO0 timings being programmed for unsupported/invalid modes
  - fix DMA timings being cleared for MWDMA3-4 and 0x25-0x39 modes
  - fix unsupported/invalid modes being set on the device

* it8213
  - clip unsupported PIO5, SWDMA0-1 and MWDMA0 modes down

* it821x
  - clip unsupported PIO5 and SWDMA0-2 modes down
  - clip DMA modes down for ATAPI devices
    (chipset in smart mode and revision 0x10 in pass-through mode)

* jmicron
  - clip unsupported SWDMA0-2 modes down
  - fix unsupported/invalid modes being set on the device

* pdc202xx_new
  - clip unsupported PIO5 and SWDMA0-2 modes down
  - fix unsupported/invalid modes being set on the device

* pdc202xx_old
  - clip unsupported PIO5 mode down
  - fix incorrect timings being set for unsupported/invalid modes
  - fix unsupported/invalid modes being set on the device

* piix
  - clip unsupported PIO5, SWDMA0-1 and MWDMA0 modes down

* sc1200
  - clip unsupported PIO5 and SWDMA0-2 modes down
  - fix unsupported/invalid modes being set on the device
  - fix BUG() on unsupported/invalid modes
    (which happened if the device accepted the setting)

* scc_pata
  - clip unsupported PIO5, SWDMA0-2 and MWDMA0-2 modes down

* serverworks
  - clip unsupported PIO5 and SWDMA0-2 modes down
  - fix DMA/UDMA timings/settings being cleared for unsupported/invalid modes
  - fix unsupported/invalid modes being set on the device

* siimage
  - clip unsupported PIO5 and SWDMA0-2 modes down
  - clip DMA modes down for ATAPI devices (SATA chipsets)

* sis5513
  - clip unsupported PIO5 mode down
  - fix BUG() on unsupported/invalid modes

* sl82c105
  - clip unsupported SWDMA0-2 modes down

* slc90e66
  - clip unsupported PIO5, SWDMA0-1 and MWDMA0 modes down

* tc86c001
  - clip unsupported PIO5 and SWDMA0-2 modes down
  - fix PIO0 timings being programmed for PIO5/0x0F/SWDMA0-2/0x13-0x19 modes
  - fix invalid 0x00 DMA timing being programmed for MWDMA3-4/0x25-0x39 modes
  - fix unsupported/invalid modes being set on the device

* triflex
  - clip unsupported PIO5 mode down

* via82cxxx
  - fix random PIO timings being set for unsupported/invalid modes
  - fix unsupported/invalid modes being set on the device

* pmac
  - clip unsupported PIO5 and SWDMA0-2 modes down

* cmd640/ht6560b
  - clip DMA modes down (if CONFIG_BLK_DEV_IDEDMA=y)
  - fix PIO5 being clipped to PIO4 (if CONFIG_BLK_DEV_IDEDMA=n)

* opti621
  - clip DMA modes down (if CONFIG_BLK_DEV_IDEDMA=y)
  - clip unsupported PIO4 to PIO3 (if CONFIG_BLK_DEV_IDEDMA=n)

While at it:

* Use ide_rate_filter() in cs5520.c::cs5520_tune_chipset().

* Remove no longer needed checks from hpt366.c::hpt3{6,7}x_tune_chipset().

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoide: add missing ide_rate_filter() calls to ->speedproc()-s
Bartlomiej Zolnierkiewicz [Thu, 11 Oct 2007 21:53:59 +0000 (23:53 +0200)]
ide: add missing ide_rate_filter() calls to ->speedproc()-s

* Fix icside, cris-ide, au1xxx-ide, amd74xx, via82cxxx and pmac host drivers
  to use ide_rate_filter().

