15 years ago[PATCH] uml skas0: stop gcc's insanity
Jeff Dike [Sun, 18 Dec 2005 16:50:39 +0000]
[PATCH] uml skas0: stop gcc's insanity

With Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>

UML skas0 stub has been miscompiling for many people (incidentally not
the authors), depending on the used GCC versions.

I think (and testing on some GCC versions shows) this patch avoids the
fundamental issue which is behind this, namely gcc using the stack when
we have just replaced it, behind gcc's back.  The remapping and storage
of the return value is hidden in a blob of asm, hopefully giving gcc no
room for creativity.

Signed-off-by: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] uml - fix some funkiness in Kconfig
Paolo 'Blaisorblade' Giarrusso [Sun, 18 Dec 2005 16:50:37 +0000]
[PATCH] uml - fix some funkiness in Kconfig

So you may have seen the miniconfig stuff wander by, which means that my
build script exits if there's a .config error, and we have this:

  fs/Kconfig:1749:warning: 'select' used by config symbol 'CIFS_UPCALL'
refer to undefined symbol 'CONNECTOR'

This makes it shut up.

Signed-off-by: Rob Landley <rob@landley.net>
[ Verified it makes sense. ]
Signed-off-by: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] uml: fix dynamic linking on some 64-bit distros
Rob Landley [Sun, 18 Dec 2005 16:50:35 +0000]
[PATCH] uml: fix dynamic linking on some 64-bit distros

With Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>

The current UML build assumes that on x86-64 systems, /lib is a symlink
to /lib64, but in some distributions (like PLD and CentOS) they are
separate directories, so the 64 bit library loader isn't found.  This
patch inserts /lib64 at the start of the rpath on x86-64 UML builds.

Signed-off-by: Rob Landley <rob@landley.net>
Signed-off-by: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] uml: arch/um/scripts/Makefile.rules - remove duplicated code
Paolo 'Blaisorblade' Giarrusso [Sun, 18 Dec 2005 16:50:32 +0000]
[PATCH] uml: arch/um/scripts/Makefile.rules - remove duplicated code

Duplicated code - the patch adding it was probably applied twice without
enough care.

Signed-off-by: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] dpt_i2o fix for deadlock condition
Salyzyn, Mark [Sun, 18 Dec 2005 03:26:30 +0000]
[PATCH] dpt_i2o fix for deadlock condition

Miquel van Smoorenburg <miquels@cistron.nl> forwarded me this fix to
resolve a deadlock condition that occurs due to the API change in
2.6.13+ kernels dropping the host locking when entering the error
handling.  They all end up calling adpt_i2o_post_wait(), which if you
call it unlocked, might return with host_lock locked anyway and that
causes a deadlock.

Signed-off-by: Mark Salyzyn <aacraid@adaptec.com>
Cc: James Bottomley <James.Bottomley@steeleye.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] i2o: Do not disable pci device when it's in use
Ben Collins [Sun, 18 Dec 2005 02:39:23 +0000]
[PATCH] i2o: Do not disable pci device when it's in use

When dpt_i2o is loaded first, i2o being loaded would cause it to call
pci_device_disable, thus breaking dpt_i2o's use of the device.  Based on
similar usage of pci_disable_device in other drivers.

Signed-off-by: Ben Collins <bcollins@ubuntu.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years agoMerge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6
Linus Torvalds [Fri, 16 Dec 2005 22:43:57 +0000]
Merge branch 'release' of git://git./linux/kernel/git/aegl/linux-2.6

15 years ago[PATCH] ppc: ppc4xx_dma DMA_MODE_{READ,WRITE} fix
Al Viro [Fri, 16 Dec 2005 22:35:23 +0000]
[PATCH] ppc: ppc4xx_dma DMA_MODE_{READ,WRITE} fix

DMA_MODE_{READ,WRITE} are declared in asm-powerpc/dma.h and their
declarations there match the definitions.  Old declarations in
ppc4xx_dma.h are not right anymore (wrong type, to start with).
Killed them, added include of asm/dma.h where needed.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] ppc: booke_wdt compile fix
Al Viro [Fri, 16 Dec 2005 22:35:28 +0000]
[PATCH] ppc: booke_wdt compile fix

booke_wdt.c had been missed in cpu_specs[] removal sweep

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] UHCI: add missing memory barriers
Alan Stern [Fri, 16 Dec 2005 19:09:01 +0000]
[PATCH] UHCI: add missing memory barriers

This patch (as617) adds a couple of memory barriers that Ben H. forgot in
his recent suspend/resume fix.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] PCI: Fix dumb bug in mmconfig fix
Andi Kleen [Fri, 16 Dec 2005 19:08:55 +0000]
[PATCH] PCI: Fix dumb bug in mmconfig fix

Use correct address when referencing mmconfig aperture while checking
for broken MCFG.  This was a typo when porting the code from 64bit to
32bit.  It caused oopses at boot on some ThinkPads.

