Thomas Gleixner [Thu, 15 Jan 2015 21:22:42 +0000 (21:22 +0000)]
x86/smpboot: Sanitize uniprocessor init
The UP related setups for local apic are mangled into smp_sanity_check().
That results in duplicate calls to disable_smp() and makes the code
hard to follow. Let smp_sanity_check() return dedicated values for the
various exit reasons and handle them at the call site.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Borislav Petkov <bp@alien8.de> Cc: Jiang Liu <jiang.liu@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Link: http://lkml.kernel.org/r/20150115211703.987833932@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Thomas Gleixner [Thu, 15 Jan 2015 21:22:40 +0000 (21:22 +0000)]
x86/smpboot: Move apic init code to apic.c
We better provide proper functions which implement the required code
flow in the apic code rather than letting the smpboot code open code
it. That allows to make more functions static and confines the APIC
functionality to apic.c where it belongs.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Borislav Petkov <bp@alien8.de> Cc: Jiang Liu <jiang.liu@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Link: http://lkml.kernel.org/r/20150115211703.907616730@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Thomas Gleixner [Thu, 15 Jan 2015 21:22:34 +0000 (21:22 +0000)]
x86/apic: Sanitize ioapic handling
We have proper stubs for the IOAPIC=n case and the setup/enable
function have the required checks inside now. Remove the ifdeffery and
the copy&pasted conditionals.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>C Acked-by: Borislav Petkov <bp@alien8.de> Cc: Jiang Liu <jiang.liu@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Link: http://lkml.kernel.org/r/20150115211703.569830549@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Thomas Gleixner [Thu, 15 Jan 2015 21:22:26 +0000 (21:22 +0000)]
x86/x2apic: Split enable and setup function
enable_x2apic() is a convoluted unreadable mess because it is used for
both enablement in early boot and for setup in cpu_init().
Split the code into x2apic_enable() for enablement and x2apic_setup()
for setup of (secondary cpus). Make use of the new state tracking to
simplify the logic.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Jiang Liu <jiang.liu@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Cc: Borislav Petkov <bp@alien8.de> Link: http://lkml.kernel.org/r/20150115211703.129287153@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Thomas Gleixner [Thu, 15 Jan 2015 21:22:22 +0000 (21:22 +0000)]
x86/x2apic: Add proper state tracking
Having 3 different variables to track the state is just silly and
error prone. Add a proper state tracking variable which covers the
three possible states: ON/OFF/DISABLED.
We cannot use x2apic_mode for this as this would require to change all
users of x2apic_mode with explicit comparisons for a state value
instead of treating it as boolean.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Borislav Petkov <bp@alien8.de> Cc: Jiang Liu <jiang.liu@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Link: http://lkml.kernel.org/r/20150115211702.955392443@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Thomas Gleixner [Thu, 15 Jan 2015 21:22:19 +0000 (21:22 +0000)]
x86/x2apic: Move code in conditional region
No point in having try_to_enable_x2apic() outside of the
CONFIG_X86_X2APIC section and having inline functions and more ifdefs
to deal with it. Move the code into the existing ifdef section and
remove the inline cruft.
Fixup the printk about not enabling interrupt remapping as suggested
by Boris.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Jiang Liu <jiang.liu@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Cc: Borislav Petkov <bp@alien8.de> Link: http://lkml.kernel.org/r/20150115211702.795388613@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Thomas Gleixner [Thu, 15 Jan 2015 21:22:17 +0000 (21:22 +0000)]
x86/apic: Check x2apic early
No point in delaying the x2apic detection for the CONFIG_X86_X2APIC=n
case to enable_IR_x2apic(). We rather detect that before we try to
setup anything there.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Jiang Liu <jiang.liu@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Cc: Borislav Petkov <bp@alien8.de> Link: http://lkml.kernel.org/r/20150115211702.702479404@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Thomas Gleixner [Thu, 15 Jan 2015 21:22:16 +0000 (21:22 +0000)]
x86/apic: Make disable x2apic work really
If x2apic_preenabled is not enabled, then disable_x2apic() is not
called from various places which results in x2apic_disabled not being
set. So other code pathes can happily reenable the x2apic.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Jiang Liu <jiang.liu@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Cc: Borislav Petkov <bp@alien8.de> Link: http://lkml.kernel.org/r/20150115211702.621431109@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Thomas Gleixner [Thu, 15 Jan 2015 21:22:14 +0000 (21:22 +0000)]
x86/ioapic: Check x2apic really
The x2apic_preenabled flag is just a horrible hack and if X2APIC
support is disabled it does not reflect the actual hardware
state. Check the hardware instead.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Jiang Liu <jiang.liu@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Cc: Borislav Petkov <bp@alien8.de> Link: http://lkml.kernel.org/r/20150115211702.541280622@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Thomas Gleixner [Thu, 15 Jan 2015 21:22:09 +0000 (21:22 +0000)]
x86/apic: Avoid open coded x2apic detection
enable_IR_x2apic() grew a open coded x2apic detection. Implement a
proper helper function which shares the code with the already existing
x2apic_enabled().
Made it use rdmsrl_safe as suggested by Boris.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Borislav Petkov <bp@alien8.de> Cc: Jiang Liu <jiang.liu@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Link: http://lkml.kernel.org/r/20150115211702.285038186@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Jiang Liu [Wed, 7 Jan 2015 07:31:38 +0000 (15:31 +0800)]
x86/apic: Only disable CPU x2apic mode when necessary
When interrupt remapping hardware is not in X2APIC, CPU X2APIC mode
will be disabled if:
1) Maximum CPU APIC ID is bigger than 255
2) hypervisior doesn't support x2apic mode.
But we should only check whether hypervisor supports X2APIC mode when
hypervisor(CONFIG_HYPERVISOR_GUEST) is enabled, otherwise X2APIC will
always be disabled when CONFIG_HYPERVISOR_GUEST is disabled and IR
doesn't work in X2APIC mode.
Signed-off-by: Jiang Liu <jiang.liu@linux.intel.com> Tested-by: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Cc: iommu@lists.linux-foundation.org Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Yinghai Lu <yinghai@kernel.org> Cc: Borislav Petkov <bp@alien8.de> Cc: David Rientjes <rientjes@google.com> Cc: HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com> Cc: Jan Beulich <JBeulich@suse.com> Cc: Richard Weinberger <richard@nod.at> Cc: Oren Twaig <oren@scalemp.com> Link: http://lkml.kernel.org/r/1420615903-28253-12-git-send-email-jiang.liu@linux.intel.com Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Jiang Liu [Wed, 7 Jan 2015 07:31:37 +0000 (15:31 +0800)]
iommu/vt-d: Allow IR works in XAPIC mode though CPU works in X2APIC mode
Currently if CPU supports X2APIC, IR hardware must work in X2APIC mode
or disabled. Change the code to support IR working in XAPIC mode when
CPU supports X2APIC. Then the CPU APIC driver will decide how to handle
such as configuration by:
1) Disabling X2APIC mode
2) Forcing X2APIC physical mode
This change also fixes a live locking when
1) BIOS enables CPU X2APIC
2) DMAR table disables X2APIC mode or IR hardware doesn't support X2APIC
with following messages:
[ 37.863463] dmar: INTR-REMAP: Request device [[f0:1f.7] fault index 2
[ 37.863463] INTR-REMAP:[fault reason 36] Detected reserved fields in the IRTE entry
[ 37.879372] dmar: INTR-REMAP: Request device [[f0:1f.7] fault index 2
[ 37.879372] INTR-REMAP:[fault reason 36] Detected reserved fields in the IRTE entry
[ 37.895282] dmar: INTR-REMAP: Request device [[f0:1f.7] fault index 2
[ 37.895282] INTR-REMAP:[fault reason 36] Detected reserved fields in the IRTE entry
[ 37.911192] dmar: INTR-REMAP: Request device [[f0:1f.7] fault index 2
Signed-off-by: Jiang Liu <jiang.liu@linux.intel.com> Tested-by: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Cc: iommu@lists.linux-foundation.org Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Yinghai Lu <yinghai@kernel.org> Cc: Borislav Petkov <bp@alien8.de> Link: http://lkml.kernel.org/r/1420615903-28253-11-git-send-email-jiang.liu@linux.intel.com Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Joerg Roedel [Wed, 7 Jan 2015 07:31:36 +0000 (15:31 +0800)]
iommu/vt-d: Allocate IRQ remapping data structures only for all IOMMUs
IRQ remapping is only supported when all IOMMUs in the
system support it. So check if all IOMMUs in the system
support IRQ remapping before doing the allocations.
