As reported by Martin J. Bligh <mbligh@google.com>, we let through some
non-slab bits to slab allocation through __get_vm_area_node when doing a
vmalloc.
I haven't been able to reproduce this, although I understand why it
happens: vmalloc allocates memory with
GFP_KERNEL | __GFP_HIGHMEM
and commit 52fd24ca1db3a741f144bbc229beefe044202cac resulted in the same
flags are passed down to cache_alloc_refill, causing the BUG. The
following patch fixes it.
Note that when calling kmalloc_node, I am masking off __GFP_HIGHMEM with
GFP_LEVEL_MASK, whereas __vmalloc_area_node does the same with
~(__GFP_HIGHMEM | __GFP_ZERO).
IMHO, using GFP_LEVEL_MASK is preferable, but either should fix this
problem.
Signed-off-by: Giridhar Pemmasani (pgiri@yahoo.com) Cc: Martin J. Bligh <mbligh@google.com> Cc: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Linus Torvalds [Sat, 28 Oct 2006 18:38:39 +0000 (11:38 -0700)]
Merge branch 'for-linus' of git://git390.osdl.marist.edu/pub/scm/linux-2.6
* 'for-linus' of git://git390.osdl.marist.edu/pub/scm/linux-2.6:
[S390] cio: Make ccw_device_register() static.
[S390] Improve AP bus device removal.
[S390] uaccess error handling.
[S390] cio: css_probe_device() must be called enabled.
[S390] Initialize interval value to 0.
[S390] sys_getcpu compat wrapper.
Mel Gorman [Sat, 28 Oct 2006 17:38:59 +0000 (10:38 -0700)]
[PATCH] Calculation fix for memory holes beyong the end of physical memory
absent_pages_in_range() made the assumption that users of the
arch-independent zone-sizing API would not care about holes beyound the end
of physical memory. This was not the case and was "fixed" in a patch
called "Account for holes that are outside the range of physical memory".
However, when given a range that started before a hole in "real" memory and
ended beyond the end of memory, it would get the result wrong. The bug is
in mainline but a patch is below.
It has been tested successfully on a number of machines and architectures.
Additional credit to Keith Mannthey for discovering the problem, helping
identify the correct fix and confirming it Worked For Him.
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Cc: keith mannthey <kmannth@us.ibm.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Alan Stern [Sat, 28 Oct 2006 17:38:58 +0000 (10:38 -0700)]
[PATCH] workqueue: update kerneldoc
This patch (as812) changes the kerneldoc comments explaining the return
values from queue_work(), queue_delayed_work(), and
queue_delayed_work_on(). The updated comments explain more accurately the
meaning of the return code and avoid suggesting that a 0 value means the
routine was unsuccessful.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Jim Houston [Sat, 28 Oct 2006 17:38:56 +0000 (10:38 -0700)]
[PATCH] time_adjust cleared before use
I notice that the code which implements adjtime clears the time_adjust
value before using it. The attached patch makes the obvious fix.
Acked-by: Roman Zippel <zippel@linux-m68k.org> Signed-off-by: Jim Houston <jim.houston@ccur.com> Cc: John Stultz <johnstul@us.ibm.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Randy Dunlap [Sat, 28 Oct 2006 17:38:55 +0000 (10:38 -0700)]
[PATCH] move SYS_HYPERVISOR inside the Generic Driver menu
Put SYS_HYPERVISOR inside the Generic Driver Config menu where it should
be. Otherwise xconfig displays it as a dangling (lost) menu item under
Device Drivers, all by itself (when all options are displayed).
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Cc: <holzheu@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Oleg Nesterov [Sat, 28 Oct 2006 17:38:54 +0000 (10:38 -0700)]
[PATCH] fill_tgid: cleanup delays accounting
fill_tgid() should skip not only an already exited group leader. If the
task has ->exit_state != 0 it already did exit_notify(), so it also did
fill_tgid_exit()->delayacct_add_tsk(->signal->stats) and we should skip it
to avoid a double accounting.
This patch doesn't close the race completely, but it cleanups the code.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: Shailabh Nagar <nagar@watson.ibm.com> Cc: Balbir Singh <balbir@in.ibm.com> Cc: Jay Lan <jlan@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Oleg Nesterov [Sat, 28 Oct 2006 17:38:54 +0000 (10:38 -0700)]
[PATCH] taskstats: don't use tasklist_lock
Remove tasklist_lock from taskstats.c. find_task_by_pid() is rcu-safe.
->siglock allows us to traverse subthread without tasklist.
Q: delay accounting looks wrong to me. If sub-thread has already called
taskstats_exit_send() but didn't call release_task(self) yet it will be
accounted twice. The window is big. No?
