linux-toradex.git/include, branch v3.4.5

memblock: free allocated memblock_reserved_regions later

2012-07-16T16:04:45+00:00

commit 29f6738609e40227dabcc63bfb3b84b3726a75bd upstream.

memblock_free_reserved_regions() calls memblock_free(), but
memblock_free() would double reserved.regions too, so we could free the
old range for reserved.regions.

Also tj said there is another bug which could be related to this.

| I don't think we're saving any noticeable
| amount by doing this "free - give it to page allocator - reserve
| again" dancing.  We should just allocate regions aligned to page
| boundaries and free them later when memblock is no longer in use.

in that case, when DEBUG_PAGEALLOC, will get panic:

     memblock_free: [0x0000102febc080-0x0000102febf080] memblock_free_reserved_regions+0x37/0x39
  BUG: unable to handle kernel paging request at ffff88102febd948
  IP: [] __next_free_mem_range+0x9b/0x155
  PGD 4826063 PUD cf67a067 PMD cf7fa067 PTE 800000102febd160
  Oops: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC
  CPU 0
  Pid: 0, comm: swapper Not tainted 3.5.0-rc2-next-20120614-sasha #447
  RIP: 0010:[]  [] __next_free_mem_range+0x9b/0x155

See the discussion at https://lkml.org/lkml/2012/6/13/469

So try to allocate with PAGE_SIZE alignment and free it later.

Reported-by: Sasha Levin 
Acked-by: Tejun Heo 
Cc: Benjamin Herrenschmidt 
Signed-off-by: Yinghai Lu 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds 
Signed-off-by: Greg Kroah-Hartman

memory hotplug: fix invalid memory access caused by stale kswapd pointer

2012-07-16T16:04:44+00:00

commit d8adde17e5f858427504725218c56aef90e90fc7 upstream.

kswapd_stop() is called to destroy the kswapd work thread when all memory
of a NUMA node has been offlined.  But kswapd_stop() only terminates the
work thread without resetting NODE_DATA(nid)->kswapd to NULL.  The stale
pointer will prevent kswapd_run() from creating a new work thread when
adding memory to the memory-less NUMA node again.  Eventually the stale
pointer may cause invalid memory access.

An example stack dump as below. It's reproduced with 2.6.32, but latest
kernel has the same issue.

  BUG: unable to handle kernel NULL pointer dereference at (null)
  IP: [] exit_creds+0x12/0x78
  PGD 0
  Oops: 0000 [#1] SMP
  last sysfs file: /sys/devices/system/memory/memory391/state
  CPU 11
  Modules linked in: cpufreq_conservative cpufreq_userspace cpufreq_powersave acpi_cpufreq microcode fuse loop dm_mod tpm_tis rtc_cmos i2c_i801 rtc_core tpm serio_raw pcspkr sg tpm_bios igb i2c_core iTCO_wdt rtc_lib mptctl iTCO_vendor_support button dca bnx2 usbhid hid uhci_hcd ehci_hcd usbcore sd_mod crc_t10dif edd ext3 mbcache jbd fan ide_pci_generic ide_core ata_generic ata_piix libata thermal processor thermal_sys hwmon mptsas mptscsih mptbase scsi_transport_sas scsi_mod
  Pid: 7949, comm: sh Not tainted 2.6.32.12-qiuxishi-5-default #92 Tecal RH2285
  RIP: 0010:exit_creds+0x12/0x78
  RSP: 0018:ffff8806044f1d78  EFLAGS: 00010202
  RAX: 0000000000000000 RBX: ffff880604f22140 RCX: 0000000000019502
  RDX: 0000000000000000 RSI: 0000000000000202 RDI: 0000000000000000
  RBP: ffff880604f22150 R08: 0000000000000000 R09: ffffffff81a4dc10
  R10: 00000000000032a0 R11: ffff880006202500 R12: 0000000000000000
  R13: 0000000000c40000 R14: 0000000000008000 R15: 0000000000000001
  FS:  00007fbc03d066f0(0000) GS:ffff8800282e0000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
  CR2: 0000000000000000 CR3: 000000060f029000 CR4: 00000000000006e0
  DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
  Process sh (pid: 7949, threadinfo ffff8806044f0000, task ffff880603d7c600)
  Stack:
   ffff880604f22140 ffffffff8103aac5 ffff880604f22140 ffffffff8104d21e
   ffff880006202500 0000000000008000 0000000000c38000 ffffffff810bd5b1
   0000000000000000 ffff880603d7c600 00000000ffffdd29 0000000000000003
  Call Trace:
    __put_task_struct+0x5d/0x97
    kthread_stop+0x50/0x58
    offline_pages+0x324/0x3da
    memory_block_change_state+0x179/0x1db
    store_mem_state+0x9e/0xbb
    sysfs_write_file+0xd0/0x107
    vfs_write+0xad/0x169
    sys_write+0x45/0x6e
    system_call_fastpath+0x16/0x1b
  Code: ff 4d 00 0f 94 c0 84 c0 74 08 48 89 ef e8 1f fd ff ff 5b 5d 31 c0 41 5c c3 53 48 8b 87 20 06 00 00 48 89 fb 48 8b bf 18 06 00 00 <8b> 00 48 c7 83 18 06 00 00 00 00 00 00 f0 ff 0f 0f 94 c0 84 c0
  RIP  exit_creds+0x12/0x78
   RSP 
  CR2: 0000000000000000