[Jiang]
1) Rebased to v3.19.
2) Remove redundant check of ecap_ir_support(iommu->ecap) in function
intel_enable_irq_remapping().
Signed-off-by: Joerg Roedel <jroedel@suse.de> Signed-off-by: Jiang Liu <jiang.liu@linux.intel.com> Tested-by: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Cc: iommu@lists.linux-foundation.org Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Yinghai Lu <yinghai@kernel.org> Cc: Borislav Petkov <bp@alien8.de> Link: http://lkml.kernel.org/r/1420615903-28253-10-git-send-email-jiang.liu@linux.intel.com Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Jiang Liu [Wed, 7 Jan 2015 07:31:34 +0000 (15:31 +0800)]
x86/apic: Handle XAPIC remap mode proper.
If remapping is in XAPIC mode, the setup code just skips X2APIC
initialization without checking max CPU APIC ID in system, which may
cause problem if system has a CPU with APIC ID bigger than 255.
Handle IR in XAPIC mode the same way as if remapping is disabled.
[ tglx: Split out from previous patch ]
Signed-off-by: Jiang Liu <jiang.liu@linux.intel.com> Cc: Tony Luck <tony.luck@intel.com> Cc: iommu@lists.linux-foundation.org Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Yinghai Lu <yinghai@kernel.org> Cc: Borislav Petkov <bp@alien8.de> Cc: David Rientjes <rientjes@google.com> Cc: HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com> Cc: Jan Beulich <JBeulich@suse.com> Cc: Richard Weinberger <richard@nod.at> Cc: Oren Twaig <oren@scalemp.com> Link: http://lkml.kernel.org/r/1420615903-28253-8-git-send-email-jiang.liu@linux.intel.com Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Jiang Liu [Wed, 7 Jan 2015 07:31:33 +0000 (15:31 +0800)]
x86/apic: Correctly detect X2APIC status in function enable_IR()
X2APIC will be disabled if user specifies "nox2apic" on kernel command
line, even when x2apic_preenabled is true. So correctly detect X2APIC
status by using x2apic_enabled() instead of x2apic_preenabled.
Signed-off-by: Jiang Liu <jiang.liu@linux.intel.com> Cc: Tony Luck <tony.luck@intel.com> Cc: iommu@lists.linux-foundation.org Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Yinghai Lu <yinghai@kernel.org> Cc: Borislav Petkov <bp@alien8.de> Cc: David Rientjes <rientjes@google.com> Cc: HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com> Cc: Jan Beulich <JBeulich@suse.com> Cc: Richard Weinberger <richard@nod.at> Cc: Oren Twaig <oren@scalemp.com> Link: http://lkml.kernel.org/r/1420615903-28253-7-git-send-email-jiang.liu@linux.intel.com Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Jiang Liu [Wed, 7 Jan 2015 07:31:31 +0000 (15:31 +0800)]
x86/apic: Panic if kernel doesn't support x2apic but BIOS has enabled x2apic
When kernel doesn't support X2APIC but BIOS has enabled X2APIC, system
may panic or hang without useful messages. On the other hand, it's
hard to dynamically disable X2APIC when CONFIG_X86_X2APIC is disabled.
So panic with a clear message in such a case.
if (x2apic_disabled) {
__disable_x2apic();
return;
}
That leaves X2APIC disabled in hardware, but x2apic_mode stays 1. So
all other code which checks x2apic_mode gets the wrong information.
Set x2apic_mode to 0 after disabling it in hardware.
This is just a hotfix. The proper solution is to rework this code so
it has seperate functions for the initial setup on the boot processor
and the secondary cpus, but that's beyond the scope of this fix.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Jiang Liu <jiang.liu@linux.intel.com> Cc: Tony Luck <tony.luck@intel.com> Cc: iommu@lists.linux-foundation.org Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Yinghai Lu <yinghai@kernel.org> Cc: Borislav Petkov <bp@alien8.de> Cc: David Rientjes <rientjes@google.com> Cc: HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com> Cc: Jan Beulich <JBeulich@suse.com> Cc: Richard Weinberger <richard@nod.at> Cc: Oren Twaig <oren@scalemp.com>
Thomas Gleixner [Wed, 7 Jan 2015 07:31:29 +0000 (15:31 +0800)]
iommu/vt-d: Move iommu preparatory allocations to irq_remap_ops.prepare
The whole iommu setup for irq remapping is a convoluted mess. The
iommu detect function gets called from mem_init() and the prepare
callback gets called from enable_IR_x2apic() for unknown reasons.
Of course AMD and Intel setup differs in nonsensical ways. Intels
prepare callback is explicit while AMDs prepare callback is implicit
in setup_irq_remapping_ops() just to be called in the prepare call
again.
Because all of this gets called from enable_IR_x2apic() and the dmar
prepare function merily parses the ACPI tables, but does not allocate
memory we end up with memory allocation from irq disabled context
later on.
AMDs iommu code at least allocates the required memory from the
prepare function. That has issues as well, but thats not scope of this
patch.
The goal of this change is to distangle the allocation from the actual
enablement. There is no point to allocate memory from irq disabled
regions with GFP_ATOMIC just because it does not matter at that point
in the boot stage. It matters with physical hotplug later on.
There is another issue with the current setup. Due to the conversion
to stacked irqdomains we end up with a call into the irqdomain
allocation code from irq disabled context, but that code does
GFP_KERNEL allocations rightfully as there is no reason to do
preperatory allocations with GFP_ATOMIC.
That change caused the allocator code to complain about GFP_KERNEL
allocations invoked in atomic context. Boris provided a temporary
hackaround which changed the GFP flags if irq_domain_add() got called
from atomic context. Not pretty and we really dont want to get this
into a mainline release for obvious reasons.
Move the ACPI table parsing and the resulting memory allocations from
the enable to the prepare function. That allows to get rid of the
horrible hackaround in irq_domain_add() later.
