On Fri, Sep 27, 2013 at 09:04:42AM +0100, Jiang Liu wrote:
diff --git a/arch/arm64/kernel/fpsimd.c b/arch/arm64/kernel/fpsimd.c
index bb785d2..12a25e5 100644
--- a/arch/arm64/kernel/fpsimd.c
+++ b/arch/arm64/kernel/fpsimd.c
@@ -80,9 +80,14 @@ void fpsimd_thread_switch(struct task_struct
Hello,
On Fri, Sep 27, 2013 at 03:16:59PM +0800, Chen Gang wrote:
Hmm... do you mean: can not evaluate an interface before implement(or
read details) them all?
No, I'm saying there are a lot more steps necessary between
recognizing that an interface needs an improvement and actually
improving
On Fri, Sep 27, 2013 at 05:14:27PM +0800, Yuanhan Liu wrote:
On Wed, Sep 25, 2013 at 09:38:38AM -0700, tip-bot for Peter Zijlstra wrote:
Commit-ID: f27dde8deef33c9e58027df11ceab2198601d6a6
Gitweb:
http://git.kernel.org/tip/f27dde8deef33c9e58027df11ceab2198601d6a6
Author: Peter
Hi
I'm trying to use -rc2 kernel however I'm getting quite often regular kernel
panic:
Here is a BUG trace from kvm running this kernel:
(I'm building kernel with some kernel debug checks)
(Kernel is used in 64bit qemu and running 32bit Debian environment)
linux-vanilla git:
On Thu, Sep 26, 2013 at 9:14 AM, Alexandre Courbot acour...@nvidia.com wrote:
Here is a proper patch series for the descriptor-based GPIO interface to be
merged into -next. It is rebased on your for-next branch and only slightly
changed with respect to the RFC I posted earlier, only more
On Fri, Sep 13, 2013 at 9:45 AM, Boris BREZILLON
b.brezil...@overkiz.com wrote:
Fix AT91_PINCTRL_DEBOUNCE_VAL dt macro typo.
Fix at91_pinctrl_mux_ops callback typos.
Signed-off-by: Boris BREZILLON b.brezil...@overkiz.com
Patch applied with Jean-Christophe's ACK.
Yours,
Linus Walleij
--
To
On Fri, Sep 13, 2013 at 9:47 AM, Boris BREZILLON
b.brezil...@overkiz.com wrote:
Replace at91_mux_get_deglitch with at91_mux_pio3_get_deglitch when using
sam9x5 (pio3) IP.
at91_mux_get_deglitch only test the activation of the Input Filter which
may be overloaded by the activation of the Input
On Fri, Sep 27, 2013 at 3:47 AM, Daniel Lezcano
daniel.lezc...@linaro.org wrote:
On 09/27/2013 04:24 AM, Rob Herring wrote:
From: Rob Herring rob.herr...@calxeda.com
Add cpuidle support for Calxeda ecx-2000 SoC.
Signed-off-by: Rob Herring rob.herr...@calxeda.com
Cc: Rafael J. Wysocki
Fixed coding style issues.
Signed-off-by: Martin Berglund mar...@rogsta.net
---
drivers/staging/sm7xxfb/sm7xxfb.c | 11 ---
1 file changed, 4 insertions(+), 7 deletions(-)
diff --git a/drivers/staging/sm7xxfb/sm7xxfb.c
b/drivers/staging/sm7xxfb/sm7xxfb.c
index 8add64b..57d8145 100644
On Tue, Aug 27, 2013 at 3:26 PM, Boris BREZILLON
b.brezil...@overkiz.com wrote:
Add missing atmel,at91sam9x5-pinctrl compatible string to the
documentation.
Signed-off-by: Boris BREZILLON b.brezil...@overkiz.com
Patch applied with Nicolas' ACK.