  This results in the following modes (from user requests) being clipped down:
  - invalid modes (values 0x46-0xFF)  [ for all hosts ]
  - unsupported by a host UDMA modes  [ for hosts which support UDMA]
  - all UDMA modes and MWDMA3-4 modes [ for hosts which don't support UDMA ]
  - invalid modes (values 0x25-0x39)  [ for hosts which don't support UDMA ]

* Host driver specific changes in behavior:

  icside:
    - no change

  cris-ide
    - clip unsupported UDMA3-6 modes down
    - fix BUG() on trying to set unsupported UDMA3-6 modes

  au1xxx-ide/pmac
    - clip unsupported UDMA modes down

  amd74xx/via82cxxx
    - clip unsupported UDMA modes down
    - fix random PIO timings being set for unsupported/invalid modes
    - fix unsupported/invalid modes being set on the device

* While at it remove no longer needed checks from pmac.c driver.

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoide: call udma_filter() before resorting to the UltraDMA mask
Sergei Shtylyov [Thu, 11 Oct 2007 21:53:59 +0000 (23:53 +0200)]
ide: call udma_filter() before resorting to the UltraDMA mask

Give the udma_filter() method call precedence over using the mode masks.

Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoide: make jmicron match vendor and device class
Tejun Heo [Thu, 11 Oct 2007 21:53:59 +0000 (23:53 +0200)]
ide: make jmicron match vendor and device class

PATA part of all current JMB controllers behave the same way and
JMicron confirms that all future ones will stay compatible.  Matching
vendor and device class is enough.

For backward compatibility, jmicron still needs to match 361,3,5,6,8
DIDs regardless of device class if libata is not configured but won't
need further DID update from this point.

Bart: cosmetic fixes to jmicron_chipset

Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Ethan Hsiao <ethanhsiao@jmicron.com>
Cc: Justin Tsai <justin@jmicron.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agopdc202xx_new: switch to using pci_get_slot() (take 2)
Sergei Shtylyov [Thu, 11 Oct 2007 21:53:58 +0000 (23:53 +0200)]
pdc202xx_new: switch to using pci_get_slot() (take 2)

Switch to using pci_get_slot() in init_setup_pdc20270() to get to the mate chip
behind DC21150 bridge as there's no need for the driver itself to walk the list
of the PCI devices (and the driver didn't check the bus # of the found device).
While at it, make it emit warning about IRQ # being fixed up (just like hpt366.c
does) and "beautify" this whole function as well as init_setup_pdc20276()...

Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoide: hook ACPI _PSx method to IDE power on/off
Shaohua Li [Thu, 11 Oct 2007 21:53:58 +0000 (23:53 +0200)]
ide: hook ACPI _PSx method to IDE power on/off

ACPI spec defines the sequence of IDE power on/off:
Powering down:
Call _GTM.
Power down drive (calls _PS3 method and turns off power planes).
Powering up:
Power up drive (calls _PS0 method if present and turns on power planes).
Call _STM passing info from _GTM (possibly modified), with ID data from
each drive.
Initialize the channel.
May modify the results of _GTF.
For each drive:
Call _GTF.
Execute task file (possibly modified).
This patch adds the missed _PS0/_PS3 methods call.

Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Acked-by: Len Brown <len.brown@intel.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoide: Platform IDE driver
Anton Vorontsov [Thu, 11 Oct 2007 21:53:58 +0000 (23:53 +0200)]
ide: Platform IDE driver

This is now very similar to pata_platform.c, they both use
same platform data structure and same resources.

To achieve that, byte_lanes_swapping platform data variable
and platform specified iops removed from that driver. It's fine,
since those were never used anyway.

pata_platform and ide_platform are carrying same driver names,
to easily switch between these drivers, without need to touch
platform code.

Bart:
- build fix from Andrew Morton <akpm@linux-foundation.org>

Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: Vitaly Bordug <vitb@kernel.crashing.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Acked-by: Alan Cox <alan@redhat.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agohpt366: MWDMA filter for SATA cards (take 2)
Sergei Shtylyov [Thu, 11 Oct 2007 21:53:58 +0000 (23:53 +0200)]
hpt366: MWDMA filter for SATA cards (take 2)

The Marvell bridge chips used on HighPoint SATA cards do not seem to support
the MWDMA modes (at least that could be seen in their so-called drivers :-),
so the driver needs to account for this -- to achieve this:

- add mdma_filter() method from the original patch by Bartlomiej Zolnierkiewicz
  with his consent;

- install the method for all chips to only return empty mask if a SATA drive
  is detected on HPT372{AN]/374 chips...

Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years agoatiixp: SB700 contains more than one IDE channel
Shane Huang [Thu, 11 Oct 2007 21:53:58 +0000 (23:53 +0200)]
atiixp: SB700 contains more than one IDE channel

SB700 supports one physical IDE channel, but SB700 SATA controller
supports combined mode.  When the SATA combined mode is enabled,
two SATA ports (port4 and port5) share one IDE channel from IDE
controller, and PATA will share the other IDE channel.

Our previous patch adding SB700 IDE device ID only supports one
IDE channel, which contains bug.  The attached patch fixes the bug.

Signed-off-by: "Shane Huang" <Shane.Huang@amd.com>
Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
16 years ago[IPv6]: Update setsockopt(IPV6_MULTICAST_IF) to support RFC 3493, try2
Brian Haley [Thu, 11 Oct 2007 21:39:29 +0000 (14:39 -0700)]
[IPv6]: Update setsockopt(IPV6_MULTICAST_IF) to support RFC 3493, try2

 From RFC 3493, Section 5.2:

       IPV6_MULTICAST_IF

          Set the interface to use for outgoing multicast packets.  The
          argument is the index of the interface to use.  If the
          interface index is specified as zero, the system selects the
          interface (for example, by looking up the address in a routing
          table and using the resulting interface).

This patch adds support for (index == 0) to reset the value to it's
original state, allowing the system to choose the best interface.  IPv4
already behaves this way.

Signed-off-by: Brian Haley <brian.haley@hp.com>
Acked-by: David L Stevens <dlstevens@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETFILTER]: x_tables: add missing ip6t_modulename aliases
Jan Engelhardt [Thu, 11 Oct 2007 21:36:40 +0000 (14:36 -0700)]
[NETFILTER]: x_tables: add missing ip6t_modulename aliases

The patch will add MODULE_ALIAS("ip6t_<modulename>") where missing,
otherwise you will get

ip6tables: No chain/target/match by that name

when xt_<modulename> is not already loaded.

Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETFILTER]: nf_conntrack_tcp: fix connection reopening
Jozsef Kadlecsik [Thu, 11 Oct 2007 21:35:52 +0000 (14:35 -0700)]
[NETFILTER]: nf_conntrack_tcp: fix connection reopening

With your description I could reproduce the bug and actually you were
completely right: the code above is incorrect. Somehow I was able to
misread RFC1122 and mixed the roles :-(:

   When a connection is >>closed actively<<, it MUST linger in
   TIME-WAIT state for a time 2xMSL (Maximum Segment Lifetime).
   However, it MAY >>accept<< a new SYN from the remote TCP to
   reopen the connection directly from TIME-WAIT state, if it:
   [...]

The fix is as follows: if the receiver initiated an active close, then the
sender may reopen the connection - otherwise try to figure out if we hold
a dead connection.

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Tested-by: Krzysztof Piotr Oledzki <ole@ans.pl>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years agobacklight: Convert corgi backlight driver into a more generic driver
Richard Purdie [Sun, 2 Sep 2007 23:27:00 +0000 (00:27 +0100)]
backlight: Convert corgi backlight driver into a more generic driver

Convert the corgi backlight driver to a more generic version
so it can be reused by other code rather than being Zaurus/PXA
specific.

Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
16 years agobacklight: Add Samsung LTV350QV LCD driver
Haavard Skinnemoen [Sun, 2 Sep 2007 22:15:49 +0000 (23:15 +0100)]
backlight: Add Samsung LTV350QV LCD driver

This patch adds support for powering on and off the Samsung LTV350QV LCD
panel via SPI.  The driver responds to framebuffer power management, it
powers off the panel on reboot/halt/poweroff.  It can also be controlled
through sysfs.  The panel is powered up when the module is loaded, and off
when the module is unloaded.  Verified on AVR32 STK1000.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
16 years agobacklight: Fix cr_bllcd allocations and error paths
Jesper Juhl [Sat, 11 Aug 2007 09:29:59 +0000 (10:29 +0100)]
backlight: Fix cr_bllcd allocations and error paths

After fixing the too small memory allocation in cr_backlight_probe()
from drivers/video/backlight/cr_bllcd.c
(commit e3bbb3f05339de438faf54124f25c92e6fe4ac2e) I noticed that the
Coverity checker also thought there were a few memory leaks in there.
I took a closer look and confirmed that there were indeed several
leaks.

