safe/jmp/linux-2.6
18 years ago[PATCH] pipe: enable atomic copying of pipe data to/from user space
Jens Axboe [Mon, 1 May 2006 18:02:05 +0000 (20:02 +0200)]
[PATCH] pipe: enable atomic copying of pipe data to/from user space

The pipe ->map() method uses kmap() to virtually map the pages, which
is both slow and has known scalability issues on SMP. This patch enables
atomic copying of pipe pages, by pre-faulting data and using kmap_atomic()
instead.

lmbench bw_pipe and lat_pipe measurements agree this is a Good Thing. Here
are results from that on a UP machine with highmem (1.5GiB of RAM), running
first a UP kernel, SMP kernel, and SMP kernel patched.

Vanilla-UP:
Pipe bandwidth: 1622.28 MB/sec
Pipe bandwidth: 1610.59 MB/sec
Pipe bandwidth: 1608.30 MB/sec
Pipe latency: 7.3275 microseconds
Pipe latency: 7.2995 microseconds
Pipe latency: 7.3097 microseconds

Vanilla-SMP:
Pipe bandwidth: 1382.19 MB/sec
Pipe bandwidth: 1317.27 MB/sec
Pipe bandwidth: 1355.61 MB/sec
Pipe latency: 9.6402 microseconds
Pipe latency: 9.6696 microseconds
Pipe latency: 9.6153 microseconds

Patched-SMP:
Pipe bandwidth: 1578.70 MB/sec
Pipe bandwidth: 1579.95 MB/sec
Pipe bandwidth: 1578.63 MB/sec
Pipe latency: 9.1654 microseconds
Pipe latency: 9.2266 microseconds
Pipe latency: 9.1527 microseconds

Signed-off-by: Jens Axboe <axboe@suse.de>
18 years ago[PATCH] splice: call handle_ra_miss() on failure to lookup page
Jens Axboe [Mon, 1 May 2006 17:59:54 +0000 (19:59 +0200)]
[PATCH] splice: call handle_ra_miss() on failure to lookup page

Notify the readahead logic of the missing page. Suggested by
Oleg Nesterov.

Signed-off-by: Jens Axboe <axboe@suse.de>
18 years ago[PATCH] Add ->splice_read/splice_write to def_blk_fops
Jens Axboe [Mon, 1 May 2006 17:59:32 +0000 (19:59 +0200)]
[PATCH] Add ->splice_read/splice_write to def_blk_fops

It can use the generic handlers.

Signed-off-by: Jens Axboe <axboe@suse.de>
18 years ago[PATCH] pipe: introduce ->pin() buffer operation
Jens Axboe [Mon, 1 May 2006 17:59:03 +0000 (19:59 +0200)]
[PATCH] pipe: introduce ->pin() buffer operation

The ->map() function is really expensive on highmem machines right now,
since it has to use the slower kmap() instead of kmap_atomic(). Splice
rarely needs to access the virtual address of a page, so it's a waste
of time doing it.

Introduce ->pin() to take over the responsibility of making sure the
page data is valid. ->map() is then reduced to just kmap(). That way we
can also share a most of the pipe buffer ops between pipe.c and splice.c

Signed-off-by: Jens Axboe <axboe@suse.de>
18 years ago[PATCH] splice: fix bugs in pipe_to_file()
Jens Axboe [Mon, 1 May 2006 17:50:48 +0000 (19:50 +0200)]
[PATCH] splice: fix bugs in pipe_to_file()

Found by Oleg Nesterov <oleg@tv-sign.ru>, fixed by me.

- Only allow full pages to go to the page cache.
- Check page != buf->page instead of using PIPE_BUF_FLAG_STOLEN.
- Remember to clear 'stolen' if add_to_page_cache() fails.

And as a cleanup on that:

- Make the bottom fall-through logic a little less convoluted. Also make
  the steal path hold an extra reference to the page, so we don't have
  to differentiate between stolen and non-stolen at the end.

Signed-off-by: Jens Axboe <axboe@suse.de>
18 years ago[PATCH] splice: fix bugs with stealing regular pipe pages
Jens Axboe [Sun, 30 Apr 2006 14:36:32 +0000 (16:36 +0200)]
[PATCH] splice: fix bugs with stealing regular pipe pages

- Check that page has suitable count for stealing in the regular pipes.
- pipe_to_file() assumes that the page is locked on succesful steal, so
  do that in the pipe steal hook
- Missing unlock_page() in add_to_page_cache() failure.

Signed-off-by: Jens Axboe <axboe@suse.de>
18 years agoMerge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc
Linus Torvalds [Sun, 30 Apr 2006 00:07:03 +0000 (17:07 -0700)]
Merge branch 'merge' of git://git./linux/kernel/git/paulus/powerpc

* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc:
  [PATCH] powerpc: update cell_defconfig
  [PATCH] spufs: Disable local interrupts for SPE hash_page calls.
  [PATCH] powerpc: Add cputable entry for POWER6
  [PATCH] ppc32 CPM_UART: Fixed odd address translations
  [PATCH] ppc32: Update board-specific code of the CPM UART users
  [PATCH] ppc32 CPM_UART: Convert to use platform devices
  [PATCH] ppc32: odd fixes and improvements in ppc_sys
  [PATCH] powerpc: Wire up *at syscalls
  [PATCH] ppc32: add 440GX erratum 440_43 workaround
  [PATCH] powerpc: Use check_legacy_ioport() on ppc32 too.
  [PATCH] powerpc64: Fix loading of modules without a .toc section
  [PATCH] sound/ppc: snd_pmac_toonie_init should be __init
  powerpc/pseries: Tell firmware our capabilities on new machines
  [PATCH] powerpc: Fix pagetable bloat for hugepages

18 years ago[PATCH] au1200fb: Remove accidentally duplicated content of au1200fb.c
Ralf Baechle [Sat, 29 Apr 2006 00:52:44 +0000 (08:52 +0800)]
[PATCH] au1200fb: Remove accidentally duplicated content of au1200fb.c

Content of file au1200fb.c was duplicated. Remove.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Antonino Daplas <adaplas@pol.net>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[SCSI] advansys driver: limp along on x86
Linus Torvalds [Sat, 29 Apr 2006 21:27:13 +0000 (14:27 -0700)]
[SCSI] advansys driver: limp along on x86

Let people enable the advansys driver on x86-32, even though it's broken
on other architectures due to missing DMA mapping infrastructure.

It's used by Jeffrey Phillips Freeman <jeffreyfreeman@syncleus.com> and
possibly others.

Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] i386: fix broken FP exception handling
Chuck Ebbert [Sat, 29 Apr 2006 18:07:49 +0000 (14:07 -0400)]
[PATCH] i386: fix broken FP exception handling

The FXSAVE information leak patch introduced a bug in FP exception
handling: it clears FP exceptions only when there are already
none outstanding.  Mikael Pettersson reported that causes problems
with the Erlang runtime and has tested this fix.

Signed-off-by: Chuck Ebbert <76306.1226@compuserve.com>
Acked-by: Mikael Pettersson <mikpe@it.uu.se>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] powerpc: update cell_defconfig
Arnd Bergmann [Sat, 29 Apr 2006 00:40:22 +0000 (02:40 +0200)]
[PATCH] powerpc: update cell_defconfig

reflect the changes to Kconfig since the last update.

Signed-off-by: Arnd Bergmann <arnd.bergmann@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[PATCH] spufs: Disable local interrupts for SPE hash_page calls.
Arnd Bergmann [Sat, 29 Apr 2006 00:40:21 +0000 (02:40 +0200)]
[PATCH] spufs: Disable local interrupts for SPE hash_page calls.

This patch disables and saves local interrupts during
hash_page processing for SPE contexts.

We have to do it explicitly in the spu_irq_class_1_bottom
function. For the interrupt handlers, we get the behaviour
implicitly by using SA_INTERRUPT to disable interrupts while
in the handler.

Signed-off-by: Arnd Bergmann <arnd.bergmann@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[PATCH] powerpc: Add cputable entry for POWER6
Anton Blanchard [Fri, 28 Apr 2006 23:51:06 +0000 (09:51 +1000)]
[PATCH] powerpc: Add cputable entry for POWER6

Add a cputable entry for the POWER6 processor.

The SIHV and SIPR bits in the mmcra have moved in POWER6, so disable
support for that until oprofile is fixed.

Also tell firmware that we know about POWER6.

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[PATCH] slab: fix crash on __drain_alien_cahce() during CPU Hotplug
shin, jacob [Fri, 28 Apr 2006 15:54:37 +0000 (10:54 -0500)]
[PATCH] slab: fix crash on __drain_alien_cahce() during CPU Hotplug

transfer_objects should only be called when all of the cpus in the
node are online.  CPU_DEAD notifier callback marks l3->shared to NULL.

Signed-off-by: Jacob Shin <jacob.shin@amd.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] asiliantfb: Add help text in Kconfig
Antonino A. Daplas [Fri, 28 Apr 2006 01:40:47 +0000 (18:40 -0700)]
[PATCH] asiliantfb: Add help text in Kconfig

Add help text in Kconfig

Signed-off-by: Antonino Daplas <adaplas@pol.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] suspend: Documentation update for IBM Thinkpad X30
Antonino A. Daplas [Fri, 28 Apr 2006 01:40:39 +0000 (18:40 -0700)]
[PATCH] suspend: Documentation update for IBM Thinkpad X30

As reported in Bugzilla Bug 6406, resume from S3 results in a blank screen.

For the IBM Thinkpad X30 using vesafb as the console driver, successful resume
from S3 requires option acpi_sleep=s3_bios,s3_mode.  Update documentation.

I would presume that, in any hardware, using vesafb as the console driver will
require as a minimum s3_mode.

Signed-off-by: Antonino Daplas <adaplas@pol.net>
Cc: <igor47@uchicago.edu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: new system calls
Martin Schwidefsky [Fri, 28 Apr 2006 01:40:33 +0000 (18:40 -0700)]
[PATCH] s390: new system calls

Add sys_set_robust_list, sys_get_robust_list, sys_splice, sys_sync_file and
sys_tee system calls.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: dasd device identifiers
Horst Hummel [Fri, 28 Apr 2006 01:40:28 +0000 (18:40 -0700)]
[PATCH] s390: dasd device identifiers

Generate new sysfs-attribute 'uid' that contains an device specific unique
identifier.  This can be used to identity multiple ALIASES of the same
physical device (PAV).  In addition the sysfs-attributes 'vendor' (containing
the manufacturer of the device) and 'alias' (identify alias or base device) is
added.  This is first part of PAV support in LPAR (also valid on zVM).

Signed-off-by: Horst Hummel <horst.hummel@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: add read_mostly optimization
Christian Borntraeger [Fri, 28 Apr 2006 01:40:24 +0000 (18:40 -0700)]
[PATCH] s390: add read_mostly optimization

Add a read_mostly section and define __read_mostly to prevent cache line
pollution due to writes for mostly read variables.  In addition fix the
incorrect alignment of the cache_line_aligned data section.  s390 has a
cacheline size of 256 bytes.

Signed-off-by: Christian Borntraeger <cborntra@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: instruction processing damage handling
Heiko Carstens [Fri, 28 Apr 2006 01:40:23 +0000 (18:40 -0700)]
[PATCH] s390: instruction processing damage handling

In case of an instruction processing damage (IPD) machine check in kernel mode
the resulting action is always to stop the kernel.  This is not necessarily
the best solution since a retry of the failing instruction might succeed.  Add
logic to retry the instruction if no more than 30 instruction processing
damage checks occured in the last 5 minutes.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: segment operation error codes
Gerald Schaefer [Fri, 28 Apr 2006 01:40:22 +0000 (18:40 -0700)]
[PATCH] s390: segment operation error codes

Print a warning with the z/VM error code if segment_load, segment_type or
segment_save fail to ease the problem determination.

Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: tape 3590 changes
Stefan Bader [Fri, 28 Apr 2006 01:40:16 +0000 (18:40 -0700)]
[PATCH] s390: tape 3590 changes

Added some changes that where proposed by Andrew Morton.  Added 3592 device
type.

Signed-off-by: Stefan Bader <shbader@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: futex atomic operations
Martin Schwidefsky [Fri, 28 Apr 2006 01:40:12 +0000 (18:40 -0700)]
[PATCH] s390: futex atomic operations

Add support for atomic futex operations.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: fix slab debugging
Christian Borntraeger [Fri, 28 Apr 2006 01:40:11 +0000 (18:40 -0700)]
[PATCH] s390: fix slab debugging

With CONFIG_SLAB_DEBUG=y networking over qeth doesn't work.  The problem is
that the qib structure embedded in the qeth_irq structure needs an alignment
of 256 but kmalloc only guarantees an alignment of 8.  When using SLAB
debugging the alignment of qeth_irq is not sufficient for the embedded qib
structure which causes all users of qdio (qeth and zfcp) to stop working.
Allocate qeth_irq structure with __get_free_page.  That wastes a small amount
of memory (~2500 bytes) per online adapter.