Yours,
Linus Walleij
--
To unsubscribe from
--
Lieber E-Mail-Benutzer;
Sie haben 23.432 überschritten Boxen setzen ihre
Web Service / Admin, und du wirst haben Probleme beim Senden und
E-Mails empfangen, während erneut überprüfen. Sie müssen aktualisieren
klicken
verknüpfen und füllen Sie die Informationen weiter unten, um Ihr Konto
zu
On Tue, Sep 24, 2013 at 9:46 AM, Axel Lin axel@ingics.com wrote:
Signed-off-by: Axel Lin axel@ingics.com
Patch applied, thanks Axel.
I should have caught this in review :-(
Yours,
Linus Walleij
--
To unsubscribe from this list: send the line unsubscribe linux-kernel in
the body of a
On Tue, Sep 24, 2013 at 9:47 AM, Axel Lin axel@ingics.com wrote:
Current code hold port-lock spinlock and then try to grab the lock again
in adi_gpio_set_value(). Fix it.
Signed-off-by: Axel Lin axel@ingics.com
Patch applied, thanks Axel.
Sonic, when you see patches like this please
On 09/27/2013 07:53 PM, Tejun Heo wrote:
Hello,
On Fri, Sep 27, 2013 at 03:16:59PM +0800, Chen Gang wrote:
Hmm... do you mean: can not evaluate an interface before implement(or
read details) them all?
No, I'm saying there are a lot more steps necessary between
recognizing that an
On Mon, Sep 23, 2013 at 11:29 AM, Sonic Zhang sonic@gmail.com wrote:
From: Sonic Zhang sonic.zh...@analog.com
Also avoid use NULL pointer in error message.
Signed-off-by: Sonic Zhang sonic.zh...@analog.com
Sonic are you respinning this with fixes to Dan's remarks?
You may need to
From: Jan Vesely jves...@redhat.com
Export the function so it can be used to predict segment counts
without calling the recalc function. This will be used in the next
patch.
Signed-off-by: Jan Vesely jves...@redhat.com
Signed-off-by: Tomas Henzl the...@redhat.com
---
block/blk-merge.c | 52
On 23:13-20130926, Eduardo Valentin wrote:
This patch changes the dtsi entry on omap4430 to contain
the thermal data. This data will enable the passive
cooling with CPUfreq cooling device at 100C and the
system will do a thermal shutdown at 125C.
Cc: Benoît Cousson bcous...@baylibre.com
On Fri, Sep 27, 2013 at 02:08:07PM +0200, Martin Berglund wrote:
@@ -508,9 +506,8 @@ static void sm7xx_set_timing(struct smtcfb_info *sfb)
/* init SEQ register SR30 - SR75 */
for (i = 0; i SIZE_SR30_SR75; i++)
- if
From: Jan Vesely jves...@redhat.com
The original behavior was to refuse all pages after the maximum number of
segments has been reached. However, some drivers (like st) craft their buffers
to potentially require exactly max segments and multiple pages in the last
segment. This patch modifies the
Signed-off-by: LABBE Corentin clabbe.montj...@gmail.com
---
drivers/hwmon/abituguru.c| 6 +++---
drivers/hwmon/abituguru3.c | 2 +-
drivers/hwmon/asus_atk0110.c | 2 +-
drivers/hwmon/max6650.c | 2 +-
drivers/hwmon/pmbus/pmbus_core.c | 2 +-
drivers/hwmon/w83791d.c
On Tue, Sep 24, 2013 at 7:47 PM, Laxman Dewangan ldewan...@nvidia.com wrote:
The AS3722 is a compact system PMU suitable for mobile phones, tablets etc.
Add a driver to support accessing the GPIO, pinmux and pin configuration
of 8 GPIO pins found on the AMS AS3722 through pin control driver
On Wed, Sep 25, 2013 at 9:48 AM, Jingoo Han jg1@samsung.com wrote:
Use devm_regulator_register() to make cleanup paths simpler.
