| Message ID | 20220518223815.809858-1-vaibhav@linux.ibm.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | memcg: provide reclaim stats via 'memory.reclaim' | expand |
On Wed, May 18, 2022 at 3:38 PM Vaibhav Jain <vaibhav@linux.ibm.com> wrote: > > [1] Provides a way for user-space to trigger proactive reclaim by introducing > a write-only memcg file 'memory.reclaim'. However reclaim stats like number > of pages scanned and reclaimed is still not directly available to the > user-space. > > This patch proposes to extend [1] to make the memcg file 'memory.reclaim' > readable which returns the number of pages scanned / reclaimed during the > reclaim process from 'struct vmpressure' associated with each memcg. This should > let user-space asses how successful proactive reclaim triggered from memcg > 'memory.reclaim' was ? Isn't this a racy read? struct vmpressure can be changed between the write and read by other reclaim operations, right? I was actually planning to send a patch that does not updated vmpressure for user-controller reclaim, similar to how PSI is handled. The interface currently returns -EBUSY if the entire amount was not reclaimed, so isn't this enough to figure out if it was successful or not? If not, we can store the scanned / reclaim counts of the last memory.reclaim invocation for the sole purpose of memory.reclaim reads. Maybe it is actually more intuitive to users to just read the amount of memory read? In a format that is similar to the one written? i.e echo "10M" > memory.reclaim cat memory.reclaim 9M > > With the patch following command flow is expected: > > # echo "1M" > memory.reclaim > > # cat memory.reclaim > scanned 76 > reclaimed 32 > > [1]: https://lore.kernel.org/r/20220425190040.2475377-1-yosryahmed@google.com > > Cc: Shakeel Butt <shakeelb@google.com> > Cc: Yosry Ahmed <yosryahmed@google.com> > Signed-off-by: Vaibhav Jain <vaibhav@linux.ibm.com> > --- > Documentation/admin-guide/cgroup-v2.rst | 15 ++++++++++++--- > mm/memcontrol.c | 14 ++++++++++++++ > 2 files changed, 26 insertions(+), 3 deletions(-) > > diff --git a/Documentation/admin-guide/cgroup-v2.rst b/Documentation/admin-guide/cgroup-v2.rst > index 27ebef2485a3..44610165261d 100644 > --- a/Documentation/admin-guide/cgroup-v2.rst > +++ b/Documentation/admin-guide/cgroup-v2.rst > @@ -1209,18 +1209,27 @@ PAGE_SIZE multiple when read back. > utility is limited to providing the final safety net. > > memory.reclaim > - A write-only nested-keyed file which exists for all cgroups. > + A nested-keyed file which exists for all cgroups. > > - This is a simple interface to trigger memory reclaim in the > - target cgroup. > + This is a simple interface to trigger memory reclaim and retrieve > + reclaim stats in the target cgroup. > > This file accepts a single key, the number of bytes to reclaim. > No nested keys are currently supported. > > + Reading the file returns number of pages scanned and number of > + pages reclaimed from the memcg. This information fetched from > + vmpressure info associated with each cgroup. > + > Example:: > > echo "1G" > memory.reclaim > > + cat memory.reclaim > + > + scanned 78 > + reclaimed 30 > + > The interface can be later extended with nested keys to > configure the reclaim behavior. For example, specify the > type of memory to reclaim from (anon, file, ..). > diff --git a/mm/memcontrol.c b/mm/memcontrol.c > index 2e2bfbed4717..9e43580a8726 100644 > --- a/mm/memcontrol.c > +++ b/mm/memcontrol.c > @@ -6423,6 +6423,19 @@ static ssize_t memory_oom_group_write(struct kernfs_open_file *of, > return nbytes; > } > > +static int memory_reclaim_show(struct seq_file *m, void *v) > +{ > + struct mem_cgroup *memcg = mem_cgroup_from_seq(m); > + struct vmpressure *vmpr = memcg_to_vmpressure(memcg); > + > + spin_lock(&vmpr->sr_lock); > + seq_printf(m, "scanned %lu\nreclaimed %lu\n", > + vmpr->scanned, vmpr->reclaimed); > + spin_unlock(&vmpr->sr_lock); > + > + return 0; > +} > + > static ssize_t memory_reclaim(struct kernfs_open_file *of, char *buf, > size_t nbytes, loff_t off) > { > @@ -6525,6 +6538,7 @@ static struct cftype memory_files[] = { > .name = "reclaim", > .flags = CFTYPE_NS_DELEGATABLE, > .write = memory_reclaim, > + .seq_show = memory_reclaim_show, > }, > { } /* terminate */ > }; > -- > 2.35.1 >
