pandora-kernel.git
16 years ago[IA64] CONFIG_SGI_SN2 - auto select NUMA and ACPI_NUMA
Jes Sorensen [Mon, 11 Feb 2008 14:10:19 +0000 (15:10 +0100)]
[IA64] CONFIG_SGI_SN2 - auto select NUMA and ACPI_NUMA

Auto select CONFIG_NUMA and CONFIG_ACPI_NUMA when picking SN2, similar
to how they are selected automatically for CONFIG_IA64_GENERIC.

Signed-off-by: Jes Sorensen <jes@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
16 years agox86: a P4 is a P6 not an i486
Hugh Dickins [Tue, 4 Mar 2008 19:33:24 +0000 (19:33 +0000)]
x86: a P4 is a P6 not an i486

P4 has been coming out as CPU_FAMILY=4 instead of 6: fix MPENTIUM4 typo.

Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agoMerge branch 'for-linus' of git://git.kernel.dk/linux-2.6-block
Linus Torvalds [Tue, 4 Mar 2008 19:37:36 +0000 (11:37 -0800)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-2.6-block

* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
  scsi: missing add of padded bytes to io completion byte count

16 years agoMerge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mfashe...
Linus Torvalds [Tue, 4 Mar 2008 19:27:32 +0000 (11:27 -0800)]
Merge branch 'upstream-linus' of git://git./linux/kernel/git/mfasheh/ocfs2

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mfasheh/ocfs2:
  [PATCH] fs/ocfs2/aops.c: Correct use of ! and &
  [2.6 patch] ocfs2: make dlm_do_assert_master() static
  [2.6 patch] make ocfs2_downconvert_thread() static
  [2.6 patch] fs/ocfs2/: possible cleanups
  [PATCH] ocfs2: le*_add_cpu conversion
  ocfs2: Fix writeout in ocfs2_data_convert_worker()
  ocfs2: Enable localalloc for local mounts

16 years agoMerge branch 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/djbw/async_tx
Linus Torvalds [Tue, 4 Mar 2008 19:26:01 +0000 (11:26 -0800)]
Merge branch 'fixes' of git://git./linux/kernel/git/djbw/async_tx

* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/djbw/async_tx:
  ioat: fix 'ack' handling, driver must ensure that 'ack' is zero
  dmaengine: fix sparse warning
  fsldma: do not cleanup descriptors in hardirq context
  dmaengine: add driver for Freescale MPC85xx DMA controller

16 years agoscsi: missing add of padded bytes to io completion byte count
Jens Axboe [Tue, 4 Mar 2008 19:22:54 +0000 (20:22 +0100)]
scsi: missing add of padded bytes to io completion byte count

Original patch from Tejun Heo <htejun@gmail.com> but should use ->extra_len
and not ->data_len, as we would then overshoot the original request size.

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mingo/linux...
Linus Torvalds [Tue, 4 Mar 2008 17:23:28 +0000 (09:23 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/mingo/linux-2.6-sched-devel

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mingo/linux-2.6-sched-devel:
  sched: revert load_balance_monitor() changes

16 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/x86/linux...
Linus Torvalds [Tue, 4 Mar 2008 17:22:32 +0000 (09:22 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/x86/linux-2.6-x86

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/x86/linux-2.6-x86:
  x86/xen: fix DomU boot problem
  x86: not set node to cpu_to_node if the node is not online
  x86, i387: fix ptrace leakage using init_fpu()

16 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/avi/kvm
Linus Torvalds [Tue, 4 Mar 2008 17:22:05 +0000 (09:22 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/avi/kvm

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/avi/kvm:
  x86: disable KVM for Voyager and friends
  KVM: VMX: Avoid rearranging switched guest msrs while they are loaded
  KVM: MMU: Fix race when instantiating a shadow pte
  KVM: Route irq 0 to vcpu 0 exclusively
  KVM: Avoid infinite-frequency local apic timer
  KVM: make MMU_DEBUG compile again
  KVM: move alloc_apic_access_page() outside of non-preemptable region
  KVM: SVM: fix Windows XP 64 bit installation crash
  KVM: remove the usage of the mmap_sem for the protection of the memory slots.
  KVM: emulate access to MSR_IA32_MCG_CTL
  KVM: Make the supported cpuid list a host property rather than a vm property
  KVM: Fix kvm_arch_vcpu_ioctl_set_sregs so that set_cr0 works properly
  KVM: SVM: set NM intercept when enabling CR0.TS in the guest
  KVM: SVM: Fix lazy FPU switching

16 years agoioat: fix 'ack' handling, driver must ensure that 'ack' is zero
Dan Williams [Sat, 1 Mar 2008 14:52:14 +0000 (07:52 -0700)]
ioat: fix 'ack' handling, driver must ensure that 'ack' is zero

Initialize 'ack' to zero in case the descriptor has been recycled.

Prevents "kernel BUG at crypto/async_tx/async_xor.c:185!"

Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Acked-by: Shannon Nelson <shannon.nelson@intel.com>
Cc: stable@kernel.org
16 years agodmaengine: fix sparse warning
Dan Williams [Sat, 1 Mar 2008 14:51:29 +0000 (07:51 -0700)]
dmaengine: fix sparse warning

include/linux/dmaengine.h:364:2: warning: returning void-valued expression

Signed-off-by: Dan Williams <dan.j.williams@intel.com>
16 years agofsldma: do not cleanup descriptors in hardirq context
Dan Williams [Sat, 1 Mar 2008 14:51:17 +0000 (07:51 -0700)]
fsldma: do not cleanup descriptors in hardirq context

"Cleaning" descriptors involves calling pending callbacks and clients
assume that their callback will only ever happen in softirq context.
Delay cleanup to the tasklet.

Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Acked-by: Zhang Wei <wei.zhang@freescale.com>
16 years agodmaengine: add driver for Freescale MPC85xx DMA controller
Zhang Wei [Sat, 1 Mar 2008 14:42:48 +0000 (07:42 -0700)]
dmaengine: add driver for Freescale MPC85xx DMA controller

The driver implements DMA engine API for Freescale MPC85xx DMA controller,
which could be used by devices in the silicon.  The driver supports the
Basic mode of Freescale MPC85xx DMA controller.  The MPC85xx processors
supported include MPC8540/60, MPC8555, MPC8548, MPC8641 and so on.

The MPC83xx(MPC8349, MPC8360) are also supported.