[akpm@linux-foundation.org: add pglist_data.kswapd locking comments]
Signed-off-by: Xishi Qiu 
Signed-off-by: Jiang Liu 
Acked-by: KAMEZAWA Hiroyuki 
Acked-by: KOSAKI Motohiro 
Acked-by: Mel Gorman 
Acked-by: David Rientjes 
Reviewed-by: Minchan Kim 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds 
Signed-off-by: Greg Kroah-Hartman

splice: fix racy pipe->buffers uses

2012-07-16T16:04:42+00:00

commit 047fe3605235888f3ebcda0c728cb31937eadfe6 upstream.

Dave Jones reported a kernel BUG at mm/slub.c:3474! triggered
by splice_shrink_spd() called from vmsplice_to_pipe()

commit 35f3d14dbbc5 (pipe: add support for shrinking and growing pipes)
added capability to adjust pipe->buffers.

Problem is some paths don't hold pipe mutex and assume pipe->buffers
doesn't change for their duration.

Fix this by adding nr_pages_max field in struct splice_pipe_desc, and
use it in place of pipe->buffers where appropriate.

splice_shrink_spd() loses its struct pipe_inode_info argument.

Reported-by: Dave Jones 
Signed-off-by: Eric Dumazet 
Cc: Jens Axboe 
Cc: Alexander Viro 
Cc: Tom Herbert 
Tested-by: Dave Jones 
Signed-off-by: Jens Axboe 
[bwh: Backported to 3.2:
 - Adjust context in vmsplice_to_pipe()
 - Update one more call to splice_shrink_spd(), from skb_splice_bits()]
Signed-off-by: Ben Hutchings 
Signed-off-by: Greg Kroah-Hartman

thp: avoid atomic64_read in pmd_read_atomic for 32bit PAE

2012-07-16T16:04:42+00:00

commit e4eed03fd06578571c01d4f1478c874bb432c815 upstream.

In the x86 32bit PAE CONFIG_TRANSPARENT_HUGEPAGE=y case while holding the
mmap_sem for reading, cmpxchg8b cannot be used to read pmd contents under
Xen.

So instead of dealing only with "consistent" pmdvals in
pmd_none_or_trans_huge_or_clear_bad() (which would be conceptually
simpler) we let pmd_none_or_trans_huge_or_clear_bad() deal with pmdvals
where the low 32bit and high 32bit could be inconsistent (to avoid having
to use cmpxchg8b).

The only guarantee we get from pmd_read_atomic is that if the low part of
the pmd was found null, the high part will be null too (so the pmd will be
considered unstable).  And if the low part of the pmd is found "stable"
later, then it means the whole pmd was read atomically (because after a
pmd is stable, neither MADV_DONTNEED nor page faults can alter it anymore,
and we read the high part after the low part).