On Thu, May 19, 2022 at 04:08:15AM +0530, Vaibhav Jain wrote: > [1] Provides a way for user-space to trigger proactive reclaim by introducing > a write-only memcg file 'memory.reclaim'. However reclaim stats like number > of pages scanned and reclaimed is still not directly available to the > user-space. > > This patch proposes to extend [1] to make the memcg file 'memory.reclaim' > readable which returns the number of pages scanned / reclaimed during the > reclaim process from 'struct vmpressure' associated with each memcg. This should > let user-space asses how successful proactive reclaim triggered from memcg > 'memory.reclaim' was ? > > With the patch following command flow is expected: > > # echo "1M" > memory.reclaim > > # cat memory.reclaim > scanned 76 > reclaimed 32 > Yosry already mentioned the race issue with the implementation and I would prefer we don't create any new dependency on vmpressure which I think we should deprecate. Anyways my question is how are you planning to use these metrics i.e. scanned & reclaimed? I wonder if the data you are interested in can be extracted without a stable interface. Have you tried BPF way to get these metrics? We already have a tracepoint in vmscan tracing the scanned and reclaimed.
Vaibhav Jain <vaibhav@linux.ibm.com> wrote: > [1] Provides a way for user-space to trigger proactive reclaim by introducing > a write-only memcg file 'memory.reclaim'. However reclaim stats like number > of pages scanned and reclaimed is still not directly available to the > user-space. > > This patch proposes to extend [1] to make the memcg file 'memory.reclaim' > readable which returns the number of pages scanned / reclaimed during the > reclaim process from 'struct vmpressure' associated with each memcg. This should > let user-space asses how successful proactive reclaim triggered from memcg > 'memory.reclaim' was ? > > With the patch following command flow is expected: > > # echo "1M" > memory.reclaim > > # cat memory.reclaim > scanned 76 > reclaimed 32 I certainly appreciate the ability for shell scripts to demonstrate cgroup operations with textual interfaces, but such interface seem like they are optimized for ease of use by developers. I wonder if for runtime production use an ioctl or netlink interface has been considered for cgroup? I don't think there are any yet, but such approaches seem like a more straightforward ways to get nontrivial input/outputs from a single call (e.g. like this proposal). And they have the benefit of not requiring ascii serialization/parsing overhead. > [1]: https://lore.kernel.org/r/20220425190040.2475377-1-yosryahmed@google.com > > Cc: Shakeel Butt <shakeelb@google.com> > Cc: Yosry Ahmed <yosryahmed@google.com> > Signed-off-by: Vaibhav Jain <vaibhav@linux.ibm.com> > --- > Documentation/admin-guide/cgroup-v2.rst | 15 ++++++++++++--- > mm/memcontrol.c | 14 ++++++++++++++ > 2 files changed, 26 insertions(+), 3 deletions(-) > > diff --git a/Documentation/admin-guide/cgroup-v2.rst b/Documentation/admin-guide/cgroup-v2.rst > index 27ebef2485a3..44610165261d 100644 > --- a/Documentation/admin-guide/cgroup-v2.rst > +++ b/Documentation/admin-guide/cgroup-v2.rst > @@ -1209,18 +1209,27 @@ PAGE_SIZE multiple when read back. > utility is limited to providing the final safety net. > > memory.reclaim > - A write-only nested-keyed file which exists for all cgroups. > + A nested-keyed file which exists for all cgroups. > > - This is a simple interface to trigger memory reclaim in the > - target cgroup. > + This is a simple interface to trigger memory reclaim and retrieve > + reclaim stats in the target cgroup. > > This file accepts a single key, the number of bytes to reclaim. > No nested keys are currently supported. > > + Reading the file returns number of pages scanned and number of > + pages reclaimed from the memcg. This information fetched from > + vmpressure info associated with each cgroup. > + > Example:: > > echo "1G" > memory.reclaim > > + cat memory.reclaim > + > + scanned 78 > + reclaimed 30 > + > The interface can be later extended with nested keys to > configure the reclaim behavior. For example, specify the > type of memory to reclaim from (anon, file, ..). > diff --git a/mm/memcontrol.c b/mm/memcontrol.c > index 2e2bfbed4717..9e43580a8726 100644 > --- a/mm/memcontrol.c > +++ b/mm/memcontrol.c > @@ -6423,6 +6423,19 @@ static ssize_t memory_oom_group_write(struct kernfs_open_file *of, > return nbytes; > } > > +static int memory_reclaim_show(struct seq_file *m, void *v) > +{ > + struct mem_cgroup *memcg = mem_cgroup_from_seq(m); > + struct vmpressure *vmpr = memcg_to_vmpressure(memcg); > + > + spin_lock(&vmpr->sr_lock); > + seq_printf(m, "scanned %lu\nreclaimed %lu\n", > + vmpr->scanned, vmpr->reclaimed); > + spin_unlock(&vmpr->sr_lock); > + > + return 0; > +} > + > static ssize_t memory_reclaim(struct kernfs_open_file *of, char *buf, > size_t nbytes, loff_t off) > { > @@ -6525,6 +6538,7 @@ static struct cftype memory_files[] = { > .name = "reclaim", > .flags = CFTYPE_NS_DELEGATABLE, > .write = memory_reclaim, > + .seq_show = memory_reclaim_show, > }, > { } /* terminate */ > };
Hi, Thanks for looking into this patch, Yosry Ahmed <yosryahmed@google.com> writes: > On Wed, May 18, 2022 at 3:38 PM Vaibhav Jain <vaibhav@linux.ibm.com> wrote: >> >> [1] Provides a way for user-space to trigger proactive reclaim by introducing >> a write-only memcg file 'memory.reclaim'. However reclaim stats like number >> of pages scanned and reclaimed is still not directly available to the >> user-space. >> >> This patch proposes to extend [1] to make the memcg file 'memory.reclaim' >> readable which returns the number of pages scanned / reclaimed during the >> reclaim process from 'struct vmpressure' associated with each memcg. This should >> let user-space asses how successful proactive reclaim triggered from memcg >> 'memory.reclaim' was ? > > Isn't this a racy read? struct vmpressure can be changed between the > write and read by other reclaim operations, right? Read/write of vmpr stats is always done in context of vmpr->sr_lock which is also the case for this patch. So not sure how the read is racy ?. > > I was actually planning to send a patch that does not updated > vmpressure for user-controller reclaim, similar to how PSI is handled. > Ok, not sure if I am inferring correctly as to how how that would be useful. Can you please provide some more context. The primary motivation for this patch was to expose the vmpressure stats to user space that are available with cgroup-v1 but not with cgroup-v2 AFAIK > The interface currently returns -EBUSY if the entire amount was not > reclaimed, so isn't this enough to figure out if it was successful or > not? Userspace may very well want to know the amount of memory that was partially reclaimed even though write to "memory.reclaim" returned '-EBUSY'. This feedback can be useful info for implementing a retry loop. > If not, we can store the scanned / reclaim counts of the last > memory.reclaim invocation for the sole purpose of memory.reclaim > reads. Sure sounds reasonable to me. > Maybe it is actually more intuitive to users to just read the > amount of memory read? In a format that is similar to the one written? > > i.e > echo "10M" > memory.reclaim > cat memory.reclaim > 9M > Agree, I will address that in v2. <snip>
