George Dunlap [Wed, 3 Apr 2013 14:46:28 +0000 (15:46 +0100)]
perf/x86: Check all MSRs before passing hw check
check_hw_exists() has a number of checks which go to two exit
paths: msr_fail and bios_fail. Checks classified as msr_fail
will cause check_hw_exists() to return false, causing the PMU
not to be used; bios_fail checks will only cause a warning to be
printed, but will return true.
The problem is that if there are both msr failures and bios
failures, and the routine hits a bios_fail check first, it will
exit early and return true, not finishing the rest of the msr
checks. If those msrs are in fact broken, it will cause them to
be used erroneously.
In the case of a Xen PV VM, the guest OS has read access to all
the MSRs, but write access is white-listed to supported
features. Writes to unsupported MSRs have no effect. The PMU
MSRs are not (typically) supported, because they are expensive
to save and restore on a VM context switch. One of the
"msr_fail" checks is supposed to detect this circumstance (ether
for Xen or KVM) and disable the harware PMU.
However, on one of my AMD boxen, there is (apparently) a broken
BIOS which triggers one of the bios_fail checks. In particular,
MSR_K7_EVNTSEL0 has the ARCH_PERFMON_EVENTSEL_ENABLE bit set.
The guest kernel detects this because it has read access to all
MSRs, and causes it to skip the rest of the checks and try to
use the non-existent hardware PMU. This minimally causes a lot
of useless instruction emulation and Xen console spam; it may
cause other issues with the watchdog as well.
This changset causes check_hw_exists() to go through all of the
msr checks, failing and returning false if any of them fail.
This makes sure that a guest running under Xen without a virtual
PMU will detect that there is no functioning PMU and not attempt
to use it.
This problem affects kernels as far back as 3.2, and should thus
be considered for backport.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com> Cc: Konrad Wilk <konrad.wilk@oracle.com> Cc: Ian Campbell <ian.campbell@citrix.com> Cc: David Vrabel <david.vrabel@citrix.com> Cc: Andrew Cooper <andrew.cooper3@citrix.com> Link: http://lkml.kernel.org/r/1365000388-32448-1-git-send-email-george.dunlap@eu.citrix.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
Jacob Shin [Fri, 19 Apr 2013 21:34:28 +0000 (16:34 -0500)]
perf/x86/amd: Add support for AMD NB and L2I "uncore" counters
Add support for AMD Family 15h [and above] northbridge
performance counters. MSRs 0xc0010240 ~ 0xc0010247 are shared
across all cores that share a common northbridge.
Add support for AMD Family 16h L2 performance counters. MSRs
0xc0010230 ~ 0xc0010237 are shared across all cores that share a
common L2 cache.
We do not enable counter overflow interrupts. Sampling mode and
per-thread events are not supported.
Signed-off-by: Jacob Shin <jacob.shin@amd.com> Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Cc: Stephane Eranian <eranian@google.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/20130419213428.GA8229@jshin-Toonie Signed-off-by: Ingo Molnar <mingo@kernel.org>
The uncore subsystem in Ivy Bridge-EP is similar to Sandy
Bridge-EP. There are some differences in config register
encoding and pci device IDs. The Ivy Bridge-EP uncore also
supports a few new events.
perf/x86/intel: Fix SNB-EP CBO and PCU uncore PMU filter management
The existing code assumes all Cbox and PCU events are using
filter, but actually the filter is event specific. Furthermore
the filter is sub-divided into multiple fields which are used
by different events.
The valid mask for both offcore_response_0 and
offcore_response_1 was wrong for SNB/SNB-EP,
IVB/IVB-EP. It was possible to write to
reserved bit and cause a GP fault crashing
the kernel.
This patch fixes the problem by correctly marking the
reserved bits in the valid mask for all the processors
mentioned above.
A distinction between desktop and server parts is introduced
because bits 24-30 are only available on the server parts.
This version of the patch is just a rebase to perf/urgent tree
and should apply to older kernels as well.
Merge branch 'uprobes/core' of git://git.kernel.org/pub/scm/linux/kernel/git/oleg/misc into perf/core
Pull uprobes updates from Oleg Nesterov:
- "uretprobes" - an optimization to uprobes, like kretprobes are an optimization
to kprobes. "perf probe -x file sym%return" now works like kretprobes.
- PowerPC fixes plus a couple of cleanups/optimizations in uprobes and trace_uprobes.
uprobes/perf: Avoid perf_trace_buf_prepare/submit if ->perf_events is empty
perf_trace_buf_prepare() + perf_trace_buf_submit() make no sense
if this task/CPU has no active counters. Change uprobe_perf_print()
to return if hlist_empty(call->perf_events).
Note: this is not uprobe-specific, we can change other users too.
Tommi Rantala [Sat, 13 Apr 2013 19:49:14 +0000 (22:49 +0300)]
perf: Treat attr.config as u64 in perf_swevent_init()
Trinity discovered that we fail to check all 64 bits of
attr.config passed by user space, resulting to out-of-bounds
access of the perf_swevent_enabled array in
sw_perf_event_destroy().