[kamalesh@linux.vnet.ibm.com: build fix]
[dan.j.williams@intel.com: merge mm fixes, rebase on async_tx-2.6.25]
Signed-off-by: Zhang Wei <wei.zhang@freescale.com>
Signed-off-by: Ebony Zhu <ebony.zhu@freescale.com>
Acked-by: Kumar Gala <galak@gate.crashing.org>
Cc: Shannon Nelson <shannon.nelson@intel.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
16 years agosched: revert load_balance_monitor() changes
Peter Zijlstra [Mon, 25 Feb 2008 16:34:02 +0000 (17:34 +0100)]
sched: revert load_balance_monitor() changes

The following commits cause a number of regressions:

  commit 58e2d4ca581167c2a079f4ee02be2f0bc52e8729
  Author: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
  Date:   Fri Jan 25 21:08:00 2008 +0100
  sched: group scheduling, change how cpu load is calculated

  commit 6b2d7700266b9402e12824e11e0099ae6a4a6a79
  Author: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
  Date:   Fri Jan 25 21:08:00 2008 +0100
  sched: group scheduler, fix fairness of cpu bandwidth allocation for task groups

Namely:
 - very frequent wakeups on SMP, reported by PowerTop users.
 - cacheline trashing on (large) SMP
 - some latencies larger than 500ms

While there is a mergeable patch to fix the latter, the former issues
are not fixable in a manner suitable for .25 (we're at -rc3 now).

Hence we revert them and try again in v2.6.26.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
CC: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
Tested-by: Alexey Zaytsev <alexey.zaytsev@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86/xen: fix DomU boot problem
Ian Campbell [Thu, 28 Feb 2008 23:16:49 +0000 (23:16 +0000)]
x86/xen: fix DomU boot problem

Construct Xen guest e820 map with a hole between 640K-1M.

It's pure luck that Xen kernels have gotten away with it in the past.

The patch below seems like the right thing to do. It certainly boots in
a domU without the DMI problem (without any of the other related patches
such as Alexander's).

Signed-off-by: Ian Campbell <ijc@hellion.org.uk>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Jeremy Fitzhardinge <jeremy@goop.org>
Tested-by: Mark McLoughlin <markmc@redhat.com>
Acked-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
16 years agox86: not set node to cpu_to_node if the node is not online
Yinghai Lu [Tue, 19 Feb 2008 23:35:54 +0000 (15:35 -0800)]
x86: not set node to cpu_to_node if the node is not online

resolve boot problem reported by Mel Gorman:

   http://lkml.org/lkml/2008/2/13/404

init_cpu_to_node will use cpu->apic (from MADT or mptable) and
apic->node(from SRAT or AMD config space with k8_bus_64.c) to have
cpu->node mapping, and later identify_cpu will overwrite them
again...(with nearby_node...)

this patch checks if the node is online, otherwise it will not
update cpu_node map. so keep cpu_node map to online node before
identify_cpu..., to prevent possible error.

Signed-off-by: Yinghai Lu <yinghai.lu@sun.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
16 years agox86, i387: fix ptrace leakage using init_fpu()
Suresh Siddha [Mon, 3 Mar 2008 21:01:08 +0000 (13:01 -0800)]
x86, i387: fix ptrace leakage using init_fpu()

This bug got introduced by the recent i387 merge:

  commit 4421011120b2304e5c248ae4165a2704588aedf1
  Author: Roland McGrath <roland@redhat.com>
  Date:   Wed Jan 30 13:31:50 2008 +0100

      x86: x86 i387 user_regset

Current usage of unlazy_fpu() in ptrace specific routines is wrong.
unlazy_fpu() will not init fpu if the task never used math. So the
ptrace calls can expose the parent tasks FPU data in some cases.

Replace it with the init_fpu() which will init the math state, if the
task never used math before.

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
16 years agoMerge branch 'for-linus' of git://git.kernel.dk/linux-2.6-block
Linus Torvalds [Tue, 4 Mar 2008 16:08:05 +0000 (08:08 -0800)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-2.6-block

* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
  block: fix blkdev_issue_flush() not detecting and passing EOPNOTSUPP back
  block: fix shadowed variable warning in blk-map.c
  block: remove extern on function definition
  cciss: remove READ_AHEAD define and use block layer defaults
  make cdrom.c:check_for_audio_disc() static
  block/genhd.c: proper externs
  unexport blk_rq_map_user_iov
  unexport blk_{get,put}_queue
  block/genhd.c: cleanups
  proper prototype for blk_dev_init()
  block/blk-tag.c should #include "blk.h"
  Fix DMA access of block device in 64-bit kernel on some non-x86 systems with 4GB or upper 4GB memory
  block: separate out padding from alignment
  block: restore the meaning of rq->data_len to the true data length
  resubmit: cciss: procfs updates to display info about many
  splice: only return -EAGAIN if there's hope of more data
  block: fix kernel-docbook parameters and files

16 years agom68k{,nommu}: Wire up new timerfd syscalls
Geert Uytterhoeven [Tue, 4 Mar 2008 08:18:16 +0000 (09:18 +0100)]
m68k{,nommu}: Wire up new timerfd syscalls

m68k{,nommu}: Wire up the new timerfd syscalls, which were introduced in
commit 4d672e7ac79b5ec5cdc90e450823441e20464691 ("timerfd: new timerfd API").

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Acked-by: Greg Ungerer <gerg@uclinux.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agom68knommu: fix fec driver interrupt races
Greg Ungerer [Tue, 4 Mar 2008 06:52:01 +0000 (16:52 +1000)]
m68knommu: fix fec driver interrupt races

The FEC driver has a common interrupt handler for all interrupt event
types. It is raised on a number of distinct interrupt vectors.
This handler can't be re-entered while processing an interrupt, so
make sure all requested vectors are flagged as IRQF_DISABLED.

Signed-off-by: Greg Ungerer <gerg@uclinux.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agom68knommu: declare do_IRQ()
Greg Ungerer [Tue, 4 Mar 2008 06:35:04 +0000 (16:35 +1000)]
m68knommu: declare do_IRQ()

Need a declaration of do_IRQ for the 68328 interrupt handling code.
It is common to all m68knommu targets, so a common declaration makes
sense.

Signed-off-by: Greg Ungerer <gerg@uclinux.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agom68knommu: remove duplicate hw_tick() code
Greg Ungerer [Tue, 4 Mar 2008 06:24:17 +0000 (16:24 +1000)]
m68knommu: remove duplicate hw_tick() code

Remove duplicate hw_tick() function from 68328 timers code.

Signed-off-by: Greg Ungerer <gerg@uclinux.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agom68knommu: update defconfig
Greg Ungerer [Tue, 4 Mar 2008 05:44:23 +0000 (15:44 +1000)]
m68knommu: update defconfig

Update the m68knommu defconfig.

