Implement cgroup2 "memory.max.effective" and "memory.swap.max.effective" files
similar to existing "cpuset.cpus.effective".
cgroup1 (by function memcg1_stat_format) contains two lines
hierarchical_memory_limit %llu
hierarchical_memsw_limit %llu
which are useful for userland to easily and performance-wise find out the
effective cgroup limits being applied. Otherwise userland has to
open+read+close the file "memory.max" and/or "memory.swap.max" in multiple
parent directories of a nested cgroup.
For cgroup1 it was implemented by:
memcg: show real limit under hierarchy mode
https://github.com/torvalds/linux/commit/fee7b548e6f2bd4bfd03a1a45d3afd593de7d5e9
Date: Wed Jan 7 18:08:26 2009 -0800
But for cgroup2 it has been missing so far.
v4:
Adjust the commit message.
v3:
memory.stat fields -> *.max.effective separate files suggested by Michal Koutny
v2:
hierarchical_memsw_limit -> hierarchical_swap_limit fix found by Waiman Long
v1:
hierarchical_memory_limit && hierarchical_memsw_limit in memory.stat
Signed-off-by: Jan Kratochvil (Azul) <[email protected]>
mm/memcontrol.c | 36 ++++++++++++++++++++++++++++++++++++
1 file changed, 36 insertions(+)
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 1ed40f9d3..8c4cb5f60 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -6845,6 +6845,19 @@ static ssize_t memory_max_write(struct kernfs_open_file *of,
return nbytes;
}
+static int memory_max_effective_show(struct seq_file *m, void *v)
+{
+ unsigned long memory;
+ struct mem_cgroup *mi;
+
+ /* Hierarchical information */
+ memory = PAGE_COUNTER_MAX;
+ for (mi = mem_cgroup_from_seq(m); mi; mi = parent_mem_cgroup(mi))
+ memory = min(memory, READ_ONCE(mi->memory.max));
+
+ return seq_puts_memcg_tunable(m, memory);
+}
+
/*
* Note: don't forget to update the 'samples/cgroup/memcg_event_listener'
* if any new events become available.
@@ -7038,6 +7051,11 @@ static struct cftype memory_files[] = {
.seq_show = memory_max_show,
.write = memory_max_write,
},
+ {
+ .name = "max.effective",
+ .flags = CFTYPE_NOT_ON_ROOT,
+ .seq_show = memory_max_effective_show,
+ },
{
.name = "events",
.flags = CFTYPE_NOT_ON_ROOT,
@@ -8040,6 +8058,19 @@ static ssize_t swap_max_write(struct kernfs_open_file *of,
return nbytes;
}
+static int swap_max_effective_show(struct seq_file *m, void *v)
+{
+ unsigned long swap;
+ struct mem_cgroup *mi;
+
+ /* Hierarchical information */
+ swap = PAGE_COUNTER_MAX;
+ for (mi = mem_cgroup_from_seq(m); mi; mi = parent_mem_cgroup(mi))
+ swap = min(swap, READ_ONCE(mi->swap.max));
+
+ return seq_puts_memcg_tunable(m, swap);
+}
+
static int swap_events_show(struct seq_file *m, void *v)
{
struct mem_cgroup *memcg = mem_cgroup_from_seq(m);
@@ -8072,6 +8103,11 @@ static struct cftype swap_files[] = {
.seq_show = swap_max_show,
.write = swap_max_write,
},
+ {
+ .name = "swap.max.effective",
+ .flags = CFTYPE_NOT_ON_ROOT,
+ .seq_show = swap_max_effective_show,
+ },
{
.name = "swap.peak",
.flags = CFTYPE_NOT_ON_ROOT,
Hello.
I have some advise about patch desription and only nitpick about
implementation.
On Wed, Feb 14, 2024 at 05:56:20AM +0800, "Jan Kratochvil (Azul)" <[email protected]> wrote:
> which are useful for userland to easily and performance-wise find out the
> effective cgroup limits being applied.
I think this is a weak reasoning for in-kernel implementation.
It may be faster but how often do you need to read that limit?
> Otherwise userland has to
Userland is at loss when running inside cgroupns with limited ancestors
out of the cgroupns. Thus I think this is the reason why kernel can
provide such an effective value. (And be subject line of the commit, I
think bringing up cgroup1 is misleading.)
> For cgroup1 it was implemented by:
> memcg: show real limit under hierarchy mode
> https://github.com/torvalds/linux/commit/fee7b548e6f2bd4bfd03a1a45d3afd593de7d5e9
> Date: Wed Jan 7 18:08:26 2009 -0800
FTR, commits are usually referenced more concisely like
fee7b548e6f2 ("memcg: show real limit under hierarchy mode")
(the document Waiman linked previously has some tips how to get this
from git).
> +static int memory_max_effective_show(struct seq_file *m, void *v)
> +{
> + unsigned long memory;
> + struct mem_cgroup *mi;
> +
> + /* Hierarchical information */
> + memory = PAGE_COUNTER_MAX;
> + for (mi = mem_cgroup_from_seq(m); mi; mi = parent_mem_cgroup(mi))
> + memory = min(memory, READ_ONCE(mi->memory.max));
root_mem_cgroup is never charged (thus limited), so you can terminate
the iteration on !mem_cgroup_is_root(mi), one level earlier
> +static int swap_max_effective_show(struct seq_file *m, void *v)
> +{
> + unsigned long swap;
> + struct mem_cgroup *mi;
> +
> + /* Hierarchical information */
> + swap = PAGE_COUNTER_MAX;
> + for (mi = mem_cgroup_from_seq(m); mi; mi = parent_mem_cgroup(mi))
> + swap = min(swap, READ_ONCE(mi->swap.max));
ditto
HTH,
Michal