Introduced in commit b0a873ebb ("perf: Register PMU
implementations").
Signed-off-by: Tommi Rantala <tt.rantala@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: davej@redhat.com Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Link: http://lkml.kernel.org/r/1365882554-30259-1-git-send-email-tt.rantala@gmail.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
Merge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Ingo Molnar:
"Misc fixes"
* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/mm: Flush lazy MMU when DEBUG_PAGEALLOC is set
x86/mm/cpa/selftest: Fix false positive in CPA self test
x86/mm/cpa: Convert noop to functional fix
x86, mm: Patch out arch_flush_lazy_mmu_mode() when running on bare metal
x86, mm, paravirt: Fix vmalloc_fault oops during lazy MMU updates
Merge branch 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull scheduler fixes from Ingo Molnar:
"Misc fixlets"
* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/cputime: Fix accounting on multi-threaded processes
sched/debug: Fix sd->*_idx limit range avoiding overflow
sched_clock: Prevent 64bit inatomicity on 32bit systems
sched: Convert BUG_ON()s in try_to_wake_up_local() to WARN_ON_ONCE()s
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Pull drm fixes from Dave Airlie:
"One fix for a hotplug locking regressions, and one fix for an oops if
you unplug the monitor at an inopportune moment on the udl device."
* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
drm/fb-helper: Fix locking in drm_fb_helper_hotplug_event
udl: handle EDID failure properly.
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs
Pull one more btrfs fix from Chris Mason:
"This has a recent fix from Josef for our tree log replay code. It
fixes problems where the inode counter for the number of bytes in the
file wasn't getting updated properly during fsync replay.
The commit did get rebased this morning, but it was only to clean up
the subject line. The code hasn't changed."
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
Btrfs: make sure nbytes are right after log replay
Merge tag 'trace-fixes-v3.9-rc-v3' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull ftrace fixes from Steven Rostedt:
"Namhyung Kim found and fixed a bug that can crash the kernel by simply
doing: echo 1234 | tee -a /sys/kernel/debug/tracing/set_ftrace_pid
Luckily, this can only be done by root, but still is a nasty bug."
* tag 'trace-fixes-v3.9-rc-v3' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
ftrace: Move ftrace_filter_lseek out of CONFIG_DYNAMIC_FTRACE section
tracing: Fix possible NULL pointer dereferences
Add file_ns_capable() helper function for open-time capability checking
Nothing is using it yet, but this will allow us to delay the open-time
checks to use time, without breaking the normal UNIX permission
semantics where permissions are determined by the opener (and the file
descriptor can then be passed to a different process, or the process can
drop capabilities).
Nicolas Ferre [Wed, 10 Apr 2013 12:36:22 +0000 (14:36 +0200)]
watchdog: Revert the AT91RM9200_WATCHDOG dependency
Compiling the at91rm9200_wdt.c driver without at91rm9200
support was leading to several errors:
drivers/built-in.o: In function `at91_wdt_close':
at91_adc.c:(.text+0xc9fe4): undefined reference to `at91_st_base'
drivers/built-in.o: In function `at91_wdt_write':
at91_adc.c:(.text+0xca004): undefined reference to `at91_st_base'
drivers/built-in.o: In function `at91wdt_shutdown':
at91_adc.c:(.text+0xca01c): undefined reference to `at91_st_base'
drivers/built-in.o: In function `at91wdt_suspend':
at91_adc.c:(.text+0xca038): undefined reference to `at91_st_base'
drivers/built-in.o: In function `at91_wdt_open':
at91_adc.c:(.text+0xca0cc): undefined reference to `at91_st_base'
drivers/built-in.o:at91_adc.c:(.text+0xca2c8): more undefined references to
`at91_st_base' follow
So, reverting the modification of the "depends" Kconfig line
introduced by patch a6a1bcd37 (watchdog: at91rm9200: add DT support)
seems to be the good solution.
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
vfs: Revert spurious fix to spinning prevention in prune_icache_sb
Revert commit 62a3ddef6181 ("vfs: fix spinning prevention in prune_icache_sb").
This commit doesn't look right: since we are looking at the tail of the
list (sb->s_inode_lru.prev) if we want to skip an inode, we should put
it back at the head of the list instead of the tail, otherwise we will
keep spinning on it.
Discovered when investigating why prune_icache_sb came top in perf
reports of a swapping load.
kobject: fix kset_find_obj() race with concurrent last kobject_put()
Anatol Pomozov identified a race condition that hits module unloading
and re-loading. To quote Anatol:
"This is a race codition that exists between kset_find_obj() and
kobject_put(). kset_find_obj() might return kobject that has refcount
equal to 0 if this kobject is freeing by kobject_put() in other
thread.