At the start of the function we allocate storage for a
'struct cr_panel' and store the pointer in a variable named 'crp'.

Then we call pci_get_device() and pci_read_config_byte() and if
either of them fail we return without freeing the memory allocated
for the 'struct cr_panel'. These two leaks are easy to fix since we
don't even use 'crp' for anything up to this point, so I simply
moved the allocation further down in the function so it only happens
just before we actually need it.

A bit further down we call backlight_device_register() and store the
result in 'crp->cr_backlight_device'. In case of error we return
'crp->cr_backlight_device' from the function, thus leaking 'crp'
itself. The same thing happens with the call to lcd_device_register().
To fix these two leaks I declare two new pointers to hold the return
values, so that in case of error we can return the pointer (as before)
but without leaking 'crp'.

This version of the patch also adds missing
backlight_device_unregister() / lcd_device_unregister() / pci_dev_put()
calls to error paths.
  Thanks to Richard Purdie <rpurdie@rpsys.net> for noticing.

Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
16 years agobacklight/leds: Make two structs static
Adrian Bunk [Sat, 11 Aug 2007 09:27:19 +0000 (10:27 +0100)]
backlight/leds: Make two structs static

This patch makes two needlessly global structs static.

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
16 years agoleds: Update Cobalt Qube series front LED support
Yoichi Yuasa [Thu, 27 Sep 2007 08:51:17 +0000 (17:51 +0900)]
leds: Update Cobalt Qube series front LED support

Update Cobalt Qube series front LED support.

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Acked-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
16 years agoleds: Add Cobalt Raq series LEDs support
Yoichi Yuasa [Thu, 27 Sep 2007 08:51:05 +0000 (17:51 +0900)]
leds: Add Cobalt Raq series LEDs support

Add Cobalt Raq series LEDs support.

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Acked-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
16 years agoleds: Rename leds-cobalt driver
Yoichi Yuasa [Thu, 20 Sep 2007 23:02:47 +0000 (00:02 +0100)]
leds: Rename leds-cobalt driver

The leds-cobalt driver only supports the Coable Qube series
(not included in Cobalt Raq series).
Rename the driver and update Kconfig/Makefile.

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Acked-by: Florian Fainelli <florian.fainelli@telecomint.eu>
Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
16 years ago[BLOCK] Fix failing compile with BLK_DEV_IO_TRACE=n
Boaz Harrosh [Thu, 11 Oct 2007 19:26:08 +0000 (21:26 +0200)]
[BLOCK] Fix failing compile with BLK_DEV_IO_TRACE=n

I get a compilation error in sglist-arch branch
with BLK_DEV_IO_TRACE=n:

  CC      block/compat_ioctl.o
/usr0/export/dev/bharrosh/git/pub/linux-2.6-block/block/compat_ioctl.c: In
function ?compat_blk_trace_setup?:
/usr0/export/dev/bharrosh/git/pub/linux-2.6-block/block/compat_ioctl.c:568:
error: expected expression before ?do?
make[2]: *** [block/compat_ioctl.o] Error 1

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agommc: don't use weight32()
Nicolas Pitre [Thu, 11 Oct 2007 05:06:03 +0000 (01:06 -0400)]
mmc: don't use weight32()

Using weight32() to determine if a value is a power of 2 is a rather
heavi weight solution.  The classic idiom is (x & (x - 1)) == 0, but
the kernel already provide a is_power_of_2 function for it.