Hi, Thanks for looking into this patch, Shakeel Butt <shakeelb@google.com> writes: > On Thu, May 19, 2022 at 04:08:15AM +0530, Vaibhav Jain wrote: >> [1] Provides a way for user-space to trigger proactive reclaim by introducing >> a write-only memcg file 'memory.reclaim'. However reclaim stats like number >> of pages scanned and reclaimed is still not directly available to the >> user-space. >> >> This patch proposes to extend [1] to make the memcg file 'memory.reclaim' >> readable which returns the number of pages scanned / reclaimed during the >> reclaim process from 'struct vmpressure' associated with each memcg. This should >> let user-space asses how successful proactive reclaim triggered from memcg >> 'memory.reclaim' was ? >> >> With the patch following command flow is expected: >> >> # echo "1M" > memory.reclaim >> >> # cat memory.reclaim >> scanned 76 >> reclaimed 32 >> > > Yosry already mentioned the race issue with the implementation and I > would prefer we don't create any new dependency on vmpressure which I > think we should deprecate. Ok, > > Anyways my question is how are you planning to use these metrics i.e. > scanned & reclaimed? I wonder if the data you are interested in can be > extracted without a stable interface. Have you tried BPF way to get > these metrics? We already have a tracepoint in vmscan tracing the > scanned and reclaimed. > Agree that there are enough static trace_mm_vmscan_ tracepoints in vmscan to get that info. Also agree that exposing nr_scanned/nr_reclaimed directly to userspace may not be a good idea but knowing the amount of memory reclaimed might be useful. With user-space triggered proactive reclaim user-space code can try to write a certain value to "memory.reclaim" in a loop till it returns '-EBUSY'. Right now there is no direct way for it to get feedback on the progress of the requested reclaim. Providing a stable interface to ascertain the progress of reclaim lets that userspace provide smaller values for proactive reclaim
Hi, Thanks for looking into this patch, Greg Thelen <gthelen@google.com> writes: > Vaibhav Jain <vaibhav@linux.ibm.com> wrote: > >> [1] Provides a way for user-space to trigger proactive reclaim by introducing >> a write-only memcg file 'memory.reclaim'. However reclaim stats like number >> of pages scanned and reclaimed is still not directly available to the >> user-space. >> >> This patch proposes to extend [1] to make the memcg file 'memory.reclaim' >> readable which returns the number of pages scanned / reclaimed during the >> reclaim process from 'struct vmpressure' associated with each memcg. This should >> let user-space asses how successful proactive reclaim triggered from memcg >> 'memory.reclaim' was ? >> >> With the patch following command flow is expected: >> >> # echo "1M" > memory.reclaim >> >> # cat memory.reclaim >> scanned 76 >> reclaimed 32 > > I certainly appreciate the ability for shell scripts to demonstrate > cgroup operations with textual interfaces, but such interface seem like > they are optimized for ease of use by developers. > Agree that directly exposing nr_scanned/reclaimed might not be a useful for users and certainly looks like a dev interface > I wonder if for runtime production use an ioctl or netlink interface has > been considered for cgroup? I don't think there are any yet, but such > approaches seem like a more straightforward ways to get nontrivial > input/outputs from a single call (e.g. like this proposal). And they > have the benefit of not requiring ascii serialization/parsing overhead. I think to a large degree eBPF and existing static tracepoints in vmscan can provide access to these metrics as Shakeel Bhat pointed to earlier. <snip>
On Thu 19-05-22 04:08:15, Vaibhav Jain wrote: > [1] Provides a way for user-space to trigger proactive reclaim by introducing > a write-only memcg file 'memory.reclaim'. However reclaim stats like number > of pages scanned and reclaimed is still not directly available to the > user-space. > > This patch proposes to extend [1] to make the memcg file 'memory.reclaim' > readable which returns the number of pages scanned / reclaimed during the > reclaim process from 'struct vmpressure' associated with each memcg. This should > let user-space asses how successful proactive reclaim triggered from memcg > 'memory.reclaim' was ? > > With the patch following command flow is expected: > > # echo "1M" > memory.reclaim > > # cat memory.reclaim > scanned 76 > reclaimed 32 Why cannot you use memory.stat? Sure it would require to iterate over the reclaimed hierarchy but the information about scanned and reclaimed pages as well as other potentially useful stats is there.
