linux-2.6
17 years agoMerge master.kernel.org:/home/rmk/linux-2.6-arm
Linus Torvalds [Sun, 4 Mar 2007 21:15:48 +0000 (13:15 -0800)] 
Merge master.kernel.org:/home/rmk/linux-2.6-arm

* master.kernel.org:/home/rmk/linux-2.6-arm: (30 commits)
  [ARM] Acorn: move the i2c bus driver into drivers/i2c
  [ARM] ARM SCSI: Don't try to dma_map_sg too many scatterlist entries
  [ARM] ARM FAS216: don't modify scsi_cmnd request_bufflen
  [ARM] rtc-pcf8583: Final fixes for this RTC on RiscPC
  [ARM] rtc-pcf8583: correct month and year offsets
  [ARM] rtc-pcf8583: don't use BCD_TO_BIN/BIN_TO_BCD
  [ARM] EBSA110: Work around build errors
  [ARM] 4241/1: Define mb() as compiler barrier on a uniprocessor system
  [ARM] 4239/1: S3C24XX: Update kconfig entries for PM
  [ARM] 4238/1: S3C24XX: docs: update suspend and resume
  [ARM] 4237/2: oprofile: Always allow backtraces on ARM
  [ARM] Yet more asm/apm-emulation.h stuff
  ARM: OMAP: Add missing get_irqnr_preamble and arch_ret_to_user for omap2
  ARM: OMAP: Use linux/delay.h not asm/delay.h
  ARM: OMAP: Remove obsolete alsa typedefs
  ARM: OMAP: omap1510->15xx conversions needed for sx1
  ARM: OMAP: Add missing includes to board-nokia770
  ARM: OMAP: Workqueue changes for board-h4.c
  ARM: OMAP: dmtimer.c omap1 register fix
  ARM: OMAP: board-nokia770: correct lcd name
  ...

17 years agoMerge branch 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus
Linus Torvalds [Sun, 4 Mar 2007 21:15:00 +0000 (13:15 -0800)] 
Merge branch 'upstream' of git://ftp.linux-mips.org/upstream-linus

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
  [MIPS] MTX1: clear PCI errors
  [MIPS] MTX1: add idsel cardbus ressources
  [MIPS] MTX1: remove unneeded settings
  [MIPS] dma_sync_sg_for_cpu is a no-op except for non-coherent R10000s.
  [MIPS] Cobalt: update reserved resources
  [MIPS] SN: PCI fixup needs to include <irq.h>.
  [MIPS] DMA: Fix a bunch of warnings due to missing inline keywords.
  [MIPS] RM: It should be #ifdef CONFIG_FOO not #if CONFIG_FOO ...
  [MIPS] Fix and cleanup the mess that a dozen prom_printf variants are.
  [MIPS] DEC: Remove redeclarations of mips_machgroup and mips_machtype.
  [MIPS] No need to write c0_compare in plat_timer_setup
  [MIPS] Convert to RTC-class ds1742 driver
  [MIPS] Oprofile: Add missing break statements.
  [MIPS] jmr3927: build fix
  [MIPS] SNI: Fix mc146818_decode_year
  [MIPS] Replace sys32_timer_create with the generic compat_sys_timer_create.
  [MIPS] Replace sys32_socketcall with the generic compat_sys_socketcall.
  [MIPS] N32 waitid is the same as o32.

17 years ago[ARM] Acorn: move the i2c bus driver into drivers/i2c
Russell King [Sun, 4 Mar 2007 20:40:50 +0000 (20:40 +0000)] 
[ARM] Acorn: move the i2c bus driver into drivers/i2c

Move the Acorn IOC/IOMD I2C bus driver from drivers/i2c, strip
out the reminants of the platform specific parts of the old
PCF8583 RTC code, and remove the old obsolete PCF8583 driver.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years ago[ARM] ARM SCSI: Don't try to dma_map_sg too many scatterlist entries
Russell King [Sun, 4 Mar 2007 20:21:38 +0000 (20:21 +0000)] 
[ARM] ARM SCSI: Don't try to dma_map_sg too many scatterlist entries

An off-by-one bug meant we were always trying to map one too many
scatterlist entries.  This was mostly harmless prior to the checks
going in to consistent_sync(), but now causes the kernel to BUG.

Also, powertec.c was missing an assignment to info->ec.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years ago[ARM] ARM FAS216: don't modify scsi_cmnd request_bufflen
Russell King [Sun, 4 Mar 2007 20:19:07 +0000 (20:19 +0000)] 
[ARM] ARM FAS216: don't modify scsi_cmnd request_bufflen

SCSI doesn't want drivers to modify request_bufflen, so keep a
driver-private copy of this in the scsi_pointer structure instead.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years ago[ARM] rtc-pcf8583: Final fixes for this RTC on RiscPC
Russell King [Sun, 4 Mar 2007 20:33:07 +0000 (20:33 +0000)] 
[ARM] rtc-pcf8583: Final fixes for this RTC on RiscPC

Replace the I2C bus address, as per drivers/acorn/char/pcf8583.c.