Should definitely go into 2.6.15.

Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] PCI express must be initialized before PCI hotplug
Milton Miller [Fri, 16 Dec 2005 19:08:48 +0000]
[PATCH] PCI express must be initialized before PCI hotplug

PCI express hotplug uses the pcieportbus driver so pcie must be
initialized before hotplug/.  This patch changes the link order.

Signed-Off-By: Milton Miller <miltonm@bga.com>
Acked-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] i2c: Fix i2c-mv64xxx compilation error
Mark A. Greer [Fri, 16 Dec 2005 19:08:43 +0000]
[PATCH] i2c: Fix i2c-mv64xxx compilation error

The busses/i2c-mv64xxx.c driver doesn't currently compile because of an
incorrect argument to dev_err().  This patch fixes that.

Signed-off-by: Mark A. Greer <mgreer@mvista.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[IA64] Add __read_mostly support for IA64
Christoph Lameter [Mon, 12 Dec 2005 17:34:32 +0000]
[IA64] Add __read_mostly support for IA64

sparc64, i386 and x86_64 have support for a special data section dedicated
to rarely updated data that is frequently read. The section was created to
avoid false sharing of those rarely read data with frequently written kernel
data.

This patch creates such a data section for ia64 and will group rarely written
data into this section.

Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>

15 years ago[IA64-SGI] change default_sn2 to NR_CPUS==1024
hawkes@sgi.com [Tue, 13 Dec 2005 21:45:44 +0000]
[IA64-SGI] change default_sn2 to NR_CPUS==1024

Change the NR_CPUS default for ia64/sn up to 1024.

Signed-off-by: John Hawkes <hawkes@sgi.com>
Signed-off-by: John Hesterberg <jh@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>

15 years ago[IA64-SGI] Missed TLB flush
Jack Steiner [Thu, 15 Dec 2005 18:41:22 +0000]
[IA64-SGI] Missed TLB flush

I see why the problem exists only on SN. SN uses a different hardware
mechanism to purge TLB entries across nodes.

It looks like there is a bug in the SN TLB flushing code. During context switch,
kernel threads inherit the mm of the task that was previously running on the
cpu. This confuses the code in sn2_global_tlb_purge().

The result is a missed TLB purge for the task that owns the "borrowed" mm.

(I hit the problem running heavy stress where kswapd was purging code pages of
a user task that woke kswapd. The user task took a SIGILL fault trying to
execute code in the page that had been ripped out from underneath it).

Signed-off-by: Jack Steiner <steiner@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>

15 years ago[IA64] uncached ref count leak
Jes Sorensen [Fri, 16 Dec 2005 16:00:03 +0000]
[IA64] uncached ref count leak

Use raw_smp_processor_id() instead of get_cpu() as we don't need the
extra features of get_cpu().

Signed-off-by: Jes Sorensen <jes@trained-monkey.org>
Signed-off-by: Tony Luck <tony.luck@intel.com>

15 years agoMake sure we copy pages inserted with "vm_insert_page()" on fork
Linus Torvalds [Fri, 16 Dec 2005 18:21:23 +0000]
Make sure we copy pages inserted with "vm_insert_page()" on fork

The logic that decides that a fork() might be able to avoid copying a VM
area when it can be re-created by page faults didn't know about the new
vm_insert_page() case.

Also make some things a bit more anal wrt VM_PFNMAP.

Pointed out by Hugh Dickins <hugh@veritas.com>

Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[IA64] disable preemption in udelay()
John Hawkes [Fri, 16 Dec 2005 18:00:24 +0000]
[IA64] disable preemption in udelay()

The udelay() inline for ia64 uses the ITC.  If CONFIG_PREEMPT is enabled
and the platform has unsynchronized ITCs and the calling task migrates
to another CPU while doing the udelay loop, then the effective delay may
be too short or very, very long.

This patch disables preemption around 100 usec chunks of the overall
desired udelay time.  This minimizes preemption-holdoffs.

udelay() is now too big to be inline, move it out of line and export it.

Signed-off-by: John Hawkes <hawkes@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>

15 years ago[PATCH] radeon drm: fix agp aperture map offset
Benjamin Herrenschmidt [Fri, 16 Dec 2005 05:52:22 +0000]
[PATCH] radeon drm: fix agp aperture map offset

This finally fixes the radeon memory mapping bug that was incorrectly
fixed by the previous patch.  This time, we use the actual vram size as
the size to calculate how far to move the AGP aperture from the
framebuffer in card's memory space.