Signed-off-by: Christian Borntraeger <cborntra@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: dasd ioctl never returns
Horst Hummel [Fri, 28 Apr 2006 01:40:10 +0000 (18:40 -0700)]
[PATCH] s390: dasd ioctl never returns

The dasd state machine is not designed to enable an unformatted device, since
'unformatted' is a final state.  The BIODASDENABLE ioctl calls
dasd_enable_device() which never returns if the device is in this special
state.  Return -EPERM in dasd_increase_state for unformatted devices to make
dasd_enable_device terminate.  Note: To get such an unformatted device online
it has to be re-analyzed.  This means that the device needs to be disabled
prior to re-enablement.

Signed-off-by: Horst Hummel <horst.hummel@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: qdio memory allocations
Andreas Herrmann [Fri, 28 Apr 2006 01:40:09 +0000 (18:40 -0700)]
[PATCH] s390: qdio memory allocations

Avoid memory allocation with GFP_KERNEL in qdio_establish/qdio_shutdown.  Use
memory pool instead.  (Otherwise this can lead to an I/O stall where qdio
waits for a free page and zfcp waits for end of error recovery in low memory
situations.)

Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: alternate signal stack handling bug
Laurent Meyer [Fri, 28 Apr 2006 01:40:07 +0000 (18:40 -0700)]
[PATCH] s390: alternate signal stack handling bug

If a signal handler has been established with the SA_ONSTACK option but no
alternate stack is provided with sigaltstack(), the kernel still tries to
install the alternate stack.  Also when setting an alternate stack with
sigalstack() and the SS_DISABLE flag, the kernel tries to install the
alternate stack on signal delivery.  Use the correct conditions sas_ss_flags()
to check if the alternate stack has to be used.

Signed-off-by: Laurent Meyer <meyerlau@fr.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: enable interrupts on error path
Stefan Bader [Fri, 28 Apr 2006 01:40:04 +0000 (18:40 -0700)]
[PATCH] s390: enable interrupts on error path

Interrupts can stay disabled if an error occurred in _chp_add().  Use
spin_unlock_irq on the error paths to reenable interrupts.

Signed-off-by: Stefan Bader <shbader@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] s390: fix I/O termination race in cio
Peter Oberparleiter [Fri, 28 Apr 2006 01:40:02 +0000 (18:40 -0700)]
[PATCH] s390: fix I/O termination race in cio

Fix a race condition in the I/O termination logic.  The race can cause I/O to
a dasd device to fail with no retry left after turning one channel path to the
device off and on multiple times.

Signed-off-by: Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] kprobe: fix resume execution on i386
Masami Hiramatsu [Fri, 28 Apr 2006 01:39:55 +0000 (18:39 -0700)]
[PATCH] kprobe: fix resume execution on i386

Fix resume_execution() to handle iret and absolute jump opcode correctly on
i386.

Signed-off-by: Masami Hiramatsu <hiramatu@sdl.hitachi.co.jp>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: "Keshavamurthy, Anil S" <anil.s.keshavamurthy@intel.com>
Cc: Prasanna S Panchamukhi <prasanna@in.ibm.com>
Cc: Jim Keniston <jkenisto@us.ibm.com>
Cc: Yumiko Sugita <sugita@sdl.hitachi.co.jp>
Cc: Satoshi Oshima <soshima@redhat.com>
Cc: Hideo Aoki <haoki@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] kprobe cleanup for VM_MASK judgement
mao, bibo [Fri, 28 Apr 2006 01:39:44 +0000 (18:39 -0700)]
[PATCH] kprobe cleanup for VM_MASK judgement

When trap happens in user space, kprobe_exceptions_notify() funtion will
skip it.  This patch deletes some unnecessary code for VM_MASK judgement in
eflags.

Signed-off-by: bibo, mao <bibo.mao@intel.com>
Cc: Masami Hiramatsu <hiramatu@sdl.hitachi.co.jp>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Acked-by: "Keshavamurthy, Anil S" <anil.s.keshavamurthy@intel.com>
Acked-by: Prasanna S Panchamukhi <prasanna@in.ibm.com>
Cc: Jim Keniston <jkenisto@us.ibm.com>
Cc: Yumiko Sugita <sugita@sdl.hitachi.co.jp>
Cc: Satoshi Oshima <soshima@redhat.com>
Cc: Hideo Aoki <haoki@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] Mark VMSPLIT EMBEDDED
Andi Kleen [Fri, 28 Apr 2006 01:39:36 +0000 (18:39 -0700)]
[PATCH] Mark VMSPLIT EMBEDDED

Running abnormal VM splits causes weird problems - people can set non-standard
splits by accident, then lots of time gets wasted diagnosing it - see the long
"[stable] 2.6.16.6 breaks java...  sort of" email thread.

So we need to make this option harder to set.  Use CONFIG_EMBEDDED for this.

CONFIG_EMBEDDED isn't really the right thing to use, but there's nothing else
obvious and avoiding these problems is more important than Kconfig purity.

Signed-off-by: Andi Kleen <ak@suse.de>
Cc: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] enable X86_PC for HOTPLUG_CPU
Ashok Raj [Fri, 28 Apr 2006 01:39:30 +0000 (18:39 -0700)]
[PATCH] enable X86_PC for HOTPLUG_CPU

CPU_HOTPLUG has race conditions when we use broadcast mode IPI.

- First we introduced no_broadcast option
    (see include/asm-i386/mach-default/mach_ipi.h)

- x86_64 solved it by using physical flat mode (same as bigsmp on i386)
  since this will not use broadcast shortcuts for IPI.

- We switched to use bigsmp on i386 so that we can have same handling as
  x86_64, but apparently this caused an error message, if kernel was
  compiled without X86_GENERICARCH, X86_BIGSMP.  The message "You have >8
  CPUS..." which was bogus and misleading, and only indicated one of the
  above ARCH wasnt selected.

So we do not switch to automatic bigsmp for HOTPLUG_CPU support in i386
until the other related config dependencies for SMP_SUSPEND etc can be done
right.

Signed-off-by: Ashok Raj <ashok.raj@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] Avoid printing pointless tsc skew msgs
Dave Jones [Fri, 28 Apr 2006 01:39:24 +0000 (18:39 -0700)]
[PATCH] Avoid printing pointless tsc skew msgs

These messages are kinda silly..

CPU#0 had 0 usecs TSC skew, fixed it up.
CPU#1 had 0 usecs TSC skew, fixed it up.

inspired from: http://bugzilla.kernel.org/attachment.cgi?id=7713&action=view

Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] re-add the OSS SOUND_CS4232 option
Adrian Bunk [Fri, 28 Apr 2006 01:39:21 +0000 (18:39 -0700)]
[PATCH] re-add the OSS SOUND_CS4232 option

A regression in the ALSA driver compared to the OSS driver was reported as
ALSA bug #1520, so let's keep the OSS driver for now.