Signed-off-by: Jingoo Han jg1@samsung.com
Acked-by: Linus Walleij linus.wall...@linaro.org
Yours,
Linus Walleij
--
To unsubscribe from this list: send the
Sorry for HTML mail, resend
On Fri, Sep 27, 2013 at 3:37 PM, Alexey Pelykh alexey.pel...@gmail.com wrote:
Hi Russel,
Probably then it would be great if you could summarize that in a patch to
replace mine, since basically mine is totally improper.
Thanks,
Alexey
On Sat, Sep 21, 2013 at
On Wed, Sep 25, 2013 at 10:12 AM, Jingoo Han jg1@samsung.com wrote:
Use devm_regulator_register() to make cleanup paths simpler,
and remove unnecessary remove().
Signed-off-by: Jingoo Han jg1@samsung.com
Acked-by: Linus Walleij linus.wall...@linaro.org
Yours,
Linus Walleij
--
To
On Fri, Sep 27, 2013 at 08:51:15AM +0200, Ingo Molnar wrote:
That message was cool and interesting back in the days when we wrote
lockdep ('hey, look ma, it really works!!'), but there hasn't been
any breakage in that area for a long time and it definitely does not
deserve one line of log spam
On Fri, 13 Sep 2013, H. Peter Anvin wrote:
crash-driver.patch
- The patch to add /dev/crash. Dave Anderson keeps making this work
(mostly).
It is also a security hole.
More than /dev/[k]mem? How so?
--
Jiri Kosina
SUSE Labs
--
To unsubscribe from this list: send the line
The test:
if (!a b)
a = b;
can be symplified in:
if (!a)
a = b;
And there's no need to test if ubi-image_seq is not null, because if it is,
it is set to image_seq.
So, we just test if image_seq is not null.
Signed-off-by: Richard Genoud richard.gen...@gmail.com
---
drivers/mtd/ubi/attach.c
Some old UBI implementations (e.g. U-Boot) have not implemented the image
sequence feature.
So, when erase blocks are written, the image sequence in the ec header
is lost (set to zero).
UBI scan_all() takes this case into account (commits
32bc4820287a1a03982979515949e8ea56eac641 and
On Friday 27 September 2013 06:10 PM, Linus Walleij wrote:
On Tue, Sep 24, 2013 at 7:47 PM, Laxman Dewangan ldewan...@nvidia.com wrote:
function which just return not supported.
- Nit cleanups.
This is getting a lot better quickly.
I'd like someone from devicet...@vger.kernel.org to say
The util-linux release v2.24-rc1 is available at
ftp://ftp.kernel.org/pub/linux/utils/util-linux/v2.24
Feedback and bug reports, as always, are welcomed.
Karel
Util-linux 2.24 Release Notes
=
Release highlights
--
fdisk(8):
- the code has
The *t variable in the fill_note_info function is only used if
siginfo-si_signo isn't 0. Moving t pointer to that inner scope.
Tested.
Signed-off-by: Geyslan G. Bem geys...@gmail.com
---
fs/binfmt_elf.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/fs/binfmt_elf.c
Hi Jan,
On 09/26/2013 10:53 PM, Jan Kara wrote:
Hello,
On Thu 26-09-13 21:32:07, Maxim Patlasov wrote:
While handling punch-hole fallocate, it's useless to truncate page cache
before removing the range from extent tree (or block map in indirect case)
because page cache can be re-populated
On Thursday 26 September 2013 09:08 PM, Stephen Warren wrote:
On 09/26/2013 06:48 AM, Laxman Dewangan wrote:
Recent movement of all configurations of pin in the single call of
pin_config_set(), it is aborting configuration if BIAS_PULL_PIN_DEFAULT
is selected as return of configuration.
The
Hi Tomasz,
On Thu, Sep 26, 2013 at 7:30 PM, Tomasz Figa t.f...@samsung.com wrote:
Hi Yadwinder,
I haven't reviewed this series yet, but let me clarify some things from
your comments.
