On Tue, May 07, 2019 at 12:27:21AM -0700, Sultan Alsawaf wrote:
> On Tue, May 07, 2019 at 09:04:30AM +0200, Greg Kroah-Hartman wrote:
> > Um, why can't "all" Android devices take the same patches that the Pixel
> > phones are using today? They should all be in the public android-common
> > kernel
On Tue, May 07, 2019 at 09:04:30AM +0200, Greg Kroah-Hartman wrote:
> Um, why can't "all" Android devices take the same patches that the Pixel
> phones are using today? They should all be in the public android-common
> kernel repositories that all Android devices should be syncing with on a
>
On Mon, May 06, 2019 at 07:16:22PM -0700, Sultan Alsawaf wrote:
> This is a complete low memory killer solution for Android that is small
> and simple. Processes are killed according to the priorities that
> Android gives them, so that the least important processes are always
>
This is a complete low memory killer solution for Android that is small
and simple. Processes are killed according to the priorities that
Android gives them, so that the least important processes are always
killed first. Processes are killed until memory deficits are satisfied,
as observed from
Hello,
Serge Semin wrote:
> memblock subsystem provides a method to optionally test the passed
> memory region in case if it was requested via special kernel boot
> argument. Lets add the function at the bottom of the arch_mem_init()
> method. Testing at this point in the boot sequence should be
memblock subsystem provides a method to optionally test the passed
memory region in case if it was requested via special kernel boot
argument. Lets add the function at the bottom of the arch_mem_init()
method. Testing at this point in the boot sequence should be safe since all
critical areas are
memblock subsystem provides a method to optionally test the passed
memory region in case if it was requested via special kernel boot
argument. Lets add the function at the bottom of the arch_mem_init()
method. Testing at this point in the boot sequence should be safe since all
critical areas are
On Mon, Apr 01, 2019 at 10:12:01PM +0200, Michal Hocko wrote:
> On Fri 29-03-19 16:16:38, Catalin Marinas wrote:
> > On Fri, Mar 29, 2019 at 01:02:37PM +0100, Michal Hocko wrote:
> > > On Thu 28-03-19 14:59:17, Catalin Marinas wrote:
> > > [...]
> > > > >From
On Fri 29-03-19 16:16:38, Catalin Marinas wrote:
> On Fri, Mar 29, 2019 at 01:02:37PM +0100, Michal Hocko wrote:
> > On Thu 28-03-19 14:59:17, Catalin Marinas wrote:
> > [...]
> > > >From 09eba8f0235eb16409931e6aad77a45a12bedc82 Mon Sep 17 00:00:00 2001
> > > From: Catalin Marinas
> > > Date:
On Fri, Mar 29, 2019 at 01:02:37PM +0100, Michal Hocko wrote:
> On Thu 28-03-19 14:59:17, Catalin Marinas wrote:
> [...]
> > >From 09eba8f0235eb16409931e6aad77a45a12bedc82 Mon Sep 17 00:00:00 2001
> > From: Catalin Marinas
> > Date: Thu, 28 Mar 2019 13:26:07 +
> > Subject: [PATCH] mm:
On Thu 28-03-19 14:59:17, Catalin Marinas wrote:
[...]
> >From 09eba8f0235eb16409931e6aad77a45a12bedc82 Mon Sep 17 00:00:00 2001
> From: Catalin Marinas
> Date: Thu, 28 Mar 2019 13:26:07 +
> Subject: [PATCH] mm: kmemleak: Use mempool allocations for kmemleak objects
>
> This patch adds
On Thu, 2019-03-28 at 15:05 +, Catalin Marinas wrote:
> > It takes 2 runs of LTP oom01 tests to disable kmemleak.
>
> What configuration are you using (number of CPUs, RAM)? I tried this on
> an arm64 guest under kvm with 4 CPUs and 512MB of RAM, together with
> fault injection on
t; > provides a good balance between enabling kmemleak in a low-memory
> > > > situation and not introducing too much hackiness into the existing
> > > > code for now.