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] fix array overrun in drivers/char/mwave/mwavedd.c
Eric Sesterhenn [Fri, 28 Apr 2006 01:39:20 +0000 (18:39 -0700)]
[PATCH] fix array overrun in drivers/char/mwave/mwavedd.c

this fixes coverity id #489.

Since the last element in the array is always ARRAY_SIZE-1 we have to check
for ipcnum >= ARRAY_SIZE()

Signed-off-by: Eric Sesterhenn <snakebyte@gmx.de>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] tipar oops fix
Daniel Drake [Fri, 28 Apr 2006 01:39:19 +0000 (18:39 -0700)]
[PATCH] tipar oops fix

If compiled into the kernel, parport_register_driver() is called before the
parport driver has been initalised.

This means that it is expected that tp_count is 0 after the
parport_register_driver() call() - tipar's attach function will not be
called until later during bootup.

Signed-off-by: Daniel Drake <dsd@gentoo.org>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] request_irq(): remove warnings from irq probing
Andrew Morton [Fri, 28 Apr 2006 01:39:18 +0000 (18:39 -0700)]
[PATCH] request_irq(): remove warnings from irq probing

- Add new SA_PROBEIRQ which suppresses the new sharing-mismatch warning.
  Some drivers like to use request_irq() to find an unused interrupt slot.

- Use it in i82365.c

- Kill unused SA_PROBE.

Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] off-by-1 in kernel/power/main.c
dean gaudet [Fri, 28 Apr 2006 01:39:17 +0000 (18:39 -0700)]
[PATCH] off-by-1 in kernel/power/main.c

There's an off-by-1 in kernel/power/main.c:state_store() ...  if your
kernel just happens to have some non-zero data at pm_states[PM_SUSPEND_MAX]
(i.e.  one past the end of the array) then it'll let you write anything you
want to /sys/power/state and in response the box will enter S5.

Signed-off-by: dean gaudet <dean@arctic.org>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] ppc32 CPM_UART: Fixed odd address translations
Vitaly Bordug [Tue, 25 Apr 2006 16:26:46 +0000 (20:26 +0400)]
[PATCH] ppc32 CPM_UART: Fixed odd address translations

Current address translation methods can produce wrong results, because
virt_to_bus and vice versa may not produce correct offsets on dma-allocated
memory. The right way is, while tracking both phys and virt address of the
window that has been allocated for boffer descriptors, and use those
numbers to compute the offset and make translation properly.

Signed-off-by: Vitaly Bordug <vbordug@ru.mvista.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[PATCH] ppc32: Update board-specific code of the CPM UART users
Vitaly Bordug [Tue, 25 Apr 2006 16:26:43 +0000 (20:26 +0400)]
[PATCH] ppc32: Update board-specific code of the CPM UART users

This has the relevant updates/additions to the BSP code so that proper
platform_info struct well be passed to the CPM UART drivers. The changes
covered mpc866ads, mpc885ads and mpc8272ads.

Signed-off-by: Vitaly Bordug <vbordug@ru.mvista.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[PATCH] ppc32 CPM_UART: Convert to use platform devices
Vitaly Bordug [Tue, 25 Apr 2006 16:26:41 +0000 (20:26 +0400)]
[PATCH] ppc32 CPM_UART: Convert to use platform devices

This is intended to make the driver code more generic and flexible,
to get rid of board-specific layouts within driver, and generic rehaul,
yet keeping compatibility with the existing stuff utilizing it, being
compatible with legacy behavior (but with complaints that legacy mode
used).

Signed-off-by: Vitaly Bordug <vbordug@ru.mvista.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[PATCH] ppc32: odd fixes and improvements in ppc_sys
Vitaly Bordug [Tue, 25 Apr 2006 16:26:33 +0000 (20:26 +0400)]
[PATCH] ppc32: odd fixes and improvements in ppc_sys

This consists of offsets fix in ..._devices.c, and update of
ppc_sys_fixup_mem_resource() function to prevent subsequent fixups

Signed-off-by: Vitaly Bordug <vbordug@ru.mvista.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[PATCH] powerpc: Wire up *at syscalls
Andreas Schwab [Thu, 27 Apr 2006 13:46:42 +0000 (15:46 +0200)]
[PATCH] powerpc: Wire up *at syscalls

Wire up *at syscalls.

This patch has been tested on ppc64 (using glibc's testsuite, both 32bit
and 64bit), and compile-tested for ppc32 (I have currently no ppc32 system
available, but I expect no problems).

Signed-off-by: Andreas Schwab <schwab@suse.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[PATCH] ppc32: add 440GX erratum 440_43 workaround
Eugene Surovegin [Tue, 25 Apr 2006 08:22:44 +0000 (01:22 -0700)]
[PATCH] ppc32: add 440GX erratum 440_43 workaround

This patch adds workaround for PPC 440GX erratum 440_43. According to
this erratum spurious MachineChecks (caused by L1 cache parity) can
happen during DataTLB miss processing. We disable L1 cache parity
checking for 440GX rev.C and rev.F

Signed-off-by: Eugene Surovegin <ebs@ebshome.net>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[PATCH] powerpc: Use check_legacy_ioport() on ppc32 too.
David Woodhouse [Mon, 24 Apr 2006 22:22:17 +0000 (23:22 +0100)]
[PATCH] powerpc: Use check_legacy_ioport() on ppc32 too.

Some people report that we die on some Macs when we are expecting to
catch machine checks after poking at some random I/O address. I'd seen
it happen on my dual G4 with serial ports until we fixed those to use
OF, but now other users are reporting it with i8042.

This expands the use of check_legacy_ioport() to avoid that situation
even on 32-bit kernels.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[PATCH] powerpc64: Fix loading of modules without a .toc section
Alan Modra [Mon, 24 Apr 2006 12:41:51 +0000 (22:11 +0930)]
[PATCH] powerpc64: Fix loading of modules without a .toc section

Normally, ppc64 module .ko files contain a table-of-contents (.toc)
section, but if the module doesn't reference any static or external
data or external procedures, it is possible for gcc/binutils to
generate a .ko that doesn't have a .toc.  Currently the module
loader refuses to load such a module, since it needs the address
of the .toc section to use in relocations.

This patch fixes the problem by using the address of the .stubs
section instead, which is an acceptable substitute in this situation.

Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[PATCH] sound/ppc: snd_pmac_toonie_init should be __init
Andreas Schwab [Sun, 23 Apr 2006 18:32:41 +0000 (20:32 +0200)]
[PATCH] sound/ppc: snd_pmac_toonie_init should be __init

snd_pmac_toonie_init is only called by __init code and calls __init code
itself.