If there are still issues with this patch, they are due to bugs in the X
driver that I'm working on fixing too.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Mark M. Hoffman <mhoffman@lightlink.com>
Cc: Paul Mackerras <paulus@samba.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] Au1550 AC'97 OSS driver spinlock fixes
Sergei Shtylylov [Thu, 15 Dec 2005 20:34:30 +0000]
[PATCH] Au1550 AC'97 OSS driver spinlock fixes

We have found some issues with Au1550 AC'97 OSS driver in 2.6
(sound/oss/au1550_ac97.c), though it also should concern 2.4 driver
(drivers/sound/au1550_psc.c).

start_dac() grabs a spinlock already held by its caller, au1550_write().
This doesn't show up with the standard UP spinlock impelmentation but when
the different one (mutex based) is in use, a lockup happens.

And the interrupt handlers also didn't grab the spinlock -- that's OK in
the usual kernel but not when the IRQ handlers are threaded.  So, they're
grabbing the spinlock now (as every correct interrupt handler should do).

Signed-off-by: Konstantin Baidarov <kbaidarov@ru.mvista.com>
Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] drivers/base/memory.c: unexport the static (sic) memory_sysdev_class
Adrian Bunk [Thu, 15 Dec 2005 20:34:29 +0000]
[PATCH] drivers/base/memory.c: unexport the static (sic) memory_sysdev_class

We can't export a static struct to modules.

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] IPMI oops fix
Paolo Galtieri [Thu, 15 Dec 2005 20:34:28 +0000]
[PATCH] IPMI oops fix

While doing some testing I discovered that if the BIOS on a board does not
properly setup the DMI information it leads to a panic in the IPMI code.

The panic is due to dereferencing a pointer which is not initialized.  The
pointer is initialized in port_setup() and/or mem_setup() and used in
init_one_smi() and cleanup_one_si(), however if either port_setup() or
mem_setup() return ENODEV the pointer does not get initialized.

Signed-off-by: Paolo Galtieri <pgaltieri@mvista.com>
Acked-by: Corey Minyard <cminyard@mvista.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] arch/powerpc/kernel/syscalls.c __user annotations
Al Viro [Thu, 15 Dec 2005 09:19:10 +0000]
[PATCH] arch/powerpc/kernel/syscalls.c __user annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] drivers/input/misc/wistron_btns.c NULL noise removal
Al Viro [Thu, 15 Dec 2005 09:19:05 +0000]
[PATCH] drivers/input/misc/wistron_btns.c NULL noise removal

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] mwave: missing __user in ioctl struct declaration
Al Viro [Thu, 15 Dec 2005 09:19:00 +0000]
[PATCH] mwave: missing __user in ioctl struct declaration

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] drivers/atm/adummy.c NULL noise removal
Al Viro [Thu, 15 Dec 2005 09:18:55 +0000]
[PATCH] drivers/atm/adummy.c NULL noise removal

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] arch/alpha/kernel/machvec_impl.h: C99 struct initializer
Al Viro [Thu, 15 Dec 2005 09:18:50 +0000]
[PATCH] arch/alpha/kernel/machvec_impl.h: C99 struct initializer

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] dst_ca __user annotations, portability fixes
Al Viro [Thu, 15 Dec 2005 09:18:45 +0000]
[PATCH] dst_ca __user annotations, portability fixes

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] ia64 sn __iomem annotations
Al Viro [Thu, 15 Dec 2005 09:18:40 +0000]
[PATCH] ia64 sn __iomem annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] Address of void __user * is void __user * *, not void * __user *
Al Viro [Thu, 15 Dec 2005 09:18:30 +0000]
[PATCH] Address of void __user * is void __user * *, not void * __user *

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] missing prototype (mm/page_alloc.c)
Al Viro [Thu, 15 Dec 2005 09:18:25 +0000]
[PATCH] missing prototype (mm/page_alloc.c)

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] __user annotations (booke_wdt.c)
Al Viro [Thu, 15 Dec 2005 09:18:20 +0000]
[PATCH] __user annotations (booke_wdt.c)

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] arcfb __user annotations
Al Viro [Thu, 15 Dec 2005 09:18:15 +0000]
[PATCH] arcfb __user annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] cyber2000fb.c __iomem annotations
Al Viro [Thu, 15 Dec 2005 09:18:10 +0000]
[PATCH] cyber2000fb.c __iomem annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] wdrtas.c: fix __user annotations
Al Viro [Thu, 15 Dec 2005 09:18:05 +0000]
[PATCH] wdrtas.c: fix __user annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] dell_rbu: NULL noise removal
Al Viro [Thu, 15 Dec 2005 09:18:00 +0000]
[PATCH] dell_rbu: NULL noise removal