[Jiang] Rebased onto v3.19
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Borislav Petkov <bp@alien8.de> Acked-and-tested-by: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Cc: iommu@lists.linux-foundation.org Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Yinghai Lu <yinghai@kernel.org> Cc: x86@kernel.org Link: http://lkml.kernel.org/r/20141205084147.313026156@linutronix.de Link: http://lkml.kernel.org/r/1420615903-28253-3-git-send-email-jiang.liu@linux.intel.com Signed-off-by: Jiang Liu <jiang.liu@linux.intel.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Thomas Gleixner [Wed, 7 Jan 2015 07:31:28 +0000 (15:31 +0800)]
iommu, x86: Restructure setup of the irq remapping feature
enable_IR_x2apic() calls setup_irq_remapping_ops() which by default
installs the intel dmar remapping ops and then calls the amd iommu irq
remapping prepare callback to figure out whether we are running on an
AMD machine with irq remapping hardware.
Right after that it calls irq_remapping_prepare() which pointlessly
checks:
if (!remap_ops || !remap_ops->prepare)
return -ENODEV;
and then calls
remap_ops->prepare()
which is silly in the AMD case as it got called from
setup_irq_remapping_ops() already a few microseconds ago.
Simplify this and just collapse everything into
irq_remapping_prepare().
The irq_remapping_prepare() remains still silly as it assigns blindly
the intel ops, but that's not scope of this patch.
The scope here is to move the preperatory work, i.e. memory
allocations out of the atomic section which is required to enable irq
remapping.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Borislav Petkov <bp@alien8.de> Acked-and-tested-by: Joerg Roedel <joro@8bytes.org> Cc: Tony Luck <tony.luck@intel.com> Cc: iommu@lists.linux-foundation.org Cc: Joerg Roedel <jroedel@suse.de> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Yinghai Lu <yinghai@kernel.org> Cc: David Rientjes <rientjes@google.com> Cc: HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com> Cc: Jan Beulich <JBeulich@suse.com> Cc: Richard Weinberger <richard@nod.at> Cc: Oren Twaig <oren@scalemp.com> Cc: x86@kernel.org Link: http://lkml.kernel.org/r/20141205084147.232633738@linutronix.de Link: http://lkml.kernel.org/r/1420615903-28253-2-git-send-email-jiang.liu@linux.intel.com Signed-off-by: Jiang Liu <jiang.liu@linux.intel.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Linus Torvalds [Sun, 11 Jan 2015 20:44:10 +0000 (12:44 -0800)]
Merge branch 'fixes' of git://ftp.arm.linux.org.uk/~rmk/linux-arm
Pull ARM fixes from Russell King:
"Three small fixes from over the Christmas period, and wiring up the
new execveat syscall for ARM"
* 'fixes' of git://ftp.arm.linux.org.uk/~rmk/linux-arm:
ARM: 8275/1: mm: fix PMD_SECT_RDONLY undeclared compile error
ARM: 8253/1: mm: use phys_addr_t type in map_lowmem() for kernel mem region
ARM: 8249/1: mm: dump: don't skip regions
ARM: wire up execveat syscall
Linus Torvalds [Sun, 11 Jan 2015 19:51:49 +0000 (11:51 -0800)]
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 fixes: group scheduling corner case fix, two deadline scheduler
fixes, effective_load() overflow fix, nested sleep fix, 6144 CPUs
system fix"
* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/fair: Fix RCU stall upon -ENOMEM in sched_create_group()
sched/deadline: Avoid double-accounting in case of missed deadlines
sched/deadline: Fix migration of SCHED_DEADLINE tasks
sched: Fix odd values in effective_load() calculations
sched, fanotify: Deal with nested sleeps
sched: Fix KMALLOC_MAX_SIZE overflow during cpumask allocation
Linus Torvalds [Sun, 11 Jan 2015 19:47:45 +0000 (11:47 -0800)]
Merge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull perf fixes from Ingo Molnar:
"Mostly tooling fixes, but also some kernel side fixes: uncore PMU
driver fix, user regs sampling fix and an instruction decoder fix that
unbreaks PEBS precise sampling"
* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/x86/uncore/hsw-ep: Handle systems with only two SBOXes
perf/x86_64: Improve user regs sampling
perf: Move task_pt_regs sampling into arch code
x86: Fix off-by-one in instruction decoder
perf hists browser: Fix segfault when showing callchain
perf callchain: Free callchains when hist entries are deleted
perf hists: Fix children sort key behavior
perf diff: Fix to sort by baseline field by default
perf list: Fix --raw-dump option
perf probe: Fix crash in dwarf_getcfi_elf
perf probe: Fix to fall back to find probe point in symbols
perf callchain: Append callchains only when requested
perf ui/tui: Print backtrace symbols when segfault occurs
perf report: Show progress bar for output resorting
Linus Torvalds [Sun, 11 Jan 2015 19:46:31 +0000 (11:46 -0800)]
Merge branch 'locking-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull locking fixes from Ingo Molnar:
"A liblockdep fix and a mutex_unlock() mutex-debugging fix"
* 'locking-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
mutex: Always clear owner field upon mutex_unlock()
tools/liblockdep: Fix debug_check thinko in mutex destroy
mm: fix corner case in anon_vma endless growing prevention
Fix for BUG_ON(anon_vma->degree) splashes in unlink_anon_vmas() ("kernel
BUG at mm/rmap.c:399!") caused by commit 7a3ef208e662 ("mm: prevent
endless growth of anon_vma hierarchy")
Anon_vma_clone() is usually called for a copy of source vma in
destination argument. If source vma has anon_vma it should be already
in dst->anon_vma. NULL in dst->anon_vma is used as a sign that it's
called from anon_vma_fork(). In this case anon_vma_clone() finds
anon_vma for reusing.
Vma_adjust() calls it differently and this breaks anon_vma reusing
logic: anon_vma_clone() links vma to old anon_vma and updates degree
counters but vma_adjust() overrides vma->anon_vma right after that. As
a result final unlink_anon_vmas() decrements degree for wrong anon_vma.
This patch assigns ->anon_vma before calling anon_vma_clone().
Signed-off-by: Konstantin Khlebnikov <koct9i@gmail.com> Reported-and-tested-by: Chris Clayton <chris2553@googlemail.com> Reported-and-tested-by: Oded Gabbay <oded.gabbay@amd.com> Reported-and-tested-by: Chih-Wei Huang <cwhuang@android-x86.org> Acked-by: Rik van Riel <riel@redhat.com> Acked-by: Vlastimil Babka <vbabka@suse.cz> Cc: Daniel Forrest <dan.forrest@ssec.wisc.edu> Cc: Michal Hocko <mhocko@suse.cz> Cc: stable@vger.kernel.org # to match back-porting of 7a3ef208e662 Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds [Sun, 11 Jan 2015 19:33:57 +0000 (11:33 -0800)]
mm: Don't count the stack guard page towards RLIMIT_STACK
Commit fee7e49d4514 ("mm: propagate error from stack expansion even for
guard page") made sure that we return the error properly for stack
growth conditions. It also theorized that counting the guard page
towards the stack limit might break something, but also said "Let's see
if anybody notices".
Somebody did notice. Apparently android-x86 sets the stack limit very
close to the limit indeed, and including the guard page in the rlimit
check causes the android 'zygote' process problems.
So this adds the (fairly trivial) code to make the stack rlimit check be
against the actual real stack size, rather than the size of the vma that
includes the guard page.