Signed-off-by: Andreas Schwab <schwab@suse.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years agopowerpc/pseries: Tell firmware our capabilities on new machines
Paul Mackerras [Fri, 28 Apr 2006 06:28:35 +0000 (16:28 +1000)]
powerpc/pseries: Tell firmware our capabilities on new machines

This adds code to call a new firmware method to tell the firmware
what machines and capabilities (such as VMX/Altivec) we support.
This will be needed on POWER5+ and POWER6 machines, and it has no
effect on past and current machines.

Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[PATCH] powerpc: Fix pagetable bloat for hugepages
David Gibson [Fri, 28 Apr 2006 05:02:51 +0000 (15:02 +1000)]
[PATCH] powerpc: Fix pagetable bloat for hugepages

At present, ARCH=powerpc kernels can waste considerable space in
pagetables when making large hugepage mappings.  Hugepage PTEs go in
PMD pages, but each PMD page maps 256M and so contains only 16
hugepage PTEs (128 bytes of data), but takes up a 1024 byte
allocation.  With CONFIG_PPC_64K_PAGES enabled (64k base page size),
the situation is worse.  Now hugepage PTEs are at the PTE page level
(also mapping 256M), so we store 16 hugepage PTEs in a 64k allocation.

The PowerPC MMU already means that any 256M region is either all
hugepage, or all normal pages.  Thus, with some care, we can use a
different allocation for the hugepage PTE tables and only allocate the
128 bytes necessary.

Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years agoMerge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6
Linus Torvalds [Fri, 28 Apr 2006 00:01:37 +0000 (17:01 -0700)]
Merge branch 'release' of git://git./linux/kernel/git/aegl/linux-2.6

* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
  [IA64] update sn2 defconfig
  [IA64] Add mca recovery failure messages
  [IA64-SGI] fix SGI Altix tioce_reserve_m32() bug
  [IA64] enable dumps to capture second page of kernel stack
  [IA64-SGI] - Reduce overhead of reading sn_topology
  [IA64-SGI] - Fix discover of nearest cpu node to IO node
  [IA64] IOC4 config option ordering
  [IA64] Setup an IA64 specific reclaim distance
  [IA64] eliminate compile time warnings
  [IA64] eliminate compile time warnings
  [IA64-SGI] SN SAL call to inject memory errors
  [IA64] - Fix MAX_PXM_DOMAINS for systems with > 256 nodes
  [IA64] Remove unused variable in sn_sal.h
  [IA64] Remove redundant NULL checks before kfree
  [IA64] wire up compat_sys_adjtimex()

18 years ago[IA64] update sn2 defconfig
Jes Sorensen [Thu, 27 Apr 2006 10:10:35 +0000 (06:10 -0400)]
[IA64] update sn2 defconfig

Update SN2 defconfig to latest kernel and add QLA FC drivers commonly
found in SN2 boxes.

Signed-off-by: Jes Sorensen <jes@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
18 years ago[IA64] Add mca recovery failure messages
Russ Anderson [Thu, 27 Apr 2006 15:07:08 +0000 (10:07 -0500)]
[IA64] Add mca recovery failure messages

When the mca recovery code encounters a condition that makes
the MCA non-recoverable, print the reason it could not recover.
This will make it easier to identify why the recovery code did
not recover.

Signed-off-by: Russ Anderson <rja@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
18 years ago[IA64-SGI] fix SGI Altix tioce_reserve_m32() bug
Mike Habeck [Wed, 26 Apr 2006 17:05:50 +0000 (12:05 -0500)]
[IA64-SGI] fix SGI Altix tioce_reserve_m32() bug

The following patch fixes a bug in the SGI Altix tioce_reserve_m32()
code.  The bug was that we could walking past the end of the CE ASIC
32/40bit PMU ATE Buffer, resulting in a PIO Reply Error.

Signed-off-by: Mike Habeck <habeck@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
18 years ago[IA64] enable dumps to capture second page of kernel stack
Cliff Wickman [Tue, 25 Apr 2006 15:47:48 +0000 (10:47 -0500)]
[IA64] enable dumps to capture second page of kernel stack

In SLES10 (2.6.16) crash dumping (in my experience, LKCD) is unable to
capture the second page of the 2-page task/stack allocation.
This is particularly troublesome for dump analysis, as the stack traceback
cannot be done.
  (A similar convention is probably needed throughout the kernel to make
   kernel multi-page allocations detectable for dumping)

Multi-page kernel allocations are represented by the single page structure
associated with the first page of the allocation.  The page structures
associated with the other pages are unintialized.

If the dumper is selecting only kernel pages it has no way to identify
any but the first page of the allocation.

The fix is to make the task/stack allocation a compound page.

Signed-off-by: Cliff Wickman <cpw@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
18 years ago[IA64-SGI] - Reduce overhead of reading sn_topology
Jack Steiner [Sat, 22 Apr 2006 14:37:19 +0000 (09:37 -0500)]
[IA64-SGI] - Reduce overhead of reading sn_topology

MPI programs using certain debug options have a long
startup time. This was traced to a "vmalloc/vfree" in
the code that reads /proc/sgi_sn/sn_topology. On large
systems, vfree requires an IPI to all cpus to do TLB
purging.

Replace the vmalloc/vfree with kmalloc/kfree. Although
the size of the structure being allocated is unknown, it
will not not exceed 96 bytes.

Signed-off-by: Jack Steiner <steiner@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
18 years ago[IA64-SGI] - Fix discover of nearest cpu node to IO node
Jack Steiner [Sat, 22 Apr 2006 14:36:07 +0000 (09:36 -0500)]
[IA64-SGI] - Fix discover of nearest cpu node to IO node

Fix a bug that causes discovery of the nearest node/cpu to
a TIO (IO node) to fail.