Also, since this driver also contains Acorn RiscPC specific code
for obtaining the current year from the SRAM (and updating the
platform specific checksum when writing new data back) this is
NOT a platform independent driver.

Document it as such, and update the dependencies to reflect this
fact.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years ago[ARM] rtc-pcf8583: correct month and year offsets
Russell King [Sun, 4 Mar 2007 20:13:13 +0000 (20:13 +0000)] 
[ARM] rtc-pcf8583: correct month and year offsets

No, today is not 4th April 3907, it's 4th March 2007.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years ago[ARM] rtc-pcf8583: don't use BCD_TO_BIN/BIN_TO_BCD
Russell King [Sun, 4 Mar 2007 20:12:07 +0000 (20:12 +0000)] 
[ARM] rtc-pcf8583: don't use BCD_TO_BIN/BIN_TO_BCD

Both BCD_TO_BIN(x) and BIN_TO_BCD(x) have an unexpected side-effect -
not only do they return the value as expected, they _modify_ their
argument in the process.

Let's play it safe and avoid these macros.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years ago[MIPS] MTX1: clear PCI errors
Florian Fainelli [Fri, 2 Mar 2007 21:08:01 +0000 (22:08 +0100)] 
[MIPS] MTX1: clear PCI errors

This patch clears PCI errors after showing more debug informations.

Signed-off-by: Florian Fainelli <florian.fainelli@int-evry.fr>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] MTX1: add idsel cardbus ressources
Florian Fainelli [Fri, 2 Mar 2007 21:07:48 +0000 (22:07 +0100)] 
[MIPS] MTX1: add idsel cardbus ressources

Adds cardbus ressources for MTX1 boards which have a PCMCIA controller.

Signed-off-by: Florian Fainelli <florian.fainelli@int-evry.fr>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] MTX1: remove unneeded settings
Florian Fainelli [Fri, 2 Mar 2007 21:07:41 +0000 (22:07 +0100)] 
[MIPS] MTX1: remove unneeded settings

This patch removes unnecessary settings at setup time.

Signed-off-by: Florian Fainelli <florian.fainelli@int-evry.fr>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] dma_sync_sg_for_cpu is a no-op except for non-coherent R10000s.
Ralf Baechle [Fri, 2 Mar 2007 11:42:11 +0000 (11:42 +0000)] 
[MIPS] dma_sync_sg_for_cpu is a no-op except for non-coherent R10000s.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] Cobalt: update reserved resources
Yoichi Yuasa [Thu, 1 Mar 2007 13:50:25 +0000 (22:50 +0900)] 
[MIPS] Cobalt: update reserved resources

This patch has removed unused timer resource.
Moreover, the name of reserved resources ware changed.

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] SN: PCI fixup needs to include <irq.h>.
Ralf Baechle [Thu, 1 Mar 2007 15:35:54 +0000 (15:35 +0000)] 
[MIPS] SN: PCI fixup needs to include <irq.h>.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] DMA: Fix a bunch of warnings due to missing inline keywords.
Ralf Baechle [Thu, 1 Mar 2007 15:30:01 +0000 (15:30 +0000)] 
[MIPS] DMA: Fix a bunch of warnings due to missing inline keywords.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] RM: It should be #ifdef CONFIG_FOO not #if CONFIG_FOO ...
Ralf Baechle [Thu, 1 Mar 2007 15:29:25 +0000 (15:29 +0000)] 
[MIPS] RM: It should be #ifdef CONFIG_FOO not #if CONFIG_FOO ...

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] Fix and cleanup the mess that a dozen prom_printf variants are.
Ralf Baechle [Thu, 1 Mar 2007 11:56:43 +0000 (11:56 +0000)] 
[MIPS] Fix and cleanup the mess that a dozen prom_printf variants are.

early_printk is a so much saner thing.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] DEC: Remove redeclarations of mips_machgroup and mips_machtype.
Ralf Baechle [Thu, 1 Mar 2007 10:47:23 +0000 (10:47 +0000)] 
[MIPS] DEC: Remove redeclarations of mips_machgroup and mips_machtype.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] No need to write c0_compare in plat_timer_setup
Atsushi Nemoto [Wed, 28 Feb 2007 16:53:13 +0000 (01:53 +0900)] 
[MIPS] No need to write c0_compare in plat_timer_setup

If R4k counter was used for hpt_timer and interrupt source,
c0_hpt_timer_init() initializes the c0_compare register.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] Convert to RTC-class ds1742 driver
Atsushi Nemoto [Wed, 28 Feb 2007 15:40:21 +0000 (00:40 +0900)] 
[MIPS] Convert to RTC-class ds1742 driver

The generic rtc-ds1742 driver can be used for RBTX4927 and JMR3927
(with __swizzle_addr trick).  This patch also removes MIPS local
DS1742 stuff.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] Oprofile: Add missing break statements.
Ralf Baechle [Wed, 28 Feb 2007 15:34:22 +0000 (15:34 +0000)] 
[MIPS] Oprofile: Add missing break statements.