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: Shailabh Nagar <nagar@watson.ibm.com> Cc: Balbir Singh <balbir@in.ibm.com> Cc: Jay Lan <jlan@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Oleg Nesterov [Sat, 28 Oct 2006 17:38:53 +0000 (10:38 -0700)]
[PATCH] taskstats: kill ->taskstats_lock in favor of ->siglock
signal_struct is (mostly) protected by ->sighand->siglock, I think we don't
need ->taskstats_lock to protect ->stats. This also allows us to simplify the
locking in fill_tgid().
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: Shailabh Nagar <nagar@watson.ibm.com> Cc: Balbir Singh <balbir@in.ibm.com> Cc: Jay Lan <jlan@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Oleg Nesterov [Sat, 28 Oct 2006 17:38:51 +0000 (10:38 -0700)]
[PATCH] taskstats_tgid_free: fix usage
taskstats_tgid_free() is called on copy_process's error path. This is wrong.
IF (clone_flags & CLONE_THREAD)
We should not clear ->signal->taskstats, current uses it,
it probably has a valid accumulated info.
ELSE
taskstats_tgid_init() set ->signal->taskstats = NULL,
there is nothing to free.
Move the callsite to __exit_signal(). We don't need any locking, entire
thread group is exiting, nobody should have a reference to soon to be
released ->signal.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: Shailabh Nagar <nagar@watson.ibm.com> Cc: Balbir Singh <balbir@in.ibm.com> Cc: Jay Lan <jlan@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
It is possible that current != tsk. Probably it was supposed
to be 'tsk->group_leader->start_time. But why we are reading
group_leader's start_time ? This accounting is per thread,
not per procees, I changed this to 'tsk->start_time.
Please corect me.
tsk->parent never == NULL, and it is unsafe to dereference it.
Both the task and it's parent may exit after the caller unlocks
tasklist_lock, the memory could be unmapped (DEBUG_SLAB).
(And we should use ->real_parent->tgid in fact).
Q: I don't understand the 'if (thread_group_leader(tsk))' check.
Why it is needed ?
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: Shailabh Nagar <nagar@watson.ibm.com> Cc: Balbir Singh <balbir@in.ibm.com> Acked-by: Jay Lan <jlan@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Michael Holzheu [Sat, 28 Oct 2006 17:38:47 +0000 (10:38 -0700)]
[PATCH] strstrip remove last blank fix
strstrip() does not remove the last blank from strings which only consist
of blanks.
Example:
char string[] = " ";
strstrip(string);
results in " ", but should produce an empty string!
The following patch solves this problem:
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Michael Holzheu <holzheu@de.ibm.com> Acked-by: Pekka Enberg <penberg@cs.helsinki.fi>
Acked-by Joern Engel <joern@wh.fh-wedel.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Stephen Rothwell [Sat, 28 Oct 2006 17:38:46 +0000 (10:38 -0700)]
[PATCH] Constify compat_get_bitmap argument
This means we can call it when the bitmap we want to fetch is declared
const.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Cc: Christoph Lameter <clameter@engr.sgi.com> Cc: Paul Jackson <pj@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
David Howells [Sat, 28 Oct 2006 17:38:46 +0000 (10:38 -0700)]
[PATCH] VFS: Fix an error in unused dentry counting
With Vasily Averin <vvs@sw.ru>
Fix an error in unused dentry counting in shrink_dcache_for_umount_subtree()
in which the count is modified without the dcache_lock held.
Signed-off-by: David Howells <dhowells@redhat.com> Cc: Vasily Averin <vvs@sw.ru> Cc: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
# If prune_dcache finds a dentry that it cannot free, it leaves it where it
# is (at the tail of the list) and exits, on the assumption that some other
# thread will be removing that dentry soon.
However as far as I see this comment is not correct: when we cannot take
s_umount rw_semaphore (for example because it was taken in do_remount) this
dentry is already extracted from dentry_unused list and we do not add it
into the list again. Therefore dentry will not be found by prune_dcache()
and shrink_dcache_sb() and will leave in memory very long time until the
partition will be unmounted.
The patch adds this dentry into tail of the dentry_unused list.
Signed-off-by: Vasily Averin <vvs@sw.ru> Cc: Neil Brown <neilb@suse.de> Acked-by: David Howells <dhowells@redhat.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Hugh Dickins [Sat, 28 Oct 2006 17:38:43 +0000 (10:38 -0700)]
[PATCH] hugetlb: fix absurd HugePages_Rsvd
If you truncated an mmap'ed hugetlbfs file, then faulted on the truncated
area, /proc/meminfo's HugePages_Rsvd wrapped hugely "negative". Reinstate my
preliminary i_size check before attempting to allocate the page (though this
only fixes the most obvious case: more work will be needed here).