On Thursday 26 of September 2013 17:38:58 Yadwinder Singh Brar wrote:
+
+/* Helper macros to define
Am 27.09.2013 14:51, schrieb Richard Genoud:
Some old UBI implementations (e.g. U-Boot) have not implemented the image
sequence feature.
So, when erase blocks are written, the image sequence in the ec header
is lost (set to zero).
UBI scan_all() takes this case into account (commits
Dne 27.9.2013 13:57, Zdenek Kabelac napsal(a):
Hi
I'm trying to use -rc2 kernel however I'm getting quite often regular kernel
panic:
Here is a BUG trace from kvm running this kernel:
(I'm building kernel with some kernel debug checks)
(Kernel is used in 64bit qemu and running 32bit Debian
Enable PMD split page table lock for X86_64 and PAE.
Signed-off-by: Kirill A. Shutemov kirill.shute...@linux.intel.com
Tested-by: Alex Thorlton athorl...@sgi.com
---
arch/x86/Kconfig | 4
arch/x86/include/asm/pgalloc.h | 11 ++-
2 files changed, 14 insertions(+), 1
2013/9/27 Richard Weinberger rich...@nod.at:
Am 27.09.2013 14:51, schrieb Richard Genoud:
Some old UBI implementations (e.g. U-Boot) have not implemented the image
sequence feature.
So, when erase blocks are written, the image sequence in the ec header
is lost (set to zero).
UBI scan_all()
Hugetlb supports multiple page sizes. We use split lock only for PMD
level, but not for PUD.
Signed-off-by: Naoya Horiguchi n-horigu...@ah.jp.nec.com
Signed-off-by: Kirill A. Shutemov kirill.shute...@linux.intel.com
Tested-by: Alex Thorlton athorl...@sgi.com
---
fs/proc/meminfo.c | 2 +-
Alex Thorlton noticed that some massively threaded workloads work poorly,
if THP enabled. This patchset fixes this by introducing split page table
lock for PMD tables. hugetlbfs is not covered yet.
This patchset is based on work by Naoya Horiguchi.
Please review and consider applying.
Changes:
With split page table lock for PMD level we can't hold
mm-page_table_lock while updating nr_ptes.
Let's convert it to atomic_t to avoid races.
Signed-off-by: Kirill A. Shutemov kirill.shute...@linux.intel.com
Tested-by: Alex Thorlton athorl...@sgi.com
---
fs/proc/task_mmu.c | 3 ++-
With split ptlock it's important to know which lock pmd_trans_huge_lock()
took. This patch adds one more parameter to the function to return the
lock.
In most places migration to new api is trivial.
Exception is move_huge_pmd(): we need to take two locks if pmd tables
are different.
Currently mm-pmd_huge_pte protected by page table lock. It will not
work with split lock. We have to have per-pmd pmd_huge_pte for proper
access serialization.
For now, let's just introduce wrapper to access mm-pmd_huge_pte.
Signed-off-by: Kirill A. Shutemov kirill.shute...@linux.intel.com
We're going to introduce split page table lock for PMD level.
Let's rename existing split ptlock for PTE level to avoid confusion.
Signed-off-by: Kirill A. Shutemov kirill.shute...@linux.intel.com
Tested-by: Alex Thorlton athorl...@sgi.com
---
arch/arm/mm/fault-armv.c | 6 +++---
With split page table lock we can't know which lock we need to take
before we find the relevant pmd.
Let's move lock taking inside the function.
Signed-off-by: Naoya Horiguchi n-horigu...@ah.jp.nec.com
Signed-off-by: Kirill A. Shutemov kirill.shute...@linux.intel.com
Tested-by: Alex Thorlton
Basic api, backed by mm-page_table_lock for now. Actual implementation
will be added later.