>
> On Thu, Mar 28, 2019 at 08:05:31AM +0200, Pekka Enberg wrote:
> > > Unfortunatel
On Thu, 2019-03-28 at 14:59 +, Catalin Marinas wrote: 2
> +/* minimum memory pool sizes */
> +#define MIN_OBJECT_POOL (NR_CPUS * 4)
> +#define MIN_SCAN_AREA_POOL (NR_CPUS * 1)
I am thinking about making those are tunable, so people could have a big pool
depends on
On Wed, Mar 27, 2019 at 02:02:27PM -0400, Qian Cai wrote:
> On 3/27/19 1:29 PM, Catalin Marinas wrote:
> > From dc4194539f8191bb754901cea74c86e7960886f8 Mon Sep 17 00:00:00 2001
> > From: Catalin Marinas
> > Date: Wed, 27 Mar 2019 17:20:57 +
> > Subject: [PATCH] mm: kmemleak: Add an emergency
On Wed, Mar 27, 2019 at 11:21:58AM -0700, Matthew Wilcox wrote:
> On Wed, Mar 27, 2019 at 05:29:57PM +, Catalin Marinas wrote:
> > On Wed, Mar 27, 2019 at 09:44:32AM +0100, Michal Hocko wrote:
> > > As long as there is an implicit __GFP_NOFAIL then kmemleak is simply
> > > broken no matter
Hi Catalin,
On 27/03/2019 2.59, Qian Cai wrote:
Unless there is a brave soul to reimplement the kmemleak to embed it's
metadata into the tracked memory itself in a foreseeable future, this
provides a good balance between enabling kmemleak in a low-memory
situation and not introducing too much
ance between enabling kmemleak in a low-memory
> > situation and not introducing too much hackiness into the existing
> > code for now.
>
> Unfortunately I am not that brave soul, but I'm wondering what the
> complication here is? It shouldn't be too hard to teach calculate_sizes(
Hi,
On 27/03/2019 2.59, Qian Cai wrote:
Unless there is a brave soul to reimplement the kmemleak to embed it's
metadata into the tracked memory itself in a foreseeable future, this
provides a good balance between enabling kmemleak in a low-memory
situation and not introducing too much hackiness
On 3/27/19 1:29 PM, Catalin Marinas wrote:
> From dc4194539f8191bb754901cea74c86e7960886f8 Mon Sep 17 00:00:00 2001
> From: Catalin Marinas
> Date: Wed, 27 Mar 2019 17:20:57 +
> Subject: [PATCH] mm: kmemleak: Add an emergency allocation pool for kmemleak
> objects
>
> This patch adds an
On Wed 27-03-19 17:29:57, Catalin Marinas wrote:
[...]
> Quick attempt below and it needs some more testing (pretty random pick
> of the EMERGENCY_POOL_SIZE value). Also, with __GFP_NOFAIL removed, are
> the other flags safe or we should trim them further?
I would be still careful about
On Wed, Mar 27, 2019 at 05:29:57PM +, Catalin Marinas wrote:
> On Wed, Mar 27, 2019 at 09:44:32AM +0100, Michal Hocko wrote:
> > As long as there is an implicit __GFP_NOFAIL then kmemleak is simply
> > broken no matter what other gfp flags you play with. Has anybody looked
> > at some sort of
ak memory scanning code, that's not actually
possible without some long periods with kmemleak_lock held. The scanner
relies on the kmemleak_object (refcounted) being around even when the
actual memory block has been freed.
> > this
> > provides a good balance between enabling kmemleak
Kmemleak could quickly fail to allocate an object structure and then
disable itself below in a low-memory situation. For example, running a
mmap() workload triggering swapping and OOM. This is especially
problematic for running things like LTP testsuite where one OOM test
case would disable
On Wed 27-03-19 09:05:31, Qian Cai wrote:
> On 3/27/19 7:44 AM, Michal Hocko wrote> What? Normal spin lock implementation
> doesn't disable interrupts. So
> > either I misunderstand what you are saying or you seem to be confused.
> > the thing is that in_atomic relies on preempt_count to work
On 3/27/19 7:44 AM, Michal Hocko wrote> What? Normal spin lock implementation
doesn't disable interrupts. So
> either I misunderstand what you are saying or you seem to be confused.