Here is timeline for the crash in case if kset_find_obj() searches for
an object tht nobody holds and other thread is doing kobject_put() on
the same kobject:
THREAD A (calls kset_find_obj()) THREAD B (calls kobject_put())
splin_lock()
atomic_dec_return(kobj->kref), counter gets zero here
... starts kobject cleanup ....
spin_lock() // WAIT thread A in kobj_kset_leave()
iterate over kset->list
atomic_inc(kobj->kref) (counter becomes 1)
spin_unlock()
spin_lock() // taken
// it does not know that thread A increased counter so it
remove obj from list
spin_unlock()
vfree(module) // frees module object with containing kobj
// kobj points to freed memory area!!
kobject_put(kobj) // OOPS!!!!
The race above happens because module.c tries to use kset_find_obj()
when somebody unloads module. The module.c code was introduced in
commit 6494a93d55fa"
Anatol supplied a patch specific for module.c that worked around the
problem by simply not using kset_find_obj() at all, but rather than make
a local band-aid, this just fixes kset_find_obj() to be thread-safe
using the proper model of refusing the get a new reference if the
refcount has already dropped to zero.
See examples of this proper refcount handling not only in the kref
documentation, but in various other equivalent uses of this pattern by
grepping for atomic_inc_not_zero().
[ Side note: the module race does indicate that module loading and
unloading is not properly serialized wrt sysfs information using the
module mutex. That may require further thought, but this is the
correct fix at the kobject layer regardless. ]
uprobes/tracing: Don't pass addr=ip to perf_trace_buf_submit()
uprobe_perf_print() passes addr=ip to perf_trace_buf_submit() for
no reason. This sets perf_sample_data->addr for PERF_SAMPLE_ADDR,
we already have perf_sample_data->ip initialized if PERF_SAMPLE_IP.
Oleg Nesterov [Sat, 30 Mar 2013 17:46:22 +0000 (18:46 +0100)]
uprobes/tracing: Make uprobe_{trace,perf}_print() uretprobe-friendly
Change uprobe_trace_print() and uprobe_perf_print() to check
is_ret_probe() and fill ring_buffer_event accordingly.
Also change uprobe_trace_func() and uprobe_perf_func() to not
_print() if is_ret_probe() is true. Note that we keep ->handler()
nontrivial even for uretprobe, we need this for filtering and for
other potential extensions.
Oleg Nesterov [Sat, 30 Mar 2013 17:25:23 +0000 (18:25 +0100)]
uprobes/tracing: Introduce is_ret_probe() and uretprobe_dispatcher()
Create the new functions we need to support uretprobes, and change
alloc_trace_uprobe() to initialize consumer.ret_handler if the new
"is_ret" argument is true. Curently this argument is always false,
so the new code is never called and is_ret_probe(tu) is false too.
Extract the output code from uprobe_trace_func() and uprobe_perf_func()
into the new helpers, they will be used by ->ret_handler() too. We also
add the unused "unsigned long func" argument in advance, to simplify the
next changes.
struct uprobe_trace_entry_head has a single member for reporting,
"unsigned long ip". If we want to support uretprobes we need to
create another struct which has "func" and "ret_ip" and duplicate
a lot of functions, like trace_kprobe.c does.
To avoid this copy-and-paste horror we turn ->ip into ->vaddr[]
and add couple of trivial helpers to calculate sizeof/data. This
uglifies the code a bit, but this allows us to avoid a lot more
complications later, when we add the support for ret-probes.
Anton Arapov [Wed, 3 Apr 2013 16:00:37 +0000 (18:00 +0200)]
uretprobes: Limit the depth of return probe nestedness
Unlike the kretprobes we can't trust userspace, thus must have
protection from user space attacks. User-space have "unlimited"
stack, and this patch limits the return probes nestedness as a
simple remedy for it.
Note that this implementation leaks return_instance on siglongjmp
until exit()/exec().
The intention is to have KISS and bare minimum solution for the
initial implementation in order to not complicate the uretprobes
code.
In the future we may come up with more sophisticated solution that
remove this depth limitation. It is not easy task and lays beyond
this patchset.
Anton Arapov [Wed, 3 Apr 2013 16:00:36 +0000 (18:00 +0200)]
uretprobes: Return probe exit, invoke handlers
Uretprobe handlers are invoked when the trampoline is hit, on completion
the trampoline is replaced with the saved return address and the uretprobe
instance deleted.
TODO: handle_trampoline() assumes that ->return_instances is always valid.
We should teach it to handle longjmp() which can invalidate the pending
return_instance's. This is nontrivial, we will try to do this in a separate
series.
When a uprobe with return probe consumer is hit, prepare_uretprobe()
function is invoked. It creates return_instance, hijacks return address
and replaces it with the trampoline.
* Return instances are kept as stack per uprobed task.
* Return instance is chained, when the original return address is
trampoline's page vaddr (e.g. recursive call of the probed function).
Anton Arapov [Wed, 3 Apr 2013 16:00:32 +0000 (18:00 +0200)]
uretprobes: Reserve the first slot in xol_vma for trampoline
Allocate trampoline page, as the very first one in uprobed
task xol area, and fill it with breakpoint opcode.
Also introduce get_trampoline_vaddr() helper, to wrap the
trampoline address extraction from area->vaddr. That removes
confusion and eases the debug experience in case ->vaddr
notion will be changed.