This was causing oprofile to fail on R10000, R12000, R14000.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] jmr3927: build fix
Atsushi Nemoto [Tue, 27 Feb 2007 16:42:19 +0000 (01:42 +0900)] 
[MIPS] jmr3927: build fix

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] SNI: Fix mc146818_decode_year
Thomas Bogendoerfer [Sat, 3 Feb 2007 13:06:10 +0000 (14:06 +0100)] 
[MIPS] SNI: Fix mc146818_decode_year

Big endian RMs uses a different mc146818_decode_year than little endian RMs
Correct mc146818_decode_year for years before 2000

Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] Replace sys32_timer_create with the generic compat_sys_timer_create.
Ralf Baechle [Tue, 27 Feb 2007 01:21:17 +0000 (01:21 +0000)] 
[MIPS] Replace sys32_timer_create with the generic compat_sys_timer_create.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] Replace sys32_socketcall with the generic compat_sys_socketcall.
Ralf Baechle [Tue, 27 Feb 2007 01:11:28 +0000 (01:11 +0000)] 
[MIPS] Replace sys32_socketcall with the generic compat_sys_socketcall.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years ago[MIPS] N32 waitid is the same as o32.
Ralf Baechle [Tue, 27 Feb 2007 00:56:37 +0000 (00:56 +0000)] 
[MIPS] N32 waitid is the same as o32.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
17 years agoKVM: Move kvmfs magic number to <linux/magic.h>
Andrew Morton [Thu, 1 Mar 2007 09:28:13 +0000 (11:28 +0200)] 
KVM: Move kvmfs magic number to <linux/magic.h>

Use the standard magic.h for kvmfs.

Cc: Avi Kivity <avi@qumranet.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Fix bogus failure in kvm.ko module initialization
Avi Kivity [Mon, 26 Feb 2007 14:29:43 +0000 (16:29 +0200)] 
KVM: Fix bogus failure in kvm.ko module initialization

A bogus 'return r' can cause an otherwise successful module load to fail.
This both denies users the use of kvm, and it also denies them the use of
their machine, as it leaves a filesystem registered with its callbacks
pointing into now-freed module memory.

Fix by returning a zero like a good module.

Thanks to Richard Lucassen <mailinglists@lucassen.org> (?) for reporting
the problem and for providing access to a machine which exhibited it.

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Remove write access permissions when dirty-page-logging is enabled
Uri Lublin [Thu, 22 Feb 2007 15:37:32 +0000 (17:37 +0200)] 
KVM: Remove write access permissions when dirty-page-logging is enabled

Enabling dirty page logging is done using KVM_SET_MEMORY_REGION ioctl.
If the memory region already exists, we need to remove write accesses,
so writes will be caught, and dirty pages will be logged.

Signed-off-by: Uri Lublin <uril@qumranet.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agokvm: move do_remove_write_access() up
Uri Lublin [Thu, 22 Feb 2007 15:15:33 +0000 (17:15 +0200)] 
kvm: move do_remove_write_access() up

To be called from kvm_vm_ioctl_set_memory_region()

Signed-off-by: Uri Lublin <uril@qumranet.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Fix dirty page log bitmap size/access calculation
Uri Lublin [Thu, 22 Feb 2007 14:43:09 +0000 (16:43 +0200)] 
KVM: Fix dirty page log bitmap size/access calculation

Since dirty_bitmap is an unsigned long array, the alignment and size need
to take that into account.

Signed-off-by: Uri Lublin <uril@qumranet.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Add missing calls to mark_page_dirty()
Uri Lublin [Wed, 21 Feb 2007 16:25:21 +0000 (18:25 +0200)] 
KVM: Add missing calls to mark_page_dirty()

A few places where we modify guest memory fail to call mark_page_dirty(),
causing live migration to fail.  This adds the missing calls.

Signed-off-by: Uri Lublin <uril@qumranet.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Bump API version
Avi Kivity [Wed, 21 Feb 2007 17:47:40 +0000 (19:47 +0200)] 
KVM: Bump API version

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Per-vcpu inodes
Avi Kivity [Wed, 21 Feb 2007 16:04:26 +0000 (18:04 +0200)] 
KVM: Per-vcpu inodes

Allocate a distinct inode for every vcpu in a VM.  This has the following
benefits:

 - the filp cachelines are no longer bounced when f_count is incremented on
   every ioctl()
 - the API and internal code are distinctly clearer; for example, on the
   KVM_GET_REGS ioctl, there is no need to copy the vcpu number from
   userspace and then copy the registers back; the vcpu identity is derived
   from the fd used to make the call

Right now the performance benefits are completely theoretical since (a) we
don't support more than one vcpu per VM and (b) virtualization hardware
inefficiencies completely everwhelm any cacheline bouncing effects.  But
both of these will change, and we need to prepare the API today.

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Move kvm_vm_ioctl_create_vcpu() around
Avi Kivity [Tue, 20 Feb 2007 16:41:05 +0000 (18:41 +0200)] 
KVM: Move kvm_vm_ioctl_create_vcpu() around

In preparation of some hacking.