Signed-off-by: Naoya Horiguchi n-horigu...@ah.jp.nec.com
Signed-off-by: Kirill A. Shutemov kirill.shute...@linux.intel.com
Tested-by: Alex Thorlton athorl...@sgi.com
---
include/linux/mm.h | 13
Only trivial cases left. Let's convert them altogether.
Signed-off-by: Naoya Horiguchi n-horigu...@ah.jp.nec.com
Signed-off-by: Kirill A. Shutemov kirill.shute...@linux.intel.com
Tested-by: Alex Thorlton athorl...@sgi.com
---
mm/huge_memory.c | 108
The basic idea is the same as with PTE level: the lock is embedded into
struct page of table's page.
We can't use mm-pmd_huge_pte to store pgtables for THP, since we don't
take mm-page_table_lock anymore. Let's reuse page-lru of table's page
for that.
pgtable_pmd_page_ctor() returns true, if
On 09/27/2013 07:35 PM, Catalin Marinas wrote:
On Fri, Sep 27, 2013 at 09:04:42AM +0100, Jiang Liu wrote:
diff --git a/arch/arm64/kernel/fpsimd.c b/arch/arm64/kernel/fpsimd.c
index bb785d2..12a25e5 100644
--- a/arch/arm64/kernel/fpsimd.c
+++ b/arch/arm64/kernel/fpsimd.c
@@ -80,9 +80,14 @@
Am 27.09.2013 15:16, schrieb Richard Genoud:
2013/9/27 Richard Weinberger rich...@nod.at:
Am 27.09.2013 14:51, schrieb Richard Genoud:
Some old UBI implementations (e.g. U-Boot) have not implemented the image
sequence feature.
So, when erase blocks are written, the image sequence in the ec
On 27-09-2013 08:24, Nishanth Menon wrote:
On 23:13-20130926, Eduardo Valentin wrote:
This patch changes the dtsi entry on omap4430 to contain
the thermal data. This data will enable the passive
cooling with CPUfreq cooling device at 100C and the
system will do a thermal shutdown at 125C.
Logging and tracking firmware bugs in the kernel has long been an issue
for system administrators. The current kernel does not have a good
uniform method of reporting firmware bugs and the code in the kernel is a
mix of printk's and WARN_ONs. This causes problems for both system
administrators
The following are a couple of fixes for ti-abb regulator based on
updated I documentation and tests.
Based on v3.12-rc2 tag. These fixes could be punted to 3.13-rc1 as the
current OMAP support in master is not ready for use yet.
Nishanth Menon (2):
regulator: ti-abb: skip optional parameter
On Thu, Sep 26, 2013 at 01:01:43PM -0400, Rhyland Klein wrote:
The tegra114 driver wasn't currently handling the cs_change
functionality. cs_change is meant to invert the decisions of whether
or not to deactivate CS after each transfer. Without cs_change, after
every transfer (other than the
On platforms like OMAP4460, LDO override is never used. Even though
efuse determines the ABB bias mode to operate at, ABB voltage is
preconfigured in internal efuse registers without the need for
LDO override for bias voltage. So skip optional parameter if
property is not present.
Signed-off-by:
On Fri, Sep 27, 2013 at 8:20 AM, Eduardo Valentin
eduardo.valen...@ti.com wrote:
On 27-09-2013 08:24, Nishanth Menon wrote:
On 23:13-20130926, Eduardo Valentin wrote:
This patch changes the dtsi entry on omap4430 to contain
the thermal data. This data will enable the passive
cooling with
As documented in Application Note SWPA117 v2.1(NDA), LDO override has a
requirement that when switching from Bias active + override active
mode(FBB/RBB) to Bypass(nominal) mode, LDO reset must be performed
*after* LDO transitions to Bypass(nominal) mode.
The same rule in reverse applies when
If another task handled a hinting fault in parallel then do not double
account for it.