On Thu, May 19, 2022 at 1:51 AM Vaibhav Jain <vaibhav@linux.ibm.com> wrote: > > Hi, > > Thanks for looking into this patch, > > Yosry Ahmed <yosryahmed@google.com> writes: > > > On Wed, May 18, 2022 at 3:38 PM Vaibhav Jain <vaibhav@linux.ibm.com> wrote: > >> > >> [1] Provides a way for user-space to trigger proactive reclaim by introducing > >> a write-only memcg file 'memory.reclaim'. However reclaim stats like number > >> of pages scanned and reclaimed is still not directly available to the > >> user-space. > >> > >> This patch proposes to extend [1] to make the memcg file 'memory.reclaim' > >> readable which returns the number of pages scanned / reclaimed during the > >> reclaim process from 'struct vmpressure' associated with each memcg. This should > >> let user-space asses how successful proactive reclaim triggered from memcg > >> 'memory.reclaim' was ? > > > > Isn't this a racy read? struct vmpressure can be changed between the > > write and read by other reclaim operations, right? > Read/write of vmpr stats is always done in context of vmpr->sr_lock > which is also the case for this patch. So not sure how the read is racy > ?. I didn't mean that you can read the value while it is being changed. I meant that between writing to memory.reclaim and reading from it, another reclaim operation could modify memcg vmpressure. A sequence like this: 1) Write to memory.reclaim 2) Kernel coincidentally runs reclaim on that memcg 3) Read from memory.reclaim The result would be that you are reading the stats of another reclaim operation, not the one invoked by writing to memory.reclaim. > > > > > I was actually planning to send a patch that does not updated > > vmpressure for user-controller reclaim, similar to how PSI is handled. > > > Ok, not sure if I am inferring correctly as to how how that would be > useful. Can you please provide some more context. IIUC vmpressure is used as an indicator for memory pressure. In my opinion it makes sense if vmpressure is not changed on reclaim operations directly invoked by the user, as they are not directly related to whether the system is under memory pressure or not. PSI is handled in a similar way. See e22c6ed90aa9 ("mm: memcontrol: don't count limit-setting reclaim as memory pressure"). > > The primary motivation for this patch was to expose the vmpressure stats > to user space that are available with cgroup-v1 but not with cgroup-v2 > AFAIK If the main goal is exposing vmpressure, regardless of proactive reclaim, this is something else. AFAIK vmpressure is not popular anymore and PSI is the more recent/better indicator. > > > The interface currently returns -EBUSY if the entire amount was not > > reclaimed, so isn't this enough to figure out if it was successful or > > not? > Userspace may very well want to know the amount of memory that was > partially reclaimed even though write to "memory.reclaim" returned > '-EBUSY'. This feedback can be useful info for implementing a retry > loop. > > > If not, we can store the scanned / reclaim counts of the last > > memory.reclaim invocation for the sole purpose of memory.reclaim > > reads. > Sure sounds reasonable to me. > > > Maybe it is actually more intuitive to users to just read the > > amount of memory read? In a format that is similar to the one written? > > > > i.e > > echo "10M" > memory.reclaim > > cat memory.reclaim > > 9M > > > Agree, I will address that in v2. > > <snip> > > -- > Cheers > ~ Vaibhav
Thanks for looking into this patch Michal, Michal Hocko <mhocko@suse.com> writes: > On Thu 19-05-22 04:08:15, Vaibhav Jain wrote: >> [1] Provides a way for user-space to trigger proactive reclaim by introducing >> a write-only memcg file 'memory.reclaim'. However reclaim stats like number >> of pages scanned and reclaimed is still not directly available to the >> user-space. >> >> This patch proposes to extend [1] to make the memcg file 'memory.reclaim' >> readable which returns the number of pages scanned / reclaimed during the >> reclaim process from 'struct vmpressure' associated with each memcg. This should >> let user-space asses how successful proactive reclaim triggered from memcg >> 'memory.reclaim' was ? >> >> With the patch following command flow is expected: >> >> # echo "1M" > memory.reclaim >> >> # cat memory.reclaim >> scanned 76 >> reclaimed 32 > > Why cannot you use memory.stat? Sure it would require to iterate over > the reclaimed hierarchy but the information about scanned and reclaimed > pages as well as other potentially useful stats is there. Agree that "memory.stat" is more suitable for scanned/reclaimed stats as it already is exposing bunch of other stats. The discussion on this patch however seems to have split into two parts: 1. Is it a good idea to expose nr_scanned/nr_reclaimed to users-space and if yes how ? IMHO, I think it will be better to expose this info via 'memory.stat' as it can be useful insight into the reclaim efficiency and vmpressure. 2. Will it be useful to provide feedback to userspace when it writes to 'memory.reclaim' on how much memory has been reclaimed ? IMHO, this will be a useful feeback to userspace to better adjust future proactive reclaim requests via 'memory.reclaim'