Reported-and-tested-by: Chih-Wei Huang <cwhuang@android-x86.org> Cc: Jay Foad <jay.foad@gmail.com> Cc: stable@kernel.org # to match back-porting of fee7e49d4514 Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds [Sat, 10 Jan 2015 05:23:27 +0000 (21:23 -0800)]
Merge tag 'sound-3.19-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Pull sound fixes from Takashi Iwai:
"All a few small regression or stable fixes: a Nvidia HDMI ID addition,
a regression fix for CAIAQ stream count, a typo fix for GPIO setup
with STAC/IDT HD-audio codecs, and a Fireworks big-endian fix"
* tag 'sound-3.19-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
ALSA: fireworks: fix an endianness bug for transaction length
ALSA: hda - Add new GPU codec ID 0x10de0072 to snd-hda
ALSA: hda - Fix wrong gpio_dir & gpio_mask hint setups for IDT/STAC codecs
ALSA: snd-usb-caiaq: fix stream count check
Linus Torvalds [Sat, 10 Jan 2015 05:18:48 +0000 (21:18 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid
Pull HID updates from Jiri Kosina:
- bounds checking fixes in logitech and roccat drivers, from Peter Wu
and Dan Carpenter
- double-kfree fix in i2c-hid driver on bus shutdown, from Mika
Westerberg
- a couple of various small driver fixes
- a few device id additions
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid:
HID: roccat: potential out of bounds in pyra_sysfs_write_settings()
HID: Add a new id 0x501a for Genius MousePen i608X
HID: logitech-hidpp: prefix the name with "Logitech"
HID: logitech-hidpp: avoid unintended fall-through
HID: Allow HID_BATTERY_STRENGTH to be enabled
HID: i2c-hid: Do not free buffers in i2c_hid_stop()
HID: add battery quirk for USB_DEVICE_ID_APPLE_ALU_WIRELESS_2011_ISO keyboard
HID: logitech-hidpp: check WTP report length
HID: logitech-dj: check report length
Linus Torvalds [Sat, 10 Jan 2015 05:13:34 +0000 (21:13 -0800)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Pull drm fixes from Dave Airlie:
"I'm briefly working between holidays and LCA, so this is close to a
couple of weeks of fixes,
Two sets of amdkfd fixes, this is a new feature this kernel, and this
pull fixes a few issues since it got merged, ordering when built-in to
kernel and also the iommu vs gpu ordering patch, it also reworks the
ioctl before the initial release.
Otherwise:
- radeon: some misc fixes all over, hdmi, 4k, dpm
- nouveau: mcp77 init fixes, oops fix, bug on fix, msi fix
- i915: power fixes, revert VGACNTR patch
Probably be quiteer next week since I'll be at LCA anyways"
* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux: (33 commits)
drm/amdkfd: rewrite kfd_ioctl() according to drm_ioctl()
drm/amdkfd: reformat IOCTL definitions to drm-style
drm/amdkfd: Do copy_to/from_user in general kfd_ioctl()
drm/radeon: integer underflow in radeon_cp_dispatch_texture()
drm/radeon: adjust default bapm settings for KV
drm/radeon: properly filter DP1.2 4k modes on non-DP1.2 hw
drm/radeon: fix sad_count check for dce3
drm/radeon: KV has three PPLLs (v2)
drm/amdkfd: unmap VMID<-->PASID when relesing VMID (non-HWS)
drm/radeon: Init amdkfd only if it was compiled
amdkfd: actually allocate longs for the pasid bitmask
drm/nouveau/nouveau: Do not BUG_ON(!spin_is_locked()) on UP
drm/nv4c/mc: disable msi
drm/nouveau/fb/ram/mcp77: enable NISO poller
drm/nouveau/fb/ram/mcp77: use carveout reg to determine size
drm/nouveau/fb/ram/mcp77: subclass nouveau_ram
drm/nouveau: wake up the card if necessary during gem callbacks
drm/nouveau/device: Add support for GK208B, resolves bug 86935
drm/nouveau: fix missing return statement in nouveau_ttm_tt_unpopulate
drm/nouveau/bios: fix oops on pre-nv50 chipsets
...
Linus Torvalds [Sat, 10 Jan 2015 04:57:58 +0000 (20:57 -0800)]
Merge tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux
Pull arm64 fixes from Will Deacon:
"Here is a handful of minor arm64 fixes discovered and fixed over the
Christmas break. The main part is adding some missing #includes that
we seem to be getting transitively but have started causing problems
in -next.
- Fix early mapping fixmap corruption by EFI runtime services
- Fix __NR_compat_syscalls off-by-one
- Add missing sanity checks for some 32-bit registers
- Add some missing #includes which we get transitively
- Remove unused prepare_to_copy() macro"
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
arm64/efi: add missing call to early_ioremap_reset()
arm64: fix missing asm/io.h include in kernel/smp_spin_table.c
arm64: fix missing asm/alternative.h include in kernel/module.c
arm64: fix missing linux/bug.h include in asm/arch_timer.h
arm64: fix missing asm/pgtable-hwdef.h include in asm/processor.h
arm64: sanity checks: add missing AArch32 registers
arm64: Remove unused prepare_to_copy()
arm64: Correct __NR_compat_syscalls for bpf
Fixes
- kgdb/kdb: Allow access on a single core, if a CPU round up is
deemed impossible, which will allow inspection of the now "trashed"
kernel
- kdb: Add enable mask for the command groups
- kdb: access controls to restrict sensitive commands"
* tag 'for_linus-3.19-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/jwessel/kgdb:
kernel/debug/debug_core.c: Logging clean-up
kgdb: timeout if secondary CPUs ignore the roundup
kdb: Allow access to sensitive commands to be restricted by default
kdb: Add enable mask for groups of commands
kdb: Categorize kdb commands (similar to SysRq categorization)
kdb: Remove KDB_REPEAT_NONE flag
kdb: Use KDB_REPEAT_* values as flags
kdb: Rename kdb_register_repeat() to kdb_register_flags()
kdb: Rename kdb_repeat_t to kdb_cmdflags_t, cmd_repeat to cmd_flags
kdb: Remove currently unused kdbtab_t->cmd_flags
Linus Torvalds [Sat, 10 Jan 2015 01:55:00 +0000 (17:55 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client
Pull two Ceph fixes from Sage Weil:
"These are both pretty trivial: a sparse warning fix and size_t printk
thing"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client:
libceph: fix sparse endianness warnings
ceph: use %zu for len in ceph_fill_inline_data()
Linus Torvalds [Sat, 10 Jan 2015 01:46:07 +0000 (17:46 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs
Pull btrfs fixes from Chris Mason:
"None of these are huge, but my commit does fix a regression from 3.18
that could cause lost files during log replay.