In the 32bit PAE x86 case, it is enough to read the low part of the pmdval
atomically to declare the pmd as "stable" and that's true for THP and no
THP, furthermore in the THP case we also have a barrier() that will
prevent any inconsistent pmdvals to be cached by a later re-read of the
*pmd.

Signed-off-by: Andrea Arcangeli 
Cc: Jonathan Nieder 
Cc: Ulrich Obergfell 
Cc: Mel Gorman 
Cc: Hugh Dickins 
Cc: Larry Woodman 
Cc: Petr Matousek 
Cc: Rik van Riel 
Cc: Jan Beulich 
Cc: KOSAKI Motohiro 
Tested-by: Andrew Jones 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds 
Signed-off-by: Greg Kroah-Hartman

mm: pmd_read_atomic: fix 32bit PAE pmd walk vs pmd_populate SMP race condition

2012-07-16T16:04:41+00:00

commit 26c191788f18129af0eb32a358cdaea0c7479626 upstream.

When holding the mmap_sem for reading, pmd_offset_map_lock should only
run on a pmd_t that has been read atomically from the pmdp pointer,
otherwise we may read only half of it leading to this crash.

PID: 11679  TASK: f06e8000  CPU: 3   COMMAND: "do_race_2_panic"
 #0 [f06a9dd8] crash_kexec at c049b5ec
 #1 [f06a9e2c] oops_end at c083d1c2
 #2 [f06a9e40] no_context at c0433ded
 #3 [f06a9e64] bad_area_nosemaphore at c043401a
 #4 [f06a9e6c] __do_page_fault at c0434493
 #5 [f06a9eec] do_page_fault at c083eb45
 #6 [f06a9f04] error_code (via page_fault) at c083c5d5
    EAX: 01fb470c EBX: fff35000 ECX: 00000003 EDX: 00000100 EBP:
    00000000
    DS:  007b     ESI: 9e201000 ES:  007b     EDI: 01fb4700 GS:  00e0
    CS:  0060     EIP: c083bc14 ERR: ffffffff EFLAGS: 00010246
 #7 [f06a9f38] _spin_lock at c083bc14
 #8 [f06a9f44] sys_mincore at c0507b7d
 #9 [f06a9fb0] system_call at c083becd
                         start           len
    EAX: ffffffda  EBX: 9e200000  ECX: 00001000  EDX: 6228537f
    DS:  007b      ESI: 00000000  ES:  007b      EDI: 003d0f00
    SS:  007b      ESP: 62285354  EBP: 62285388  GS:  0033
    CS:  0073      EIP: 00291416  ERR: 000000da  EFLAGS: 00000286

This should be a longstanding bug affecting x86 32bit PAE without THP.
Only archs with 64bit large pmd_t and 32bit unsigned long should be
affected.

With THP enabled the barrier() in pmd_none_or_trans_huge_or_clear_bad()
would partly hide the bug when the pmd transition from none to stable,
by forcing a re-read of the *pmd in pmd_offset_map_lock, but when THP is
enabled a new set of problem arises by the fact could then transition
freely in any of the none, pmd_trans_huge or pmd_trans_stable states.
So making the barrier in pmd_none_or_trans_huge_or_clear_bad()
unconditional isn't good idea and it would be a flakey solution.

This should be fully fixed by introducing a pmd_read_atomic that reads
the pmd in order with THP disabled, or by reading the pmd atomically
with cmpxchg8b with THP enabled.

Luckily this new race condition only triggers in the places that must
already be covered by pmd_none_or_trans_huge_or_clear_bad() so the fix
is localized there but this bug is not related to THP.

NOTE: this can trigger on x86 32bit systems with PAE enabled with more
than 4G of ram, otherwise the high part of the pmd will never risk to be
truncated because it would be zero at all times, in turn so hiding the
SMP race.

This bug was discovered and fully debugged by Ulrich, quote:

----
[..]
pmd_none_or_trans_huge_or_clear_bad() loads the content of edx and
eax.

    496 static inline int pmd_none_or_trans_huge_or_clear_bad(pmd_t
    *pmd)
    497 {
    498         /* depend on compiler for an atomic pmd read */
    499         pmd_t pmdval = *pmd;

                                // edi = pmd pointer
0xc0507a74 :   mov    0x8(%esp),%edi
...
                                // edx = PTE page table high address
0xc0507a84 :   mov    0x4(%edi),%edx
...
                                // eax = PTE page table low address
0xc0507a8e :   mov    (%edi),%eax

[..]