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Rename some kvm_dev_ioctl_*() functions to kvm_vm_ioctl_*()
Avi Kivity [Tue, 20 Feb 2007 16:27:58 +0000 (18:27 +0200)] 
KVM: Rename some kvm_dev_ioctl_*() functions to kvm_vm_ioctl_*()

This reflects the changed scope, from device-wide to single vm (previously
every device open created a virtual machine).

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Create an inode per virtual machine
Avi Kivity [Wed, 21 Feb 2007 17:28:04 +0000 (19:28 +0200)] 
KVM: Create an inode per virtual machine

This avoids having filp->f_op and the corresponding inode->i_fop different,
which is a little unorthodox.

The ioctl list is split into two: global kvm ioctls and per-vm ioctls.  A new
ioctl, KVM_CREATE_VM, is used to create VMs and return the VM fd.

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Add internal filesystem for generating inodes
Avi Kivity [Tue, 20 Feb 2007 12:07:37 +0000 (14:07 +0200)] 
KVM: Add internal filesystem for generating inodes

The kvmfs inodes will represent virtual machines and vcpus, as necessary,
reducing cacheline bouncing due to inodes and filps being shared.

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: More 0 -> NULL conversions
Avi Kivity [Mon, 19 Feb 2007 12:37:48 +0000 (14:37 +0200)] 
KVM: More 0 -> NULL conversions

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: SVM: intercept SMI to handle it at host level
Joerg Roedel [Mon, 19 Feb 2007 12:37:47 +0000 (14:37 +0200)] 
KVM: SVM: intercept SMI to handle it at host level

This patch changes the SVM code to intercept SMIs and handle it
outside the guest.

Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: svm: init cr0 with the wp bit set
Avi Kivity [Mon, 19 Feb 2007 12:37:47 +0000 (14:37 +0200)] 
KVM: svm: init cr0 with the wp bit set

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Wire up hypercall handlers to a central arch-independent location
Avi Kivity [Mon, 19 Feb 2007 12:37:47 +0000 (14:37 +0200)] 
KVM: Wire up hypercall handlers to a central arch-independent location

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Add hypercall host support for svm
Avi Kivity [Mon, 19 Feb 2007 12:37:47 +0000 (14:37 +0200)] 
KVM: Add hypercall host support for svm

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Add host hypercall support for vmx
Ingo Molnar [Mon, 19 Feb 2007 12:37:47 +0000 (14:37 +0200)] 
KVM: Add host hypercall support for vmx

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: add MSR based hypercall API
Ingo Molnar [Mon, 19 Feb 2007 12:37:47 +0000 (14:37 +0200)] 
KVM: add MSR based hypercall API

This adds a special MSR based hypercall API to KVM. This is to be
used by paravirtual kernels and virtual drivers.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Use page_private()/set_page_private() apis
Markus Rechberger [Mon, 19 Feb 2007 12:37:47 +0000 (14:37 +0200)] 
KVM: Use page_private()/set_page_private() apis

Besides using an established api, this allows using kvm in older kernels.

Signed-off-by: Markus Rechberger <markus.rechberger@amd.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Use ARRAY_SIZE macro instead of manual calculation.
Ahmed S. Darwish [Mon, 19 Feb 2007 12:37:46 +0000 (14:37 +0200)] 
KVM: Use ARRAY_SIZE macro instead of manual calculation.

Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Signed-off-by: Dor Laor <dor.laor@qumranet.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: vmx: hack set_cr0_no_modeswitch() to actually do modeswitch
Joerg Roedel [Mon, 19 Feb 2007 12:37:46 +0000 (14:37 +0200)] 
KVM: vmx: hack set_cr0_no_modeswitch() to actually do modeswitch

The whole thing is rotten, but this allows vmx to boot with the guest reboot
fix.

Signed-off-by: Markus Rechberger <markus.rechberger@amd.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Cosmetics
Avi Kivity [Mon, 19 Feb 2007 12:37:46 +0000 (14:37 +0200)] 
KVM: Cosmetics

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: Move virtualization deactivation from CPU_DEAD state to CPU_DOWN_PREPARE
Jeremy Katz [Mon, 19 Feb 2007 12:37:46 +0000 (14:37 +0200)] 
KVM: Move virtualization deactivation from CPU_DEAD state to CPU_DOWN_PREPARE

This gives it more chances of surviving suspend.

Signed-off-by: Jeremy Katz <katzj@redhat.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years agoKVM: mmu: add missing dirty page tracking cases
Avi Kivity [Mon, 19 Feb 2007 12:37:46 +0000 (14:37 +0200)] 
KVM: mmu: add missing dirty page tracking cases

We fail to mark a page dirty in three cases:

- setting the accessed bit in a pte
- setting the dirty bit in a pte
- emulating a write into a pagetable

This fix adds the missing cases.