Not-signed-off-by: Peter Zijlstra
---
mm/huge_memory.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 860a368..5c37cd2 100644
---
Weighing in at 63 patches makes the term basic in the series title is
a misnomer.
This series still has roughly the same goals as previous versions. It
reduces overhead of automatic balancing through scan rate reduction
and the avoidance of TLB flushes. It selects a preferred node and moves
tasks
From: Rik van Riel r...@redhat.com
Short spikes of CPU load can lead to a task being migrated
away from its preferred node for temporary reasons.
It is important that the task is migrated back to where it
belongs, in order to avoid migrating too much memory to its
new location, and generally
From: Rik van Riel r...@redhat.com
When a task is already running on its preferred node, increment
numa_migrate_seq to indicate that the task is settled if migration is
temporarily disabled, and memory should migrate towards it.
[mgor...@suse.de: Only increment migrate_seq if migration
On 27/09/13 12:34, Mark Rutland wrote:
+ - rx-mode: Can be infrared or uhf. rx-mode should be present iff
+the rx pins are wired up.
I'm unsure on this. What if the device has multiple receivers that can
be independently configured? What if it supports something other than
infrared or
This patch tracks what nodes numa hinting faults were incurred on.
This information is later used to schedule a task on the node storing
the pages most frequently faulted by the task.
Signed-off-by: Mel Gorman mgor...@suse.de
---
include/linux/sched.h | 2 ++
kernel/sched/core.c | 3 +++
A tasks preferred node is selected based on the number of faults
recorded for a node but the actual task_numa_migate() conducts a global
search regardless of the preferred nid. This patch checks if the
preferred nid has capacity and if so, searches for a CPU within that
node. This avoids a global
Base page PMD faulting is meant to batch handle NUMA hinting faults from
PTEs. However, even is no PTE faults would ever be handled within a
range the kernel still traps PMD hinting faults. This patch avoids the
overhead.
Signed-off-by: Mel Gorman mgor...@suse.de
---
mm/mprotect.c | 7 +--
1
From: Peter Zijlstra pet...@infradead.org
And here's a little something to make sure not the whole world ends up
in a single group.
As while we don't migrate shared executable pages, we do scan/fault on
them. And since everybody links to libc, everybody ends up in the same
group.
From: Peter Zijlstra pet...@infradead.org
(This changelog needs more work, it's currently inaccurate and it's not
clear at exactly what point rt env-fbq_type is true for the logic to
kick in)
This patch classifies scheduler domains and runqueues into FBQ (cannot
guess what this expands to)
From: Rik van Riel r...@redhat.com
Shared faults can lead to lots of unnecessary page migrations,
slowing down the system, and causing private faults to hit the
per-pgdat migration ratelimit.
This patch adds sysctl numa_balancing_migrate_deferred, which specifies
how many shared page migrations
From: Rik van Riel r...@redhat.com
Due to the way the pid is truncated, and tasks are moved between
CPUs by the scheduler, it is possible for the current task_numa_fault
to group together tasks that do not actually share memory together.
This patch adds a few easy sanity checks to
From: Rik van Riel r...@redhat.com
Adjust numa_scan_period in task_numa_placement, depending on how much
useful work the numa code can do. The more local faults there are in a
given scan window the longer the period (and hence the slower the scan rate)
during the next window. If there are
As Peter says If you're going to hold locks you can also do away with all
that atomic_long_*() nonsense. Lock aquisition moved slightly to protect
the updates.
Signed-off-by: Mel Gorman mgor...@suse.de
---
kernel/sched/fair.c | 49 -
1 file
Am 27.09.2013 14:51, schrieb Richard Genoud:
The test:
if (!a b)
a = b;
can be symplified in:
if (!a)
a = b;
And there's no need to test if ubi-image_seq is not null, because if it is,
it is set to image_seq.
So, we just test if image_seq is not null.