> the thing is that in_atomic relies on preempt_count to work properly and
> if you have CONFIG_PREEMPT_COUNT=n then
On Wed 27-03-19 07:34:32, Qian Cai wrote:
> On 3/27/19 4:44 AM, Michal Hocko wrote:
> >> diff --git a/mm/kmemleak.c b/mm/kmemleak.c
> >> index a2d894d3de07..7f4545ab1f84 100644
> >> --- a/mm/kmemleak.c
> >> +++ b/mm/kmemleak.c
> >> @@ -580,7 +580,16 @@ static struct kmemleak_object
On 3/27/19 4:44 AM, Michal Hocko wrote:
>> diff --git a/mm/kmemleak.c b/mm/kmemleak.c
>> index a2d894d3de07..7f4545ab1f84 100644
>> --- a/mm/kmemleak.c
>> +++ b/mm/kmemleak.c
>> @@ -580,7 +580,16 @@ static struct kmemleak_object *create_object(unsigned
>> long ptr, size_t size,
>> struct
On Tue 26-03-19 20:59:48, Qian Cai wrote:
[...]
> Unless there is a brave soul to reimplement the kmemleak to embed it's
> metadata into the tracked memory itself in a foreseeable future, this
> provides a good balance between enabling kmemleak in a low-memory
> situation and not int
Kmemleak could quickly fail to allocate an object structure and then
disable itself below in a low-memory situation. For example, running a
mmap() workload triggering swapping and OOM. This is especially
problematic for running things like LTP testsuite where one OOM test
case would disable
.
I suppose this is not going to work well, as DMA_API_DEBUG use a similar
approach [1] but I still saw it is struggling in a low-memory situation and
disable itself occasionally.
[1] https://lkml.org/lkml/2018/12/10/383
ata into the tracked memory itself in a foreseeable future, this
> > > provides a good balance between enabling kmemleak in a low-memory
> > > situation and not introducing too much hackiness into the existing
> > > code for now.
> >
> > I don't understand k
On 3/26/19 12:00 PM, Christopher Lameter wrote:
>> + */
>> +gfp = (in_atomic() || irqs_disabled()) ? GFP_ATOMIC :
>> + gfp_kmemleak_mask(gfp) | __GFP_DIRECT_RECLAIM;
>> +object = kmem_cache_alloc(object_cache, gfp);
>> +}
>> +
>> if
> provides a good balance between enabling kmemleak in a low-memory
> > situation and not introducing too much hackiness into the existing
> > code for now.
>
> I don't understand kmemleak. Kirill pointed me at this a few days ago:
>
> https://gist.github.com/kiryl/3225
On Tue, Mar 26, 2019 at 11:43:38AM -0400, Qian Cai wrote:
> Kmemleak could quickly fail to allocate an object structure and then
> disable itself in a low-memory situation. For example, running a mmap()
> workload triggering swapping and OOM. This is especially problematic for
> ru
On Tue, Mar 26, 2019 at 11:43:38AM -0400, Qian Cai wrote:
> Unless there is a brave soul to reimplement the kmemleak to embed it's
> metadata into the tracked memory itself in a foreseeable future, this
> provides a good balance between enabling kmemleak in a low-memory
&g
On Tue, 26 Mar 2019, Qian Cai wrote:
> + if (!object) {
> + /*
> + * The tracked memory was allocated successful, if the kmemleak
> + * object failed to allocate for some reasons, it ends up with
> + * the whole kmemleak disabled, so let it
Kmemleak could quickly fail to allocate an object structure and then
disable itself in a low-memory situation. For example, running a mmap()
workload triggering swapping and OOM. This is especially problematic for