Signed-off-by: Jack Steiner <steiner@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/gregkh/pci-2.6
Linus Torvalds [Thu, 27 Apr 2006 21:26:03 +0000 (14:26 -0700)]
Merge /pub/scm/linux/kernel/git/gregkh/pci-2.6

* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/pci-2.6:
  [PATCH] PCI quirk: VIA IRQ fixup should only run for VIA southbridges
  [PATCH] PCI: fix potential resource leak in drivers/pci/msi.c
  [PATCH] PCI: Documentation: no more device ids
  [PATCH] PCI: fix via irq SATA patch

18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/gregkh/usb-2.6
Linus Torvalds [Thu, 27 Apr 2006 21:25:46 +0000 (14:25 -0700)]
Merge /pub/scm/linux/kernel/git/gregkh/usb-2.6

* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/usb-2.6:
  [PATCH] USB: ftdi_sio: add support for ASK RDR 400 series card reader
  [PATCH] USB: ftdi_sio: Adds support for iPlus device.
  [PATCH] USB: ftdi_sio vendor code for RR-CirKits LocoBuffer USB
  [PATCH] USB: Use new PCI_CLASS_SERIAL_USB_* defines
  [PATCH] USB: net2280: set driver data before it is used
  [PATCH] USB: net2280: check for shared IRQs
  [PATCH] USB: net2280: send 0-length packets for ep0
  [PATCH] USB: net2280: Handle STALLs for 0-length control-IN requests
  [PATCH] USB: storage: atmel unusual dev update
  [PATCH] USB: Storage: unusual devs update
  [PATCH] USB: add new iTegno usb CDMA 1x card support for pl2303
  [PATCH] USB: Resource leak fix for whiteheat driver

18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/gregkh/driver-2.6
Linus Torvalds [Thu, 27 Apr 2006 21:25:30 +0000 (14:25 -0700)]
Merge /pub/scm/linux/kernel/git/gregkh/driver-2.6

* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/driver-2.6:
  [PATCH] Added URI of "linux kernel development process"
  [PATCH] Kobject: possible cleanups
  [PATCH] Fix OCFS2 warning when DEBUG_FS is not enabled
  [PATCH] Kobject: fix build error
  [PATCH] Frame buffer: remove cmap sysfs interface

18 years agoMerge branch 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus
Linus Torvalds [Thu, 27 Apr 2006 21:24:49 +0000 (14:24 -0700)]
Merge branch 'upstream' of git://ftp.linux-mips.org/upstream-linus

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
  [MIPS] Fix branch emulation for floating-point exceptions.
  [MIPS] Cleanup inode->r_dev usage.
  [MIPS] Update MIPS defconfigs.
  [MIPS] Get rid of CONFIG_ADVANCED.
  [MIPS] Kconfig: Clarify description of CROSSCOMPILE.
  [MIPS] 24K LV: Add core card id.
  [MIPS] Sparse: fix sparse for 64-bit kernels.
  [MIPS] Use __ffs() instead of ffs() in ip32_irq0().
  [MIPS] Fix bitops for MIPS32/MIPS64 CPUs.
  [MIPS] Fix ip27 build.
  [MIPS] Oprofile: fix sparse warning.
  [MIPS] Fix oprofile module unloading

18 years agoMerge branch 'splice' of git://brick.kernel.dk/data/git/linux-2.6-block
Linus Torvalds [Thu, 27 Apr 2006 21:24:20 +0000 (14:24 -0700)]
Merge branch 'splice' of git://brick.kernel.dk/data/git/linux-2.6-block

* 'splice' of git://brick.kernel.dk/data/git/linux-2.6-block:
  [PATCH] splice: make the read-side do batched page lookups
  [PATCH] Add find_get_pages_contig(): contiguous variant of find_get_pages()
  [PATCH] splice: switch to using page_cache_readahead()

18 years ago[PATCH] Added URI of "linux kernel development process"
Paolo Ciarrocchi [Tue, 25 Apr 2006 20:47:51 +0000 (22:47 +0200)]
[PATCH] Added URI of "linux kernel development process"

Signed-off-by: Paolo Ciarrocchi <paolo.ciarrocchi@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] Kobject: possible cleanups
Adrian Bunk [Sat, 22 Apr 2006 10:14:44 +0000 (12:14 +0200)]
[PATCH] Kobject: possible cleanups

This patch contains the following possible cleanups:
- #if 0 the following unused global function:
  - subsys_remove_file()
- remove the following unused EXPORT_SYMBOL's:
  - kset_find_obj
  - subsystem_init
- remove the following unused EXPORT_SYMBOL_GPL:
  - kobject_add_dir

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] Fix OCFS2 warning when DEBUG_FS is not enabled
Jean Delvare [Wed, 19 Apr 2006 04:30:22 +0000 (21:30 -0700)]
[PATCH] Fix OCFS2 warning when DEBUG_FS is not enabled

Fix the following warning which happens when OCFS2_FS is enabled but
DEBUG_FS isn't:

fs/ocfs2/dlmglue.c: In function `ocfs2_dlm_init_debug':
fs/ocfs2/dlmglue.c:2036: warning: passing arg 5 of `debugfs_create_file' discards qualifiers from pointer target type

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: Arjan van de Ven <arjan@infradead.org>
Cc: Joel Becker <Joel.Becker@oracle.com>
Acked-by: Mark Fasheh <mark.fasheh@oracle.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] Kobject: fix build error
Kay Sievers [Tue, 25 Apr 2006 13:37:26 +0000 (15:37 +0200)]
[PATCH] Kobject: fix build error

This fixes a build error for various odd combinations of CONFIG_HOTPLUG
and CONFIG_NET.

Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
Cc: Nigel Cunningham <ncunningham@cyclades.com>
Cc: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] Frame buffer: remove cmap sysfs interface
Jon Smirl [Wed, 12 Apr 2006 23:43:35 +0000 (19:43 -0400)]
[PATCH] Frame buffer: remove cmap sysfs interface

Remove it as it does not work properly due to sysfs core changes.

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI quirk: VIA IRQ fixup should only run for VIA southbridges
Chris Wedgwood [Wed, 19 Apr 2006 06:57:09 +0000 (23:57 -0700)]
[PATCH] PCI quirk: VIA IRQ fixup should only run for VIA southbridges

Alan Cox pointed out that the VIA 'IRQ fixup' was erroneously running
on my system which has no VIA southbridge (but I do have a VIA IEEE
1394 device).

This should address that.  I also changed "Via IRQ" to "VIA IRQ"
(initially I read Via as a capitalized via (by way/means of).

Signed-off-by: Chris Wedgwood <cw@f00f.org>
Acked-by: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: fix potential resource leak in drivers/pci/msi.c
Jesper Juhl [Mon, 17 Apr 2006 02:02:54 +0000 (04:02 +0200)]
[PATCH] PCI: fix potential resource leak in drivers/pci/msi.c

The coverity checker spotted (as entry #599) that we might leak `entry' in
drivers/pci/msi.c::msix_capability_init()
This patch should take care of that.

Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: Documentation: no more device ids
Ingo Oeser [Tue, 18 Apr 2006 09:20:55 +0000 (11:20 +0200)]
[PATCH] PCI: Documentation: no more device ids

Document that we don't like to add more PCI device ids
but are happy to accept PCI vendor ids for linux/include/pci_ids.h

Original text from Jeff Garzik.