Enclose return probes implementation, introduce ->ret_handler() and update
existing code to rely on ->handler() *and* ->ret_handler() for uprobe and
uretprobe respectively.
Josef Bacik [Fri, 5 Apr 2013 20:50:09 +0000 (20:50 +0000)]
Btrfs: make sure nbytes are right after log replay
While trying to track down a tree log replay bug I noticed that fsck was always
complaining about nbytes not being right for our fsynced file. That is because
the new fsync stuff doesn't wait for ordered extents to complete, so the inodes
nbytes are not necessarily updated properly when we log it. So to fix this we
need to set nbytes to whatever it is on the inode that is on disk, so when we
replay the extents we can just add the bytes that are being added as we replay
the extent. This makes it work for the case that we have the wrong nbytes or
the case that we logged everything and nbytes is actually correct. With this
I'm no longer getting nbytes errors out of btrfsck.
Cc: stable@vger.kernel.org Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Ingo guesses this got introduced by commit 611ae8e3f520 ("x86/tlb:
enable tlb flush range support for x86") since that code started to free
unused pagetables.
On x86-32 PAE kernels, that new code has the potential to free an entire
PMD page and will clear one of the four page-directory-pointer-table
(aka pgd_t entries).
The hardware aggressively "caches" these top-level entries and invlpg
does not actually affect the CPU's copy. If we clear one we *HAVE* to
do a full TLB flush, otherwise we might continue using a freed pmd page.
(note, we do this properly on the population side in pud_populate()).
This patch tracks whenever we clear one of these entries in the 'struct
mmu_gather', and ensures that we follow up with a full tlb flush.
BTW, I disassembled and checked that:
if (tlb->fullmm == 0)
and
if (!tlb->fullmm && !tlb->need_flush_all)
generate essentially the same code, so there should be zero impact there
to the !PAE case.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Cc: Peter Anvin <hpa@zytor.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Artem S Tashkinov <t.artem@mailcity.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull SCSI target fixes from Nicholas Bellinger:
"Here are remaining target-pending items for v3.9-rc7 code.
The tcm_vhost patches are more than I'd usually include in a -rc7
pull, but are changes required for v3.9 to work correctly with the
pending vhost-scsi-pci QEMU upstream series merge. (Paolo CC'ed)
Plus Asias's conversion to use vhost_virtqueue->private_data + RCU for
managing vhost-scsi endpoints has gotten alot of review + testing over
the past weeks, and MST has ACKed the full series.
Also, there is a target patch to fix a long-standing bug within
control CDB handling with Standby/Offline/Transition ALUA port access
states, that had been incorrectly rejecting the control CDBs required
for LUN scan to work during these port group states. CC'ing to
stable."
* git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending:
target: Fix incorrect fallthrough of ALUA Standby/Offline/Transition CDBs
tcm_vhost: Send bad target to guest when cmd fails
tcm_vhost: Add vhost_scsi_send_bad_target() helper
tcm_vhost: Fix tv_cmd leak in vhost_scsi_handle_vq
tcm_vhost: Remove double check of response
tcm_vhost: Initialize vq->last_used_idx when set endpoint
tcm_vhost: Use vq->private_data to indicate if the endpoint is setup
tcm_vhost: Use ACCESS_ONCE for vs->vs_tpg[target] access
Merge tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Pull SCSI fixes from James Bottomley:
"This is a set of ten bug fixes (and two consisting of copyright year
update and version number change) pretty much all of which involve
either a crash or a hang except the removal of the random sleep from
the qla2xxx driver (which is a coding error so bad, we want it gone
before anyone has a chance to copy it)."
* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
[SCSI] lpfc: fix potential NULL pointer dereference in lpfc_sli4_rq_put()
[SCSI] libsas: fix handling vacant phy in sas_set_ex_phy()
[SCSI] ibmvscsi: Fix slave_configure deadlock
[SCSI] qla2xxx: Update the driver version to 8.04.00.13-k.
[SCSI] qla2xxx: Remove debug code that msleeps for random duration.
[SCSI] qla2xxx: Update copyright dates information in LICENSE.qla2xxx file.
[SCSI] qla2xxx: Fix crash during firmware dump procedure.
[SCSI] Revert "qla2xxx: Add setting of driver version string for vendor application."
[SCSI] ipr: dlpar failed when adding an adapter back
[SCSI] ipr: fix addition of abort command to HRRQ free queue
[SCSI] st: Take additional queue ref in st_probe
[SCSI] libsas: use right function to alloc smp response
[SCSI] ipr: ipr_test_msi() fails when running with msi-x enabled adapter
ftrace: Move ftrace_filter_lseek out of CONFIG_DYNAMIC_FTRACE section
As ftrace_filter_lseek is now used with ftrace_pid_fops, it needs to
be moved out of the #ifdef CONFIG_DYNAMIC_FTRACE section as the
ftrace_pid_fops is defined when DYNAMIC_FTRACE is not.