This also adds Dave Sterba to the list of Btrfs maintainers. It
doesn't mean we're doing things differently, but Dave has really been
helping with the maintainer workload for years"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
Btrfs: don't delay inode ref updates during log replay
Btrfs: correctly get tree level in tree_backref_for_extent
Btrfs: call inode_dec_link_count() on mkdir error path
Btrfs: abort transaction if we don't find the block group
Btrfs, scrub: uninitialized variable in scrub_extent_for_parity()
Btrfs: add more maintainers
Linus Torvalds [Fri, 9 Jan 2015 23:10:59 +0000 (15:10 -0800)]
Merge branch 'akpm' (patches from Andrew)
Merge misc fixes from Andrew Morton:
"12 fixes"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
mm, vmscan: prevent kswapd livelock due to pfmemalloc-throttled process being killed
memcg: fix destination cgroup leak on task charges migration
mm: memcontrol: switch soft limit default back to infinity
mm/debug_pagealloc: remove obsolete Kconfig options
vfs: renumber FMODE_NONOTIFY and add to uniqueness check
arch/blackfin/mach-bf533/boards/stamp.c: add linux/delay.h
ocfs2: fix the wrong directory passed to ocfs2_lookup_ino_from_name() when link file
MAINTAINERS: update rydberg's addresses
mm: protect set_page_dirty() from ongoing truncation
mm: prevent endless growth of anon_vma hierarchy
exit: fix race between wait_consider_task() and wait_task_zombie()
ocfs2: remove bogus check in dlm_process_recovery_data
In v3.19-rc3 tree when CONFIG_ARM_LPAE and CONFIG_DEBUG_RODATA are enabled
image failed to compile with the following error:
arch/arm/mm/init.c:661:14: error: ‘PMD_SECT_RDONLY’ undeclared here (not in a function)
It seems that '80d6b0c ARM: mm: allow text and rodata sections to be read-only'
and 'ded9477 ARM: 8109/1: mm: Modify pte_write and pmd_write logic for LPAE'
commits crossed. 80d6b0c uses PMD_SECT_RDONLY macro but ded9477 renames it
and uses software bits L_PMD_SECT_RDONLY instead.
Fix is to use L_PMD_SECT_RDONLY instead PMD_SECT_RDONLY as ded9477 does in
another places.
Signed-off-by: Victor Kamensky <victor.kamensky@linaro.org> Acked-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Dan Carpenter [Fri, 9 Jan 2015 12:32:31 +0000 (15:32 +0300)]
HID: roccat: potential out of bounds in pyra_sysfs_write_settings()
This is a static checker fix. We write some binary settings to the
sysfs file. One of the settings is the "->startup_profile". There
isn't any checking to make sure it fits into the
pyra->profile_settings[] array in the profile_activated() function.
I added a check to pyra_sysfs_write_settings() in both places because
I wasn't positive that the other callers were correct.
Cc: <stable@vger.kernel.org> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Chris Wilson [Tue, 6 Jan 2015 10:29:35 +0000 (10:29 +0000)]
mutex: Always clear owner field upon mutex_unlock()
Currently if DEBUG_MUTEXES is enabled, the mutex->owner field is only
cleared iff debug_locks is active. This exposes a race to other users of
the field where the mutex->owner may be still set to a stale value,
potentially upsetting mutex_spin_on_owner() among others.
References: https://bugs.freedesktop.org/show_bug.cgi?id=87955 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Davidlohr Bueso <dave@stgolabs.net> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Linus Torvalds <torvalds@linux-foundation.org> Link: http://lkml.kernel.org/r/1420540175-30204-1-git-send-email-chris@chris-wilson.co.uk Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tetsuo Handa [Thu, 25 Dec 2014 06:51:21 +0000 (15:51 +0900)]
sched/fair: Fix RCU stall upon -ENOMEM in sched_create_group()
When alloc_fair_sched_group() in sched_create_group() fails,
free_sched_group() is called, and free_fair_sched_group() is called by
free_sched_group(). Since destroy_cfs_bandwidth() is called by
free_fair_sched_group() without calling init_cfs_bandwidth(),
RCU stall occurs at hrtimer_cancel():
Luca Abeni [Wed, 17 Dec 2014 10:50:32 +0000 (11:50 +0100)]
sched/deadline: Avoid double-accounting in case of missed deadlines
The dl_runtime_exceeded() function is supposed to ckeck if
a SCHED_DEADLINE task must be throttled, by checking if its
current runtime is <= 0. However, it also checks if the
scheduling deadline has been missed (the current time is
larger than the current scheduling deadline), further
decreasing the runtime if this happens.
This "double accounting" is wrong:
- In case of partitioned scheduling (or single CPU), this
happens if task_tick_dl() has been called later than expected
(due to small HZ values). In this case, the current runtime is
also negative, and replenish_dl_entity() can take care of the
deadline miss by recharging the current runtime to a value smaller
than dl_runtime
- In case of global scheduling on multiple CPUs, scheduling
deadlines can be missed even if the task did not consume more
runtime than expected, hence penalizing the task is wrong
This patch fix this problem by throttling a SCHED_DEADLINE task
only when its runtime becomes negative, and not modifying the runtime
Luca Abeni [Wed, 17 Dec 2014 10:50:31 +0000 (11:50 +0100)]
sched/deadline: Fix migration of SCHED_DEADLINE tasks
According to global EDF, tasks should be migrated between runqueues
without checking if their scheduling deadlines and runtimes are valid.
However, SCHED_DEADLINE currently performs such a check:
a migration happens doing:
which ends up calling dequeue_task_dl(), setting the new CPU, and then
calling enqueue_task_dl().
enqueue_task_dl() then calls enqueue_dl_entity(), which calls
update_dl_entity(), which can modify scheduling deadline and runtime,
breaking global EDF scheduling.
As a result, some of the properties of global EDF are not respected:
for example, a taskset {(30, 80), (40, 80), (120, 170)} scheduled on
two cores can have unbounded response times for the third task even
if 30/80+40/80+120/170 = 1.5809 < 2
This can be fixed by invoking update_dl_entity() only in case of
wakeup, or if this is a new SCHED_DEADLINE task.
Yuyang Du [Fri, 19 Dec 2014 00:29:56 +0000 (08:29 +0800)]
sched: Fix odd values in effective_load() calculations
In effective_load, we have (long w * unsigned long tg->shares) / long W,
when w is negative, it is cast to unsigned long and hence the product is
insanely large. Fix this by casting tg->shares to long.
Reported-by: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Yuyang Du <yuyang.du@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Dave Jones <davej@redhat.com> Cc: Andrey Ryabinin <a.ryabinin@samsung.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Link: http://lkml.kernel.org/r/20141219002956.GA25405@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
Peter Zijlstra [Tue, 16 Dec 2014 15:28:38 +0000 (16:28 +0100)]
sched, fanotify: Deal with nested sleeps
As per e23738a7300a ("sched, inotify: Deal with nested sleeps").
fanotify_read is a wait loop with sleeps in. Wait loops rely on
task_struct::state and sleeps do too, since that's the only means of
actually sleeping. Therefore the nested sleeps destroy the wait loop
state and the wait loop breaks the sleep functions that assume
TASK_RUNNING (mutex_lock).
Fix this by using the new woken_wake_function and wait_woken() stuff,
which registers wakeups in wait and thereby allows shrinking the
task_state::state changes to the actual sleep part.
Reported-by: Yuanhan Liu <yuanhan.liu@linux.intel.com> Reported-by: Sedat Dilek <sedat.dilek@gmail.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Takashi Iwai <tiwai@suse.de> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Eric Paris <eparis@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Eric Paris <eparis@redhat.com> Link: http://lkml.kernel.org/r/20141216152838.GZ3337@twins.programming.kicks-ass.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
Andy Lutomirski [Sun, 4 Jan 2015 18:36:20 +0000 (10:36 -0800)]
perf/x86_64: Improve user regs sampling
Perf reports user regs for kernel-mode samples so that samples can
be backtraced through user code. The old code was very broken in
syscall context, resulting in useless backtraces.