Signed-off-by: Hugh Dickins <hugh@veritas.com> Cc: Adam Litke <agl@us.ibm.com> Cc: David Gibson <david@gibson.dropbear.id.au> Cc: "Chen, Kenneth W" <kenneth.w.chen@intel.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Hugh Dickins [Sat, 28 Oct 2006 17:38:43 +0000 (10:38 -0700)]
[PATCH] hugetlb: fix prio_tree unit
hugetlb_vmtruncate_list was misconverted to prio_tree: its prio_tree is in
units of PAGE_SIZE (PAGE_CACHE_SIZE) like any other, not HPAGE_SIZE (whereas
its radix_tree is kept in units of HPAGE_SIZE, otherwise slots would be
absurdly sparse).
At first I thought the error benign, just calling __unmap_hugepage_range on
more vmas than necessary; but on 32-bit machines, when the prio_tree is
searched correctly, it happens to ensure the v_offset calculation won't
overflow. As it stood, when truncating at or beyond 4GB, it was liable to
discard pages COWed from lower offsets; or even to clear pmd entries of
preceding vmas, triggering exit_mmap's BUG_ON(nr_ptes).
Signed-off-by: Hugh Dickins <hugh@veritas.com> Cc: Adam Litke <agl@us.ibm.com> Cc: David Gibson <david@gibson.dropbear.id.au> Cc: "Chen, Kenneth W" <kenneth.w.chen@intel.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Hugh Dickins [Sat, 28 Oct 2006 17:38:41 +0000 (10:38 -0700)]
[PATCH] hugetlb: fix size=4G parsing
On 32-bit machines, mount -t hugetlbfs -o size=4G gave a 0GB filesystem,
size=5G gave a 1GB filesystem etc: there's no point in masking size with
HPAGE_MASK just before shifting its lower bits away, and since HPAGE_MASK is a
UL, that removed all the higher bits of the unsigned long long size.
Signed-off-by: Hugh Dickins <hugh@veritas.com> Cc: Adam Litke <agl@us.ibm.com> Cc: David Gibson <david@gibson.dropbear.id.au> Cc: "Chen, Kenneth W" <kenneth.w.chen@intel.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Randy Dunlap [Sat, 28 Oct 2006 17:38:40 +0000 (10:38 -0700)]
[PATCH] cciss: fix printk format warning
Fix printk format warnings:
drivers/block/cciss.c:2000: warning: long long int format, long unsigned int arg (arg 2)
drivers/block/cciss.c:2035: warning: long long int format, long unsigned int arg (arg 2)
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Acked-by: Mike Miller <mike.miller@hp.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
[PATCH] __vmalloc with GFP_ATOMIC causes 'sleeping from invalid context'
If __vmalloc is called to allocate memory with GFP_ATOMIC in atomic
context, the chain of calls results in __get_vm_area_node allocating memory
for vm_struct with GFP_KERNEL, causing the 'sleeping from invalid context'
warning. This patch fixes it by passing the gfp flags along so
__get_vm_area_node allocates memory for vm_struct with the same flags.
Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Pavel Emelianov [Sat, 28 Oct 2006 17:38:33 +0000 (10:38 -0700)]
[PATCH] Fix potential OOPs in blkdev_open()
blkdev_open() calls bc_acquire() to get a struct block_device. Since
bc_acquire() may return NULL when system is out of memory an appropriate
check is required.
Signed-off-by: Pavel Emelianov <xemul@openvz.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Yasunori Goto [Sat, 28 Oct 2006 17:38:32 +0000 (10:38 -0700)]
[PATCH] memory hotplug: __GFP_NOWARN is better for __kmalloc_section_memmap()
Add __GFP_NOWARN flag to calling of __alloc_pages() in
__kmalloc_section_memmap(). It can reduce noisy failure message.
In ia64, section size is 1 GB, this means that order 8 pages are necessary
for each section's memmap. It is often very hard requirement under heavy
memory pressure as you know. So, __alloc_pages() gives up allocation and
shows many noisy stack traces which means no page for each sections.
(Current my environment shows 32 times of stack trace....)
But, __kmalloc_section_memmap() calls vmalloc() after failure of it, and it
can succeed allocation of memmap. So, its stack trace warning becomes just
noisy. I suppose it shouldn't be shown.