Signed-off-by: Greg Ungerer <gerg@uclinux.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agoBuild fix for drivers/s390/char/defkeymap.c
Tony Breeds [Tue, 4 Mar 2008 05:05:06 +0000 (16:05 +1100)]
Build fix for drivers/s390/char/defkeymap.c

Commit 5ce2087ed0eb424e0889bdc9102727f65d2ecdde (Fix default compose
table initialization) left a trailing quote.

  CC      drivers/s390/char/defkeymap.o
drivers/s390/char/defkeymap.c:155: error: missing terminating ' character
drivers/s390/char/defkeymap.c:156: error: syntax error before ';' token
make[3]: *** [drivers/s390/char/defkeymap.o] Error 1

Fix that.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/sfrench/cifs-2.6
Linus Torvalds [Tue, 4 Mar 2008 16:00:34 +0000 (08:00 -0800)]
Merge git://git./linux/kernel/git/sfrench/cifs-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/sfrench/cifs-2.6:
  [CIFS] remove unused variable
  [CIFS] consolidate duplicate code in posix/unix inode handling
  [CIFS] fix build break when proc disabled
  [CIFS] factoring out common code in get_inode_info functions
  [CIFS] fix prepath conversion when server supports posix paths
  [CIFS] Only convert / when server does not support posix paths
  [CIFS] Fix mixed case name in structure dfs_info3_param
  [CIFS] fixup prefixpaths which contain multiple path components
  [CIFS] fix typo
  [CIFS] patch to fix incorrect encoding of number of aces on set mode
  [CIFS] Fix typo in quota operations
  [CIFS] clean up some hard to read ifdefs
  [CIFS] reduce checkpatch warnings
  [CIFS] fix warning in cifs_spnego.c

16 years agofreezer vs stopped or traced
Roland McGrath [Tue, 4 Mar 2008 04:22:05 +0000 (20:22 -0800)]
freezer vs stopped or traced

This changes the "freezer" code used by suspend/hibernate in its treatment
of tasks in TASK_STOPPED (job control stop) and TASK_TRACED (ptrace) states.

As I understand it, the intent of the "freezer" is to hold all tasks
from doing anything significant.  For this purpose, TASK_STOPPED and
TASK_TRACED are "frozen enough".  It's possible the tasks might resume
from ptrace calls (if the tracer were unfrozen) or from signals
(including ones that could come via timer interrupts, etc).  But this
doesn't matter as long as they quickly block again while "freezing" is
in effect.  Some minor adjustments to the signal.c code make sure that
try_to_freeze() very shortly follows all wakeups from both kinds of
stop.  This lets the freezer code safely leave stopped tasks unmolested.

Changing this fixes the longstanding bug of seeing after resuming from
suspend/hibernate your shell report "[1] Stopped" and the like for all
your jobs stopped by ^Z et al, as if you had freshly fg'd and ^Z'd them.
It also removes from the freezer the arcane special case treatment for
ptrace'd tasks, which relied on intimate knowledge of ptrace internals.

Signed-off-by: Roland McGrath <roland@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agox86: disable KVM for Voyager and friends
Randy Dunlap [Wed, 20 Feb 2008 17:20:08 +0000 (09:20 -0800)]
x86: disable KVM for Voyager and friends

Most classic Pentiums don't have hardware virtualization extension,
and building kvm with Voyager, Visual Workstation, or NUMAQ
generates spurious failures.

Signed-off-by: Avi Kivity <avi@qumranet.com>
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
16 years agoKVM: VMX: Avoid rearranging switched guest msrs while they are loaded
Avi Kivity [Wed, 27 Feb 2008 14:06:57 +0000 (16:06 +0200)]
KVM: VMX: Avoid rearranging switched guest msrs while they are loaded

KVM tries to run as much as possible with the guest msrs loaded instead of
host msrs, since switching msrs is very expensive.  It also tries to minimize
the number of msrs switched according to the guest mode; for example,
MSR_LSTAR is needed only by long mode guests.  This optimization is done by
setup_msrs().

However, we must not change which msrs are switched while we are running with
guest msr state:

 - switch to guest msr state
 - call setup_msrs(), removing some msrs from the list
 - switch to host msr state, leaving a few guest msrs loaded

An easy way to trigger this is to kexec an x86_64 linux guest.  Early during
setup, the guest will switch EFER to not include SCE.  KVM will stop saving
MSR_LSTAR, and on the next msr switch it will leave the guest LSTAR loaded.
The next host syscall will end up in a random location in the kernel.

Fix by reloading the host msrs before changing the msr list.

Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years agoKVM: MMU: Fix race when instantiating a shadow pte
Avi Kivity [Tue, 26 Feb 2008 20:12:10 +0000 (22:12 +0200)]
KVM: MMU: Fix race when instantiating a shadow pte

For improved concurrency, the guest walk is performed concurrently with other
vcpus.  This means that we need to revalidate the guest ptes once we have
write-protected the guest page tables, at which point they can no longer be
modified.

The current code attempts to avoid this check if the shadow page table is not
new, on the assumption that if it has existed before, the guest could not have
modified the pte without the shadow lock.  However the assumption is incorrect,
as the racing vcpu could have modified the pte, then instantiated the shadow
page, before our vcpu regains control:

  vcpu0        vcpu1

  fault
  walk pte

               modify pte
               fault in same pagetable
               instantiate shadow page

  lookup shadow page
  conclude it is old
  instantiate spte based on stale guest pte

We could do something clever with generation counters, but a test run by
Marcelo suggests this is unnecessary and we can just do the revalidation
unconditionally.  The pte will be in the processor cache and the check can
be quite fast.

Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years agoKVM: Route irq 0 to vcpu 0 exclusively
Avi Kivity [Mon, 25 Feb 2008 08:28:31 +0000 (10:28 +0200)]
KVM: Route irq 0 to vcpu 0 exclusively

Some Linux versions allow the timer interrupt to be processed by more than
one cpu, leading to hangs due to tsc instability.  Work around the issue
by only disaptching the interrupt to vcpu 0.

Problem analyzed (and patch tested) by Sheng Yang.

Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years agoKVM: Avoid infinite-frequency local apic timer
Avi Kivity [Sun, 24 Feb 2008 12:37:50 +0000 (14:37 +0200)]
KVM: Avoid infinite-frequency local apic timer

If the local apic initial count is zero, don't start a an hrtimer with infinite
frequency, locking up the host.

Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years agoKVM: make MMU_DEBUG compile again
Marcelo Tosatti [Thu, 14 Feb 2008 23:25:39 +0000 (21:25 -0200)]
KVM: make MMU_DEBUG compile again

the cr3 variable is now inside the vcpu->arch structure.

Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years agoKVM: move alloc_apic_access_page() outside of non-preemptable region
Marcelo Tosatti [Thu, 14 Feb 2008 23:21:43 +0000 (21:21 -0200)]
KVM: move alloc_apic_access_page() outside of non-preemptable region

alloc_apic_access_page() can sleep, while vmx_vcpu_setup is called
inside a non preemptable region. Move it after put_cpu().

Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years agoKVM: SVM: fix Windows XP 64 bit installation crash
Joerg Roedel [Wed, 13 Feb 2008 15:30:28 +0000 (16:30 +0100)]
KVM: SVM: fix Windows XP 64 bit installation crash

While installing Windows XP 64 bit wants to access the DEBUGCTL and the last
branch record (LBR) MSRs. Don't allowing this in KVM causes the installation to
crash. This patch allow the access to these MSRs and fixes the issue.

Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Markus Rechberger <markus.rechberger@amd.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years agoKVM: remove the usage of the mmap_sem for the protection of the memory slots.
Izik Eidus [Sun, 10 Feb 2008 16:04:15 +0000 (18:04 +0200)]
KVM: remove the usage of the mmap_sem for the protection of the memory slots.

This patch replaces the mmap_sem lock for the memory slots with a new
kvm private lock, it is needed beacuse untill now there were cases where
kvm accesses user memory while holding the mmap semaphore.

Signed-off-by: Izik Eidus <izike@qumranet.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years agoblock: fix blkdev_issue_flush() not detecting and passing EOPNOTSUPP back
Jens Axboe [Tue, 4 Mar 2008 10:47:46 +0000 (11:47 +0100)]
block: fix blkdev_issue_flush() not detecting and passing EOPNOTSUPP back

This is important to eg dm, that tries to decide whether to stop using
barriers or not.

Tested as working by Anders Henke <anders.henke@1und1.de>

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agoblock: fix shadowed variable warning in blk-map.c
Harvey Harrison [Tue, 4 Mar 2008 10:31:22 +0000 (11:31 +0100)]
block: fix shadowed variable warning in blk-map.c

Introduced between 2.6.25-rc2 and -rc3
block/blk-map.c:154:14: warning: symbol 'bio' shadows an earlier one
block/blk-map.c:110:13: originally declared here

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agoblock: remove extern on function definition
Harvey Harrison [Tue, 4 Mar 2008 10:30:18 +0000 (11:30 +0100)]
block: remove extern on function definition

Intoduced between 2.6.25-rc2 and -rc3
block/blk-settings.c:319:12: warning: function 'blk_queue_dma_drain' with external linkage has definition

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agocciss: remove READ_AHEAD define and use block layer defaults
Mike Miller [Tue, 4 Mar 2008 10:25:15 +0000 (11:25 +0100)]
cciss: remove READ_AHEAD define and use block layer defaults

This patch removes the #define READ_AHEAD 1024 from the driver and uses the
block layer defaults, instead. We have found that under certain workloads
the setting can cause a disk connected to the e200 controller to go offline.
If the disk hiccups the link may try to downshift but the controller is
never notified that the link successfully completed the renegotiation.
We've also found that performance using the block layer default of 32 pages
was on par with the 1024 setting. We tried setting it to zero at one time
based on info from our firmware guys but that killed performance. Turns out
we were talking about 2 different read ahead settings.
Please consider this for inclusion.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agomake cdrom.c:check_for_audio_disc() static
Adrian Bunk [Tue, 4 Mar 2008 10:23:51 +0000 (11:23 +0100)]
make cdrom.c:check_for_audio_disc() static

This patch makes the needlessly global check_for_audio_disc() static.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agoblock/genhd.c: proper externs
Adrian Bunk [Tue, 4 Mar 2008 10:23:50 +0000 (11:23 +0100)]
block/genhd.c: proper externs

This patch adds proper externs for two structs in include/linux/genhd.h

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agounexport blk_rq_map_user_iov
Adrian Bunk [Tue, 4 Mar 2008 10:23:48 +0000 (11:23 +0100)]
unexport blk_rq_map_user_iov

This patch removes the unused export of blk_rq_map_user_iov.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agounexport blk_{get,put}_queue
Adrian Bunk [Tue, 4 Mar 2008 10:23:47 +0000 (11:23 +0100)]
unexport blk_{get,put}_queue

This patch removes the unused exports of blk_{get,put}_queue.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agoblock/genhd.c: cleanups
Adrian Bunk [Tue, 4 Mar 2008 10:23:46 +0000 (11:23 +0100)]
block/genhd.c: cleanups

This patch contains the following cleanups:
- make the needlessly global struct disk_type static
- #if 0 the unused genhd_media_change_notify()

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agoproper prototype for blk_dev_init()
Adrian Bunk [Tue, 4 Mar 2008 10:23:45 +0000 (11:23 +0100)]
proper prototype for blk_dev_init()

This patch adds a proper prototye for blk_dev_init() in block/blk.h

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agoblock/blk-tag.c should #include "blk.h"
Adrian Bunk [Tue, 4 Mar 2008 10:23:44 +0000 (11:23 +0100)]
block/blk-tag.c should #include "blk.h"

Every file should include the headers containing the externs for its
global functions (in this case for __blk_queue_free_tags()).

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agoFix DMA access of block device in 64-bit kernel on some non-x86 systems with 4GB...
Yang Shi [Tue, 4 Mar 2008 10:20:51 +0000 (11:20 +0100)]
Fix DMA access of block device in 64-bit kernel on some non-x86 systems with 4GB or upper 4GB memory

For some non-x86 systems with 4GB or upper 4GB memory,
we need increase the range of addresses that can be
used for direct DMA in 64-bit kernel.

Signed-off-by: Yang Shi <yang.shi@windriver.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agoblock: separate out padding from alignment
Tejun Heo [Tue, 4 Mar 2008 10:18:17 +0000 (11:18 +0100)]
block: separate out padding from alignment

Block layer alignment was used for two different purposes - memory
alignment and padding.  This causes problems in lower layers because
drivers which only require memory alignment ends up with adjusted
rq->data_len.  Separate out padding such that padding occurs iff
driver explicitly requests it.

Tomo: restorethe code to update bio in blk_rq_map_user
      introduced by the commit 40b01b9bbdf51ae543a04744283bf2d56c4a6afa
      according to padding alignment.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agoblock: restore the meaning of rq->data_len to the true data length
FUJITA Tomonori [Tue, 4 Mar 2008 10:17:11 +0000 (11:17 +0100)]
block: restore the meaning of rq->data_len to the true data length

The meaning of rq->data_len was changed to the length of an allocated
buffer from the true data length. It breaks SG_IO friends and
bsg. This patch restores the meaning of rq->data_len to the true data
length and adds rq->extra_len to store an extended length (due to
drain buffer and padding).