running things like LTP testsuite where one OOM test case would disable
the whole
On Wed, Mar 20, 2019 at 12:10:23AM +0100, Christian Brauner wrote:
> On Tue, Mar 19, 2019 at 03:48:32PM -0700, Daniel Colascione wrote:
> > On Tue, Mar 19, 2019 at 3:14 PM Christian Brauner
> > wrote:
> > > So I dislike the idea of allocating new inodes from the procfs super
> > > block. I would
On Tue, Mar 19, 2019 at 03:48:32PM -0700, Daniel Colascione wrote:
> On Tue, Mar 19, 2019 at 3:14 PM Christian Brauner
> wrote:
> > So I dislike the idea of allocating new inodes from the procfs super
> > block. I would like to avoid pinning the whole pidfd concept exclusively
> > to proc. The
On Tue, Mar 19, 2019 at 3:14 PM Christian Brauner wrote:
> So I dislike the idea of allocating new inodes from the procfs super
> block. I would like to avoid pinning the whole pidfd concept exclusively
> to proc. The idea is that the pidfd API will be useable through procfs
> via open("/proc/")
On Tue, Mar 19, 2019 at 11:14:17PM +0100, Christian Brauner wrote:
[snip]
> >
> > ---8<---
> >
> > From: Joel Fernandes
> > Subject: [PATCH] Partial skeleton prototype of pidfd_wait frontend
> >
> > Signed-off-by: Joel Fernandes
> > ---
> >
On Mon, Mar 18, 2019 at 07:50:52PM -0400, Joel Fernandes wrote:
> On Mon, Mar 18, 2019 at 01:29:51AM +0100, Christian Brauner wrote:
> > On Sun, Mar 17, 2019 at 08:40:19AM -0700, Daniel Colascione wrote:
> > > On Sun, Mar 17, 2019 at 4:42 AM Christian Brauner
> > > wrote:
> > > >
> > > > On Sat,
On Mon, Mar 18, 2019 at 01:29:51AM +0100, Christian Brauner wrote:
> On Sun, Mar 17, 2019 at 08:40:19AM -0700, Daniel Colascione wrote:
> > On Sun, Mar 17, 2019 at 4:42 AM Christian Brauner
> > wrote:
> > >
> > > On Sat, Mar 16, 2019 at 09:53:06PM -0400, Joel Fernandes wrote:
> > > > On Sat, Mar
On Sun, Mar 17, 2019 at 08:40:19AM -0700, Daniel Colascione wrote:
> On Sun, Mar 17, 2019 at 4:42 AM Christian Brauner
> wrote:
> >
> > On Sat, Mar 16, 2019 at 09:53:06PM -0400, Joel Fernandes wrote:
> > > On Sat, Mar 16, 2019 at 12:37:18PM -0700, Suren Baghdasaryan wrote:
> > > > On Sat, Mar
On Sun, Mar 17, 2019 at 10:16 AM Serge E. Hallyn wrote:
>
> On Sun, Mar 17, 2019 at 10:11:10AM -0700, Daniel Colascione wrote:
> > On Sun, Mar 17, 2019 at 9:35 AM Serge E. Hallyn wrote:
> > >
> > > On Sun, Mar 17, 2019 at 12:42:40PM +0100, Christian Brauner wrote:
> > > > On Sat, Mar 16, 2019 at
On Sun, Mar 17, 2019 at 10:11:10AM -0700, Daniel Colascione wrote:
> On Sun, Mar 17, 2019 at 9:35 AM Serge E. Hallyn wrote:
> >
> > On Sun, Mar 17, 2019 at 12:42:40PM +0100, Christian Brauner wrote:
> > > On Sat, Mar 16, 2019 at 09:53:06PM -0400, Joel Fernandes wrote:
> > > > On Sat, Mar 16, 2019
On Sun, Mar 17, 2019 at 9:35 AM Serge E. Hallyn wrote:
>
> On Sun, Mar 17, 2019 at 12:42:40PM +0100, Christian Brauner wrote:
> > On Sat, Mar 16, 2019 at 09:53:06PM -0400, Joel Fernandes wrote:
> > > On Sat, Mar 16, 2019 at 12:37:18PM -0700, Suren Baghdasaryan wrote:
> > > > On Sat, Mar 16, 2019
On Sun, Mar 17, 2019 at 12:42:40PM +0100, Christian Brauner wrote:
> On Sat, Mar 16, 2019 at 09:53:06PM -0400, Joel Fernandes wrote:
> > On Sat, Mar 16, 2019 at 12:37:18PM -0700, Suren Baghdasaryan wrote:
> > > On Sat, Mar 16, 2019 at 11:57 AM Christian Brauner
> > > wrote:
> > > >
> > > > On
On Sun, Mar 17, 2019 at 4:42 AM Christian Brauner wrote:
>
> On Sat, Mar 16, 2019 at 09:53:06PM -0400, Joel Fernandes wrote:
> > On Sat, Mar 16, 2019 at 12:37:18PM -0700, Suren Baghdasaryan wrote:
> > > On Sat, Mar 16, 2019 at 11:57 AM Christian Brauner
> > > wrote:
> > > >
> > > > On Sat, Mar
On Sat, Mar 16, 2019 at 09:53:06PM -0400, Joel Fernandes wrote:
> On Sat, Mar 16, 2019 at 12:37:18PM -0700, Suren Baghdasaryan wrote:
> > On Sat, Mar 16, 2019 at 11:57 AM Christian Brauner
> > wrote:
> > >
> > > On Sat, Mar 16, 2019 at 11:00:10AM -0700, Daniel Colascione wrote:
> > > > On Sat,
On Sat, Mar 16, 2019 at 12:37:18PM -0700, Suren Baghdasaryan wrote:
> On Sat, Mar 16, 2019 at 11:57 AM Christian Brauner
> wrote:
> >
> > On Sat, Mar 16, 2019 at 11:00:10AM -0700, Daniel Colascione wrote:
> > > On Sat, Mar 16, 2019 at 10:31 AM Suren Baghdasaryan
> > > wrote:
> > > >
> > > > On
On Sat, Mar 16, 2019 at 11:57 AM Christian Brauner wrote:
>
> On Sat, Mar 16, 2019 at 11:00:10AM -0700, Daniel Colascione wrote:
> > On Sat, Mar 16, 2019 at 10:31 AM Suren Baghdasaryan
> > wrote:
> > >
> > > On Fri, Mar 15, 2019 at 11:49 AM Joel Fernandes
> > > wrote:
> > > >
> > > > On Fri,
On Sat, Mar 16, 2019 at 11:00:10AM -0700, Daniel Colascione wrote:
> On Sat, Mar 16, 2019 at 10:31 AM Suren Baghdasaryan wrote:
> >
> > On Fri, Mar 15, 2019 at 11:49 AM Joel Fernandes
> > wrote:
> > >
> > > On Fri, Mar 15, 2019 at 07:24:28PM +0100, Christian Brauner wrote:
> > > [..]
> > > > >
On Sat, Mar 16, 2019 at 10:31 AM Suren Baghdasaryan wrote:
>
> On Fri, Mar 15, 2019 at 11:49 AM Joel Fernandes
> wrote:
> >
> > On Fri, Mar 15, 2019 at 07:24:28PM +0100, Christian Brauner wrote:
> > [..]
> > > > why do we want to add a new syscall (pidfd_wait) though? Why not just
> > > > use
On Fri, Mar 15, 2019 at 11:49 AM Joel Fernandes wrote:
>
> On Fri, Mar 15, 2019 at 07:24:28PM +0100, Christian Brauner wrote:
> [..]
> > > why do we want to add a new syscall (pidfd_wait) though? Why not just use
> > > standard poll/epoll interface on the proc fd like Daniel was suggesting.
> > >
On Fri, Mar 15, 2019 at 07:24:28PM +0100, Christian Brauner wrote:
[..]
> > why do we want to add a new syscall (pidfd_wait) though? Why not just use
> > standard poll/epoll interface on the proc fd like Daniel was suggesting.
> > AFAIK, once the proc file is opened, the struct pid is essentially
that we're finally getting it.
> > > Adding new system calls is not expensive; going to great lengths to
> > > avoid adding one is like calling a helicopter to avoid crossing the
> > > street. I don't think we should present an abuse of the debugging and
> > >
> > performance monitoring infrastructure as an alternative to a robust
> > and desperately-needed bit of core functionality that's neither hard
> > to add nor complex to implement nor expensive to use.
> >
> > Regarding the proposal for a new kernel-side lmkd: when poss
d present an abuse of the debugging and
> performance monitoring infrastructure as an alternative to a robust
> and desperately-needed bit of core functionality that's neither hard
> to add nor complex to implement nor expensive to use.