Signed-off-by: Richard
NUMA hinting faults will not migrate a shared executable page mapped by
multiple processes on the grounds that the data is probably in the CPU
cache already and the page may just bounce between tasks running on multipl
nodes. Even if the migration is avoided, there is still the overhead of
From: Rik van Riel r...@redhat.com
With the scan rate code working (at least for multi-instance specjbb),
the large hammer that is sched: Do not migrate memory immediately after
switching node can be replaced with something smarter. Revert temporarily
migration disabling and all traces of
Scan rate is altered based on whether shared/private faults dominated.
task_numa_group() may detect false sharing but that information is not
taken into account when adapting the scan rate. Take it into account.
Signed-off-by: Mel Gorman mgor...@suse.de
---
kernel/sched/fair.c | 8 ++--
1
From: Rik van Riel r...@redhat.com
This patch separately considers task and group affinities when searching
for swap candidates during task NUMA placement. If tasks are not part of
a group or the same group then the task weights are considered.
Otherwise the group weights are compared.
From: Rik van Riel r...@redhat.com
This patch separately considers task and group affinities when
searching for swap candidates during NUMA placement. If tasks
are part of the same group, or no group at all, the task weights
are considered.
Some hysteresis is added to prevent tasks within one
Hi Jason,
On 09/18/2013 10:32 PM, Jason Gunthorpe wrote:
On Wed, Sep 18, 2013 at 03:15:17PM -0400, Jason Cooper wrote:
+ Jason Gunthorpe
Thanks, looks interesting, we could possibly use this interface if it
met our needs..
On Wed, Sep 18, 2013 at 05:56:39PM +0200, Michal Simek wrote:
With scan rate adaptions based on whether the workload has properly
converged or not there should be no need for the scan period reset
hammer. Get rid of it.
Signed-off-by: Mel Gorman mgor...@suse.de
---
Documentation/sysctl/kernel.txt | 11 +++
include/linux/mm_types.h| 3 ---
With the THP migration races closed it is still possible to occasionally
see corruption. The problem is related to handling PMD pages in batch.
When a page fault is handled it can be assumed that the page being
faulted will also be flushed from the TLB. The same flushing does not
happen when
Hi
With recent build of 3.12-rc2 I'm getting this warning report from kernel:
(hw Lenovo T61, C2D, 4GB Ram)
(repost since linux-kernel@ rejected my gmail email)
e1000e :00:19.0: irq 46 for MSI/MSI-X
[ cut here ]
WARNING: CPU: 1 PID: 301 at fs/sysfs/dir.c:526
* Borislav Petkov b...@alien8.de wrote:
On Fri, Sep 27, 2013 at 08:51:15AM +0200, Ingo Molnar wrote:
That message was cool and interesting back in the days when we wrote
lockdep ('hey, look ma, it really works!!'), but there hasn't been
any breakage in that area for a long time and it
From: Rik van Riel r...@redhat.com
After page migration, the new page has the nidpid unset. This makes
every fault on a recently migrated page look like a first numa fault,
leading to another page migration.
Copying over the nidpid at page migration time should prevent erroneous
migrations of
Having multiple tasks in a group go through task_numa_placement
simultaneously can lead to a task picking a wrong node to run on, because
the group stats may be in the middle of an update. This patch avoids
parallel updates by holding the numa_group lock during placement
decisions.
Signed-off-by:
From: Rik van Riel r...@redhat.com
It is possible for a task in a numa group to call exec, and
have the new (unrelated) executable inherit the numa group
association from its former self.
This has the potential to break numa grouping, and is trivial
to fix.