Cc: stable@vger.kernel.org Cc: Namhyung Kim <namhyung@kernel.org> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Namhyung Kim [Thu, 11 Apr 2013 06:55:01 +0000 (15:55 +0900)]
tracing: Fix possible NULL pointer dereferences
Currently set_ftrace_pid and set_graph_function files use seq_lseek
for their fops. However seq_open() is called only for FMODE_READ in
the fops->open() so that if an user tries to seek one of those file
when she open it for writing, it sees NULL seq_file and then panic.
It can be easily reproduced with following command:
$ cd /sys/kernel/debug/tracing
$ echo 1234 | sudo tee -a set_ftrace_pid
In this example, GNU coreutils' tee opens the file with fopen(, "a")
and then the fopen() internally calls lseek().
Link: http://lkml.kernel.org/r/1365663302-2170-1-git-send-email-namhyung@kernel.org Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: stable@vger.kernel.org Signed-off-by: Namhyung Kim <namhyung@kernel.org> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Merge tag 'sound-3.9' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Pull sound fixes from Takashi Iwai:
"This contains a few small ASoC fixes (wm8903, wm5102, samsung-i2s,
tegra, and soc-compress) and an endian fix for NI USB-audio devices,
update for Mark's e-mail address.
No scary changes, AFAIS."
* tag 'sound-3.9' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
MAINTAINERS: Update e-mail address
ASoC: wm5102: Correct lookup of arizona struct in SYSCLK event
ASoC: wm8903: Fix the bypass to HP/LINEOUT when no DAC or ADC is running
ALSA: usb-audio: fix endianness bug in snd_nativeinstruments_*
ASoC: tegra: Don't claim to support PCM pause and resume
ASoC: Samsung: set drvdata before adding secondary device
ASoC: Samsung: return error if drvdata is not set
ASoC: compress: Cancel delayed power down if needed
ASoC: core: Fix to check return value of snd_soc_update_bits_locked()
Merge tag 'asoc-v3.9-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus
ASoC: Updates for v3.9
A few updates, more than I'd like, fixing some relatively small issues
but mostly driver specific ones. Nothing wildly exciting so if it
doesn't make v3.9 it won't be the end of the world but it'd be nice.
Boris Ostrovsky [Thu, 11 Apr 2013 17:59:52 +0000 (13:59 -0400)]
x86/mm: Flush lazy MMU when DEBUG_PAGEALLOC is set
When CONFIG_DEBUG_PAGEALLOC is set page table updates made by
kernel_map_pages() are not made visible (via TLB flush)
immediately if lazy MMU is on. In environments that support lazy
MMU (e.g. Xen) this may lead to fatal page faults, for example,
when zap_pte_range() needs to allocate pages in
__tlb_remove_page() -> tlb_next_batch().
Daniel Vetter [Thu, 11 Apr 2013 14:26:55 +0000 (14:26 +0000)]
drm/fb-helper: Fix locking in drm_fb_helper_hotplug_event
Driver's and ->fill_modes functions are allowed to grab crtc mutexes
(for e.g. load detect). Hence we need to first only grab the general
kms mutex, and only in a second step grab all locks to do the
modesets.
This prevents a deadlock on my gm45 in the tv load detect code called
by drm_helper_probe_single_connector_modes.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Dave Airlie <airlied@redhat.com>
Merge branch 'fixes' of git://git.infradead.org/users/vkoul/slave-dma
Pull slave-dmaengine fixes from Vinod Koul:
"The first one fixes issue in pl330 to check for DT compatible and
the second one fixes omap-dma to start without delay"
* 'fixes' of git://git.infradead.org/users/vkoul/slave-dma:
dmaengine: omap-dma: Start DMA without delay for cyclic channels
DMA: PL330: Add check if device tree compatible
Merge tag 'pm-3.9-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management fixes from Rafael Wysocki:
- System reboot/halt fix related to CPU offline ordering from Huacai
Chen.
- intel_pstate driver fix for a delay time computation error
occasionally crashing systems using it from Dirk Brandewie.
* tag 'pm-3.9-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
PM / reboot: call syscore_shutdown() after disable_nonboot_cpus()
cpufreq / intel_pstate: Set timer timeout correctly
Merge tag 'regmap-v3.9-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap
Pull regmap revert from Mark Brown:
"regmap: Back out work buffer fix
This reverts commit bc8ce4 (regmap: don't corrupt work buffer in
_regmap_raw_write()) since it turns out that it can cause issues when
taken in isolation from the other changes in -next that lead to its
discovery. On the basis that nobody noticed the problems for quite
some time without that subsequent work let's drop it from v3.9."
* tag 'regmap-v3.9-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap:
regmap: Back out work buffer fix
Merge tag 'arm-soc-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc
Pull ARM SoC bug fixes from Arnd Bergmann:
"A little later during the week than the last few pull requests, since
there was very little that came in before 3.9-rc6. At least things
have calmed down again here.
Some important bug fixes that came in over the last 10 days, mostly
mvebu and imx:
- Multiple regressions on i.mx following the conversion of the clock
code, hopefully the last we are seeing of those.
- a regression in the mvebu irq handling code
- An incorrect register offset in the rewritten s3c24xx irq code.