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] cm4000_cs: __user annotations
Al Viro [Thu, 15 Dec 2005 09:17:55 +0000]
[PATCH] cm4000_cs: __user annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] i386,amd64: ioremap.c __iomem annotations
Al Viro [Thu, 15 Dec 2005 09:17:50 +0000]
[PATCH] i386,amd64: ioremap.c __iomem annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] i386,amd64: mmconfig __iomem annotations
Al Viro [Thu, 15 Dec 2005 09:17:44 +0000]
[PATCH] i386,amd64: mmconfig __iomem annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] em28xx: %zd for size_t
Al Viro [Thu, 15 Dec 2005 09:17:39 +0000]
[PATCH] em28xx: %zd for size_t

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] auerswald.c: %zd for size_t
Al Viro [Thu, 15 Dec 2005 09:17:34 +0000]
[PATCH] auerswald.c: %zd for size_t

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] s2io: __iomem annotations for recent changes
Al Viro [Thu, 15 Dec 2005 09:17:29 +0000]
[PATCH] s2io: __iomem annotations for recent changes

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] xfs: missing gfp_t annotations
Al Viro [Thu, 15 Dec 2005 09:17:24 +0000]
[PATCH] xfs: missing gfp_t annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] iscsi gfp_t annotations
Al Viro [Thu, 15 Dec 2005 09:17:19 +0000]
[PATCH] iscsi gfp_t annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] arch/sparc/kernel/led.c __user annotations
Al Viro [Tue, 6 Dec 2005 11:04:55 +0000]
[PATCH] arch/sparc/kernel/led.c __user annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] sun4c_memerr_reg __iomem annotations
Al Viro [Tue, 6 Dec 2005 10:56:39 +0000]
[PATCH] sun4c_memerr_reg __iomem annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] sparc: NULL noise removal (ebus.c)
Al Viro [Sun, 4 Dec 2005 23:48:45 +0000]
[PATCH] sparc: NULL noise removal (ebus.c)

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] sparc/kernel/time: __iomem annotations
Al Viro [Tue, 6 Dec 2005 10:55:44 +0000]
[PATCH] sparc/kernel/time: __iomem annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] sbus/char/uctrl: missing prototypes and NULL noise removal
Al Viro [Tue, 6 Dec 2005 10:54:54 +0000]
[PATCH] sbus/char/uctrl: missing prototypes and NULL noise removal

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] sparc: jsflash __user annotations
Al Viro [Tue, 6 Dec 2005 10:51:43 +0000]
[PATCH] sparc: jsflash __user annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] sparc: vfc __iomem annotations and fixes
Al Viro [Tue, 6 Dec 2005 10:50:56 +0000]
[PATCH] sparc: vfc __iomem annotations and fixes

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] fix iomem annotations in sparc32 pcic code
Al Viro [Tue, 6 Dec 2005 10:44:18 +0000]
[PATCH] fix iomem annotations in sparc32 pcic code

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/bart/ide-2.6
Linus Torvalds [Thu, 15 Dec 2005 02:58:46 +0000]
Merge /pub/scm/linux/kernel/git/bart/ide-2.6

15 years ago[PATCH] reiserfs: close open transactions on error path
Jeff Mahoney [Wed, 14 Dec 2005 19:38:05 +0000]
[PATCH] reiserfs: close open transactions on error path

The following patch fixes a bug where if the journal is aborted, it can
leave a transaction open.  The result will be a BUG when another code
path attempts to start a transaction and will get a "nesting into
different fs" error, since current->journal_info will be left non-NULL.

Original fix against SUSE kernel by Chris Mason <mason@suse.com>

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] reiserfs: skip commit on io error
Jeff Mahoney [Wed, 14 Dec 2005 19:38:36 +0000]
[PATCH] reiserfs: skip commit on io error

This should have been part of the original io error patch, but got
dropped somewhere along the way.

It's extremely important when handling the i/o error in the journal to
not commit the transaction with corrupt data.  This patch adds that code
back in.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] Input: ALPS - correctly report button presses on Fujitsu Siemens S6010
Vojtech Pavlik [Thu, 15 Dec 2005 02:47:00 +0000]
[PATCH] Input: ALPS - correctly report button presses on Fujitsu Siemens S6010

Without this patch Forward and Backward buttons on the touchpad do not
generate any events.

Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years agoMove size optimization option outside of EMBEDDED menu, mark it EXPERIMENTAL
Linus Torvalds [Thu, 15 Dec 2005 02:52:21 +0000]
Move size optimization option outside of EMBEDDED menu, mark it EXPERIMENTAL

Also, disable on sparc64 - a number of people report breakage.  Probably
a compiler bug, but it's quite possible that it tickles some latent
kernel problem too.

It still defaults to 'y' everywhere else (when enabled through
EXPERIMENTAL), and Dave Jones points out that Fedora (and RHEL4) has
been building with size optimizations for a long time on x86, x86-64,
ia64, s390, s390x, ppc32 and ppc64.  So it is really only moderately
experimental, but the sparc64 breakage certainly shows that it can
trigger "issues".

Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
Linus Torvalds [Thu, 15 Dec 2005 02:46:41 +0000]
Merge /pub/scm/linux/kernel/git/davem/net-2.6

15 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/jejb/scsi-rc-fixes-2.6
Linus Torvalds [Thu, 15 Dec 2005 02:40:02 +0000]
Merge /linux/kernel/git/jejb/scsi-rc-fixes-2.6

15 years ago[PATCH] ide-cd: remove write-only cmd field from struct cdrom_info
Bartlomiej Zolnierkiewicz [Thu, 15 Dec 2005 01:20:49 +0000]
[PATCH] ide-cd: remove write-only cmd field from struct cdrom_info

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

15 years ago[PATCH] ide: cleanup ide_driver_t
Bartlomiej Zolnierkiewicz [Thu, 15 Dec 2005 01:20:15 +0000]
[PATCH] ide: cleanup ide_driver_t

Remove unused fields: ioctl, ata[pi]_prebuilder.

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

15 years ago[PATCH] ide: cleanup ide.h
Bartlomiej Zolnierkiewicz [Thu, 15 Dec 2005 01:19:57 +0000]
[PATCH] ide: cleanup ide.h

Remove:
* stale comment
* unused HOST() macro
* unused ata_{error,control}_t types
* unused atapi_select_t type
* ide_init_subdrivers() prototype

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

15 years ago[PATCH] ide-disk: flush cache after calling del_gendisk()
Bartlomiej Zolnierkiewicz [Thu, 15 Dec 2005 01:19:20 +0000]
[PATCH] ide-disk: flush cache after calling del_gendisk()

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

15 years ago[PATCH] ide: AU1200 IDE update
Jordan Crouse [Thu, 15 Dec 2005 01:17:46 +0000]
[PATCH] ide: AU1200 IDE update

Changes here include removing all of CONFIG_PM while it is being repeatedly
smacked with a lead pipe, moving the BURSTMODE param to a #define (it should
be defined almost always anyway), fixing the rqsize stuff, pulling ide_ioreg_t,
and general cleanups and whatnot.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

15 years ago[PATCH] ide: core modifications for AU1200
Jordan Crouse [Thu, 15 Dec 2005 01:16:18 +0000]
[PATCH] ide: core modifications for AU1200

bart: slightly modified by me

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

15 years ago[PATCH] ide: MPC8xx IDE depends on IDE=y && BLK_DEV_IDE=y
Marcelo Tosatti [Thu, 15 Dec 2005 01:12:53 +0000]
[PATCH] ide: MPC8xx IDE depends on IDE=y && BLK_DEV_IDE=y

The following patch adds a dependancy on IDE=y && BLK_DEV_IDE=y
for the MPC8xx IDE driver.

The code is not modular at the moment (init called from platform setup code).

Signed-off-by: Marcelo Tosatti <marcelo.tosatti@cyclades.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

15 years ago[PATCH] via82cxxx IDE: Add VT8251 ISA bridge
Daniel Drake [Thu, 15 Dec 2005 01:11:55 +0000]
[PATCH] via82cxxx IDE: Add VT8251 ISA bridge

Some motherboards (such as the Asus P5V800-MX) ship a
PCI_DEVICE_ID_VIA_82C586_1 IDE controller alongside a VT8251 southbridge.

This southbridge is currently unrecognised in the via82cxxx IDE driver,
preventing those users from getting DMA access to disks.

Signed-off-by: Daniel Drake <dsd@gentoo.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

15 years ago[PATCH] sgiioc4: check for no hwifs available
Jeremy Higdon [Thu, 15 Dec 2005 01:10:35 +0000]
[PATCH] sgiioc4: check for no hwifs available

Add a check to the sgiioc4 driver for the case where all available
ide_hwifs structures are in use.