Signed-off-by: Nicolas Pitre <nico@cam.org>
Signed-off-by: Pierre Ossman <drzeus@drzeus.cx>
16 years agox86: fixup the x86 namespace change in scripts/namespace.pl
Thomas Gleixner [Thu, 11 Oct 2007 15:56:02 +0000 (17:56 +0200)]
x86: fixup the x86 namespace change in scripts/namespace.pl

Let namespace.pl look at the correct files.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86: Fix the $(ARCH) dependent help output in the top Makefile
Thomas Gleixner [Thu, 11 Oct 2007 15:53:52 +0000 (17:53 +0200)]
x86: Fix the $(ARCH) dependent help output in the top Makefile

Change the $(ARCH) dependency to $(SRCARCH) to honor the x86
namespace for i386 and x86_64.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years ago[AVR32] Fix random segfault with preemption
Philippe Rétornaz [Wed, 10 Oct 2007 22:52:24 +0000 (18:52 -0400)]
[AVR32] Fix random segfault with preemption

As explained on:
http://www.avrfreaks.net/index.php?nameÿphpBB2&fileÿewtopic&tS307
If the current process is preempted before it can copy RAR_SUP and
RSR_SUP both register are lost and the process will segfault as soon
as it return from the syscall since the return adress will be
corrupted.

This patch disable IRQ as soon as we enter the syscall path and
reenable them when the copy is done.

In the interrupt handlers, check if we are interrupting the srrf
instruction, if so disable interrupts and return. The interrupt
handler will be re-called immediatly when the interrupts are
reenabled.

After some stressing workload:
 - find / > /dev/null in loop
 - top (in ssh)
 - ping -f avr32

The segfaults are not seen anymore.

Signed-off-by: Philippe Rétornaz <philippe.retornaz@epfl.ch>
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] Don't use __builtin_xchg()
Haavard Skinnemoen [Wed, 3 Oct 2007 13:05:20 +0000 (15:05 +0200)]
[AVR32] Don't use __builtin_xchg()

The implementation of __builtin_xchg() in at least some versions of
avr32 gcc is buggy. Rather than find out exactly which versions that
have this bug, let's just avoid the problem altogether by implementing
xchg() in inline assembly.

Also, in most architectures, xchg() seems to imply a memory barrier,
while the existing avr32 implementation did not. This patch also fixes
that discrepancy.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] ngw100 i2c-gpio tweaks
David Brownell [Tue, 25 Sep 2007 14:17:48 +0000 (07:17 -0700)]
[AVR32] ngw100 i2c-gpio tweaks

Make the NGW100 bitbang i2c use open drain signaling.

Also, speed it up, so it's closer to 100 kHz ... the code paths seem
to be long enough that the udelay isn't dominating bit times.  The
peak bit rate I observed was around 125 kHz, but that's with large
delays (usually before ACK/NAK) which hold the overall rate down to
around 80 kHz (call it 100 usec/byte on average).

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] Ignore a few irrelevant syscalls
Haavard Skinnemoen [Sat, 22 Sep 2007 21:31:39 +0000 (23:31 +0200)]
[AVR32] Ignore a few irrelevant syscalls

Ignore a few syscalls that are irrelevant because they're either old,
depends on NUMA or depends on SMP.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] SMC configuration in clock cycles
Kristoffer Nyborg Gregertsen [Fri, 17 Aug 2007 14:59:57 +0000 (16:59 +0200)]
[AVR32] SMC configuration in clock cycles

This patch makes the SMC configuration take timings in clock cycles
instead of nanoseconds. A function to calculate timings in clock
cycles is added.

This patch removes the rounding troubles of the previous SMC
configuration method.

[hskinnemoen@atmel.com: fix atstk1002/atngw100 flash config]
Signed-off-by: Kristoffer Nyborg Gregertsen <gregerts@stud.ntnu.no>
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] Drop support for redundant "keepinitrd" boot-time parm.
Robert P. J. Day [Wed, 19 Sep 2007 12:43:42 +0000 (08:43 -0400)]
[AVR32] Drop support for redundant "keepinitrd" boot-time parm.

Given the existing "retain_initrd" boot-time parameter defined in
init/initramfs.c, there appears to be no need for the equivalent
"keepinitrd" parameter.