On Fri 20-05-22 10:45:43, Vaibhav Jain wrote: > > Thanks for looking into this patch Michal, > > Michal Hocko <mhocko@suse.com> writes: > > > On Thu 19-05-22 04:08:15, Vaibhav Jain wrote: > >> [1] Provides a way for user-space to trigger proactive reclaim by introducing > >> a write-only memcg file 'memory.reclaim'. However reclaim stats like number > >> of pages scanned and reclaimed is still not directly available to the > >> user-space. > >> > >> This patch proposes to extend [1] to make the memcg file 'memory.reclaim' > >> readable which returns the number of pages scanned / reclaimed during the > >> reclaim process from 'struct vmpressure' associated with each memcg. This should > >> let user-space asses how successful proactive reclaim triggered from memcg > >> 'memory.reclaim' was ? > >> > >> With the patch following command flow is expected: > >> > >> # echo "1M" > memory.reclaim > >> > >> # cat memory.reclaim > >> scanned 76 > >> reclaimed 32 > > > > Why cannot you use memory.stat? Sure it would require to iterate over > > the reclaimed hierarchy but the information about scanned and reclaimed > > pages as well as other potentially useful stats is there. > > Agree that "memory.stat" is more suitable for scanned/reclaimed stats as > it already is exposing bunch of other stats. > > The discussion on this patch however seems to have split into two parts: > > 1. Is it a good idea to expose nr_scanned/nr_reclaimed to users-space > and if yes how ? > > IMHO, I think it will be better to expose this info via 'memory.stat' as it > can be useful insight into the reclaim efficiency and vmpressure. We already do that with some more metrics pgrefill 9801926 pgscan 27329762 pgsteal 22715987 pgactivate 250691267 pgdeactivate 9521843 pglazyfree 0 pglazyfreed 0 > 2. Will it be useful to provide feedback to userspace when it writes to > 'memory.reclaim' on how much memory has been reclaimed ? > > IMHO, this will be a useful feeback to userspace to better adjust future > proactive reclaim requests via 'memory.reclaim' How precise this information should be? A very simplistic approach would be cp memory.stat stats.before echo $WHATEVER > memory.reclaim cp memory.stat stats.after This will obviously contain also activity outside of the explicitly triggered reclaim (racing background/direct reclaim) but isn't that what actually matters? Are there any cases where the only metric you care about is the triggered reclaim in isolation?
On Fri, May 20, 2022 at 12:29 AM Michal Hocko <mhocko@suse.com> wrote: > > On Fri 20-05-22 10:45:43, Vaibhav Jain wrote: > > > > Thanks for looking into this patch Michal, > > > > Michal Hocko <mhocko@suse.com> writes: > > > > > On Thu 19-05-22 04:08:15, Vaibhav Jain wrote: > > >> [1] Provides a way for user-space to trigger proactive reclaim by introducing > > >> a write-only memcg file 'memory.reclaim'. However reclaim stats like number > > >> of pages scanned and reclaimed is still not directly available to the > > >> user-space. > > >> > > >> This patch proposes to extend [1] to make the memcg file 'memory.reclaim' > > >> readable which returns the number of pages scanned / reclaimed during the > > >> reclaim process from 'struct vmpressure' associated with each memcg. This should > > >> let user-space asses how successful proactive reclaim triggered from memcg > > >> 'memory.reclaim' was ? > > >> > > >> With the patch following command flow is expected: > > >> > > >> # echo "1M" > memory.reclaim > > >> > > >> # cat memory.reclaim > > >> scanned 76 > > >> reclaimed 32 > > > > > > Why cannot you use memory.stat? Sure it would require to iterate over > > > the reclaimed hierarchy but the information about scanned and reclaimed > > > pages as well as other potentially useful stats is there. > > > > Agree that "memory.stat" is more suitable for scanned/reclaimed stats as > > it already is exposing bunch of other stats. > > > > The discussion on this patch however seems to have split into two parts: > > > > 1. Is it a good idea to expose nr_scanned/nr_reclaimed to users-space > > and if yes how ? > > > > IMHO, I think it will be better to expose this info via 'memory.stat' as it > > can be useful insight into the reclaim efficiency and vmpressure. > > We already do that with some more metrics > pgrefill 9801926 > pgscan 27329762 > pgsteal 22715987 > pgactivate 250691267 > pgdeactivate 9521843 > pglazyfree 0 > pglazyfreed 0 > > > 2. Will it be useful to provide feedback to userspace when it writes to > > 'memory.reclaim' on how much memory has been reclaimed ? > > > > IMHO, this will be a useful feeback to userspace to better adjust future > > proactive reclaim requests via 'memory.reclaim' > > How precise this information should be? A very simplistic approach would > be > cp memory.stat stats.before > echo $WHATEVER > memory.reclaim > cp memory.stat stats.after > > This will obviously contain also activity outside of the explicitly > triggered reclaim (racing background/direct reclaim) but isn't that what > actually matters? Are there any cases where the only metric you care > about is the triggered reclaim in isolation? I think it might be useful to have a dedicated entry in memory.stat for proactively reclaimed memory. A case where this would be useful is tuning and evaluating userspace proactive reclaimers. For instance, if a userspace agent is asking the kernel to reclaim 100M, but it could only reclaim 10M, then most probably the proactive reclaimer is not using a good methodology to figure out how much memory do we need to reclaim. IMO this is more useful, and a superset of just reading the last reclaim request status through memory.reclaim (read stat before and after). Additionally, things get complicated if the userspace agent is multi-threaded. For a cumulative entry in memory.stat, it shouldn't matter by a lot as we are looking at the total for all threads cumulatively anyway. If we are only reading the memory reclaimed in the last request (through memory.reclaim), then we can easily get the results of a request that happened on a different thread. > > -- > Michal Hocko > SUSE Labs
On Mon, May 23, 2022 at 03:50:34PM -0700, Yosry Ahmed wrote: > I think it might be useful to have a dedicated entry in memory.stat > for proactively reclaimed memory. A case where this would be useful is > tuning and evaluating userspace proactive reclaimers. For instance, if > a userspace agent is asking the kernel to reclaim 100M, but it could > only reclaim 10M, then most probably the proactive reclaimer is not > using a good methodology to figure out how much memory do we need to > reclaim. > > IMO this is more useful, and a superset of just reading the last > reclaim request status through memory.reclaim (read stat before and > after). +1
On Tue, May 24, 2022 at 4:45 AM Johannes Weiner <hannes@cmpxchg.org> wrote: > > On Mon, May 23, 2022 at 03:50:34PM -0700, Yosry Ahmed wrote: > > I think it might be useful to have a dedicated entry in memory.stat > > for proactively reclaimed memory. A case where this would be useful is > > tuning and evaluating userspace proactive reclaimers. For instance, if > > a userspace agent is asking the kernel to reclaim 100M, but it could > > only reclaim 10M, then most probably the proactive reclaimer is not > > using a good methodology to figure out how much memory do we need to > > reclaim. > > > > IMO this is more useful, and a superset of just reading the last > > reclaim request status through memory.reclaim (read stat before and > > after). > > +1 It might also be useful to have a breakdown of this by memory type: file, anon, or shrinkers. It would also fit in nicely with a potential type=file/anon/shrinker argument to memory.reclaim. Thoughts on this?
On Tue 24-05-22 12:01:01, Yosry Ahmed wrote: > On Tue, May 24, 2022 at 4:45 AM Johannes Weiner <hannes@cmpxchg.org> wrote: > > > > On Mon, May 23, 2022 at 03:50:34PM -0700, Yosry Ahmed wrote: > > > I think it might be useful to have a dedicated entry in memory.stat > > > for proactively reclaimed memory. A case where this would be useful is > > > tuning and evaluating userspace proactive reclaimers. For instance, if > > > a userspace agent is asking the kernel to reclaim 100M, but it could > > > only reclaim 10M, then most probably the proactive reclaimer is not > > > using a good methodology to figure out how much memory do we need to > > > reclaim. > > > > > > IMO this is more useful, and a superset of just reading the last > > > reclaim request status through memory.reclaim (read stat before and > > > after). > > > > +1 > > It might also be useful to have a breakdown of this by memory type: > file, anon, or shrinkers. > > It would also fit in nicely with a potential type=file/anon/shrinker > argument to memory.reclaim. Thoughts on this? Can we start simple and see what real usecases actually will need?