Please note that the PMD is not read atomically. These are two "mov"
instructions where the high order bits of the PMD entry are fetched
first. Hence, the above machine code is prone to the following race.

-  The PMD entry {high|low} is 0x0000000000000000.
   The "mov" at 0xc0507a84 loads 0x00000000 into edx.

-  A page fault (on another CPU) sneaks in between the two "mov"
   instructions and instantiates the PMD.

-  The PMD entry {high|low} is now 0x00000003fda38067.
   The "mov" at 0xc0507a8e loads 0xfda38067 into eax.
----

Reported-by: Ulrich Obergfell 
Signed-off-by: Andrea Arcangeli 
Cc: Mel Gorman 
Cc: Hugh Dickins 
Cc: Larry Woodman 
Cc: Petr Matousek 
Cc: Rik van Riel 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds 
Signed-off-by: Greg Kroah-Hartman

SUNRPC: new svc_bind() routine introduced

2012-07-16T16:04:39+00:00

upstream commit 9793f7c88937e7ac07305ab1af1a519225836823.

This new routine is responsible for service registration in a specified
network context.

The idea is to separate service creation from per-net operations.

Note also: since registering service with svc_bind() can fail, the
service will be destroyed and during destruction it will try to
unregister itself from rpcbind. In this case unregistration has to be
skipped.

Signed-off-by: Stanislav Kinsbursky 
Signed-off-by: J. Bruce Fields 
Signed-off-by: Greg Kroah-Hartman

Lockd: pass network namespace to creation and destruction routines

2012-07-16T16:04:39+00:00

upstream commit e3f70eadb7dddfb5a2bb9afff7abfc6ee17a29d0.

v2: dereference of most probably already released nlm_host removed in
nlmclnt_done() and reclaimer().

These routines are called from locks reclaimer() kernel thread. This thread
works in "init_net" network context and currently relays on persence on lockd
thread and it's per-net resources. Thus lockd_up() and lockd_down() can't relay
on current network context. So let's pass corrent one into them.

Signed-off-by: Stanislav Kinsbursky 
Signed-off-by: J. Bruce Fields 
Signed-off-by: Greg Kroah-Hartman

PCI: EHCI: fix crash during suspend on ASUS computers

2012-07-16T16:04:39+00:00

commit dbf0e4c7257f8d684ec1a3c919853464293de66e upstream.

Quite a few ASUS computers experience a nasty problem, related to the
EHCI controllers, when going into system suspend.  It was observed
that the problem didn't occur if the controllers were not put into the
D3 power state before starting the suspend, and commit
151b61284776be2d6f02d48c23c3625678960b97 (USB: EHCI: fix crash during
suspend on ASUS computers) was created to do this.

It turned out this approach messed up other computers that didn't have
the problem -- it prevented USB wakeup from working.  Consequently
commit c2fb8a3fa25513de8fedb38509b1f15a5bbee47b (USB: add
NO_D3_DURING_SLEEP flag and revert 151b61284776be2) was merged; it
reverted the earlier commit and added a whitelist of known good board
names.

Now we know the actual cause of the problem.  Thanks to AceLan Kao for
tracking it down.

According to him, an engineer at ASUS explained that some of their
BIOSes contain a bug that was added in an attempt to work around a
problem in early versions of Windows.  When the computer goes into S3
suspend, the BIOS tries to verify that the EHCI controllers were first
quiesced by the OS.  Nothing's wrong with this, but the BIOS does it
by checking that the PCI COMMAND registers contain 0 without checking
the controllers' power state.  If the register isn't 0, the BIOS
assumes the controller needs to be quiesced and tries to do so.  This
involves making various MMIO accesses to the controller, which don't
work very well if the controller is already in D3.  The end result is
a system hang or memory corruption.

Since the value in the PCI COMMAND register doesn't matter once the
controller has been suspended, and since the value will be restored
anyway when the controller is resumed, we can work around the BIOS bug
simply by setting the register to 0 during system suspend.  This patch
(as1590) does so and also reverts the second commit mentioned above,
which is now unnecessary.