This patch also removes the code to update bio in blk_rq_map_user
introduced by the commit 40b01b9bbdf51ae543a04744283bf2d56c4a6afa.
The commit adjusts bio according to memory alignment
(queue_dma_alignment). However, memory alignment is NOT padding
alignment. This adjustment also breaks SG_IO friends and bsg. Padding
alignment needs to be fixed in a proper way (by a separate patch).

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Jens Axboe <axboe@carl.home.kernel.dk>
16 years agoresubmit: cciss: procfs updates to display info about many
Mike Miller [Thu, 21 Feb 2008 07:54:03 +0000 (08:54 +0100)]
resubmit: cciss: procfs updates to display info about many

volumes

This patch allows us to display information about all of the logical volumes
configured on a particular controller without stepping on memory even when
there are many volumes (128 or more) configured.
Please consider this for inclusion.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agosplice: only return -EAGAIN if there's hope of more data
Jens Axboe [Wed, 20 Feb 2008 09:34:51 +0000 (10:34 +0100)]
splice: only return -EAGAIN if there's hope of more data

sys_tee() currently is a bit eager in returning -EAGAIN, it may do so
even if we don't have a chance of anymore data becoming available. So
improve the logic and only return -EAGAIN if we have an attached writer
to the input pipe.

Reported by Johann Felix Soden <johfel@gmx.de> and
Patrick McManus <mcmanus@ducksong.com>.

Tested-by: Johann Felix Soden <johfel@users.sourceforge.net>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agoblock: fix kernel-docbook parameters and files
Randy Dunlap [Wed, 20 Feb 2008 08:01:22 +0000 (09:01 +0100)]
block: fix kernel-docbook parameters and files

kernel-doc for block/:
- add missing parameters
- fix one function's parameter list (remove blank line)
- add 2 source files to docbook for non-exported kernel-doc functions

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
16 years agoMerge branch 'master' of /pub/scm/linux/kernel/git/torvalds/linux-2.6
Steve French [Tue, 4 Mar 2008 04:22:00 +0000 (04:22 +0000)]
Merge branch 'master' of /linux/kernel/git/torvalds/linux-2.6

16 years ago[PATCH] fs/ocfs2/aops.c: Correct use of ! and &
Julia Lawall [Tue, 26 Feb 2008 20:45:56 +0000 (21:45 +0100)]
[PATCH] fs/ocfs2/aops.c: Correct use of ! and &

In commit e6bafba5b4765a5a252f1b8d31cbf6d2459da337, a bug was fixed that
involved converting !x & y to !(x & y).  The code below shows the same
pattern, and thus should perhaps be fixed in the same way.

This is not tested and clearly changes the semantics, so it is only
something to consider.

Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
16 years ago[2.6 patch] ocfs2: make dlm_do_assert_master() static
Adrian Bunk [Sun, 17 Feb 2008 08:20:41 +0000 (10:20 +0200)]
[2.6 patch] ocfs2: make dlm_do_assert_master() static

This patch makes the needlessly global dlm_do_assert_master() static.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
16 years ago[2.6 patch] make ocfs2_downconvert_thread() static
Adrian Bunk [Sun, 17 Feb 2008 08:20:38 +0000 (10:20 +0200)]
[2.6 patch] make ocfs2_downconvert_thread() static

This patch makes the needlessly global ocfs2_downconvert_thread()
static.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
16 years ago[2.6 patch] fs/ocfs2/: possible cleanups
Adrian Bunk [Mon, 28 Jan 2008 22:11:41 +0000 (00:11 +0200)]
[2.6 patch] fs/ocfs2/: possible cleanups

This patch contains the following cleanups that are now possible:
- make the following needlessly global functions static:
  - dlmglue.c:ocfs2_process_blocked_lock()
  - heartbeat.c:ocfs2_node_map_init()
- #if 0 the following unused global function plus support functions:
  - heartbeat.c:ocfs2_node_map_is_only()

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
16 years ago[PATCH] ocfs2: le*_add_cpu conversion
Marcin Slusarz [Tue, 12 Feb 2008 23:06:18 +0000 (00:06 +0100)]
[PATCH] ocfs2: le*_add_cpu conversion

replace all:
little_endian_variable = cpu_to_leX(leX_to_cpu(little_endian_variable) +
expression_in_cpu_byteorder);
with:
leX_add_cpu(&little_endian_variable, expression_in_cpu_byteorder);
generated with semantic patch

Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
16 years agoocfs2: Fix writeout in ocfs2_data_convert_worker()
Mark Fasheh [Fri, 29 Feb 2008 01:16:03 +0000 (17:16 -0800)]
ocfs2: Fix writeout in ocfs2_data_convert_worker()

Commit f1f540688eae66c274ff1c1133b5d9c687b28f58 "optimized"
ocfs2_data_convert_worker() to "only do work for regular files".
Unfortunately, I left out a '!', which casued it to *skip* regular files.
This was hidden from testing until recently because the default data
journaling mode (data=ordered) doesn't exercise this code.

Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Signed-off-by: Joel Becker <joel.becker@oracle.com>
16 years agoocfs2: Enable localalloc for local mounts
Sunil Mushran [Wed, 6 Feb 2008 20:11:17 +0000 (12:11 -0800)]
ocfs2: Enable localalloc for local mounts

Commit 2fbe8d1ebe004425b4f7b8bba345623d2280be82 disabled localalloc
for local mounts. This caused issues as ocfs2 uses localalloc to
provide write locality. This patch enables localalloc for local mounts.

Signed-off-by: Sunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
16 years agoMerge branch 'slab-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/christoph/vm
Linus Torvalds [Mon, 3 Mar 2008 23:00:09 +0000 (15:00 -0800)]
Merge branch 'slab-linus' of git://git./linux/kernel/git/christoph/vm

* 'slab-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/christoph/vm:
  slub: fix possible NULL pointer dereference
  slub: Add kmalloc_large_node() to support kmalloc_node fallback
  slub: look up object from the freelist once
  slub: Fix up comments
  slub: Rearrange #ifdef CONFIG_SLUB_DEBUG in calculate_sizes()
  slub: Remove BUG_ON() from ksize and omit checks for !SLUB_DEBUG
  slub: Use the objsize from the kmem_cache_cpu structure
  slub: Remove useless checks in alloc_debug_processing
  slub: Remove objsize check in kmem_cache_flags()
  slub: rename slab_objects to show_slab_objects
  Revert "unique end pointer" patch
  slab: avoid double initialization & do initialization in 1 place

16 years agoexit_notify: fix kill_orphaned_pgrp() usage with mt exit
Oleg Nesterov [Sun, 2 Mar 2008 18:44:44 +0000 (21:44 +0300)]
exit_notify: fix kill_orphaned_pgrp() usage with mt exit

1. exit_notify() always calls kill_orphaned_pgrp(). This is wrong, we
   should do this only when the whole process exits.