On Wed, May 25, 2022 at 1:59 AM Michal Hocko <mhocko@suse.com> wrote: > > On Tue 24-05-22 12:01:01, Yosry Ahmed wrote: > > On Tue, May 24, 2022 at 4:45 AM Johannes Weiner <hannes@cmpxchg.org> wrote: > > > > > > On Mon, May 23, 2022 at 03:50:34PM -0700, Yosry Ahmed wrote: > > > > I think it might be useful to have a dedicated entry in memory.stat > > > > for proactively reclaimed memory. A case where this would be useful is > > > > tuning and evaluating userspace proactive reclaimers. For instance, if > > > > a userspace agent is asking the kernel to reclaim 100M, but it could > > > > only reclaim 10M, then most probably the proactive reclaimer is not > > > > using a good methodology to figure out how much memory do we need to > > > > reclaim. > > > > > > > > IMO this is more useful, and a superset of just reading the last > > > > reclaim request status through memory.reclaim (read stat before and > > > > after). > > > > > > +1 > > > > It might also be useful to have a breakdown of this by memory type: > > file, anon, or shrinkers. > > > > It would also fit in nicely with a potential type=file/anon/shrinker > > argument to memory.reclaim. Thoughts on this? > > Can we start simple and see what real usecases actually will need? Agreed. Let's start with a single proactively reclaimed memory stat and then add subcategories if/when needed. > -- > Michal Hocko > SUSE Labs
diff --git a/Documentation/admin-guide/cgroup-v2.rst b/Documentation/admin-guide/cgroup-v2.rst index 27ebef2485a3..44610165261d 100644 --- a/Documentation/admin-guide/cgroup-v2.rst +++ b/Documentation/admin-guide/cgroup-v2.rst @@ -1209,18 +1209,27 @@ PAGE_SIZE multiple when read back. utility is limited to providing the final safety net. memory.reclaim - A write-only nested-keyed file which exists for all cgroups. + A nested-keyed file which exists for all cgroups. - This is a simple interface to trigger memory reclaim in the - target cgroup. + This is a simple interface to trigger memory reclaim and retrieve + reclaim stats in the target cgroup. This file accepts a single key, the number of bytes to reclaim. No nested keys are currently supported. + Reading the file returns number of pages scanned and number of + pages reclaimed from the memcg. This information fetched from + vmpressure info associated with each cgroup. + Example:: echo "1G" > memory.reclaim + cat memory.reclaim + + scanned 78 + reclaimed 30 + The interface can be later extended with nested keys to configure the reclaim behavior. For example, specify the type of memory to reclaim from (anon, file, ..). diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 2e2bfbed4717..9e43580a8726 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -6423,6 +6423,19 @@ static ssize_t memory_oom_group_write(struct kernfs_open_file *of, return nbytes; } +static int memory_reclaim_show(struct seq_file *m, void *v) +{ + struct mem_cgroup *memcg = mem_cgroup_from_seq(m); + struct vmpressure *vmpr = memcg_to_vmpressure(memcg); + + spin_lock(&vmpr->sr_lock); + seq_printf(m, "scanned %lu\nreclaimed %lu\n", + vmpr->scanned, vmpr->reclaimed); + spin_unlock(&vmpr->sr_lock); + + return 0; +} + static ssize_t memory_reclaim(struct kernfs_open_file *of, char *buf, size_t nbytes, loff_t off) { @@ -6525,6 +6538,7 @@ static struct cftype memory_files[] = { .name = "reclaim", .flags = CFTYPE_NS_DELEGATABLE, .write = memory_reclaim, + .seq_show = memory_reclaim_show, }, { } /* terminate */ };
[1] Provides a way for user-space to trigger proactive reclaim by introducing a write-only memcg file 'memory.reclaim'. However reclaim stats like number of pages scanned and reclaimed is still not directly available to the user-space. This patch proposes to extend [1] to make the memcg file 'memory.reclaim' readable which returns the number of pages scanned / reclaimed during the reclaim process from 'struct vmpressure' associated with each memcg. This should let user-space asses how successful proactive reclaim triggered from memcg 'memory.reclaim' was ? With the patch following command flow is expected: # echo "1M" > memory.reclaim # cat memory.reclaim scanned 76 reclaimed 32 [1]: https://lore.kernel.org/r/20220425190040.2475377-1-yosryahmed@google.com Cc: Shakeel Butt <shakeelb@google.com> Cc: Yosry Ahmed <yosryahmed@google.com> Signed-off-by: Vaibhav Jain <vaibhav@linux.ibm.com> --- Documentation/admin-guide/cgroup-v2.rst | 15 ++++++++++++--- mm/memcontrol.c | 14 ++++++++++++++ 2 files changed, 26 insertions(+), 3 deletions(-)