Signed-off-by: Avi Kivity <avi@qumranet.com>
17 years ago[ARM] EBSA110: Work around build errors
Russell King [Sat, 3 Mar 2007 11:54:19 +0000 (11:54 +0000)] 
[ARM] EBSA110: Work around build errors

Work around EBSA110 build errors by selecting NO_IOPORT.  EBSA110
can't support an IO port to MMIO mapping mechanism because the
MMIO and IO port spaces have quite different and complex addressing
requirements.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years agoMerge branch 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik...
Linus Torvalds [Sat, 3 Mar 2007 01:58:52 +0000 (17:58 -0800)] 
Merge branch 'upstream-linus' of /linux/kernel/git/jgarzik/libata-dev

* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  libata: add CONFIG_PM to libata core layer
  libata: add missing CONFIG_PM in LLDs
  libata: add missing PM callbacks
  pata_qdi: Fix initialisation
  [libata] pata_cmd64x: fix driver description in comments
  [libata] pata_{legacy,sc1200,sl82c105}: add missing hooks
  [libata] change master/slave IDENTIFY order
  libata-core: Fix simplex handling

17 years ago[netdrvr] tulip, de2104x: fix typo: s/__sparc_/__sparc__/
Jeff Garzik [Sat, 3 Mar 2007 01:21:56 +0000 (20:21 -0500)] 
[netdrvr] tulip, de2104x: fix typo: s/__sparc_/__sparc__/

Noticed by Doug Nazar (via David Miller).

Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agomv643xx_eth: move mac_addr inside mv643xx_eth_platform_data
Dale Farnsworth [Thu, 1 Mar 2007 23:31:48 +0000 (16:31 -0700)] 
mv643xx_eth: move mac_addr inside mv643xx_eth_platform_data

The information contained within platform_data should be self-contained.
Replace the pointer to a MAC address with the actual MAC address in
struct mv643xx_eth_platform_data.

Signed-off-by: Dale Farnsworth <dale@farnsworth.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agoFix mv643xx_eth compilation.
Dave Jones [Wed, 28 Feb 2007 20:41:39 +0000 (15:41 -0500)] 
Fix mv643xx_eth compilation.

Commit 908b637fe793165b6aecdc875cdca67c4959a1ad removed ETH_DMA_ALIGN
but missed a usage of it in a macro, which broke the build.

Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agoqla3xxx: bugfix for line omitted in previous patch.
Ron Mercer [Thu, 1 Mar 2007 00:42:17 +0000 (16:42 -0800)] 
qla3xxx: bugfix for line omitted in previous patch.

This missing line caused transmit errors on the Qlogic 4032 chip.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agoNetXen: Fix second rmmod failure observed on PowerPC machines.
Linsys Contractor Mithlesh Thukral [Wed, 28 Feb 2007 13:16:40 +0000 (05:16 -0800)] 
NetXen: Fix second rmmod failure observed on PowerPC machines.

Signed-off by: Mithlesh Thukral <mithlesh@netxen.com>

Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agoNetXen: Updates, removal of unsupported features and minor bug fixes.
Linsys Contractor Mithlesh Thukral [Wed, 28 Feb 2007 13:14:07 +0000 (05:14 -0800)] 
NetXen: Updates, removal of unsupported features and minor bug fixes.

Signed-off-by: Mithlesh Thukral <mithlesh@netxen.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agocxgb3 - Tag driver version
Divy Le Ray [Wed, 28 Feb 2007 03:28:14 +0000 (19:28 -0800)] 
cxgb3 - Tag driver version

This patch adds a "-ko" tag to the driver version.

Signed-off-by: Divy Le Ray <divy@chelsio.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agomyri10ge: fix copyright and license
Brice Goglin [Tue, 27 Feb 2007 16:18:40 +0000 (17:18 +0100)] 
myri10ge: fix copyright and license

Fix copyright and license ("regents" should not have ever been used).

Signed-off-by: Brice Goglin <brice@myri.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agojmr3927: do not call tc35815_killall().
Ralf Baechle [Tue, 27 Feb 2007 18:15:01 +0000 (18:15 +0000)] 
jmr3927: do not call tc35815_killall().

No need to stop tc35815 before resetting the board.  This fixes the
build of tc35815 as a module.  This also means there is no caller of
tc35815_killall left, so remove that function also.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agoMerge branch 'upstream-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/linvil...
Jeff Garzik [Sat, 3 Mar 2007 00:40:24 +0000 (19:40 -0500)] 
Merge branch 'upstream-fixes' of git://git./linux/kernel/git/linville/wireless-2.6 into upstream-fixes

17 years agolibata: add CONFIG_PM to libata core layer
Tejun Heo [Fri, 2 Mar 2007 08:32:47 +0000 (17:32 +0900)] 
libata: add CONFIG_PM to libata core layer

Conditionalize all PM related stuff in libata core layer using
CONFIG_PM.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agolibata: add missing CONFIG_PM in LLDs
Tejun Heo [Fri, 2 Mar 2007 08:31:26 +0000 (17:31 +0900)] 
libata: add missing CONFIG_PM in LLDs

Add missing #ifdef CONFIG_PM conditionals around all PM related parts
in libata LLDs.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agolibata: add missing PM callbacks
Tejun Heo [Fri, 2 Mar 2007 08:30:01 +0000 (17:30 +0900)] 
libata: add missing PM callbacks

Some LLDs were missing scsi device PM callbacks while having host/port
suspend support.  Add missing ones.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years ago[AOE]: Add get_unaligned() calls where needed.
David S. Miller [Fri, 2 Mar 2007 02:30:08 +0000 (18:30 -0800)] 
[AOE]: Add get_unaligned() calls where needed.