- Two bugs in setting up the iomega_ix2_200 machine
- Turning on an extra bus clock on imx
- A MAINTAINERS file entry for Roland Stigge"
* tag 'arm-soc-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:
arm: mvebu: Fix the irq map function in SMP mode
Fix GE0/GE1 init on ix2-200 as GE0 has no PHY
ARM: S3C24XX: Fix interrupt pending register offset of the EINT controller
ARM: S3C24XX: Correct NR_IRQS definition for s3c2440
ARM i.MX6: Fix ldb_di clock selection
ARM: imx: provide twd clock lookup from device tree
ARM: imx35 Bugfix admux clock
ARM: clk-imx35: Bugfix iomux clock
ARM: mxs: Slow down the I2C clock speed
MAINTAINERS: Add maintainer for LPC32xx
ARM: Kirkwood: Fix typo in the definition of ix2-200 rebuild LED
We actually have to pass chip as the host_data parameter of
irq_domain_add_simple() as later on, it is used to initialize chip_data
in pca953x_gpio_irq_map(). Failing to do so is leading to a NULL pointer
dereference after calling irq_data_get_irq_chip_data() in
pca953x_irq_mask(), pca953x_irq_unmask(), pca953x_irq_bus_lock(),
pca953x_irq_bus_sync_unlock() and pca953x_irq_set_type().
Merge tag 'mvebu_fixes_for_v3.9_round3' of git://git.infradead.org/users/jcooper/linux into fixes
From Jason Cooper <jason@lakedaemon.net>:
mvebu fixes for v3.9 round 3
- Kirkwood
- a couple of small fixes for the Iomega ix2-200 board (ether and led)
- mvebu
- allow GPIO button to work on Mirabox when running SMP
* tag 'mvebu_fixes_for_v3.9_round3' of git://git.infradead.org/users/jcooper/linux:
arm: mvebu: Fix the irq map function in SMP mode
Fix GE0/GE1 init on ix2-200 as GE0 has no PHY
ARM: Kirkwood: Fix typo in the definition of ix2-200 rebuild LED
target: Fix incorrect fallthrough of ALUA Standby/Offline/Transition CDBs
This patch fixes a bug where a handful of informational / control CDBs
that should be allowed during ALUA access state Standby/Offline/Transition
where incorrectly returning CHECK_CONDITION + ASCQ_04H_ALUA_TG_PT_*.
This includes INQUIRY + REPORT_LUNS, which would end up preventing LUN
registration when LUN scanning occured during these ALUA access states.
Cc: Hannes Reinecke <hare@suse.de> Cc: <stable@vger.kernel.org> Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
Andrea Arcangeli [Wed, 10 Apr 2013 13:28:25 +0000 (15:28 +0200)]
x86/mm/cpa: Convert noop to functional fix
Commit:
a8aed3e0752b ("x86/mm/pageattr: Prevent PSE and GLOABL leftovers to confuse pmd/pte_present and pmd_huge")
introduced a valid fix but one location that didn't trigger the bug that
lead to finding those (small) problems, wasn't updated using the
right variable.
The wrong variable was also initialized for no good reason, that
may have been the source of the confusion. Remove the noop
initialization accordingly.
Commit a8aed3e0752b also erroneously removed one canon_pgprot pass meant
to clear pmd bitflags not supported in hardware by older CPUs, that
automatically gets corrected by this patch too by applying it to the right
variable in the new location.
Reported-by: Stefan Bader <stefan.bader@canonical.com> Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Borislav Petkov <bp@alien8.de> Cc: Andy Whitcroft <apw@canonical.com> Cc: Mel Gorman <mgorman@suse.de> Link: http://lkml.kernel.org/r/1365600505-19314-1-git-send-email-aarcange@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
Merge tag 'tty-3.9-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty
Pull tty/serial fixes from Greg Kroah-Hartman:
"Here are 4 small tty/serial fixes for 3.9.
One fixes a bug where we broke the documentation build, and the others
fix reported problems in some serial drivers.
All have been in linux-next for a while"
* tag 'tty-3.9-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
tty: mxser: fix cycle termination condition in mxser_probe() and mxser_module_init()
Revert "tty/8250_pnp: serial port detection regression since v3.7"
OMAP/serial: Revert bad fix of Rx FIFO threshold granularity
tty: Documentation: fix a path in a DocBook template
Merge tag 'stable/for-linus-3.9-rc6-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen
Pull Xen fixes from Konrad Rzeszutek Wilk:
"Two bug-fixes:
- Early bootup issue found on DL380 machines
- Fix for the timer interrupt not being processed right awaym leading
to quite delayed time skew on certain workloads"
* tag 'stable/for-linus-3.9-rc6-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
xen/mmu: On early bootup, flush the TLB when changing RO->RW bits Xen provided pagetables.
xen/events: Handle VIRQ_TIMER before any other hardirq in event loop.
Merge tag 'trace-fixes-3.9-rc-v2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull tracing fix from Steven Rostedt:
"Namhyung Kim fixed a long standing bug that can cause a kernel panic.