Randy Dunlap [Sat, 28 Oct 2006 17:38:32 +0000 (10:38 -0700)]
[PATCH] md: fix printk format warnings, seen on powerpc64:
drivers/md/raid1.c:1479: warning: long long unsigned int format, long unsigned int arg (arg 4)
drivers/md/raid10.c:1475: warning: long long unsigned int format, long unsigned int arg (arg 4)
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
NeilBrown [Sat, 28 Oct 2006 17:38:31 +0000 (10:38 -0700)]
[PATCH] md: fix up maintenance of ->degraded in multipath
A recent fix which made sure ->degraded was initialised properly exposed a
second bug - ->degraded wasn't been updated when drives failed or were
hot-added.
Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
NeilBrown [Sat, 28 Oct 2006 17:38:30 +0000 (10:38 -0700)]
[PATCH] md: simplify checking of available size when resizing an array
When "mdadm --grow --size=xxx" is used to resize an array (use more or less of
each device), we check the new siza against the available space in each
device.
We already have that number recorded in rdev->size, so calculating it is
pointless (and wrong in one obscure case).
Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
bibo,mao [Sat, 28 Oct 2006 17:38:29 +0000 (10:38 -0700)]
[PATCH] fix efi_memory_present_wrapper()
efi_memory_present_wrapper() parameter start/end is physical address, but
function memory_present parameter is PFN, this patch converts physical
address to PFN.
Signed-off-by: bibo, mao <bibo.mao@intel.com> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Eric Sandeen [Sat, 28 Oct 2006 17:38:28 +0000 (10:38 -0700)]
[PATCH] jbd2: journal_dirty_data re-check for unmapped buffers
When running several fsx's and other filesystem stress tests, we found
cases where an unmapped buffer was still being sent to submit_bh by the
ext3 dirty data journaling code.
I saw this happen in two ways, both related to another thread doing a
truncate which would unmap the buffer in question.
Either we would get into journal_dirty_data with a bh which was already
unmapped (although journal_dirty_data_fn had checked for this earlier, the
state was not locked at that point), or it would get unmapped in the middle
of journal_dirty_data when we dropped locks to call sync_dirty_buffer.
By re-checking for mapped state after we've acquired the bh state lock, we
should avoid these races. If we find a buffer which is no longer mapped,
we essentially ignore it, because journal_unmap_buffer has already decided
that this buffer can go away.
I've also added tracepoints in these two cases, and made a couple other
tracepoint changes that I found useful in debugging this.
Signed-off-by: Eric Sandeen <esandeen@redhat.com> Cc: <linux-ext4@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Eric Sandeen [Sat, 28 Oct 2006 17:38:27 +0000 (10:38 -0700)]
[PATCH] jbd: journal_dirty_data re-check for unmapped buffers
When running several fsx's and other filesystem stress tests, we found
cases where an unmapped buffer was still being sent to submit_bh by the
ext3 dirty data journaling code.
I saw this happen in two ways, both related to another thread doing a
truncate which would unmap the buffer in question.
Either we would get into journal_dirty_data with a bh which was already
unmapped (although journal_dirty_data_fn had checked for this earlier, the
state was not locked at that point), or it would get unmapped in the middle
of journal_dirty_data when we dropped locks to call sync_dirty_buffer.
By re-checking for mapped state after we've acquired the bh state lock, we
should avoid these races. If we find a buffer which is no longer mapped,
we essentially ignore it, because journal_unmap_buffer has already decided
that this buffer can go away.
I've also added tracepoints in these two cases, and made a couple other
tracepoint changes that I found useful in debugging this.
Signed-off-by: Eric Sandeen <esandeen@redhat.com> Cc: <linux-ext4@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Randy Dunlap [Sat, 28 Oct 2006 17:38:26 +0000 (10:38 -0700)]
[PATCH] ext4: fix printk format warnings
fs/ext4/resize.c:72: warning: long long unsigned int format, __u64 arg (arg 4)
fs/ext4/resize.c:76: warning: long long unsigned int format, __u64 arg (arg 4)
fs/ext4/resize.c:81: warning: long long unsigned int format, __u64 arg (arg 4)
fs/ext4/resize.c:85: warning: long long unsigned int format, __u64 arg (arg 4)
fs/ext4/resize.c:89: warning: long long unsigned int format, __u64 arg (arg 4)
fs/ext4/resize.c:89: warning: long long unsigned int format, __u64 arg (arg 5)
fs/ext4/resize.c:93: warning: long long unsigned int format, __u64 arg (arg 4)
fs/ext4/resize.c:93: warning: long long unsigned int format, __u64 arg (arg 5)
fs/ext4/resize.c:98: warning: long long unsigned int format, __u64 arg (arg 4)
fs/ext4/resize.c:103: warning: long long unsigned int format, __u64 arg (arg 4)
fs/ext4/resize.c:109: warning: long long unsigned int format, __u64 arg (arg 4)
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Martin Bligh [Sat, 28 Oct 2006 17:38:25 +0000 (10:38 -0700)]
[PATCH] Use min of two prio settings in calculating distress for reclaim
If try_to_free_pages / balance_pgdat are called with a gfp_mask specifying
GFP_IO and/or GFP_FS, they will reclaim the requisite number of pages, and the
reset prev_priority to DEF_PRIORITY (or to some other high (ie: unurgent)
value).