Based upon a report by Andrew Walrond.

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[RADEON]: Fix blanking return value.
David S. Miller [Fri, 2 Mar 2007 02:29:14 +0000 (18:29 -0800)] 
[RADEON]: Fix blanking return value.

If you'll recall, over a year ago, I pointed out that the current
Radeon driver erroneously returns -EINVAL for valid blanking codes,
here is a link to that thread:

http://lkml.org/lkml/2006/1/28/6

No other driver does this, and it confuses the X server into thinking
that the device does not support blanking properly.

I looked again and there is simply no reason for the Radeon driver to
return -EINVAL for FB_BLANK_NORMAL.  It claims it wants to do this in
order to convince fbcon to blank in software, right here:

if (fb_blank(info, blank))
fbcon_generic_blank(vc, info, blank);

to software blank the screen.  But it only causes that to happen
in the FB_BLANK_NORMAL case.

That makes no sense because the Radeon code does this:

val |= CRTC_DISPLAY_DIS;

in the FB_BLANK_NORMAL case so should be blanking the hardware, and
there is therefore no reason to SW blank by returning -EINVAL.

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Antonino Daplas <adaplas@gmail.com>
17 years ago[SPARC]: Provide 'get_property()' alias for of_get_property().
David S. Miller [Fri, 2 Mar 2007 02:10:25 +0000 (18:10 -0800)] 
[SPARC]: Provide 'get_property()' alias for of_get_property().

Another powerpc compatibility item, this will allow us to share
more code with them.

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[SPARC]: Provide pci_device_to_OF_node() just like powerpc.
David S. Miller [Fri, 2 Mar 2007 02:09:18 +0000 (18:09 -0800)] 
[SPARC]: Provide pci_device_to_OF_node() just like powerpc.

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[SPARC]: Handle unresolvable resources better in of_device.c
David S. Miller [Fri, 2 Mar 2007 02:08:05 +0000 (18:08 -0800)] 
[SPARC]: Handle unresolvable resources better in of_device.c

Just leave them as zero if we couldn't calculate it.

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[SPARC]: Fix bus handling in build_device_resources().
David S. Miller [Thu, 1 Mar 2007 07:20:12 +0000 (23:20 -0800)] 
[SPARC]: Fix bus handling in build_device_resources().

We mistakedly modify 'bus' in the innermost loop.  What
should happen is that at each register index iteration,
we start with the same 'bus'.

So preserve it's value at the top level, and use a loop
local variable 'dbus' for iteration.

This bug causes registers other than the first to be
decoded improperly.

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years agopata_qdi: Fix initialisation
Alan Cox [Fri, 2 Mar 2007 14:56:35 +0000 (14:56 +0000)] 
pata_qdi: Fix initialisation

The QDI init code contains some bugs which mean it only works if you have
a test setup that causes both a successful and failed probe. Fix this

Found by Philip Guo

(Who found it working on code analysis tools not running VLB IDE
controllers)

Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years ago[libata] pata_cmd64x: fix driver description in comments
Jeff Garzik [Fri, 2 Mar 2007 23:17:22 +0000 (18:17 -0500)] 
[libata] pata_cmd64x: fix driver description in comments

Trivial comment fix, taken out of a larger Alan Cox patch.

Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years ago[libata] pata_{legacy,sc1200,sl82c105}: add missing hooks
Jeff Garzik [Fri, 2 Mar 2007 23:09:05 +0000 (18:09 -0500)] 
[libata] pata_{legacy,sc1200,sl82c105}: add missing hooks

Alan Cox noticed several hooks in pata_* drivers were missing, when
he authored his ->cable_detect hook patches.  This patch extracts
just those fixes from Alan's patches, adding the necessary hooks
(usually ->freeze, ->thaw, and ->post_internal_cmd) to the drivers.

Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years ago[libata] change master/slave IDENTIFY order
Jeff Garzik [Fri, 2 Mar 2007 22:47:28 +0000 (17:47 -0500)] 
[libata] change master/slave IDENTIFY order

2.6.21-rc has horrible problems with libata and PATA cable types (and
thus speeds). This occurs because Tejun fixed a pile of other bugs and
we now do cable detect enforcement for drive side detection properly.

Unfortunately we don't do the process around cable detection right. Tejun
identified the problem and pointed to the right Annex in the spec, this patch
implements the needed changes.

The basic requirement is that we have to identify the slave before the
master.

The patch switches the identify order so that we can do the drive side
detection correctly.