Signed-off-by: Robert P. J. Day <rpjday@mindspring.com>
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] Make dma_sync_*_for_cpu no-ops
Haavard Skinnemoen [Fri, 7 Sep 2007 15:00:25 +0000 (16:00 +0100)]
[AVR32] Make dma_sync_*_for_cpu no-ops

I don't think the dma_sync_*_for_cpu ever did anything useful. We
flush the relevant cache lines when mapping the buffer or when calling
dma_sync_*_for_device(), and the CPU isn't allowed to touch the buffer
after that.

In other words, if these functions actually have anything to flush
from the caches, we're already in trouble.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] Remove unneeded 8K alignment of .text section
Haavard Skinnemoen [Mon, 17 Sep 2007 09:13:45 +0000 (11:13 +0200)]
[AVR32] Remove unneeded 8K alignment of .text section

__init_end, which comes immediately before .text, is already page
aligned, and that should be more than enough for the .text section.

The reason why we need to align the .text section is because the
interrupt handler offset is ORed with EVBA, so we need to provide
enough alignment of EVBA that this OR operation works as an ADD.

Currently, the last interrupt handler is not nearly a full page away
from EVBA, so it won't be a problem.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] Kill a few hardcoded constants in vmlinux.lds
Haavard Skinnemoen [Mon, 17 Sep 2007 09:08:28 +0000 (11:08 +0200)]
[AVR32] Kill a few hardcoded constants in vmlinux.lds

Use PAGE_SIZE, THREAD_SIZE and L1_CACHE_BYTES instead of harcoded
constants in places where that's what we really mean.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] rename vmlinux.lds
Sam Ravnborg [Sat, 15 Sep 2007 20:47:02 +0000 (22:47 +0200)]
[AVR32] rename vmlinux.lds

Rename vmlinux.lds to a .S file to match other architectures.
Simplify Makefile to match the rename and deleted the unused
USE_STANDARD_AS_RULE

Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] fix command line parsing in early_parse_fbmem
Matteo Vit [Wed, 29 Aug 2007 08:19:40 +0000 (10:19 +0200)]
[AVR32] fix command line parsing in early_parse_fbmem

Signed-off-by: Matteo Vit - Dave S.r.l. <matteo.vit@dave.eu>
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] checkstack support
Haavard Skinnemoen [Wed, 4 Jan 2006 15:58:49 +0000 (16:58 +0100)]
[AVR32] checkstack support

Add regexes to recognize stack frame adjustments in AVR32 code.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[QETH]: fix qeth_main.c
Andrew Morton [Thu, 11 Oct 2007 09:44:23 +0000 (02:44 -0700)]
[QETH]: fix qeth_main.c

drivers/s390/net/qeth_main.c: In function 'qeth_hard_header_parse':
drivers/s390/net/qeth_main.c:6584: error: 'dev' undeclared (first use in this function)
drivers/s390/net/qeth_main.c:6584: error: (Each undeclared identifier is reported only once
drivers/s390/net/qeth_main.c:6584: error: for each function it appears in.)

Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[AVR32] Wire up USBA device
Haavard Skinnemoen [Thu, 14 Jun 2007 15:37:31 +0000 (17:37 +0200)]
[AVR32] Wire up USBA device

Implement at32_add_device_usba() and use it to wire up the USBA device
on ATSTK1000 and ATNGW100.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] add multidrive support for pio driver
Matteo Vit [Thu, 9 Aug 2007 12:55:34 +0000 (14:55 +0200)]
[AVR32] add multidrive support for pio driver

This patch add multidrive support for pio driver

Signed-off-by: Matteo Vit - Dave S.r.l. <matteo.vit@dave.eu>
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] /sys/kernel/debug/at32ap_clk
David Brownell [Fri, 10 Aug 2007 03:56:07 +0000 (20:56 -0700)]
[AVR32] /sys/kernel/debug/at32ap_clk