2. exit_notify() uses "current" as "ignored_task", obviously wrong.
   Use ->group_leader instead.

Test case:

void hup(int sig)
{
printf("HUP received\n");
}

void *tfunc(void *arg)
{
sleep(2);
printf("sub-thread exited\n");
return NULL;
}

int main(int argc, char *argv[])
{
if (!fork()) {
signal(SIGHUP, hup);
kill(getpid(), SIGSTOP);
exit(0);
}

pthread_t thr;
pthread_create(&thr, NULL, tfunc, NULL);

sleep(1);
printf("main thread exited\n");
syscall(__NR_exit, 0);

return 0;
}

output:

main thread exited
HUP received
Hangup

With this patch the output is:

main thread exited
sub-thread exited
HUP received

Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agowill_become_orphaned_pgrp: partially fix insufficient ->exit_state check
Oleg Nesterov [Sun, 2 Mar 2008 18:44:42 +0000 (21:44 +0300)]
will_become_orphaned_pgrp: partially fix insufficient ->exit_state check

p->exit_state != 0 doesn't mean this process is dead, it may have
sub-threads.  Change the code to use "p->exit_state && thread_group_empty(p)"
instead.

Without this patch, ^Z doesn't deliver SIGTSTP to the foreground process
if the main thread has exited.

However, the new check is not perfect either.  There is a window when
exit_notify() drops tasklist and before release_task().  Suppose that
the last (non-leader) thread exits.  This means that entire group exits,
but thread_group_empty() is not true yet.

As Eric pointed out, is_global_init() is wrong as well, but I did not
dare to do other changes.

Just for the record, has_stopped_jobs() is absolutely wrong too.  But we
can't fix it now, we should first fix SIGNAL_STOP_STOPPED issues.

Even with this patch ^Z doesn't play well with the dead main thread.
The task is stopped correctly but do_wait(WSTOPPED) won't see it.  This
is another unrelated issue, will be (hopefully) fixed separately.

Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agointroduce kill_orphaned_pgrp() helper
Oleg Nesterov [Sun, 2 Mar 2008 18:44:40 +0000 (21:44 +0300)]
introduce kill_orphaned_pgrp() helper

Factor out the common code in reparent_thread() and exit_notify().

No functional changes.

Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agoFix default compose table initialization
Samuel Thibault [Mon, 3 Mar 2008 01:23:49 +0000 (01:23 +0000)]
Fix default compose table initialization

Oddly enough, unsigned int c = '\300'; puts a "negative" value in c, not
0300...  This fixes the default unicode compose table by using integers
instead of character constants.

Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agoslub: fix possible NULL pointer dereference
Cyrill Gorcunov [Sun, 2 Mar 2008 20:28:24 +0000 (23:28 +0300)]
slub: fix possible NULL pointer dereference

This patch fix possible NULL pointer dereference if kzalloc
failed. To be able to return proper error code the function
return type is changed to ssize_t (according to callees and
sysfs definitions).

Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
16 years agoslub: Add kmalloc_large_node() to support kmalloc_node fallback
Christoph Lameter [Sat, 1 Mar 2008 21:56:40 +0000 (13:56 -0800)]
slub: Add kmalloc_large_node() to support kmalloc_node fallback

Slub is missing some NUMA support for large kmallocs. Provide that.

Reviewed-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
16 years agoslub: look up object from the freelist once
Pekka J Enberg [Sat, 1 Mar 2008 21:43:54 +0000 (13:43 -0800)]
slub: look up object from the freelist once

We only need to look up object from c->page->freelist once in
__slab_alloc().

Signed-off-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
16 years agoslub: Fix up comments
Christoph Lameter [Sat, 16 Feb 2008 07:45:26 +0000 (23:45 -0800)]
slub: Fix up comments

Provide comments and fix up various spelling / style issues.

Signed-off-by: Christoph Lameter <clameter@sgi.com>
16 years agoslub: Rearrange #ifdef CONFIG_SLUB_DEBUG in calculate_sizes()
Christoph Lameter [Sat, 16 Feb 2008 07:45:25 +0000 (23:45 -0800)]
slub: Rearrange #ifdef CONFIG_SLUB_DEBUG in calculate_sizes()

Group SLUB_DEBUG code together to reduce the number of #ifdefs. Move some
debug checks under the #ifdef.

Reviewed-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
16 years agoslub: Remove BUG_ON() from ksize and omit checks for !SLUB_DEBUG
Christoph Lameter [Sat, 16 Feb 2008 07:45:25 +0000 (23:45 -0800)]
slub: Remove BUG_ON() from ksize and omit checks for !SLUB_DEBUG

The BUG_ONs are useless since the pointer derefs will lead to
NULL deref errors anyways. Some of the checks are not necessary
if no debugging is possible.

Signed-off-by: Christoph Lameter <clameter@sgi.com>
16 years agoslub: Use the objsize from the kmem_cache_cpu structure
Christoph Lameter [Sat, 16 Feb 2008 07:45:25 +0000 (23:45 -0800)]
slub: Use the objsize from the kmem_cache_cpu structure

No need to access the kmem_cache structure. We have the same value
in kmem_cache_cpu.

Reviewed-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
16 years agoslub: Remove useless checks in alloc_debug_processing
Christoph Lameter [Sat, 16 Feb 2008 07:45:24 +0000 (23:45 -0800)]
slub: Remove useless checks in alloc_debug_processing

Alloc debug processing is never called with a NULL object pointer.
No reason to check for NULL.

Reviewed-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
16 years agoslub: Remove objsize check in kmem_cache_flags()
Christoph Lameter [Sat, 16 Feb 2008 07:45:24 +0000 (23:45 -0800)]
slub: Remove objsize check in kmem_cache_flags()

There is no page->offset anymore and also no associated limit on the number
of objects. The page->offset field was removed for 2.6.24. So the check
in kmem_cache_flags() is now also obsolete (should have been dropped
earlier, somehow a hunk vanished).

Reviewed-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-by: Christoph Lameter <clameter@sgi.com>
16 years agoslub: rename slab_objects to show_slab_objects
Christoph Lameter [Fri, 15 Feb 2008 23:22:21 +0000 (15:22 -0800)]
slub: rename slab_objects to show_slab_objects

The sysfs callback is better named show_slab_objects since it is always
called from the xxx_show callbacks. We need the name for other purposes
later.

Reviewed-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
16 years agoRevert "unique end pointer" patch
Christoph Lameter [Sat, 1 Mar 2008 21:40:44 +0000 (13:40 -0800)]
Revert "unique end pointer" patch

This only made sense for the alternate fastpath which was reverted last week.