[NOTE: patch and description extracted from a larger work written
and signed-off-by Alan Cox]

Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agolibata-core: Fix simplex handling
Alan [Thu, 1 Mar 2007 17:36:46 +0000 (17:36 +0000)] 
libata-core: Fix simplex handling

The initial simplex handling code is fooled if you suspend and resume.
This also causes problems with some single channel controllers which
claim to be simplex.

The fix is fairly simple, instead of keeping a flag to remember if we
gave away the simplex channel we remember the actual owner. As the owner
is always part of the host_set we don't even need a refcount.

Knowing the owner also means we can reassign simplex DMA channels in
future hotplug code etc if we need to

Signed-off-by: Alan Cox <alan@redhat.com>
(and a signed-off for the patch I sent before while I remember)
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agoMerge branch 'omap-fixes' of master.kernel.org:/pub/scm/linux/kernel/git/tmlind/linux...
Russell King [Fri, 2 Mar 2007 11:59:16 +0000 (11:59 +0000)] 
Merge branch 'omap-fixes' of /linux/kernel/git/tmlind/linux-omap-2.6

17 years ago[ARM] 4241/1: Define mb() as compiler barrier on a uniprocessor system
Catalin Marinas [Wed, 28 Feb 2007 11:30:38 +0000 (12:30 +0100)] 
[ARM] 4241/1: Define mb() as compiler barrier on a uniprocessor system

Currently, the mb() is defined as a DMB operation on ARMv6, even for
UP systems. This patch defines mb() as a compiler barrier only. For
the SMP case, the smp_* variants should be used anyway and the patch
defines them as DMB.

Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years ago[ARM] 4239/1: S3C24XX: Update kconfig entries for PM
Ben Dooks [Tue, 27 Feb 2007 23:23:20 +0000 (00:23 +0100)] 
[ARM] 4239/1: S3C24XX: Update kconfig entries for PM

Update help text with location of documentation
and duplicate the note on the speed of CRC

Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years ago[ARM] 4238/1: S3C24XX: docs: update suspend and resume
Ben Dooks [Tue, 27 Feb 2007 23:16:26 +0000 (00:16 +0100)] 
[ARM] 4238/1: S3C24XX: docs: update suspend and resume

Remove some of the explicit use of S3C2410 where
it is generic to all the S3C24XX series. Add more
info on the CRC code, and add an example of using
IRQ_EINT0 to resume from suspend

Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years ago[ARM] 4237/2: oprofile: Always allow backtraces on ARM
Richard Purdie [Tue, 27 Feb 2007 11:09:33 +0000 (12:09 +0100)] 
[ARM] 4237/2: oprofile: Always allow backtraces on ARM

Always allow backtrace when using oprofile on ARM, even if a PMU
isn't present.

Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years ago[ARM] Yet more asm/apm-emulation.h stuff
Russell King [Tue, 27 Feb 2007 12:10:07 +0000 (12:10 +0000)] 
[ARM] Yet more asm/apm-emulation.h stuff

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years agoARM: OMAP: Add missing get_irqnr_preamble and arch_ret_to_user for omap2
Tony Lindgren [Thu, 1 Mar 2007 12:46:43 +0000 (04:46 -0800)] 
ARM: OMAP: Add missing get_irqnr_preamble and arch_ret_to_user for omap2

Only entries for omap1 were added earlier.

Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: Use linux/delay.h not asm/delay.h
Tony Lindgren [Mon, 12 Feb 2007 18:50:53 +0000 (10:50 -0800)] 
ARM: OMAP: Use linux/delay.h not asm/delay.h

Use linux/delay.h not asm/delay.h

Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: Remove obsolete alsa typedefs
Tony Lindgren [Mon, 12 Feb 2007 18:16:43 +0000 (10:16 -0800)] 
ARM: OMAP: Remove obsolete alsa typedefs

Remove obsolete alsa typedefs

Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: omap1510->15xx conversions needed for sx1
Vladimir Ananiev [Mon, 11 Dec 2006 21:30:21 +0000 (13:30 -0800)] 
ARM: OMAP: omap1510->15xx conversions needed for sx1

Convert 1510->15xx in generic omap code, so that sx1 can work.

Signed-off-by: Pavel Machek <pavel@suse.cz>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: Add missing includes to board-nokia770
Dirk Behme [Fri, 26 Jan 2007 00:29:42 +0000 (16:29 -0800)] 
ARM: OMAP: Add missing includes to board-nokia770

Add missing includes to board-nokia770 to make it
compile again.

Signed-off-by: Dirk Behme <dirk.behme_at_gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: Workqueue changes for board-h4.c
Dirk Behme [Fri, 26 Jan 2007 00:29:17 +0000 (16:29 -0800)] 
ARM: OMAP: Workqueue changes for board-h4.c

Workqueue changes for board-h4.c

Signed-off-by: Dirk Behme <dirk.behme_at_gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: dmtimer.c omap1 register fix
Matthew Percival [Fri, 26 Jan 2007 00:24:29 +0000 (16:24 -0800)] 
ARM: OMAP: dmtimer.c omap1 register fix

When I went to use dmtimer7 it did not seem to work.  I noticed that
the base addresses for dmtimers 7 and 8 were set wrong.  A simple patch
to correct a small error.  Confirmed to fix the problem on an OSK.