The new code, in contrast, is still dangerously racy, but it should
at least work most of the time.
Tested-by: Jiri Olsa <jolsa@kernel.org> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: chenggang.qcg@taobao.com Cc: Wu Fengguang <fengguang.wu@intel.com> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arjan van de Ven <arjan@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Link: http://lkml.kernel.org/r/243560c26ff0f739978e2459e203f6515367634d.1420396372.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
Andy Lutomirski [Sun, 4 Jan 2015 18:36:19 +0000 (10:36 -0800)]
perf: Move task_pt_regs sampling into arch code
On x86_64, at least, task_pt_regs may be only partially initialized
in many contexts, so x86_64 should not use it without extra care
from interrupt context, let alone NMI context.
This will allow x86_64 to override the logic and will supply some
scratch space to use to make a cleaner copy of user regs.
Tested-by: Jiri Olsa <jolsa@kernel.org> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Cc: chenggang.qcg@taobao.com Cc: Wu Fengguang <fengguang.wu@intel.com> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arjan van de Ven <arjan@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Jean Pihet <jean.pihet@linaro.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Mark Salter <msalter@redhat.com> Cc: Russell King <linux@arm.linux.org.uk> Cc: Will Deacon <will.deacon@arm.com> Cc: linux-arm-kernel@lists.infradead.org Link: http://lkml.kernel.org/r/e431cd4c18c2e1c44c774f10758527fb2d1025c4.1420396372.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
Peter Zijlstra [Tue, 16 Dec 2014 10:46:14 +0000 (11:46 +0100)]
x86: Fix off-by-one in instruction decoder
Stephane reported that the PEBS fixup was broken by the recent commit to
the instruction decoder. The thing had an off-by-one which resulted in
not being able to decode the last instruction and always bail.
Reported-by: Stephane Eranian <eranian@google.com> Fixes: 6ba48ff46f76 ("x86: Remove arbitrary instruction size limit in instruction decoder") Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: stable@vger.kernel.org # 3.18 Cc: <ak@linux.intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Liang Kan <kan.liang@intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: Jim Keniston <jkenisto@us.ibm.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Link: http://lkml.kernel.org/r/20141216104614.GV3337@twins.programming.kicks-ass.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
Ingo Molnar [Fri, 9 Jan 2015 10:04:13 +0000 (11:04 +0100)]
Merge tag 'perf-urgent-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/urgent
Pull perf/urgent fixes from Arnaldo Carvalho de Melo:
- Free callchains when hist entries are deleted, plugging a massive leak in
'top -g', where hist_entries (and its callchains) are decayed over time. (Namhyung Kim)
- Fix segfault when showing callchain in the hists browser (report & top) (Namhyung Kim)
- Fix children sort key behavior, and also the 'perf test 32' test that
was failing due to reliance on undefined behaviour (Namhyung Kim)
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org>
Vlastimil Babka [Thu, 8 Jan 2015 22:32:40 +0000 (14:32 -0800)]
mm, vmscan: prevent kswapd livelock due to pfmemalloc-throttled process being killed
Charles Shirron and Paul Cassella from Cray Inc have reported kswapd
stuck in a busy loop with nothing left to balance, but
kswapd_try_to_sleep() failing to sleep. Their analysis found the cause
to be a combination of several factors:
1. A process is waiting in throttle_direct_reclaim() on pgdat->pfmemalloc_wait
2. The process has been killed (by OOM in this case), but has not yet been
scheduled to remove itself from the waitqueue and die.
3. kswapd checks for throttled processes in prepare_kswapd_sleep():
if (waitqueue_active(&pgdat->pfmemalloc_wait)) {
wake_up(&pgdat->pfmemalloc_wait);
return false; // kswapd will not go to sleep
}
However, for a process that was already killed, wake_up() does not remove
the process from the waitqueue, since try_to_wake_up() checks its state
first and returns false when the process is no longer waiting.
4. kswapd is running on the same CPU as the only CPU that the process is
allowed to run on (through cpus_allowed, or possibly single-cpu system).
5. CONFIG_PREEMPT_NONE=y kernel is used. If there's nothing to balance, kswapd
encounters no voluntary preemption points and repeatedly fails
prepare_kswapd_sleep(), blocking the process from running and removing
itself from the waitqueue, which would let kswapd sleep.
So, the source of the problem is that we prevent kswapd from going to
sleep until there are processes waiting on the pfmemalloc_wait queue,
and a process waiting on a queue is guaranteed to be removed from the
queue only when it gets scheduled. This was done to make sure that no
process is left sleeping on pfmemalloc_wait when kswapd itself goes to
sleep.
However, it isn't necessary to postpone kswapd sleep until the
pfmemalloc_wait queue actually empties. To prevent processes from being
left sleeping, it's actually enough to guarantee that all processes
waiting on pfmemalloc_wait queue have been woken up by the time we put
kswapd to sleep.
This patch therefore fixes this issue by substituting 'wake_up' with
'wake_up_all' and removing 'return false' in the code snippet from
prepare_kswapd_sleep() above. Note that if any process puts itself in
the queue after this waitqueue_active() check, or after the wake up
itself, it means that the process will also wake up kswapd - and since
we are under prepare_to_wait(), the wake up won't be missed. Also we
update the comment prepare_kswapd_sleep() to hopefully more clearly
describe the races it is preventing.
Fixes: 5515061d22f0 ("mm: throttle direct reclaimers if PF_MEMALLOC reserves are low and swap is backed by network storage") Signed-off-by: Vlastimil Babka <vbabka@suse.cz> Signed-off-by: Vladimir Davydov <vdavydov@parallels.com> Cc: Mel Gorman <mgorman@suse.de> Cc: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Acked-by: Rik van Riel <riel@redhat.com> Cc: <stable@vger.kernel.org> [3.6+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
memcg: fix destination cgroup leak on task charges migration
We are supposed to take one css reference per each memory page and per
each swap entry accounted to a memory cgroup. However, during task
charges migration we take a reference to the destination cgroup twice
per each swap entry: first in mem_cgroup_do_precharge()->try_charge()
and then in mem_cgroup_move_swap_account(), permanently leaking the
destination cgroup.
The hunk taking the second reference seems to be a leftover from the
pre-00501b531c472 ("mm: memcontrol: rewrite charge API") era. Remove it
to fix the leak.
Fixes: e8ea14cc6ead (mm: memcontrol: take a css reference for each charged page) Signed-off-by: Vladimir Davydov <vdavydov@parallels.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Johannes Weiner [Thu, 8 Jan 2015 22:32:35 +0000 (14:32 -0800)]
mm: memcontrol: switch soft limit default back to infinity
Commit 3e32cb2e0a12 ("mm: memcontrol: lockless page counters")
accidentally switched the soft limit default from infinity to zero,
which turns all memcgs with even a single page into soft limit excessors
and engages soft limit reclaim on all of them during global memory
pressure. This makes global reclaim generally more aggressive, but also
inverts the meaning of existing soft limit configurations where unset
soft limits are usually more generous than set ones.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Acked-by: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
David Drysdale [Thu, 8 Jan 2015 22:32:29 +0000 (14:32 -0800)]
vfs: renumber FMODE_NONOTIFY and add to uniqueness check
Fix clashing values for O_PATH and FMODE_NONOTIFY on sparc. The
clashing O_PATH value was added in commit 5229645bdc35 ("vfs: add
nonconflicting values for O_PATH") but this can't be changed as it is
user-visible.