Mathieu is working on a new version that addresses the fastpath issues but that
new code first needs to go through mm and it is not clear if we need the
unique end pointers with his new scheme.

Reviewed-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
16 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6
Christoph Lameter [Mon, 3 Mar 2008 19:18:08 +0000 (11:18 -0800)]
Merge branch 'master' of git://git./linux/kernel/git/torvalds/linux-2.6

16 years agoMerge branch 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm
Linus Torvalds [Mon, 3 Mar 2008 18:47:52 +0000 (10:47 -0800)]
Merge branch 'for-linus' of /home/rmk/linux-2.6-arm

* 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm:
  [ARM] Fix freeing of page tables for ARM in free_pgd_slow

16 years agodocbook: fix fusion source files
Randy Dunlap [Sat, 1 Mar 2008 06:03:27 +0000 (22:03 -0800)]
docbook: fix fusion source files

Fix docbook problems in fusion source files.
These cause the generated docbook to be incorrect.

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agodocbook: fix kernel-api source files
Randy Dunlap [Sat, 1 Mar 2008 06:03:15 +0000 (22:03 -0800)]
docbook: fix kernel-api source files

Fix docbook problems in kernel-api.tmpl.
These cause the generated docbook to be incorrect.

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agodocbook: fix usb source files
Randy Dunlap [Sat, 1 Mar 2008 06:03:07 +0000 (22:03 -0800)]
docbook: fix usb source files

Fix docbook problems in USB source files.
These cause the generated docbook to be incorrect.

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agodocbook: fix scsi source file
Randy Dunlap [Sat, 1 Mar 2008 06:02:50 +0000 (22:02 -0800)]
docbook: fix scsi source file

Fix docbook problem in SCSI source files.
These cause the generated docbook to be incorrect.

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agodocbook: fix rapidio source files
Randy Dunlap [Sat, 1 Mar 2008 06:02:40 +0000 (22:02 -0800)]
docbook: fix rapidio source files

Fix docbook problems in rapidio source files.
These cause the generated docbook to be incorrect.

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agodocbook: fix filesystems.tmpl source files
Randy Dunlap [Sat, 1 Mar 2008 06:02:31 +0000 (22:02 -0800)]
docbook: fix filesystems.tmpl source files

Fix docbook problems in filesystems.tmpl.
These cause the generated docbook to be incorrect.

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/x86/linux-2.6-x86
Linus Torvalds [Mon, 3 Mar 2008 18:36:50 +0000 (10:36 -0800)]
Merge git://git./linux/kernel/git/x86/linux-2.6-x86

* git://git.kernel.org/pub/scm/linux/kernel/git/x86/linux-2.6-x86:
  x86: revert "x86: fix pmd_bad and pud_bad to support huge pages"
  x86: revert "x86: CPA: avoid split of alias mappings"

16 years agoMerge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc
Linus Torvalds [Mon, 3 Mar 2008 18:35:38 +0000 (10:35 -0800)]
Merge branch 'merge' of git://git./linux/kernel/git/paulus/powerpc

* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc: (24 commits)
  [POWERPC] Convert the cell IOMMU fixed mapping to 16M IOMMU pages
  [POWERPC] Allow for different IOMMU page sizes in cell IOMMU code
  [POWERPC] Cell IOMMU: n_pte_pages is in 4K page units, not IOMMU_PAGE_SIZE
  [POWERPC] Split setup of IOMMU stab and ptab, allocate dynamic/fixed ptabs separately
  [POWERPC] Move allocation of cell IOMMU pad page
  [POWERPC] Remove unused pte_offset variable
  [POWERPC] Use it_offset not pte_offset in cell IOMMU code
  [POWERPC] Clearup cell IOMMU fixed mapping terminology
  [POWERPC] enable hardware watchpoints on cell blades
  [POWERPC] move celleb DABRX definitions
  [POWERPC] OProfile: enable callgraph support for Cell
  [POWERPC] spufs: fix use time accounting on SPE-overcommit
  [POWERPC] spufs: serialize SLB invalidation against SLB loading
  [POWERPC] spufs: invalidate SLB translation before adding a new entry
  [POWERPC] spufs: synchronize IRQ when disabling
  [POWERPC] spufs: fix order of sputrace thread IDs
  [POWERPC] Xilinx: hwicap cleanup
  [POWERPC] 4xx: Use correct board info structure in cuboot wrappers
  [POWERPC] spufs: fix invalid scheduling of forgotten contexts
  [POWERPC] 44x: add missing define TARGET_4xx and TARGET_440GX to cuboot-taishan
  ...

16 years agoAllow ARG_MAX execve string space even with a small stack limit
Linus Torvalds [Mon, 3 Mar 2008 18:12:14 +0000 (10:12 -0800)]
Allow ARG_MAX execve string space even with a small stack limit

The new code that removed the limitation on the execve string size
(which was historically 32 pages) replaced it with a much softer limit
based on RLIMIT_STACK which is usually much larger than the traditional
limit.  See commit b6a2fea39318e43fee84fa7b0b90d68bed92d2ba ("mm:
variable length argument support") for details.

However, if you have a small stack limit (perhaps because you need lots
of stacks in a threaded environment), the new heuristic of allowing up
to 1/4th of RLIMIT_STACK to be used for argument and environment strings
could actually be smaller than the old limit.

So just say that it's ok to have up to ARG_MAX strings regardless of the
value of RLIMIT_STACK, and check the rlimit only when going over that
traditional limit.

(Of course, if you actually have a *really* small stack limit, the whole
stack itself will be limited before you hit ARG_MAX, but that has always
been true and is clearly the right behaviour anyway).

Acked-by: Carlos O'Donell <carlos@codesourcery.com>
Cc: Michael Kerrisk <michael.kerrisk@googlemail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Ollie Wild <aaw@google.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agoRevert "x86: fix pmd_bad and pud_bad to support huge pages"
Linus Torvalds [Mon, 3 Mar 2008 18:02:44 +0000 (10:02 -0800)]
Revert "x86: fix pmd_bad and pud_bad to support huge pages"

This reverts commit cded932b75ab0a5f9181ee3da34a0a488d1a14fd.

Arjan bisected down a boot-time hang to this, saying:
  ".. it prevents the kernel to finish booting on my (Penryn based)
   laptop.  The boot stops right after freeing the init memory."

and while it's not clear exactly what triggers it, at this stage we're
better off just reverting it while Ingo tries to figure out what went
wrong.