Signed-off-by: Jeremy Higdon <jeremy@sgi.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

15 years ago[VLAN]: Fix hardware rx csum errors
Stephen Hemminger [Thu, 15 Dec 2005 00:23:16 +0000]
[VLAN]: Fix hardware rx csum errors

Receiving VLAN packets over a device (without VLAN assist) that is
doing hardware checksumming (CHECKSUM_HW), causes errors because the
VLAN code forgets to adjust the hardware checksum.

Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

15 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/torvalds/linux-2.6
David S. Miller [Thu, 15 Dec 2005 00:21:48 +0000]
Merge /pub/scm/linux/kernel/git/torvalds/linux-2.6

15 years agoMerge master.kernel.org:/home/rmk/linux-2.6-arm
Linus Torvalds [Wed, 14 Dec 2005 23:48:06 +0000]
Merge master.kernel.org:/home/rmk/linux-2.6-arm

15 years agoMerge master.kernel.org:/home/rmk/linux-2.6-mmc
Linus Torvalds [Wed, 14 Dec 2005 23:46:46 +0000]
Merge master.kernel.org:/home/rmk/linux-2.6-mmc

15 years ago[ARM] 3205/1: Handle new EABI relocations when loading kernel modules.
Daniel Jacobowitz [Wed, 14 Dec 2005 22:04:22 +0000]
[ARM] 3205/1: Handle new EABI relocations when loading kernel modules.

Patch from Daniel Jacobowitz

Handle new EABI relocations when loading kernel modules.  This is
necessary for CONFIG_AEABI kernels, and also for some broken
(since fixed) old ABI toolchains.

Signed-off-by: Daniel Jacobowitz <dan@codesourcery.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

15 years ago[GRE]: Fix hardware checksum modification
Herbert Xu [Wed, 14 Dec 2005 20:55:24 +0000]
[GRE]: Fix hardware checksum modification

The skb_postpull_rcsum introduced a bug to the checksum modification.
Although the length pulled is offset bytes, the origin of the pulling
is the GRE header, not the IP header.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

15 years ago[IA64] fix for SET_PERSONALITY when CONFIG_IA32_SUPPORT is not set.
Robin Holt [Wed, 14 Dec 2005 12:58:05 +0000]
[IA64] fix for SET_PERSONALITY when CONFIG_IA32_SUPPORT is not set.

Missed this when fixing the SET_PERSONALITY change.

Signed-off-by: Robin Holt <holt@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>

15 years ago[MMC] Explain the internals of mmc_power_up()
Russell King [Wed, 14 Dec 2005 14:57:35 +0000]
[MMC] Explain the internals of mmc_power_up()

It seems that people get confused about what is happening in
mmc_power_up().  Add a comment to make it clear why we have
a two stage process.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

15 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/davem/tg3-2.6
Linus Torvalds [Wed, 14 Dec 2005 07:08:24 +0000]
Merge /pub/scm/linux/kernel/git/davem/tg3-2.6

15 years ago[PKT_SCHED]: Disable debug tracing logs by default in packet action API.
David S. Miller [Wed, 14 Dec 2005 06:59:50 +0000]
[PKT_SCHED]: Disable debug tracing logs by default in packet action API.

Noticed by Andi Kleen.

Signed-off-by: David S. Miller <davem@davemloft.net>

15 years ago[IPV6] addrconf: Do not print device pointer in privacy log message.
David S. Miller [Wed, 14 Dec 2005 06:59:36 +0000]
[IPV6] addrconf: Do not print device pointer in privacy log message.

Noticed by Andi Kleen, it is pointless to emit the device
structure pointer in the kernel logs like this.

Signed-off-by: David S. Miller <davem@davemloft.net>

15 years ago[SCSI] Consolidate REQ_BLOCK_PC handling path (fix ipod panic)
James Bottomley [Fri, 9 Dec 2005 22:45:22 +0000]
[SCSI] Consolidate REQ_BLOCK_PC handling path (fix ipod panic)

This follows on from Jens' patch and consolidates all of the ULD
separate handlers for REQ_BLOCK_PC into a single call which has his
fix for our direction bug.

Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>

15 years ago[PATCH] hid-core: Zero-pad truncated reports
Adam Kropelin [Wed, 14 Dec 2005 01:03:39 +0000]
[PATCH] hid-core: Zero-pad truncated reports

When it detects a truncated report, hid-core emits a warning and then
processes the report as usual.  This is good because it allows buggy
devices to still get data thru to userspace.  However, the missing bytes of
the report should be cleared before processing, otherwise userspace will be
handed partially-uninitialized data.

This fixes Debian tracker bug #330487.