>
> Regarding the proposal for a new kernel-
On Fri, Mar 15, 2019 at 9:43 AM Steven Rostedt wrote:
>
> On Thu, 14 Mar 2019 21:36:43 -0700
> Daniel Colascione wrote:
>
> > On Thu, Mar 14, 2019 at 8:16 PM Steven Rostedt wrote:
> > >
> > > On Thu, 14 Mar 2019 13:49:11 -0700
> > > Sultan Alsawaf wrote:
> > >
> > > > Perhaps I'm missing
On Thu, 14 Mar 2019 21:36:43 -0700
Daniel Colascione wrote:
> On Thu, Mar 14, 2019 at 8:16 PM Steven Rostedt wrote:
> >
> > On Thu, 14 Mar 2019 13:49:11 -0700
> > Sultan Alsawaf wrote:
> >
> > > Perhaps I'm missing something, but if you want to know when a process has
> > > died
> > > after
gt; > performance monitoring infrastructure as an alternative to a robust
> > and desperately-needed bit of core functionality that's neither hard
> > to add nor complex to implement nor expensive to use.
> >
> > Regarding the proposal for a new kernel-side lmkd: when
an abuse of the debugging and
> performance monitoring infrastructure as an alternative to a robust
> and desperately-needed bit of core functionality that's neither hard
> to add nor complex to implement nor expensive to use.
>
> Regarding the proposal for a new kernel-side lm
roposal for a new kernel-side lmkd: when possible, the
> kernel should provide mechanism, not policy. Putting the low memory
> killer back into the kernel after we've spent significant effort
> making it possible for userspace to do that job. Compared to kernel
> code, more easily under
kd: when possible, the
kernel should provide mechanism, not policy. Putting the low memory
killer back into the kernel after we've spent significant effort
making it possible for userspace to do that job. Compared to kernel
code, more easily understood, more easily debuggable, more easily
updat
On Thu, Mar 14, 2019 at 11:16:41PM -0400, Steven Rostedt wrote:
> How would you implement such a method in userspace? kill() doesn't take
> any parameters but the pid of the process you want to send a signal to,
> and the signal to send. This would require a new system call, and be
> quite a bit
On Thu, Mar 14, 2019 at 10:54:48PM -0400, Joel Fernandes wrote:
> I'm not sure if that makes much semantic sense for how the signal handling is
> supposed to work. Imagine a parent sends SIGKILL to its child, and then does
> a wait(2). Because the SIGKILL blocks in your idea, then the wait cannot
On Thu, 14 Mar 2019 13:49:11 -0700
Sultan Alsawaf wrote:
> Perhaps I'm missing something, but if you want to know when a process has died
> after sending a SIGKILL to it, then why not just make the SIGKILL optionally
> block until the process has died completely? It'd be rather trivial to just
>
On Thu, Mar 14, 2019 at 01:49:11PM -0700, Sultan Alsawaf wrote:
> On Thu, Mar 14, 2019 at 10:47:17AM -0700, Joel Fernandes wrote:
> > About the 100ms latency, I wonder whether it is that high because of
> > the way Android's lmkd is observing that a process has died. There is
> > a gap between
On Thu, Mar 14, 2019 at 10:47:17AM -0700, Joel Fernandes wrote:
> About the 100ms latency, I wonder whether it is that high because of
> the way Android's lmkd is observing that a process has died. There is
> a gap between when a process memory is freed and when it disappears
> from the
Hi Tim,
Thanks for the detailed and excellent write-up. It will serve as a
good future reference for low memory killer requirements. I made some
comments below on the "how to kill" part.