Requested-by: Arjan van de Ven <arjan@linux.intel.com>
Cc: Hans Rosenfeld <hans.rosenfeld@amd.com>
Cc: Nish Aravamudan <nish.aravamudan@gmail.com>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
16 years agox86: revert "x86: fix pmd_bad and pud_bad to support huge pages"
Ingo Molnar [Mon, 3 Mar 2008 12:53:58 +0000 (13:53 +0100)]
x86: revert "x86: fix pmd_bad and pud_bad to support huge pages"

revert commit cded932b75ab0a5f9181ee3da34a0a488d1a14fd,
"x86: fix pmd_bad and pud_bad to support huge pages", it causes
a bootup hang, as reported and bisected by Arjan van de Ven.

Bisected-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agox86: revert "x86: CPA: avoid split of alias mappings"
Rafael J. Wysocki [Mon, 3 Mar 2008 00:17:37 +0000 (01:17 +0100)]
x86: revert "x86: CPA: avoid split of alias mappings"

Revert:

  commit 8be8f54bae3453588011cad06363813a5293af53
  Author: Thomas Gleixner <tglx@linutronix.de>
  Date:   Sat Feb 23 20:43:21 2008 +0100

      x86: CPA: avoid split of alias mappings

because it clearly mishandles the case when __change_page_attr(), called
from __change_page_attr_set_clr(), changes cpa->processed to 1 and
cpa_process_alias(cpa) is executed right after that.

This crashes my x86-64 test box early in the boot process
(ref. http://bugzilla.kernel.org/show_bug.cgi?id=10140#c4).

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
16 years agoMerge branch 'for-2.6.25' of master.kernel.org:/pub/scm/linux/kernel/git/arnd/cell...
Paul Mackerras [Mon, 3 Mar 2008 10:31:09 +0000 (21:31 +1100)]
Merge branch 'for-2.6.25' of /linux/kernel/git/arnd/cell-2.6 into merge

16 years agoKVM: emulate access to MSR_IA32_MCG_CTL
Joerg Roedel [Mon, 11 Feb 2008 19:28:27 +0000 (20:28 +0100)]
KVM: emulate access to MSR_IA32_MCG_CTL

Injecting an GP when accessing this MSR lets Windows crash when running some
stress test tools in KVM.  So this patch emulates access to this MSR.

Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Markus Rechberger <markus.rechberger@amd.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years agoKVM: Make the supported cpuid list a host property rather than a vm property
Avi Kivity [Mon, 11 Feb 2008 16:37:23 +0000 (18:37 +0200)]
KVM: Make the supported cpuid list a host property rather than a vm property

One of the use cases for the supported cpuid list is to create a "greatest
common denominator" of cpu capabilities in a server farm.  As such, it is
useful to be able to get the list without creating a virtual machine first.

Since the code does not depend on the vm in any way, all that is needed is
to move it to the device ioctl handler.  The capability identifier is also
changed so that binaries made against -rc1 will fail gracefully.

Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years agoKVM: Fix kvm_arch_vcpu_ioctl_set_sregs so that set_cr0 works properly
Paul Knowles [Wed, 6 Feb 2008 11:02:35 +0000 (11:02 +0000)]
KVM: Fix kvm_arch_vcpu_ioctl_set_sregs so that set_cr0 works properly

Whilst working on getting a VM to initialize in to IA32e mode I found
this issue. set_cr0 relies on comparing the old cr0 to the new one to
work correctly.  Move the assignment below so the compare can work.

Signed-off-by: Paul Knowles <paul@transitive.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years agoKVM: SVM: set NM intercept when enabling CR0.TS in the guest
Joerg Roedel [Tue, 29 Jan 2008 12:01:27 +0000 (13:01 +0100)]
KVM: SVM: set NM intercept when enabling CR0.TS in the guest

Explicitly enable the NM intercept in svm_set_cr0 if we enable TS in the guest
copy of CR0 for lazy FPU switching. This fixes guest SMP with Linux under SVM.
Without that patch Linux deadlocks or panics right after trying to boot the
other CPUs.

Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Markus Rechberger <markus.rechberger@amd.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years agoKVM: SVM: Fix lazy FPU switching
Joerg Roedel [Mon, 21 Jan 2008 12:09:33 +0000 (13:09 +0100)]
KVM: SVM: Fix lazy FPU switching

If the guest writes to cr0 and leaves the TS flag at 0 while vcpu->fpu_active
is also 0, the TS flag in the guest's cr0 gets lost. This leads to corrupt FPU
state an causes Windows Vista 64bit to crash very soon after boot.  This patch
fixes this bug.

Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Markus Rechberger <markus.rechberger@amd.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
16 years ago[POWERPC] Convert the cell IOMMU fixed mapping to 16M IOMMU pages
Michael Ellerman [Fri, 29 Feb 2008 07:33:29 +0000 (18:33 +1100)]
[POWERPC] Convert the cell IOMMU fixed mapping to 16M IOMMU pages

The only tricky part is we need to adjust the PTE insertion loop to
cater for holes in the page table. The PTEs for each segment start on
a 4K boundary, so with 16M pages we have 16 PTEs per segment and then
a gap to the next 4K page boundary.

It might be possible to allocate the PTEs for each segment separately,
saving the memory currently filling the gaps. However we'd need to
check that's OK with the hardware, and that it actually saves memory.

Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
16 years ago[POWERPC] Allow for different IOMMU page sizes in cell IOMMU code
Michael Ellerman [Fri, 29 Feb 2008 07:33:27 +0000 (18:33 +1100)]
[POWERPC] Allow for different IOMMU page sizes in cell IOMMU code

Make some preliminary changes to cell_iommu_alloc_ptab() to allow it to
take the page size as a parameter rather than assuming IOMMU_PAGE_SIZE.

Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
16 years ago[POWERPC] Cell IOMMU: n_pte_pages is in 4K page units, not IOMMU_PAGE_SIZE
Michael Ellerman [Fri, 29 Feb 2008 07:33:26 +0000 (18:33 +1100)]
[POWERPC] Cell IOMMU: n_pte_pages is in 4K page units, not IOMMU_PAGE_SIZE

We use n_pte_pages to calculate the stride through the page tables, but
we also use it to set the NPPT value in the segment table entry. That is
defined as the number of 4K pages per segment, so we should calculate
it as such regardless of the IOMMU page size.

Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
16 years ago[POWERPC] Split setup of IOMMU stab and ptab, allocate dynamic/fixed ptabs separately
Michael Ellerman [Fri, 29 Feb 2008 07:33:25 +0000 (18:33 +1100)]
[POWERPC] Split setup of IOMMU stab and ptab, allocate dynamic/fixed ptabs separately

Currently the cell IOMMU code allocates the entire IOMMU page table in a
contiguous chunk. This is nice and tidy, but for machines with larger
amounts of RAM the page table allocation can fail due to it simply being
too large.

So split the segment table and page table setup routine, and arrange to
have the dynamic and fixed page tables allocated separately.

Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>