If the function profiler fails to allocate memory for everything, it
will do a double free on the same pointer which can cause a panic"
* tag 'trace-fixes-3.9-rc-v2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
tracing: Fix double free when function profile init failed
1) cfg80211_conn_scan() must be called with the sched_scan_mutex, fix
from Artem Savkov.
2) Fix regression in TCP ICMPv6 processing, we do not want to treat
redirects as socket errors, from Christoph Paasch.
3) Fix several recvmsg() msg_name kernel memory leaks into userspace,
in ATM, AX25, Bluetooth, CAIF, IRDA, s390 IUCV, L2TP, LLC, Netrom,
NFC, Rose, TIPC, and VSOCK. From Mathias Krause and Wei Yongjun.
4) Fix AF_IUCV handling of segmented SKBs in recvmsg(), from Ursula
Braun and Eric Dumazet.
5) CAN gw.c code does kfree() on SLAB cache memory, use
kmem_cache_free() instead. Fix from Wei Yongjun.
6) Fix LSM regression on TCP SYN/ACKs, some LSMs such as SELINUX want
an skb->sk socket context available for these packets, but nothing
else requires it. From Eric Dumazet and Paul Moore.
7) Fix ipv4 address lifetime processing so that we don't perform
sleepable acts inside of rcu_read_lock() sections, do them in an
rtnl_lock() section instead. From Jiri Pirko.
8) mvneta driver accidently sets HW features after device registry, it
should do so beforehand. Fix from Willy Tarreau.
9) Fix bonding unload races more correctly, from Nikolay Aleksandrov
and Veaceslav Falico.
10) rtnl_dump_ifinfo() and rtnl_calcit() invoke nlmsg_parse() with wrong
header size argument. Fix from Michael Riesch.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (44 commits)
lsm: add the missing documentation for the security_skb_owned_by() hook
bnx2x: Prevent null pointer dereference in AFEX mode
e100: Add dma mapping error check
selinux: add a skb_owned_by() hook
can: gw: use kmem_cache_free() instead of kfree()
netrom: fix invalid use of sizeof in nr_recvmsg()
qeth: fix qeth_wait_for_threads() deadlock for OSN devices
af_iucv: fix recvmsg by replacing skb_pull() function
rtnetlink: Call nlmsg_parse() with correct header length
bonding: fix bonding_masters race condition in bond unloading
Revert "bonding: remove sysfs before removing devices"
net: mvneta: enable features before registering the driver
hyperv: Fix RNDIS send_completion code path
hyperv: Fix a kernel warning from netvsc_linkstatus_callback()
net: ipv4: fix schedule while atomic bug in check_lifetime()
net: ipv4: reset check_lifetime_work after changing lifetime
bnx2x: Fix KR2 rapid link flap
sctp: remove 'sridhar' from maintainers list
VSOCK: Fix missing msg_namelen update in vsock_stream_recvmsg()
VSOCK: vmci - fix possible info leak in vmci_transport_dgram_dequeue()
...
cifs: Allow passwords which begin with a delimitor
Fixes a regression in cifs_parse_mount_options where a password
which begins with a delimitor is parsed incorrectly as being a blank
password.
Signed-off-by: Sachin Prabhu <sprabhu@redhat.com> Acked-by: Jeff Layton <jlayton@redhat.com> Cc: <stable@vger.kernel.org> Signed-off-by: Steve French <sfrench@us.ibm.com>
bnx2x: Prevent null pointer dereference in AFEX mode
The cnic module is responsible for initializing various bnx2x structs
via callbacks provided by the bnx2x module.
One such struct is the queue object for the FCoE queue.
If a device is working in AFEX mode and its configuration allows FCoE yet
the cnic module is not loaded, it's very likely a null pointer dereference
will occur, as the bnx2x will erroneously access the FCoE's queue object.
Prevent said access until cnic properly registers itself.
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com> Signed-off-by: Ariel Elior <ariele@broadcom.com> Signed-off-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Easy fix, modify the cb paramter to e100_exec_cb to return an error, and do the
dma_mapping_error check in the obvious place
This was reported previously here:
http://article.gmane.org/gmane.linux.network/257893
But nobody stepped up and fixed it.
CC: Josh Boyer <jwboyer@redhat.com> CC: e1000-devel@lists.sourceforge.net Signed-off-by: Neil Horman <nhorman@tuxdriver.com> Reported-by: Michal Jaegermann <michal@harddata.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Boris Ostrovsky [Sat, 23 Mar 2013 13:36:36 +0000 (09:36 -0400)]
x86, mm: Patch out arch_flush_lazy_mmu_mode() when running on bare metal
Invoking arch_flush_lazy_mmu_mode() results in calls to
preempt_enable()/disable() which may have performance impact.
Since lazy MMU is not used on bare metal we can patch away
arch_flush_lazy_mmu_mode() so that it is never called in such
environment.