FMODE_NONOTIFY is only used internally in the kernel, but it is in the
same numbering space as the other O_* flags, as indicated by the comment
at the top of include/uapi/asm-generic/fcntl.h (and its use in
fs/notify/fanotify/fanotify_user.c). So renumber it to avoid the clash.
All of this has happened before (commit 12ed2e36c98a: "fanotify:
FMODE_NONOTIFY and __O_SYNC in sparc conflict"), and all of this will
happen again -- so update the uniqueness check in fcntl_init() to
include __FMODE_NONOTIFY.
Signed-off-by: David Drysdale <drysdale@google.com> Acked-by: David S. Miller <davem@davemloft.net> Acked-by: Jan Kara <jack@suse.cz> Cc: Heinrich Schuchardt <xypron.glpk@gmx.de> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Stephen Rothwell <sfr@canb.auug.org.au> Cc: Eric Paris <eparis@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Xue jiufei [Thu, 8 Jan 2015 22:32:23 +0000 (14:32 -0800)]
ocfs2: fix the wrong directory passed to ocfs2_lookup_ino_from_name() when link file
In ocfs2_link(), the parent directory inode passed to function
ocfs2_lookup_ino_from_name() is wrong. Parameter dir is the parent of
new_dentry not old_dentry. We should get old_dir from old_dentry and
lookup old_dentry in old_dir in case another node remove the old dentry.
With this change, hard linking works again, when paths are relative with
at least one subdirectory. This is how the problem was reproducable:
# mkdir a
# mkdir b
# touch a/test
# ln a/test b/test
ln: failed to create hard link `b/test' => `a/test': No such file or directory
However when creating links in the same dir, it worked well.
Now the link gets created.
Fixes: 0e048316ff57 ("ocfs2: check existence of old dentry in ocfs2_link()") Signed-off-by: joyce.xue <xuejiufei@huawei.com> Reported-by: Szabo Aron - UBIT <aron@ubit.hu> Cc: Mark Fasheh <mfasheh@suse.com> Cc: Joel Becker <jlbec@evilplan.org> Tested-by: Aron Szabo <aron@ubit.hu> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Henrik Rydberg [Thu, 8 Jan 2015 22:32:21 +0000 (14:32 -0800)]
MAINTAINERS: update rydberg's addresses
My ISP finally gave up on the old mail address, so I am moving things
over to bitmath.org instead. Also change the status fields to better
reflect reality.
Signed-off-by: Henrik Rydberg <rydberg@bitmath.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Johannes Weiner [Thu, 8 Jan 2015 22:32:18 +0000 (14:32 -0800)]
mm: protect set_page_dirty() from ongoing truncation
Tejun, while reviewing the code, spotted the following race condition
between the dirtying and truncation of a page:
__set_page_dirty_nobuffers() __delete_from_page_cache()
if (TestSetPageDirty(page))
page->mapping = NULL
if (PageDirty())
dec_zone_page_state(page, NR_FILE_DIRTY);
dec_bdi_stat(mapping->backing_dev_info, BDI_RECLAIMABLE);
if (page->mapping)
account_page_dirtied(page)
__inc_zone_page_state(page, NR_FILE_DIRTY);
__inc_bdi_stat(mapping->backing_dev_info, BDI_RECLAIMABLE);
which results in an imbalance of NR_FILE_DIRTY and BDI_RECLAIMABLE.
Dirtiers usually lock out truncation, either by holding the page lock
directly, or in case of zap_pte_range(), by pinning the mapcount with
the page table lock held. The notable exception to this rule, though,
is do_wp_page(), for which this race exists. However, do_wp_page()
already waits for a locked page to unlock before setting the dirty bit,
in order to prevent a race where clear_page_dirty() misses the page bit
in the presence of dirty ptes. Upgrade that wait to a fully locked
set_page_dirty() to also cover the situation explained above.
Afterwards, the code in set_page_dirty() dealing with a truncation race
is no longer needed. Remove it.
Reported-by: Tejun Heo <tj@kernel.org> Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Reviewed-by: Jan Kara <jack@suse.cz> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Constantly forking task causes unlimited grow of anon_vma chain. Each
next child allocates new level of anon_vmas and links vma to all
previous levels because pages might be inherited from any level.
This patch adds heuristic which decides to reuse existing anon_vma
instead of forking new one. It adds counter anon_vma->degree which
counts linked vmas and directly descending anon_vmas and reuses anon_vma
if counter is lower than two. As a result each anon_vma has either vma
or at least two descending anon_vmas. In such trees half of nodes are
leafs with alive vmas, thus count of anon_vmas is no more than two times
bigger than count of vmas.
This heuristic reuses anon_vmas as few as possible because each reuse
adds false aliasing among vmas and rmap walker ought to scan more ptes
when it searches where page is might be mapped.
Link: http://lkml.kernel.org/r/20120816024610.GA5350@evergreen.ssec.wisc.edu Fixes: 5beb49305251 ("mm: change anon_vma linking to fix multi-process server scalability issue")
[akpm@linux-foundation.org: fix typo, per Rik] Signed-off-by: Konstantin Khlebnikov <koct9i@gmail.com> Reported-by: Daniel Forrest <dan.forrest@ssec.wisc.edu> Tested-by: Michal Hocko <mhocko@suse.cz> Tested-by: Jerome Marchand <jmarchan@redhat.com> Reviewed-by: Michal Hocko <mhocko@suse.cz> Reviewed-by: Rik van Riel <riel@redhat.com> Cc: <stable@vger.kernel.org> [2.6.34+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Thu, 8 Jan 2015 22:32:12 +0000 (14:32 -0800)]
exit: fix race between wait_consider_task() and wait_task_zombie()
wait_consider_task() checks EXIT_ZOMBIE after EXIT_DEAD/EXIT_TRACE and
both checks can fail if we race with EXIT_ZOMBIE -> EXIT_DEAD/EXIT_TRACE
change in between, gcc needs to reload p->exit_state after
security_task_wait(). In this case ->notask_error will be wrongly
cleared and do_wait() can hang forever if it was the last eligible
child.
Many thanks to Arne who carefully investigated the problem.
Note: this bug is very old but it was pure theoretical until commit b3ab03160dfa ("wait: completely ignore the EXIT_DEAD tasks"). Before
this commit "-O2" was probably enough to guarantee that compiler won't
read ->exit_state twice.
Signed-off-by: Oleg Nesterov <oleg@redhat.com> Reported-by: Arne Goedeke <el@laramies.com> Tested-by: Arne Goedeke <el@laramies.com> Cc: <stable@vger.kernel.org> [3.15+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Joseph Qi [Thu, 8 Jan 2015 22:32:09 +0000 (14:32 -0800)]
ocfs2: remove bogus check in dlm_process_recovery_data
In dlm_process_recovery_data, only when dlm_new_lock failed the ret will
be set to -ENOMEM. And in this case, newlock is definitely NULL. So
test newlock is meaningless, remove it.
Signed-off-by: Joseph Qi <joseph.qi@huawei.com> Reviewed-by: Alex Chen <alex.chen@huawei.com> Reviewed-by: Mark Fasheh <mfasheh@suse.de> Cc: Joel Becker <jlbec@evilplan.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Michal Marek [Thu, 8 Jan 2015 15:16:39 +0000 (16:16 +0100)]
Makefile: include arch/*/include/generated/uapi before .../generated
The introduction of the uapi directories in v3.7-rc1 moved some of the
generated headers from arch/*/include/generated to the uapi directory,
keeping the #include directives intact.