Signed-off-by: Rik van Riel
This patch replaces find_idlest_cpu_node with task_numa_find_cpu.
find_idlest_cpu_node has two critical limitations. It does not take the
scheduling class into account when calculating the load and it is unsuitable
for using when comparing loads between NUMA nodes.
task_numa_find_cpu uses similar
From: Ingo Molnar mi...@kernel.org
Signed-off-by: Ingo Molnar mi...@kernel.org
Signed-off-by: Peter Zijlstra pet...@infradead.org
Link: http://lkml.kernel.org/n/tip-5giqjcqnc93a89q01ymtj...@git.kernel.org
---
include/linux/sched.h | 6 ++
kernel/sched/debug.c | 60
This patch uses the fraction of faults on a particular node for both task
and group, to figure out the best node to place a task. If the task and
group statistics disagree on what the preferred node should be then a full
rescan will select the node with the best combined weight.
Signed-off-by:
From: Rik van Riel r...@redhat.com
A newly spawned thread inside a process should stay on the same
NUMA node as its parent. This prevents processes from being torn
across multiple NUMA nodes every time they spawn a new thread.
Signed-off-by: Rik van Riel r...@redhat.com
Signed-off-by: Mel Gorman
When a preferred node is selected for a tasks there is an attempt to migrate
the task to a CPU there. This may fail in which case the task will only
migrate if the active load balancer takes action. This may never happen if
the conditions are not right. This patch will check at NUMA hinting fault
From: Rik van Riel r...@redhat.com
The load balancer will spread workloads across multiple NUMA nodes,
in order to balance the load on the system. This means that sometimes
a task's preferred node has available capacity, but moving the task
there will not succeed, because that would create too
This patch implements a system-wide search for swap/migration candidates
based on total NUMA hinting faults. It has a balance limit, however it
doesn't properly consider total node balance.
In the old scheme a task selected a preferred node based on the highest
number of private faults recorded
From: Rik van Riel r...@redhat.com
The load balancer can move tasks between nodes and does not take NUMA
locality into account. With automatic NUMA balancing this may result in the
tasks working set being migrated to the new node. However, as the fault
buffer will still store faults from the old
From: Peter Zijlstra pet...@infradead.org
Use the new stop_two_cpus() to implement migrate_swap(), a function that
flips two tasks between their respective cpus.
I'm fairly sure there's a less crude way than employing the stop_two_cpus()
method, but everything I tried either got horribly fragile
From: Peter Zijlstra pet...@infradead.org
Change the per page last fault tracking to use cpu,pid instead of
nid,pid. This will allow us to try and lookup the alternate task more
easily. Note that even though it is the cpu that is store in the page
flags that the mpol_misplaced decision is still
On 09/26/2013 05:50 AM, Thierry Reding wrote:
On Thu, Sep 26, 2013 at 03:26:13PM +0300, Tomi Valkeinen wrote:
On 26/09/13 15:02, Thierry Reding wrote:
On Thu, Sep 26, 2013 at 01:13:18PM +0300, Tomi Valkeinen wrote:
On 11/09/13 14:40, Mike Dunn wrote:
On 09/10/2013 10:21 AM, Thierry Reding
From: Peter Zijlstra pet...@infradead.org
While parallel applications tend to align their data on the cache
boundary, they tend not to align on the page or THP boundary.
Consequently tasks that partition their data can still false-share
pages presenting a problem for optimal NUMA placement.
This
Ideally it would be possible to distinguish between NUMA hinting faults that
are private to a task and those that are shared. If treated identically
there is a risk that shared pages bounce between nodes depending on
the order they are referenced by tasks. Ultimately what is desirable is
that task
This patch selects a preferred node for a task to run on based on the
NUMA hinting faults. This information is later used to migrate tasks
towards the node during balancing.
Signed-off-by: Mel Gorman mgor...@suse.de
---
include/linux/sched.h | 1 +
kernel/sched/core.c | 1 +
From: Peter Zijlstra pet...@infradead.org
Introduce stop_two_cpus() in order to allow controlled swapping of two
tasks. It repurposes the stop_machine() state machine but only stops
the two cpus which we can do with on-stack structures and avoid
machine wide synchronization issues.
The ordering
701 - 800 of 1156 matches
Mail list logo