Signed-off-by: Ingo Oeser <netdev@axxeo.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: fix via irq SATA patch
Greg Kroah-Hartman [Thu, 20 Apr 2006 06:50:55 +0000 (15:50 +0900)]
[PATCH] PCI: fix via irq SATA patch

This device id improperly got added to the VIA chipset list with a
previous patch.  Remove it as it is not correct.

Cc: Grzegorz Janoszka <Grzegorz@Janoszka.pl>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] x86/PAE: Fix pte_clear for the >4GB RAM case
Zachary Amsden [Thu, 27 Apr 2006 18:32:29 +0000 (11:32 -0700)]
[PATCH] x86/PAE: Fix pte_clear for the >4GB RAM case

Proposed fix for ptep_get_and_clear_full PAE bug.  Pte_clear had the same bug,
so use the same fix for both.  Turns out pmd_clear had it as well, but pgds
are not affected.

The problem is rather intricate.  Page table entries in PAE mode are 64-bits
wide, but the only atomic 8-byte write operation available in 32-bit mode is
cmpxchg8b, which is expensive (at least on P4), and thus avoided.  But it can
happen that the processor may prefetch entries into the TLB in the middle of an
operation which clears a page table entry.  So one must always clear the P-bit
in the low word of the page table entry first when clearing it.

Since the sequence *ptep = __pte(0) leaves the order of the write dependent on
the compiler, it must be coded explicitly as a clear of the low word followed
by a clear of the high word.  Further, there must be a write memory barrier
here to enforce proper ordering by the compiler (and, in the future, by the
processor as well).

On > 4GB memory machines, the implementation of pte_clear for PAE was clearly
deficient, as it could leave virtual mappings of physical memory above 4GB
aliased to memory below 4GB in the TLB.  The implementation of
ptep_get_and_clear_full has a similar bug, although not nearly as likely to
occur, since the mappings being cleared are in the process of being destroyed,
and should never be dereferenced again.

But, as luck would have it, it is possible to trigger bugs even without ever
dereferencing these bogus TLB mappings, even if the clear is followed fairly
soon after with a TLB flush or invalidation.  The problem is that memory above
4GB may now be aliased into the first 4GB of memory, and in fact, may hit a
region of memory with non-memory semantics.  These regions include AGP and PCI
space.  As such, these memory regions are not cached by the processor.  This
introduces the bug.

The processor can speculate memory operations, including memory writes, as long
as they are committed with the proper ordering.  Speculating a memory write to
a linear address that has a bogus TLB mapping is possible.  Normally, the
speculation is harmless.  But for cached memory, it does leave the falsely
speculated cacheline unmodified, but in a dirty state.  This cache line will be
eventually written back.  If this cacheline happens to intersect a region of
memory that is not protected by the cache coherency protocol, it can corrupt
data in I/O memory, which is generally a very bad thing to do, and can cause
total system failure or just plain undefined behavior.

These bugs are extremely unlikely, but the severity is of such magnitude, and
the fix so simple that I think fixing them immediately is justified.  Also,
they are nearly impossible to debug.

Signed-off-by: Zachary Amsden <zach@vmware.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] USB: ftdi_sio: add support for ASK RDR 400 series card reader
Ian Abbott [Wed, 12 Apr 2006 14:20:35 +0000 (15:20 +0100)]
[PATCH] USB: ftdi_sio: add support for ASK RDR 400 series card reader

This patch adds support for an ASK RDR 400 series contactless card
reader <http://www.ask.fr/uk/products_and_services/terminals.html> to
the ftdi_sio driver's device ID table.  The product ID was supplied by
Adriano Couto on the ftdi-usb-sio-devel list.

Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: ftdi_sio: Adds support for iPlus device.
Luiz Fernando N. Capitulino [Tue, 11 Apr 2006 18:52:41 +0000 (15:52 -0300)]
[PATCH] USB: ftdi_sio: Adds support for iPlus device.

Adds support in ftdi_sio usbserial driver for USB modems sold by
Plus GSM Company in Poland.

Signed-off-by: Luiz Fernando Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: ftdi_sio vendor code for RR-CirKits LocoBuffer USB
Nathan Bronson [Mon, 10 Apr 2006 04:05:09 +0000 (00:05 -0400)]
[PATCH] USB: ftdi_sio vendor code for RR-CirKits LocoBuffer USB

This patch adds recognition of the RR-CirKits LocoBuffer USB
to the existing FTDI driver.  http://www.rr-cirkits.com

Signed-off-by: Nathan Bronson <ngb@sns-usa.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: Use new PCI_CLASS_SERIAL_USB_* defines
Jean Delvare [Sun, 9 Apr 2006 18:07:35 +0000 (20:07 +0200)]
[PATCH] USB: Use new PCI_CLASS_SERIAL_USB_* defines

We could use the recently added PCI_CLASS_SERIAL_USB_UHCI,
PCI_CLASS_SERIAL_USB_OHCI and PCI_CLASS_SERIAL_USB_EHCI defines in
more places, for slightly shorter and clearer code.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: net2280: set driver data before it is used
Alan Stern [Fri, 14 Apr 2006 20:46:28 +0000 (16:46 -0400)]
[PATCH] USB: net2280: set driver data before it is used

This patch (as671) fixes a bug in the error pathway for the net2280
probe routine.  A failure during probe will cause the driver to call
pci_get_drvdata before the corresponding pci_set_drvdata has been set.
The patch also does a kzalloc conversion.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: net2280: check for shared IRQs
Alan Stern [Fri, 14 Apr 2006 20:44:11 +0000 (16:44 -0400)]
[PATCH] USB: net2280: check for shared IRQs

This patch (as670) adds a check for whether a shared IRQ was actually
generated by the net2280 device.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: net2280: send 0-length packets for ep0
Alan Stern [Fri, 14 Apr 2006 20:42:03 +0000 (16:42 -0400)]
[PATCH] USB: net2280: send 0-length packets for ep0

This patch (as669) fixes a bug in the net2280 driver.  Now it will
properly send zero-length packets on ep0 until the control status stage
occurs.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: net2280: Handle STALLs for 0-length control-IN requests
Alan Stern [Fri, 14 Apr 2006 20:40:00 +0000 (16:40 -0400)]
[PATCH] USB: net2280: Handle STALLs for 0-length control-IN requests

This patch (as668) fixes a typo in net2280.  The handler for 0-length
control-IN requests should check that the endpoint _isn't_ halted before
sending a 0-length packet.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: storage: atmel unusual dev update
Olivier Blondeau [Mon, 17 Apr 2006 02:19:25 +0000 (19:19 -0700)]
[PATCH] USB: storage: atmel unusual dev update

Originally submitted by Olivier Blondeau <zeitoun@gmail.com>, with re-diffing
by me. Adds a new atmel unusual_dev entry.