However, another reclaimer without those gfp_mask flags set (say, GFP_NOIO)
may still be struggling to reclaim pages. The concurrent overwrite of
zone->prev_priority will cause this GFP_NOIO thread to unexpectedly cease
deactivating mapped pages, thus causing reclaim difficulties.
Fix this is to key the distress calculation not off zone->prev_priority, but
also take into account the local caller's priority by using
min(zone->prev_priority, sc->priority)
Signed-off-by: Martin J. Bligh <mbligh@google.com> Cc: Nick Piggin <nickpiggin@yahoo.com.au> Cc: <stable@kernel.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Martin Bligh [Sat, 28 Oct 2006 17:38:24 +0000 (10:38 -0700)]
[PATCH] vmscan: Fix temp_priority race
The temp_priority field in zone is racy, as we can walk through a reclaim
path, and just before we copy it into prev_priority, it can be overwritten
(say with DEF_PRIORITY) by another reclaimer.
The same bug is contained in both try_to_free_pages and balance_pgdat, but
it is fixed slightly differently. In balance_pgdat, we keep a separate
priority record per zone in a local array. In try_to_free_pages there is
no need to do this, as the priority level is the same for all zones that we
reclaim from.
Impact of this bug is that temp_priority is copied into prev_priority, and
setting this artificially high causes reclaimers to set distress
artificially low. They then fail to reclaim mapped pages, when they are,
in fact, under severe memory pressure (their priority may be as low as 0).
This causes the OOM killer to fire incorrectly.
From: Andrew Morton <akpm@osdl.org>
__zone_reclaim() isn't modifying zone->prev_priority. But zone->prev_priority
is used in the decision whether or not to bring mapped pages onto the inactive
list. Hence there's a risk here that __zone_reclaim() will fail because
zone->prev_priority ir large (ie: low urgency) and lots of mapped pages end up
stuck on the active list.
Fix that up by decreasing (ie making more urgent) zone->prev_priority as
__zone_reclaim() scans the zone's pages.
This bug perhaps explains why ZONE_RECLAIM_PRIORITY was created. It should be
possible to remove that now, and to just start out at DEF_PRIORITY?
Cc: Nick Piggin <nickpiggin@yahoo.com.au> Cc: Christoph Lameter <clameter@engr.sgi.com> Cc: <stable@kernel.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/pci-2.6:
PCI: Remove quirk_via_abnormal_poweroff
PCI: reset pci device state to unknown state for resume
PCI: x86-64: mmconfig missing printk levels
PCI: fix pci_fixup_video as it blows up on sparc64
acpiphp: fix latch status
Andrew Morton [Fri, 27 Oct 2006 18:42:37 +0000 (11:42 -0700)]
[PATCH] drivers: wait for threaded probes between initcall levels
The multithreaded-probing code has a problem: after one initcall level (eg,
core_initcall) has been processed, we will then start processing the next
level (postcore_initcall) while the kernel threads which are handling
core_initcall are still executing. This breaks the guarantees which the
layered initcalls previously gave us.
IOW, we want to be multithreaded _within_ an initcall level, but not between
different levels.
Fix that up by causing the probing code to wait for all outstanding probes at
one level to complete before we start processing the next level.
Karsten Wiese [Fri, 20 Oct 2006 21:45:36 +0000 (14:45 -0700)]
PCI: Remove quirk_via_abnormal_poweroff
My K8T800 mobo resumes fine from suspend to ram with and without patch
applied against 2.6.18.
quirk_via_abnormal_poweroff makes some boards not boot 2.6.18, so IMO patch
should go to head, 2.6.18.2 and everywhere "ACPI: ACPICA 20060623" has been
applied.