When debugfs is available, /sys/kernel/debug/at32ap_clk will provide a
dump of the power manager registers and of the current clock tree.  This
can help sorting out various surprises, and when making runtime PM work.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years ago[AVR32] Move AT32_PM_BASE definition into pm.h
Haavard Skinnemoen [Thu, 16 Aug 2007 09:12:47 +0000 (11:12 +0200)]
[AVR32] Move AT32_PM_BASE definition into pm.h

We don't want to redefine this in every file that needs to access
the PM.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agox86: remove migration makefile tweaks
Thomas Gleixner [Thu, 11 Oct 2007 09:20:06 +0000 (11:20 +0200)]
x86: remove migration makefile tweaks

The makefile tweaks to allow the build during the move of
code from arch/i386 and arch/x86_64 are not longer necessary.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86: Add feature-removal entry for bzImage symlinks
Thomas Gleixner [Thu, 11 Oct 2007 09:20:05 +0000 (11:20 +0200)]
x86: Add feature-removal entry for bzImage symlinks

The i386/x86_64 merge provides a symlink to the old bzImage
location. This is just a transitional migration helper.
Remove it after a 6 month grace period.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386/x86_64: move headers to include/asm-x86
Thomas Gleixner [Thu, 11 Oct 2007 09:20:03 +0000 (11:20 +0200)]
i386/x86_64: move headers to include/asm-x86

Move the headers to include/asm-x86 and fixup the
header install make rules

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86: sanitize pathes arch/x86/kernel/cpu/Makefile
Thomas Gleixner [Thu, 11 Oct 2007 09:17:26 +0000 (11:17 +0200)]
x86: sanitize pathes arch/x86/kernel/cpu/Makefile

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86: sanitize pathes arch/x86/kernel/Makefile_64
Thomas Gleixner [Thu, 11 Oct 2007 09:17:25 +0000 (11:17 +0200)]
x86: sanitize pathes arch/x86/kernel/Makefile_64

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86: sanitize pathes arch/x86/kernel/Makefile_32
Thomas Gleixner [Thu, 11 Oct 2007 09:17:25 +0000 (11:17 +0200)]
x86: sanitize pathes arch/x86/kernel/Makefile_32

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move kernel
Thomas Gleixner [Thu, 11 Oct 2007 09:17:24 +0000 (11:17 +0200)]
x86_64: move kernel

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move ia32
Thomas Gleixner [Thu, 11 Oct 2007 09:17:21 +0000 (11:17 +0200)]
x86_64: move ia32

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move oprofile
Thomas Gleixner [Thu, 11 Oct 2007 09:17:19 +0000 (11:17 +0200)]
x86_64: move oprofile

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move mm
Thomas Gleixner [Thu, 11 Oct 2007 09:17:18 +0000 (11:17 +0200)]
x86_64: move mm

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move boot
Thomas Gleixner [Thu, 11 Oct 2007 09:17:16 +0000 (11:17 +0200)]
x86_64: move boot

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move boot
Thomas Gleixner [Thu, 11 Oct 2007 09:17:15 +0000 (11:17 +0200)]
x86_64: move boot

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move boot
Thomas Gleixner [Thu, 11 Oct 2007 09:17:13 +0000 (11:17 +0200)]
x86_64: move boot

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move pci
Thomas Gleixner [Thu, 11 Oct 2007 09:17:12 +0000 (11:17 +0200)]
x86_64: move pci

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move vdso
Thomas Gleixner [Thu, 11 Oct 2007 09:17:10 +0000 (11:17 +0200)]
x86_64: move vdso

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move lib
Thomas Gleixner [Thu, 11 Oct 2007 09:17:08 +0000 (11:17 +0200)]
x86_64: move lib

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move kernel/cpufreq
Thomas Gleixner [Thu, 11 Oct 2007 09:17:06 +0000 (11:17 +0200)]
x86_64: move kernel/cpufreq

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move kernel/acpi
Thomas Gleixner [Thu, 11 Oct 2007 09:17:05 +0000 (11:17 +0200)]
x86_64: move kernel/acpi

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: move crypto
Thomas Gleixner [Thu, 11 Oct 2007 09:17:03 +0000 (11:17 +0200)]
x86_64: move crypto