Signed-off-by: Phil Dibowitz <phil@ipom.com>
18 years ago[PATCH] USB: Storage: unusual devs update
Phil Dibowitz [Mon, 17 Apr 2006 02:18:36 +0000 (19:18 -0700)]
[PATCH] USB: Storage: unusual devs update

This patch removes the Protocol portion of the Iomega Click! device as it's not
needed. Not-needed message reported by Kenneth Crudup <kenny@panix.com>

Signed-off-by: Phil Dibowitz <phil@ipom.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: add new iTegno usb CDMA 1x card support for pl2303
Wang Jun [Wed, 19 Apr 2006 08:32:07 +0000 (16:32 +0800)]
[PATCH] USB: add new iTegno usb CDMA 1x card support for pl2303

Add new iTegno usb CDMA 1x card (usbid '0eba:2080') support to pl2303 driver

Signed-off-by: Wang Jun <wangjun1974@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: Resource leak fix for whiteheat driver
Jesper Juhl [Sun, 23 Apr 2006 17:59:23 +0000 (19:59 +0200)]
[PATCH] USB: Resource leak fix for whiteheat driver

We may return from drivers/usb/serial/whiteheat.c::whiteheat_attach()
without freeing `result' if we leave via the no_firmware: label.

Spotted by the coverity checker as #670

Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[MIPS] Fix branch emulation for floating-point exceptions.
Ralf Baechle [Wed, 26 Apr 2006 20:33:03 +0000 (21:33 +0100)]
[MIPS] Fix branch emulation for floating-point exceptions.

In the branch emulation for floating-point exceptions, __compute_return_epc
must determine for bc1f et al which condition code bit to test. This is
based on bits <4:2> of the rt field. The switch statement to distinguish
bc1f et al needs to use only the two low bits of rt, but the old code tests
on the whole rt field.  This patch masks off the proper bits.

Signed-off-by: Win Treese <treese@acm.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Cleanup inode->r_dev usage.
Ralf Baechle [Mon, 24 Apr 2006 16:15:10 +0000 (17:15 +0100)]
[MIPS] Cleanup inode->r_dev usage.

imajor()/iminor() should be used instead of accessing r_dev directly.

Based on patch from Eric Sesterhenn (snakebyte@gmx.de).

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Update MIPS defconfigs.
Ralf Baechle [Mon, 24 Apr 2006 14:00:59 +0000 (15:00 +0100)]
[MIPS] Update MIPS defconfigs.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Get rid of CONFIG_ADVANCED.
Ralf Baechle [Mon, 24 Apr 2006 13:58:53 +0000 (14:58 +0100)]
[MIPS] Get rid of CONFIG_ADVANCED.

It's been a horrible source of confusion and let users to shoot themselves
into both feet with uzis to no end.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Kconfig: Clarify description of CROSSCOMPILE.
Ralf Baechle [Mon, 24 Apr 2006 12:08:41 +0000 (13:08 +0100)]
[MIPS] Kconfig: Clarify description of CROSSCOMPILE.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] 24K LV: Add core card id.
Chris Dearman [Fri, 14 Apr 2006 23:31:16 +0000 (00:31 +0100)]
[MIPS] 24K LV: Add core card id.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Sparse: fix sparse for 64-bit kernels.
Atsushi Nemoto [Fri, 14 Apr 2006 15:04:18 +0000 (00:04 +0900)]
[MIPS] Sparse: fix sparse for 64-bit kernels.

This commit breaks sparse for 64bit kernel.  The -m64 option is
required.  Also, some macro values (such as _MIPS_TUNE, etc.)  contain
double-quote characters so it would be better quoting arguments by
single-quote characters.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Use __ffs() instead of ffs() in ip32_irq0().
Atsushi Nemoto [Mon, 17 Apr 2006 12:24:49 +0000 (21:24 +0900)]
[MIPS] Use __ffs() instead of ffs() in ip32_irq0().

With recent rewrite for generic bitops, ffs() is defined the same way
as the libc and compiler built-in routines (returns int instead of
unsigned long).  Use __ffs() for 64bit value.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Fix bitops for MIPS32/MIPS64 CPUs.
Atsushi Nemoto [Mon, 17 Apr 2006 12:19:12 +0000 (21:19 +0900)]
[MIPS] Fix bitops for MIPS32/MIPS64 CPUs.

With recent rewrite for generic bitops, fls() for 32bit kernel with
MIPS64_CPU is broken.  Also, ffs(), fls() should be defined the same
way as the libc and compiler built-in routines (returns int instead of
unsigned long).

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Fix ip27 build.
Atsushi Nemoto [Wed, 19 Apr 2006 15:12:05 +0000 (00:12 +0900)]
[MIPS] Fix ip27 build.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Oprofile: fix sparse warning.
Ralf Baechle [Mon, 24 Apr 2006 09:50:02 +0000 (10:50 +0100)]
[MIPS] Oprofile: fix sparse warning.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Fix oprofile module unloading
Ralf Baechle [Mon, 24 Apr 2006 09:48:54 +0000 (10:48 +0100)]
[MIPS] Fix oprofile module unloading

When unloading oprofile module with timer-mode, oprofile_arch_exit
dereferences a NULL pointer.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[PATCH] splice: make the read-side do batched page lookups
Jens Axboe [Thu, 27 Apr 2006 09:05:22 +0000 (11:05 +0200)]
[PATCH] splice: make the read-side do batched page lookups

Use the new find_get_pages_contig() to potentially look up the entire
splice range in one single call. This speeds up generic_file_splice_read()
quite a bit.

Signed-off-by: Jens Axboe <axboe@suse.de>
18 years ago[PATCH] Add find_get_pages_contig(): contiguous variant of find_get_pages()
Jens Axboe [Thu, 27 Apr 2006 06:46:01 +0000 (08:46 +0200)]
[PATCH] Add find_get_pages_contig(): contiguous variant of find_get_pages()

find_get_pages_contig() will break out if we hit a hole in the page cache.
From Andrew Morton, small modifications and documentation by me.

Signed-off-by: Jens Axboe <axboe@suse.de>
18 years ago[PATCH] splice: switch to using page_cache_readahead()
Jens Axboe [Thu, 27 Apr 2006 06:44:27 +0000 (08:44 +0200)]
[PATCH] splice: switch to using page_cache_readahead()

Avoids doing useless work, when the file is fully cached.

Signed-off-by: Jens Axboe <axboe@suse.de>
18 years agoLinux v2.6.17-rc3 v2.6.17-rc3
Linus Torvalds [Thu, 27 Apr 2006 02:19:25 +0000 (19:19 -0700)]
Linux v2.6.17-rc3