Remove quirk_via_abnormal_poweroff
Obsoleted by "ACPI: ACPICA 20060623":
<snip>
Implemented support for "ignored" bits in the ACPI
registers. According to the ACPI specification, these
bits should be preserved when writing the registers via
a read/modify/write cycle. There are 3 bits preserved
in this manner: PM1_CONTROL[0] (SCI_EN), PM1_CONTROL[9],
and PM1_STATUS[11].
http://bugzilla.kernel.org/show_bug.cgi?id=3691
</snip>
Signed-off-by: Karsten Wiese <fzu@wemgehoertderstaat.de> Cc: Bob Moore <robert.moore@intel.com> Cc: "Brown, Len" <len.brown@intel.com> Acked-by: Dave Jones <davej@redhat.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Shaohua Li [Fri, 20 Oct 2006 21:45:32 +0000 (14:45 -0700)]
PCI: reset pci device state to unknown state for resume
Considering below scenario:
1.Unload a PCI device's driver, the device ->current remains in PCI_D0.
2.Do suspend/resume circle. After that, BIOS puts the device to D3.
3.Reload the device driver. The calling pci_set_power_state in the
driver can't change the state to D0, as set_power_state thinks the
device is already in D0.
A bug is reported at http://bugzilla.kernel.org/show_bug.cgi?id=6024
Pat attached a patch at
http://marc.theaimsgroup.com/?l=linux-pci&m=114049761428561&w=2 for this
issue, but it's lost. As pci_set_power_state can handle D3 -> D0
correctly (restore config space), I simplified Patrick's patch.
Signed-off-by: Shaohua Li <shaohua.li@intel.com> Cc: Patrick Mochel <mochel@digitalimplant.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Ralph Wuerthner [Fri, 27 Oct 2006 10:39:26 +0000 (12:39 +0200)]
[S390] Improve AP bus device removal.
Added a call to device_unregister() in ap_scan_bus() to actively
remove unavailable AP bus devices with every bus scan. Previously
devices were only removed in ap_queue_message() or __ap_poll_all().
Signed-off-by: Ralph Wuerthner <rwuerthn@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Paul Mundt [Fri, 27 Oct 2006 10:39:10 +0000 (12:39 +0200)]
[S390] sys_getcpu compat wrapper.
Looking at the new syscall additions, I noticed that
sys_getcpu_wrapper wraps in to sys_tee, in what appears to be
a copy and paste error. Switch it to point to sys_getcpu..
Signed-off-by: Paul Mundt <lethal@linux-sh.org> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Gavin McCullagh [Thu, 26 Oct 2006 06:05:52 +0000 (23:05 -0700)]
[TCP] H-TCP: fix integer overflow
When using H-TCP with a single flow on a 500Mbit connection (or less
actually), alpha can exceed 65000, so alpha needs to be a u32.
Signed-off-by: Gavin McCullagh <gavin.mccullagh@nuim.ie> Signed-off-by: Doug Leith <doug.leith@nuim.ie> Signed-off-by: David S. Miller <davem@davemloft.net>
Doug Leith observed a discrepancy between the version of CUBIC described
in the papers and the version in 2.6.18. A math error related to scaling
causes Cubic to grow too slowly.
Patch is from "Sangtae Ha" <sha2@ncsu.edu>. I validated that
it does fix the problems.
See the following to show behavior over 500ms 100 Mbit link.
Sync atstk1002_defconfig with latest git, turn off non-existent
drivers and enable a few more userspace-visible options like
SysV IPC and inotify support.
[PATCH] AVR32: Use __raw MMIO access for internal peripherals
The read[bwl] and write[bwl] functions are meant for accessing PCI
devices. How this is achieved on AVR32 is unknown, as there are no
systems with a PCI bridge available yet.
On-chip peripheral access, however, should not depend on how we end
up implementing PCI access, so using __raw_read[bwl]/__raw_write[bwl]
is the right thing to do for on-chip peripherals. This patch converts
the drivers for the static memory controller, interrupt controller,
PIO controller and system manager to use __raw MMIO access.
Ben Nizette [Tue, 24 Oct 2006 08:12:43 +0000 (10:12 +0200)]
[PATCH] AVR32: add io{read,write}{8,16,32}{be,} support
A number of new drivers require io{read,write}{8,16,32}{be,} family of io
operations. These are provided for the AVR32 by this patch in the form of
a series of macros.
Access to the (memory mapped) io space through these macros is defined to
be little endian only as little endian devices (such as PCI) are the main
consumer of IO access. If high speed access is required,
io{read,write}{16,32}be macros are supplied to perform native big endian
access to this io space.
[PATCH] AVR32: Fix oversize immediates in atomic.h
When calling e.g. atomic_sub_return with a large constant, the
compiler may output an immediate that is too large for the sub
instruction in the middle of the loop.
Fix this by explicitly specifying the number of bits allowed in the
constraint. Also stop atomic_add_return() and friends from falling
back to their respective "sub" variants if the constant is too large
to fit in an immediate.