Signed-off-by: Adam Kropelin <akropel1@rochester.rr.com>
Cc: Vojtech Pavlik <vojtech@suse.cz>
Acked-by: Dmitry Torokhov <dtor_core@ameritech.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] fbdev: make pxafb more robust to errors with CONFIG_FB_PXA_PARAMETERS
Ole Reinhardt [Wed, 14 Dec 2005 01:03:38 +0000]
[PATCH] fbdev: make pxafb more robust to errors with CONFIG_FB_PXA_PARAMETERS

pxafb.c runs into an oops if CONFIG_FB_PXA_PARAMETERS is enabled and no
parameters are set in command line.  The following patch avoids this
problem.

Signed-off-by: Nicolas Pitre <nico@cam.org>
Cc: "Antonino A. Daplas" <adaplas@pol.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] Fix calculation of grow_pgdat_span() in mm/memory_hotplug.c
Yasunori Goto [Wed, 14 Dec 2005 01:03:37 +0000]
[PATCH] Fix calculation of grow_pgdat_span() in mm/memory_hotplug.c

The calculation for node_spanned_pages at grow_pgdat_span() is clearly
wrong.  This is patch for it.

(Please see grow_zone_span() to compare. It is correct.)

Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Acked-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] Dmitry Torokhov is input subsystem maintainer
Vojtech Pavlik [Wed, 14 Dec 2005 01:03:36 +0000]
[PATCH] Dmitry Torokhov is input subsystem maintainer

I haven't been very actively maintaining the input layer in past months,
mostly because of my lack of time to concentrate on that.  For that reason,
I've decided to pass the maintainership of the Linux Input Layer to Dmitry
Torokhov, whom I trust to do the job very well.

Signed-off-by: Vojtech Pavlik <vojtech@suse.cz>
Cc: Dmitry Torokhov <dtor_core@ameritech.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[TG3]: Fix low power state
Michael Chan [Wed, 14 Dec 2005 05:15:53 +0000]
[TG3]: Fix low power state

Fix the following bugs in tg3_set_power_state():

1. Both WOL and ASF flags require switching to aux power.

2. Add a missing handshake with firmware to enable WOL.

3. Turn off the PHY if both WOL and ASF are disabled.

4. Add nvram arbitration before halting the firmware.

5. Fix tg3_setup_copper_phy() to switch to 100Mbps when
   changing to low power state.

Update revision and date.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

15 years ago[TG3]: Fix 5704 single-port mode
Michael Chan [Wed, 14 Dec 2005 05:09:54 +0000]
[TG3]: Fix 5704 single-port mode

If the dual-port 5704 is configured as a single-port device with
only one PCI function, it would trigger a BUG() condition in
tg3_find_5704_peer(). This fixes the problem by returning its
own pdev if the peer cannot be found.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

15 years ago[TG3]: Fix suspend and resume
Michael Chan [Wed, 14 Dec 2005 05:08:58 +0000]
[TG3]: Fix suspend and resume

Fix tg3_suspend() and tg3_resume() by clearing and setting the
TG3_FLAG_INIT_COMPLETE flag when appropriate. tg3_set_power_state()
looks at TG3_FLAG_INIT_COMPLETE on the peer device to determine
when to appropriately switch to aux power.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

15 years ago[TG3]: Fix nvram arbitration bugs.
Michael Chan [Wed, 14 Dec 2005 05:08:21 +0000]
[TG3]: Fix nvram arbitration bugs.

The nvram arbitration rules were not strictly followed in a few places
and this could lead to reading corrupted values from the nvram.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

15 years ago[SCSI] Negotiate correctly with async-only devices
Matthew Wilcox [Wed, 30 Nov 2005 04:08:40 +0000]
[SCSI] Negotiate correctly with async-only devices

When we got a device only capable of async, we would zero out goal->period
which would cause us to try PPR negotiations.  Leave goal->period alone,
and check goal->offset before doing PPR.  Kudos to Daniel Forsgren for
figuring this out.

Signed-off-by: Matthew Wilcox <matthew@wil.cx>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>

15 years agoMerge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6
Linus Torvalds [Tue, 13 Dec 2005 20:36:54 +0000]
Merge branch 'release' of git://git./linux/kernel/git/aegl/linux-2.6

15 years agoExpose "Optimize for size" option for everybody
Linus Torvalds [Tue, 13 Dec 2005 19:39:05 +0000]
Expose "Optimize for size" option for everybody

Let's put my money where my mouth is.  Smaller code is almost always
faster, if only because a single I$ miss ends up leaving a lot of cycles
to make up for.  And system software - kernels in particular - are known
for taking more cache misses than most other kinds.