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move kernel
Thomas Gleixner [Thu, 11 Oct 2007 09:17:01 +0000 (11:17 +0200)]
i386: move kernel

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move kernel/cpu
Thomas Gleixner [Thu, 11 Oct 2007 09:16:58 +0000 (11:16 +0200)]
i386: move kernel/cpu

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move video
Thomas Gleixner [Thu, 11 Oct 2007 09:16:56 +0000 (11:16 +0200)]
i386: move video

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move oprofile
Thomas Gleixner [Thu, 11 Oct 2007 09:16:55 +0000 (11:16 +0200)]
i386: move oprofile

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move mach-voyager
Thomas Gleixner [Thu, 11 Oct 2007 09:16:53 +0000 (11:16 +0200)]
i386: move mach-voyager

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move xen
Thomas Gleixner [Thu, 11 Oct 2007 09:16:51 +0000 (11:16 +0200)]
i386: move xen

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move mach-es7000
Thomas Gleixner [Thu, 11 Oct 2007 09:16:50 +0000 (11:16 +0200)]
i386: move mach-es7000

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move mach-default
Thomas Gleixner [Thu, 11 Oct 2007 09:16:48 +0000 (11:16 +0200)]
i386: move mach-default

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move mm
Thomas Gleixner [Thu, 11 Oct 2007 09:16:47 +0000 (11:16 +0200)]
i386: move mm

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move boot
Thomas Gleixner [Thu, 11 Oct 2007 09:16:45 +0000 (11:16 +0200)]
i386: move boot

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move boot
Thomas Gleixner [Thu, 11 Oct 2007 09:16:43 +0000 (11:16 +0200)]
i386: move boot

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move boot
Thomas Gleixner [Thu, 11 Oct 2007 09:16:41 +0000 (11:16 +0200)]
i386: move boot

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move mach-generic
Thomas Gleixner [Thu, 11 Oct 2007 09:16:39 +0000 (11:16 +0200)]
i386: move mach-generic

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move mach-visws
Thomas Gleixner [Thu, 11 Oct 2007 09:16:38 +0000 (11:16 +0200)]
i386: move mach-visws

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move pci
Thomas Gleixner [Thu, 11 Oct 2007 09:16:36 +0000 (11:16 +0200)]
i386: move pci

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move power
Thomas Gleixner [Thu, 11 Oct 2007 09:16:34 +0000 (11:16 +0200)]
i386: move power

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move lib
Thomas Gleixner [Thu, 11 Oct 2007 09:16:33 +0000 (11:16 +0200)]
i386: move lib

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move math-emu
Thomas Gleixner [Thu, 11 Oct 2007 09:16:31 +0000 (11:16 +0200)]
i386: move math-emu

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move kernel/cpu/mtrr
Thomas Gleixner [Thu, 11 Oct 2007 09:16:28 +0000 (11:16 +0200)]
i386: move kernel/cpu/mtrr

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move kernel/cpu/cpufreq
Thomas Gleixner [Thu, 11 Oct 2007 09:16:27 +0000 (11:16 +0200)]
i386: move kernel/cpu/cpufreq

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move kernel/cpu/mcheck
Thomas Gleixner [Thu, 11 Oct 2007 09:16:25 +0000 (11:16 +0200)]
i386: move kernel/cpu/mcheck

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move kernel/acpi
Thomas Gleixner [Thu, 11 Oct 2007 09:16:23 +0000 (11:16 +0200)]
i386: move kernel/acpi

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoi386: move crypto
Thomas Gleixner [Thu, 11 Oct 2007 09:16:21 +0000 (11:16 +0200)]
i386: move crypto

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: prepare shared mm/Makefile
Thomas Gleixner [Thu, 11 Oct 2007 09:16:20 +0000 (11:16 +0200)]
x86_64: prepare shared mm/Makefile

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86_64: prepare shared mm/fault.c
Thomas Gleixner [Thu, 11 Oct 2007 09:16:19 +0000 (11:16 +0200)]
x86_64: prepare shared mm/fault.c

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>