[PATCH] AVR32: Don't try to iounmap P2 segment addresses
While ioremap() will happily map a physical address through the
P2 (uncached) segment when appropriate, iounmap() doesn't know how
to handle those mappings.
This patch makes iounmap() do the right thing, i.e. nothing, for
such mappings.
Linus Torvalds [Thu, 26 Oct 2006 03:22:04 +0000 (20:22 -0700)]
Merge branch 'intelfb-patches' of master.kernel.org:/pub/scm/linux/kernel/git/airlied/intelfb-2.6
* 'intelfb-patches' of master.kernel.org:/pub/scm/linux/kernel/git/airlied/intelfb-2.6:
Remove unnecessary check in drivers/video/intelfb/intelfbhw.c
intel fb: switch to pci_get API
Linus Torvalds [Thu, 26 Oct 2006 03:21:43 +0000 (20:21 -0700)]
Merge branch 'drm-patches' of master.kernel.org:/pub/scm/linux/kernel/git/airlied/drm-2.6
* 'drm-patches' of master.kernel.org:/pub/scm/linux/kernel/git/airlied/drm-2.6:
drm: ioremap balanced with iounmap for drivers/char/drm
drm: fix error returns, sysfs error handling
fix return code in error case.
drm: mga: set dev_priv_size
drm: savage: dev->agp_buffer_map is not initialized for AGP DMA on savages
drm: radeon: only allow specific type-3 packetss through verifier
Linus Torvalds [Thu, 26 Oct 2006 03:21:16 +0000 (20:21 -0700)]
Merge branch 'for-linus' of git://one.firstfloor.org/home/andi/git/linux-2.6
* 'for-linus' of git://one.firstfloor.org/home/andi/git/linux-2.6:
[PATCH] x86-64: Only look at per_cpu data for online cpus.
[PATCH] x86-64: Simplify the vector allocator.
Linus Torvalds [Thu, 26 Oct 2006 03:20:03 +0000 (20:20 -0700)]
Merge branch 'merge' of master.kernel.org:/pub/scm/linux/kernel/git/paulus/powerpc
* 'merge' of master.kernel.org:/pub/scm/linux/kernel/git/paulus/powerpc:
[POWERPC] Make sure __cpu_preinit_ppc970 gets called on 970GX processors
[POWERPC] Fix CHRP platforms with only 8259
[POWERPC] IPIC: Fix spinlock recursion in set_irq_handler
[POWERPC] Fix the UCC rx/tx clock of QE
[POWERPC] cell: update defconfig
[POWERPC] spufs: fix another off-by-one bug in spufs_mbox_read
[POWERPC] spufs: fix signal2 file to report signal2
[POWERPC] Fix device_is_compatible() const warning
[POWERPC] Cell timebase bug workaround
[POWERPC] Support feature fixups in modules
[POWERPC] Support feature fixups in vdso's
[POWERPC] Support nested cpu feature sections
[POWERPC] Consolidate feature fixup code
[POWERPC] Fix hang in start_ldr if _end or _edata is unaligned
[POWERPC] Fix spelling errors in ucc_fast.c and ucc_slow.c
[POWERPC] Don't require execute perms on wrapper when building zImage.initrd
[POWERPC] Add 970GX cputable entry
[POWERPC] Fix build breakage with CONFIG_PPC32
[POWERPC] Fix compiler warning message on get_property call
[POWERPC] Simplify stolen time calculation
Hannes Reinecke [Mon, 23 Oct 2006 13:26:37 +0000 (15:26 +0200)]
[SCSI] aic79xx: Print out signalling
This is a cross-port of a similar patch for aic7xxx;
only it's a bit simpler here as we don't support HVD
and all controller actually implement this register.
I hope.
Signed-off-by: Hannes Reinecke <hare@suse.de> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Hannes Reinecke [Mon, 23 Oct 2006 13:25:36 +0000 (15:25 +0200)]
[SCSI] aic7xxx: Remove slave_destroy
This is a cross-port from aic79xx; we still hit the occasional
BUG_ON in slave_destroy. And again we don't really need the
slave_destroy callback nor the ahc_linux_target structure
at all.
Signed-off-by: Hannes Reinecke <hare@suse.de> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Hannes Reinecke [Mon, 23 Oct 2006 13:22:37 +0000 (15:22 +0200)]
[SCSI] aic79xx: Fixup external device reset
Whenever an external device is resetted we really have to take
care to keep the channel in sync. Just notifying SCSI-ML and retry
is not enough as we have to make sure the SCSI bus is not getting
confused, either.