On my random config, this made the kernel about 10% smaller, and lmbench
seems to say that it's pretty uniformly faster too. Your milage may vary.

Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[IA64] Split 16-bit severity field in sal_log_record_header
Tony Luck [Tue, 13 Dec 2005 18:41:49 +0000]
[IA64] Split 16-bit severity field in sal_log_record_header

ERR_SEVERITY item is defined as a 8 bits item in SAL documentation
($B.2.1 rev december 2003), but as an u16 in sal.h.
This has the side effect that current code in mca.c may not call
ia64_sal_clear_state_info() upon receiving corrected platform errors
if there are bits set in the validation byte.  Reported by Xavier Bru.

Signed-off-by: Tony Luck <tony.luck@intel.com>

15 years ago[PATCH] allow KOBJECT_UEVENT=n only if EMBEDDED
Adrian Bunk [Tue, 13 Dec 2005 01:58:07 +0000]
[PATCH] allow KOBJECT_UEVENT=n only if EMBEDDED

KOBJECT_UEVENT=n seems to be a common pitfall for udev users in 2.6.14 .

-mm already contains a bigger patch removing this option that is IMHO
too big for being applied now to 2.6.15-rc.

This patch simply allows KOBJECT_UEVENT=n only if EMBEDDED.

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years agoMerge branch 'upstream-fixes' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik...
Linus Torvalds [Tue, 13 Dec 2005 16:53:56 +0000]
Merge branch 'upstream-fixes' of /linux/kernel/git/jgarzik/libata-dev

15 years ago[netdrvr skge] fix build
Jeff Garzik [Tue, 13 Dec 2005 16:35:22 +0000]
[netdrvr skge] fix build

15 years ago[libata] mark certain hardware (or drivers) with a no-atapi flag
Jeff Garzik [Tue, 13 Dec 2005 07:29:45 +0000]
[libata] mark certain hardware (or drivers) with a no-atapi flag

Some hardware does not support the PACKET command at all.
Other hardware supports ATAPI, but the driver does something nasty such
as calling BUG() when an ATAPI command is issued.

For these such cases, we mark them with a new flag, ATA_FLAG_NO_ATAPI.

Initial version contributed by Ben Collins.

15 years ago[PATCH] fbdev: Fix incorrect unaligned access in little-endian machines
Antonino A. Daplas [Tue, 13 Dec 2005 06:17:21 +0000]
[PATCH] fbdev: Fix incorrect unaligned access in little-endian machines

The drawing function cfbfillrect does not work correctly when access is not
unsigned-long aligned.  It manifests as extra lines of pixels that are not
complete drawn.  Reversing the shift operator solves the problem, so I would
presume that this bug would manifest only on little endian machines.  The
function cfbcopyarea may also have this bug.

Aligned access should present no problems.

Signed-off-by: Antonino Daplas <adaplas@pol.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] fbdev: Shift pixel value before entering loop in cfbimageblit
Antonino A. Daplas [Tue, 13 Dec 2005 06:17:20 +0000]
[PATCH] fbdev: Shift pixel value before entering loop in cfbimageblit

In slow imageblit, the pixel value is shifted by a certain amount (dependent
on the bpp and endianness) for each iteration.  This is inefficient.  Better
do the shifting once before going into the loop.

Signed-off-by: Antonino Daplas <adaplas@pol.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

15 years ago[PATCH] fbdev: fix switch to KD_TEXT, enhanced version
Knut Petersen [Tue, 13 Dec 2005 06:17:19 +0000]
[PATCH] fbdev: fix switch to KD_TEXT, enhanced version

Every framebuffer driver relies on the assumption that the set_par()
function of the driver is called before drawing functions and other
functions dependent on the hardware state are executed.

Whenever you switch from X to a framebuffer console for the very first
time, there is a chance that a broken X system has _not_ set the mode to
KD_GRAPHICS, thus the vt and framebuffer code executes a screen redraw and
several other functions before a set_par() is executed.  This is believed
to be not a bug of linux but a bug of X/xdm.  At least some X releases used
by SuSE and Debian show this behaviour.

There was a 2nd case, but that has been fixed by Antonino Daplas on
10-dec-2005.

This patch allows drivers to set a flag to inform fbcon_switch() that they
prefer a set_par() call on every console switch, working around the
problems caused by the broken X releases.

The flag will be used by the next release of cyblafb and might help other
drivers that assume a hardware state different to the one used by X.

As the default behaviour does not change, this patch should be acceptable
to everybody.

Signed-off-by: Knut Petersen <Knut_Petersen@t-online.de>
Acked-by: "Antonino A. Daplas" <adaplas@pol.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>