[ hpa: the previous patch "Fix vmalloc_fault oops during lazy MMU
updates" may cause a minor performance regression on
bare metal. This patch resolves that performance regression. It is
somewhat unclear to me if this is a good -stable candidate. ]
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com> Link: http://lkml.kernel.org/r/1364045796-10720-2-git-send-email-konrad.wilk@oracle.com Tested-by: Josh Boyer <jwboyer@redhat.com> Tested-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Acked-by: Borislav Petkov <bp@suse.de> Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Cc: <stable@vger.kernel.org> SEE NOTE ABOVE
Samu Kallio [Sat, 23 Mar 2013 13:36:35 +0000 (09:36 -0400)]
x86, mm, paravirt: Fix vmalloc_fault oops during lazy MMU updates
In paravirtualized x86_64 kernels, vmalloc_fault may cause an oops
when lazy MMU updates are enabled, because set_pgd effects are being
deferred.
One instance of this problem is during process mm cleanup with memory
cgroups enabled. The chain of events is as follows:
- zap_pte_range enables lazy MMU updates
- zap_pte_range eventually calls mem_cgroup_charge_statistics,
which accesses the vmalloc'd mem_cgroup per-cpu stat area
- vmalloc_fault is triggered which tries to sync the corresponding
PGD entry with set_pgd, but the update is deferred
- vmalloc_fault oopses due to a mismatch in the PUD entries
Merge tag 'nfs-for-3.9-5' of git://git.linux-nfs.org/projects/trondmy/linux-nfs
Pull another nfs fixlet from Trond Myklebust:
"I suddenly noticed that a one-line issue that I _thought_ I had fixed
with the nfs41_walk_client_list patch was apparently still there in
the pull request I sent earlier today. I'm very sorry for not
catching that in time.
- Fix a brain fart in nfs41_walk_client_list"
* tag 'nfs-for-3.9-5' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
NFSv4: Doh! Typo in the fix to nfs41_walk_client_list
This patch fix the regression introduced by the commit 3202bf0157ccb
"arm: mvebu: Improve the SMP support of the interrupt controller":
GPIO IRQ were no longer delivered to the CPUs.
To be delivered to a CPU an interrupt must be enabled at CPU level and
at interrupt source level. Before the offending patch, all the
interrupts were enabled at source level during map() function. Mask()
and unmask() was done by handling the per-CPU part. It was fine when
running in UP with only one CPU.
The offending patch added support for SMP, in this case mask() and
unmask() was done by handling the interrupt source level part. The
per-CPU level part was handled by the affinity API to select the CPU
which will receive the interrupt. (Due to some hardware limitation
only one CPU at a time can received a given interrupt).
For "normal" interrupt __setup_irq() was called when an irq was
registered. irq_set_affinity() is called from this function, which
enabled the interrupt on one of the CPUs. Whereas for GPIO IRQ which
were chained interrupts, the irq_set_affinity() was never called and
none of the CPUs was selected to receive the interrupt.
With this patch all the interrupt are enable on the current CPU during
map() function. Enabling the interrupts on a CPU doesn't depend
anymore on irq_set_affinity() and then the chained irq are not anymore
a special case. However the CPU which will receive the irq can still
be modify later using irq_set_affinity().
Tested with Mirabox (A370) and Openblocks AX3 (AXP), rootfs mounted
over NFS, compiled with CONFIG_SMP=y/N.
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com> Reported-by: Ryan Press <ryan@presslab.us> Investigated-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com> Tested-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com> Tested-by: Ryan Press <ryan@presslab.us> Signed-off-by: Jason Cooper <jason@lakedaemon.net>
Merge tag 'nfs-for-3.9-4' of git://git.linux-nfs.org/projects/trondmy/linux-nfs
Pull NFS client bugfixes from Trond Myklebust:
- fix for memory corruption issues in nfs4[01]_walk_client_list (stable)
- fix for an Oopsable bug in rpc_clone_client (stable)
- another state manager deadlock in the NFSv4 open code
- memory leaks in nfs4_discover_server_trunking and rpc_new_client
* tag 'nfs-for-3.9-4' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
NFSv4: Fix another potential state manager deadlock
SUNRPC: Fix a potential memory leak in rpc_new_client
NFSv4/4.1: Fix bugs in nfs4[01]_walk_client_list
NFSv4: Fix a memory leak in nfs4_discover_server_trunking
SUNRPC: Remove extra xprt_put()
These recently documented events have restrictions to counter
0-3 and counter 2 respectively. The perf scheduler needs to know
that to schedule them correctly.
Peter Ujfalusi [Tue, 9 Apr 2013 14:33:06 +0000 (16:33 +0200)]
dmaengine: omap-dma: Start DMA without delay for cyclic channels
cyclic DMA is only used by audio which needs DMA to be started without a
delay.
If the DMA for audio is started using the tasklet we experience random
channel switch (to be more precise: channel shift).
Reported-by: Peter Meerwald <pmeerw@pmeerw.net> CC: stable@vger.kernel.org # v3.7+ Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com> Acked-by: Santosh Shilimkar <santosh.shilimkar@ti.com> Acked-by: Russell King <rmk+kernel@arm.linux.org.uk> Signed-off-by: Vinod Koul <vinod.koul@intel.com>