This creates a problem when bisecting, because the unversioned files are
not cleaned automatically by git and the compiler might include stale
headers as a result. Instead of cleaning them in the Makefiles, promote
arch/*/include/generated/uapi in the search path. Under normal
circumstances, there is no overlap between this uapi subdirectory and
its parent, so the include choices remain the same. We keep
arch/*/include/generated/uapi in the USERINCLUDE variable so that it is
usable standalone.
Note that we cannot completely swap the order of the uapi and
kernel-only directories, since the headers in include/uapi/asm-generic
are meant to be wrapped by their include/asm-generic counterparts when
building kernel code.
Reported-by: "Nicholas A. Bellinger" <nab@linux-iscsi.org> Reported-by: David Drysdale <dmd@lurklurk.org> Signed-off-by: Michal Marek <mmarek@suse.cz> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds [Thu, 8 Jan 2015 22:19:44 +0000 (14:19 -0800)]
Merge tag 'pinctrl-v3.19-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl
Pull pinctrl fixes from Linus Walleij:
"Allright allright I've been lazy over christmas and New Years. Here
are a few collected pin control fixes eventually. Details:
A set of assorted pin control fixes for the Rockchip and STi drivers"
* tag 'pinctrl-v3.19-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:
pinctrl: st: Add irq_disable hook to st_gpio_irqchip
pinctrl: st: avoid multiple mutex lock
pinctrl: rockchip: Fix enable/disable/mask/unmask
pinctrl: rockchip: Handle wakeup pins
Linus Torvalds [Thu, 8 Jan 2015 22:11:03 +0000 (14:11 -0800)]
Merge tag 'pm+acpi-3.19-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management and ACPI fixes from Rafael Wysocki:
"These are an ACPI device power management initialization fix (-stable
material), two commits renaming stuff in the ACPI processor driver to
make it more suitable for ARM64 processors and a new ACPI backlight
blacklist entry.
Specifics:
- Fix ACPI power management intialization for device objects
corresponding to devices that are not present at the init time (the
_STA control method returns 0 for them) and therefore should not be
regarded as power manageable (Rafael J Wysocki).
- Rename a structure field and two functions used by the ACPI
processor driver to make them less tied to architectures that use
APICs (both x86 and ia64) and more suitable for ARM64 processors
(Hanjun Guo).
- Add a disable_native_backlight quirk for Dell XPS15 L521X designed
in an unusual way preventing native backlight from working on that
machine (Hans de Goede)"
* tag 'pm+acpi-3.19-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
ACPI / video: Add disable_native_backlight quirk for Dell XPS15 L521X
ACPI / processor: Rename acpi_(un)map_lsapic() to acpi_(un)map_cpu()
ACPI / processor: Convert apic_id to phys_id to make it arch agnostic
ACPI / PM: Fix PM initialization for devices that are not present
Linus Torvalds [Thu, 8 Jan 2015 21:52:16 +0000 (13:52 -0800)]
Merge tag 'keys-fixes-20150107' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs
Pull keyrings fixes from David Howells:
"Two fixes:
- Fix for the order in which things are done during key garbage
collection to prevent named keyrings causing a crash
[CVE-2014-9529].
- Fix assoc_array to explicitly #include rcupdate.h to prevent
compilation errors under certain circumstances"
* tag 'keys-fixes-20150107' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs:
assoc_array: Include rcupdate.h for call_rcu() definition
KEYS: close race between key lookup and freeing
Linus Torvalds [Thu, 8 Jan 2015 21:05:56 +0000 (13:05 -0800)]
Merge tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost
pull virtio/vhost fixes from Michael Tsirkin:
"This fixes a couple of bugs triggered by hot-unplug of virtio devices,
as well as a regression in vhost-net"
* tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:
vhost/net: length miscalculation
virtio_pci: document why we defer kfree
virtio_pci: defer kfree until release callback
virtio_pci: device-specific release callback
virtio: make del_vqs idempotent
Linus Torvalds [Thu, 8 Jan 2015 20:07:34 +0000 (12:07 -0800)]
Merge tag 'iommu-fixes-v3.19-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu
Pull IOMMU fixes from Joerg Roedel:
"Including:
- a domain structure leak fix in the Intel VT-d driver
- compile error fix for the VMSA IPMMU driver because of the
IOMMU_EXEC -> IOMMU_NOEXEC conversion
- two small cleanups as an aftermath of the merge window and the
domain-leak fix"
* tag 'iommu-fixes-v3.19-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu:
iommu/rockchip: Drop owner assignment from platform_drivers
iommu/vt-d: Remove dead code in device_notifier
iommu/vt-d: Fix dmar_domain leak in iommu_attach_device
iommu/ipmmu-vmsa: Change IOMMU_EXEC to IOMMU_NOEXEC
Pull crypto fixes from Herbert Xu:
"This fixes a build problem with sha-mb with old toolchains and an
implementation bug in the ctr(aes)/by8 branch of aesni-intel that's
enabled when AVX is available"
* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
crypto: sha-mb - Add avx2_supported check.
crypto: aesni - fix "by8" variant for 128 bit keys
Namhyung Kim [Tue, 30 Dec 2014 05:38:13 +0000 (14:38 +0900)]
perf callchain: Free callchains when hist entries are deleted
Markus reported that "perf top -g" can leak ~300MB per second on his
machine. This is partly because it missed to free callchains when hist
entries are deleted. Fix it.
Reported-by: Markus Trippelsdorf <markus@trippelsdorf.de> Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/20141230053813.GD6081@sejong Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Namhyung Kim [Tue, 23 Dec 2014 04:36:21 +0000 (13:36 +0900)]
perf hists: Fix children sort key behavior
When perf report --children resorts output fields, it tries to put
caller above the callee. But this was only meaningful for a same thread
and doing this requires callchain enabled. So fix its check before
comparing the callchain depth.
This also changes the hist accumulation tests: In test 3, xmalloc in
bash thread should be above than other perf threads due to alphabetical
order of comm string. Also it's under page_fault in bash thread since
alphabetical order of dso name. The sys_perf_event_open in perf thread
is put on the last line since it's self overhead is 0.
In test 4, the sys_perf_event_open is put above other perf entries that
have same children overhead since its callchain depth is smaller.
Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1419309381-2593-1-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Ard Biesheuvel [Thu, 8 Jan 2015 09:54:58 +0000 (09:54 +0000)]
arm64/efi: add missing call to early_ioremap_reset()
The early ioremap support introduced by patch bf4b558eba92
("arm64: add early_ioremap support") failed to add a call to
early_ioremap_reset() at an appropriate time. Without this call,
invocations of early_ioremap etc. that are done too late will go
unnoticed and may cause corruption.
This is exactly what happened when the first user of this feature
was added in patch f84d02755f5a ("arm64: add EFI runtime services").
The early mapping of the EFI memory map is unmapped during an early
initcall, at which time the early ioremap support is long gone.
Fix by adding the missing call to early_ioremap_reset() to
setup_arch(), and move the offending early_memunmap() to right after
the point where the early mapping of the EFI memory map is last used.
projects / linux-beck.git / log