On Tue, Mar 12, 2019 at 10:17 AM Tim Murray wrote:
>
> On Tue, Mar 12, 2019 at 9:37 AM Sulta
On Tue, Mar 12, 2019 at 7:43 PM Tim Murray wrote:
>
> On Tue, Mar 12, 2019 at 10:45 AM Sultan Alsawaf
> wrote:
> >
> > On Tue, Mar 12, 2019 at 10:17:43AM -0700, Tim Murray wrote:
> > > Knowing whether a SIGKILL'd process has finished reclaiming is as far
> > > as I know not possible without
On Tue, Mar 12, 2019 at 10:45 AM Sultan Alsawaf wrote:
>
> On Tue, Mar 12, 2019 at 10:17:43AM -0700, Tim Murray wrote:
> > Knowing whether a SIGKILL'd process has finished reclaiming is as far
> > as I know not possible without something like procfds. That's where
> > the 100ms timeout in lmkd
On Tue, Mar 12, 2019 at 10:17:43AM -0700, Tim Murray wrote:
> Knowing whether a SIGKILL'd process has finished reclaiming is as far
> as I know not possible without something like procfds. That's where
> the 100ms timeout in lmkd comes in. lowmemorykiller and lmkd both
> attempt to wait up to
On Tue, Mar 12, 2019 at 9:58 AM Michal Hocko wrote:
>
> On Tue 12-03-19 09:37:41, Sultan Alsawaf wrote:
> > I have not had a chance to look at PSI yet, but
> > unless a PSI-enabled solution allows allocations to reach the same point as
> > when
> > the OOM killer is invoked (which is
On Tue, Mar 12, 2019 at 9:37 AM Sultan Alsawaf wrote:
>
> On Tue, Mar 12, 2019 at 09:05:32AM +0100, Michal Hocko wrote:
> > The only way to control the OOM behavior pro-actively is to throttle
> > allocation speed. We have memcg high limit for that purpose. Along with
> > PSI, I can imagine a
On Tue 12-03-19 09:37:41, Sultan Alsawaf wrote:
> I have not had a chance to look at PSI yet, but
> unless a PSI-enabled solution allows allocations to reach the same point as
> when
> the OOM killer is invoked (which is contradictory to what it sets out to do),
> then it cannot take advantage of
On Tue 12-03-19 09:37:41, Sultan Alsawaf wrote:
> On Tue, Mar 12, 2019 at 09:05:32AM +0100, Michal Hocko wrote:
> > The only way to control the OOM behavior pro-actively is to throttle
> > allocation speed. We have memcg high limit for that purpose. Along with
> > PSI, I can imagine a reasonably
On Tue, Mar 12, 2019 at 09:05:32AM +0100, Michal Hocko wrote:
> The only way to control the OOM behavior pro-actively is to throttle
> allocation speed. We have memcg high limit for that purpose. Along with
> PSI, I can imagine a reasonably working user space early oom
> notifications and
On Tue 12-03-19 16:33:15, Michal Hocko wrote:
> On Tue 12-03-19 08:25:41, Matthew Wilcox wrote:
> > On Tue, Mar 12, 2019 at 09:05:32AM +0100, Michal Hocko wrote:
> > > On Mon 11-03-19 15:15:35, Suren Baghdasaryan wrote:
> > > > Yeah, killing speed is a well-known problem which we are considering
>
On Tue 12-03-19 08:25:41, Matthew Wilcox wrote:
> On Tue, Mar 12, 2019 at 09:05:32AM +0100, Michal Hocko wrote:
> > On Mon 11-03-19 15:15:35, Suren Baghdasaryan wrote:
> > > Yeah, killing speed is a well-known problem which we are considering
> > > in LMKD. For example the recent LMKD change to
On Tue, Mar 12, 2019 at 09:05:32AM +0100, Michal Hocko wrote:
> On Mon 11-03-19 15:15:35, Suren Baghdasaryan wrote:
> > Yeah, killing speed is a well-known problem which we are considering
> > in LMKD. For example the recent LMKD change to assign process being
> > killed to a cpuset cgroup
On Tue, Mar 12, 2019 at 1:05 AM Michal Hocko wrote:
>
> On Mon 11-03-19 15:15:35, Suren Baghdasaryan wrote:
> > On Mon, Mar 11, 2019 at 1:46 PM Sultan Alsawaf
> > wrote:
> > >
> > > On Mon, Mar 11, 2019 at 01:10:36PM -0700, Suren Baghdasaryan wrote:
> > > > The idea seems interesting although I
On Mon 11-03-19 15:15:35, Suren Baghdasaryan wrote:
> On Mon, Mar 11, 2019 at 1:46 PM Sultan Alsawaf wrote:
> >
> > On Mon, Mar 11, 2019 at 01:10:36PM -0700, Suren Baghdasaryan wrote:
> > > The idea seems interesting although I need to think about this a bit
> > > more. Killing processes based on
On Mon, Mar 11, 2019 at 03:15:35PM -0700, Suren Baghdasaryan wrote:
> This what LMKD currently is - a userspace RT process.
> My point was that this page allocation queue that you implemented
> can't be implemented in userspace, at least not without extensive
> communication with kernel.