So whenever we detect an external reset we rewrite the command to
TUR, disable packetized command and notify the internal engine
that an abort has happened. This way we trigger a proper bus
reset sequence and all devices will be renegotiated properly.
Kudos to Justin Gibbs and Luben Tuikov for this idea.
Signed-off-by: Hannes Reinecke <hare@suse.de> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Randy Dunlap [Tue, 24 Oct 2006 04:47:13 +0000 (21:47 -0700)]
[SCSI] lpfc: fix printk format warning
Fix printk format warning:
drivers/scsi/lpfc/lpfc_attr.c:597: warning: long long unsigned int format, uint64_t arg (arg 4)
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Acked-by: James Smart <James.Smart@Emulex.Com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Adrian Bunk [Fri, 20 Oct 2006 21:48:07 +0000 (14:48 -0700)]
[SCSI] aic79xx: make ahd_set_tags() static
Signed-off-by: Adrian Bunk <bunk@stusta.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Acked-by: Hannes Reinecke <hare@suse.de> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Adrian Bunk [Fri, 20 Oct 2006 21:47:57 +0000 (14:47 -0700)]
[SCSI] aic7xxx: cleanups
- make needlessly global code static
- #if 0 the following unused global functions:
- aic79xx_core.c: ahd_print_scb
- aic79xx_core.c: ahd_suspend
- aic79xx_core.c: ahd_resume
- aic79xx_core.c: ahd_dump_scbs
- aic79xx_osm.c: ahd_softc_comp
Signed-off-by: Adrian Bunk <bunk@stusta.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Acked-by: Hannes Reinecke <hare@suse.de> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Hannes Reinecke [Fri, 20 Oct 2006 07:58:47 +0000 (09:58 +0200)]
[SCSI] scsi_debug: support REPORT TARGET PORT GROUPS
This patch adds support for REPORT TARGET PORT GROUPS. This is used
eg for the multipathing priority callout to determine the path
priority.
With this patch multipath-tools can use the existing mpath_prio_alua
callout to exercise the path priority grouping.
Signed-off-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Douglas Gilbert <dougg@torque.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Mike Christie [Mon, 16 Oct 2006 22:09:42 +0000 (18:09 -0400)]
[SCSI] libiscsi: fix logout pdu processing
According to the iscsi RFC, we cannot send other requests if
we have sent a logout pdu. This patch enforces this requirement
by blocking the session and suspending the send thread. Userspace
decides if we restart the connection or if we just free everything.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Mike Christie [Mon, 16 Oct 2006 22:09:40 +0000 (18:09 -0400)]
[SCSI] libiscsi: fix missed iscsi_task_put in xmit error path
from bhalevy@gmail.com:
It looks like change 652 to libiscsi.c added some dead code around line
670
if (rc) {
spin_unlock_bh(&conn->session->lock);
goto again;
}
since 5 lines above we goto again if (rc).
It looks like the previous if (rc) should go away if we want to put the
ctask before
breaking out of the while loop with "goto again" (see following patch).
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Mike Christie [Mon, 16 Oct 2006 22:09:39 +0000 (18:09 -0400)]
[SCSI] libiscsi: fix oops in connection create failure path
If connection creation fails we end up calling list_del
on a invalid struct. This then causes an oops. We are not
acutally using the lists (old MCS code we thought might
be useful elsewhere) so this patch just removes that
code.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Mike Christie [Mon, 16 Oct 2006 22:09:38 +0000 (18:09 -0400)]
[SCSI] iscsi class: fix slab corruption during restart
The transport class recv mempools are causing slab corruption.
We could hack around netlink's lack of mempool support like dm,
but it is just too ulgy (dm's hack is ugly enough :) when you need
to support broadcast.
This patch removes the recv pools. We have not used them even when
we were allocting 20 MB per session and the system only had 64 MBs.
And we have no pools on the send side and have been ok there. When
Peter's work gets merged we can use that since the network guys
are in favor of that approach and are not going to add mempools
everywhere.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Alan Cox [Mon, 16 Oct 2006 15:55:46 +0000 (16:55 +0100)]
[SCSI] Switch fdomain to the pci_get API
Doesn't make the hardware hot pluggable but does ensure the driver won't
crash when another device is hot-unplugged at the wrong moment. Soon I
propose to deprecate pci_find_device() and some of its friends.
Signed-off-by: Alan Cox <alan@redhat.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
It is known that 2 LSI Logic MegaRAID SATA RAID Controllers (150-4 and
150-6) don't support 64-bit DMA. Unfortunately currently this check is
wrong and driver sets 64-bit DMA mode for these devices.
projects / linux-2.6 / log