Signed-off-by: Matthew Percival <matthew@capgo.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: board-nokia770: correct lcd name
Arnaud Patard [Thu, 7 Dec 2006 01:14:06 +0000 (17:14 -0800)] 
ARM: OMAP: board-nokia770: correct lcd name

Some time ago, the 'lcd_lph8923' device was renamed to 'lcd_mipid' but
the board-nokia770.c file was not updated accordingly, leading to not
working lcd.

This one-liner fixe the trouble.

Signed-off-by: Arnaud Patard <arnaud.patard@rtp-net.org>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: omap GP timer: HZ != 100
David Brownell [Thu, 7 Dec 2006 01:14:04 +0000 (17:14 -0800)] 
ARM: OMAP: omap GP timer: HZ != 100

Teach OMAP2 gp timer that HZ isn't always 100.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: Fix warning in mach-omap1
Dirk Behme [Thu, 7 Dec 2006 01:14:04 +0000 (17:14 -0800)] 
ARM: OMAP: Fix warning in mach-omap1

Fix warning

arch/arm/mach-omap1/pm.c: In function 'omap_pm_init':
arch/arm/mach-omap1/pm.c:765: warning: ignoring return value
of 'subsys_create_file', declared with attribute
warn_unused_result

Signed-off-by: Dirk Behme <dirk.behme_at_gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: Fix CONFIG_DEBUG_LL
Dirk Behme [Thu, 7 Dec 2006 01:14:02 +0000 (17:14 -0800)] 
ARM: OMAP: Fix CONFIG_DEBUG_LL

Fix broken CONFIG_DEBUG_LL. In case of low level debugging
reconfigure some clocks early.

See

http://source.mvista.com/git/gitweb.cgi?p=linux-omap-2.6.git;a=commitdiff;h=3bfb289ccc6c4624fd5ff0381546935e105f4093

too.

Signed-off-by: Dirk Behme <dirk.behme_at_gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: Fix warning in mach-omap2
Dirk Behme [Thu, 7 Dec 2006 01:14:00 +0000 (17:14 -0800)] 
ARM: OMAP: Fix warning in mach-omap2

Fix warning

arch/arm/mach-omap2/timer-gp.c: In function
'omap2_gp_timer_init':
arch/arm/mach-omap2/timer-gp.c:70: warning: implicit
declaration of function 'setup_irq'

Signed-off-by: Dirk Behme <dirk.behme_at_gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: Fix warnings in plat-omap
Dirk Behme [Thu, 7 Dec 2006 01:14:00 +0000 (17:14 -0800)] 
ARM: OMAP: Fix warnings in plat-omap

Fix warnings

arch/arm/plat-omap/dmtimer.c: In function
'omap_dm_timer_modify_idlect_mask':
arch/arm/plat-omap/dmtimer.c:317: warning: no return
statement in function returning non-void

arch/arm/plat-omap/mailbox.c: In function 'omap_mbox_init':
arch/arm/plat-omap/mailbox.c:231: warning: ignoring return
value of 'class_device_create_file', declared with attribute
warn_unused_result

Signed-off-by: Dirk Behme <dirk.behme_at_gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: No IRQF_TRIGGER set_type function for IRQ 353 (MPUIO)
Dirk Behme [Thu, 7 Dec 2006 01:13:57 +0000 (17:13 -0800)] 
ARM: OMAP: No IRQF_TRIGGER set_type function for IRQ 353 (MPUIO)

No IRQF_TRIGGER set_type function for IRQ 353 (MPUIO)

Signed-off-by: Dirk Behme <dirk.behme_at_gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: OMAP310 Serial
Marek Vašut [Thu, 7 Dec 2006 01:13:56 +0000 (17:13 -0800)] 
ARM: OMAP: OMAP310 Serial

This makes serial usable also on omap310, not only 1510.
(changing 1510->15xx)

Signed-off-by: Marek Vašut <marek.vasut@gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: Proper handling of DMA4_IRQSTATUS_L0
Timo Teras [Thu, 7 Dec 2006 01:13:52 +0000 (17:13 -0800)] 
ARM: OMAP: Proper handling of DMA4_IRQSTATUS_L0

The register bits are reset by writing one. Remove the unneeded reads and
fix writes to not clear too many bits.

Signed-off-by: Timo Teras <timo.teras@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
17 years agoARM: OMAP: Fix warning in clock.c
Dirk Behme [Thu, 7 Dec 2006 01:13:51 +0000 (17:13 -0800)] 
ARM: OMAP: Fix warning in clock.c

Fix warning:

arch/arm/mach-omap1/clock.c: In function
'omap1_clk_enable_generic':
arch/arm/mach-omap1/clock.c:499: warning: 'return' with no
value, in function returning non-void

Signed-off-by: Dirk Behme <dirk.behme_at_gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>