Oh,
On Mon, Mar 11, 2019 at 1:46 PM Sultan Alsawaf wrote:
>
> On Mon, Mar 11, 2019 at 01:10:36PM -0700, Suren Baghdasaryan wrote:
> > The idea seems interesting although I need to think about this a bit
> > more. Killing processes based on failed page allocation might backfire
> > during transient
On Mon, Mar 11, 2019 at 05:11:25PM -0400, Joel Fernandes wrote:
> But the point is that a transient temporary memory spike should not be a
> signal to kill _any_ process. The reaction to kill shouldn't be so
> spontaneous that unwanted tasks are killed because the system went into
> panic mode.
On Mon, Mar 11, 2019 at 01:46:26PM -0700, Sultan Alsawaf wrote:
> On Mon, Mar 11, 2019 at 01:10:36PM -0700, Suren Baghdasaryan wrote:
> > The idea seems interesting although I need to think about this a bit
> > more. Killing processes based on failed page allocation might backfire
> > during
On Mon, Mar 11, 2019 at 01:10:36PM -0700, Suren Baghdasaryan wrote:
> The idea seems interesting although I need to think about this a bit
> more. Killing processes based on failed page allocation might backfire
> during transient spikes in memory usage.
This issue could be alleviated if tasks
Hi Sultan,
On Mon, Mar 11, 2019 at 10:58 AM Sultan Alsawaf wrote:
>
> On Mon, Mar 11, 2019 at 06:43:20PM +0100, Michal Hocko wrote:
> > I am sorry but we are not going to maintain two different OOM
> > implementations in the kernel. From a quick look the implementation is
> > quite a hack which
On Mon, Mar 11, 2019 at 06:43:20PM +0100, Michal Hocko wrote:
> I am sorry but we are not going to maintain two different OOM
> implementations in the kernel. From a quick look the implementation is
> quite a hack which is not really suitable for anything but a very
> specific usecase. E.g.
On Sun 10-03-19 13:34:03, Sultan Alsawaf wrote:
> From: Sultan Alsawaf
>
> This is a complete low memory killer solution for Android that is small
> and simple. It kills the largest, least-important processes it can find
> whenever a page allocation has completely failed (rig
On Mon, Mar 11, 2019 at 12:32:33PM -0400, Joel Fernandes wrote:
> On Sun, Mar 10, 2019 at 01:34:03PM -0700, Sultan Alsawaf wrote:
> [...]
> >
> > /* Perform scheduler related setup. Assign this task to a CPU. */
> > retval = sched_fork(clone_flags, p);
> > diff --git a/mm/page_alloc.c
On Sun, Mar 10, 2019 at 01:34:03PM -0700, Sultan Alsawaf wrote:
[...]
>
> /* Perform scheduler related setup. Assign this task to a CPU. */
> retval = sched_fork(clone_flags, p);
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 3eb01dedf..fd0d697c6 100644
> ---
On Sun, Mar 10, 2019 at 10:03:35PM +0100, Greg Kroah-Hartman wrote:
> On Sun, Mar 10, 2019 at 01:34:03PM -0700, Sultan Alsawaf wrote:
> > From: Sultan Alsawaf
> >
> > This is a complete low memory killer solution for Android that is small
> > and simple. It kills t
On Sun, Mar 10, 2019 at 01:34:03PM -0700, Sultan Alsawaf wrote:
> From: Sultan Alsawaf
>
> This is a complete low memory killer solution for Android that is small
> and simple. It kills the largest, least-important processes it can find
> whenever a page allocation has completel
From: Sultan Alsawaf
This is a complete low memory killer solution for Android that is small
and simple. It kills the largest, least-important processes it can find
whenever a page allocation has completely failed (right after direct
reclaim). Processes are killed according to the priorities
object(unsigned long ptr, size_t size,
>>>>>> struct rb_node **link, *rb_parent;
>>>>>>
>>>>>> object = kmem_cache_alloc(object_cache, gfp_kmemleak_mask(gfp));
>>>>>> +#ifdef CONFIG_PREEMPT_COUNT
>>
ink, *rb_parent;
>>>>>
>>>>> object = kmem_cache_alloc(object_cache, gfp_kmemleak_mask(gfp));
>>>>> +#ifdef CONFIG_PREEMPT_COUNT
>>>>> + if (!object) {
>>>>> + /* last-ditch effort in a low-memory situation */
>>
101 - 200 of 637 matches
Mail list logo