In theory we could do this for every PCI device.  However to avoid
introducing new problems, the patch restricts itself to EHCI host
controllers.

Finally the affected systems can suspend with USB wakeup working
properly.

Reference: https://bugzilla.kernel.org/show_bug.cgi?id=37632
Reference: https://bugzilla.kernel.org/show_bug.cgi?id=42728
Based-on-patch-by: AceLan Kao 
Signed-off-by: Alan Stern 
Tested-by: Dâniel Fraga 
Tested-by: Javier Marcet 
Tested-by: Andrey Rahmatullin 
Tested-by: Oleksij Rempel 
Tested-by: Pavel Pisa 
Acked-by: Bjorn Helgaas 
Acked-by: Rafael J. Wysocki 
Signed-off-by: Greg Kroah-Hartman

SCSI: libsas: fix taskfile corruption in sas_ata_qc_fill_rtf

2012-07-16T16:04:37+00:00

commit 6ef1b512f4e6f936d89aa20be3d97a7ec7c290ac upstream.

fill_result_tf() grabs the taskfile flags from the originating qc which
sas_ata_qc_fill_rtf() promptly overwrites.  The presence of an
ata_taskfile in the sata_device makes it tempting to just copy the full
contents in sas_ata_qc_fill_rtf().  However, libata really only wants
the fis contents and expects the other portions of the taskfile to not
be touched by ->qc_fill_rtf.  To that end store a fis buffer in the
sata_device and use ata_tf_from_fis() like every other ->qc_fill_rtf()
implementation.

Reported-by: Praveen Murali 
Tested-by: Praveen Murali 
Signed-off-by: Dan Williams 
Signed-off-by: James Bottomley 
Signed-off-by: Greg Kroah-Hartman

SCSI: Fix NULL dereferences in scsi_cmd_to_driver

2012-07-16T16:04:30+00:00

commit 222a806af830fda34ad1f6bc991cd226916de060 upstream.

Avoid crashing if the private_data pointer happens to be NULL. This has
been seen sometimes when a host reset happens, notably when there are
many LUNs:

host3: Assigned Port ID 0c1601
scsi host3: libfc: Host reset succeeded on port (0c1601)
BUG: unable to handle kernel NULL pointer dereference at 0000000000000350
IP: [] scsi_send_eh_cmnd+0x58/0x3a0

Process scsi_eh_3 (pid: 4144, threadinfo ffff88030920c000, task ffff880326b160c0)
Stack:
 000000010372e6ba 0000000000000282 000027100920dca0 ffffffffa0038ee0
 0000000000000000 0000000000030003 ffff88030920dc80 ffff88030920dc80
 00000002000e0000 0000000a00004000 ffff8803242f7760 ffff88031326ed80
Call Trace:
 [] ? lock_timer_base+0x70/0x70
 [] scsi_eh_tur+0x3e/0xc0
 [] scsi_eh_test_devices+0x76/0x170
 [] scsi_eh_host_reset+0x85/0x160
 [] scsi_eh_ready_devs+0x91/0x110
 [] scsi_unjam_host+0xed/0x1f0
 [] scsi_error_handler+0x1a8/0x200
 [] ? scsi_unjam_host+0x1f0/0x1f0
 [] kthread+0x9e/0xb0
 [] kernel_thread_helper+0x4/0x10
 [] ? kthread_freezable_should_stop+0x70/0x70
 [] ? gs_change+0x13/0x13
Code: 25 28 00 00 00 48 89 45 c8 31 c0 48 8b 87 80 00 00 00 48 8d b5 60 ff ff ff 89 d1 48 89 fb 41 89 d6 4c 89 fa 48 8b 80 b8 00 00 00
 <48> 8b 80 50 03 00 00 48 8b 00 48 89 85 38 ff ff ff 48 8b 07 4c
RIP  [] scsi_send_eh_cmnd+0x58/0x3a0
 RSP 
CR2: 0000000000000350


Signed-off-by: Mark Rustad 
Tested-by: Marcus Dennis 
Signed-off-by: James Bottomley 
Signed-off-by: Greg Kroah-Hartman