Currently memcg stats show several types of kernel memory:
kernel stack, page tables, sock, vmalloc, and slab.
However, there are other allocations with __GFP_ACCOUNT
(or supersets such as GFP_KERNEL_ACCOUNT) that are not accounted
in any of those stats, a few examples are:
- various kvm allocations (e.g. allocated pages to create vcpus)
- io_uring
- tmp_page in pipes during pipe_write()
- bpf ringbuffers
- unix sockets
Keeping track of the total kernel memory is essential for the ease of
migration from cgroup v1 to v2 as there are large discrepancies between
v1's kmem.usage_in_bytes and the sum of the available kernel memory stats
in v2. Adding separate memcg stats for all __GFP_ACCOUNT kernel
allocations is an impractical maintenance burden as there a lot of those
all over the kernel code, with more use cases likely to show up in the
future.
Therefore, add a "kernel" memcg stat that is analogous to kmem
page counter, with added benefits such as using rstat infrastructure
which aggregates stats more efficiently. Additionally, this provides a
lighter alternative in case the legacy kmem is deprecated in the future
Signed-off-by: Yosry Ahmed <[email protected]>
Acked-by: Shakeel Butt <[email protected]>
Acked-by: Johannes Weiner <[email protected]>
Reported-by: kernel test robot <[email protected]>
---
Changelog since v1:
- Moved "kernel" stat ahead of other subset kernel stats.
- Renamed mem_cgroup_kmem_record() to memcg_account_kmem(), following
Johannes's review to avoid the line wrap, but keeping a memcg_ prefix
to stay consistent with other static functions in the file.
- Fixed a build error when CONFIG_MEMCG_KMEM is not set (added an empty
version if the config is not set).
---
Documentation/admin-guide/cgroup-v2.rst | 5 +++++
include/linux/memcontrol.h | 1 +
mm/memcontrol.c | 27 +++++++++++++++++++------
3 files changed, 27 insertions(+), 6 deletions(-)
diff --git a/Documentation/admin-guide/cgroup-v2.rst b/Documentation/admin-guide/cgroup-v2.rst
index 5aa368d165da..69d7a6983f78 100644
--- a/Documentation/admin-guide/cgroup-v2.rst
+++ b/Documentation/admin-guide/cgroup-v2.rst
@@ -1301,6 +1301,11 @@ PAGE_SIZE multiple when read back.
Amount of memory used to cache filesystem data,
including tmpfs and shared memory.
+ kernel (npn)
+ Amount of total kernel memory, including
+ (kernel_stack, pagetables, percpu, vmalloc, slab) in
+ addition to other kernel memory use cases.
+
kernel_stack
Amount of memory allocated to kernel stacks.
diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index b72d75141e12..fa51986365a4 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -34,6 +34,7 @@ enum memcg_stat_item {
MEMCG_SOCK,
MEMCG_PERCPU_B,
MEMCG_VMALLOC,
+ MEMCG_KMEM,
MEMCG_NR_STAT,
};
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 09d342c7cbd0..c61dc03a3333 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -1371,6 +1371,7 @@ struct memory_stat {
static const struct memory_stat memory_stats[] = {
{ "anon", NR_ANON_MAPPED },
{ "file", NR_FILE_PAGES },
+ { "kernel", MEMCG_KMEM },
{ "kernel_stack", NR_KERNEL_STACK_KB },
{ "pagetables", NR_PAGETABLE },
{ "percpu", MEMCG_PERCPU_B },
@@ -2114,6 +2115,7 @@ static DEFINE_MUTEX(percpu_charge_mutex);
static void drain_obj_stock(struct obj_stock *stock);
static bool obj_stock_flush_required(struct memcg_stock_pcp *stock,
struct mem_cgroup *root_memcg);
+static void memcg_account_kmem(struct mem_cgroup *memcg, int nr_pages);
#else
static inline void drain_obj_stock(struct obj_stock *stock)
@@ -2124,6 +2126,9 @@ static bool obj_stock_flush_required(struct memcg_stock_pcp *stock,
{
return false;
}
+static void memcg_account_kmem(struct mem_cgroup *memcg, int nr_pages)
+{
+}
#endif
/**
@@ -2979,6 +2984,18 @@ static void memcg_free_cache_id(int id)
ida_simple_remove(&memcg_cache_ida, id);
}
+static void memcg_account_kmem(struct mem_cgroup *memcg, int nr_pages)
+{
+ mod_memcg_state(memcg, MEMCG_KMEM, nr_pages);
+ if (!cgroup_subsys_on_dfl(memory_cgrp_subsys)) {
+ if (nr_pages > 0)
+ page_counter_charge(&memcg->kmem, nr_pages);
+ else
+ page_counter_uncharge(&memcg->kmem, -nr_pages);
+ }
+}
+
+
/*
* obj_cgroup_uncharge_pages: uncharge a number of kernel pages from a objcg
* @objcg: object cgroup to uncharge
@@ -2991,8 +3008,7 @@ static void obj_cgroup_uncharge_pages(struct obj_cgroup *objcg,
memcg = get_mem_cgroup_from_objcg(objcg);
- if (!cgroup_subsys_on_dfl(memory_cgrp_subsys))
- page_counter_uncharge(&memcg->kmem, nr_pages);
+ memcg_account_kmem(memcg, -nr_pages);
refill_stock(memcg, nr_pages);
css_put(&memcg->css);
@@ -3018,8 +3034,7 @@ static int obj_cgroup_charge_pages(struct obj_cgroup *objcg, gfp_t gfp,
if (ret)
goto out;
- if (!cgroup_subsys_on_dfl(memory_cgrp_subsys))
- page_counter_charge(&memcg->kmem, nr_pages);
+ memcg_account_kmem(memcg, nr_pages);
out:
css_put(&memcg->css);
@@ -6801,8 +6816,8 @@ static void uncharge_batch(const struct uncharge_gather *ug)
page_counter_uncharge(&ug->memcg->memory, ug->nr_memory);
if (do_memsw_account())
page_counter_uncharge(&ug->memcg->memsw, ug->nr_memory);
- if (!cgroup_subsys_on_dfl(memory_cgrp_subsys) && ug->nr_kmem)
- page_counter_uncharge(&ug->memcg->kmem, ug->nr_kmem);
+ if (ug->nr_kmem)
+ memcg_account_kmem(ug->memcg, -ug->nr_kmem);
memcg_oom_recover(ug->memcg);
}
--
2.35.0.rc2.247.g8bbb082509-goog
On Thu, Feb 03, 2022 at 07:38:56PM +0000, Yosry Ahmed wrote:
> Currently memcg stats show several types of kernel memory:
> kernel stack, page tables, sock, vmalloc, and slab.
> However, there are other allocations with __GFP_ACCOUNT
> (or supersets such as GFP_KERNEL_ACCOUNT) that are not accounted
> in any of those stats, a few examples are:
> - various kvm allocations (e.g. allocated pages to create vcpus)
> - io_uring
> - tmp_page in pipes during pipe_write()
> - bpf ringbuffers
> - unix sockets
>
> Keeping track of the total kernel memory is essential for the ease of
> migration from cgroup v1 to v2 as there are large discrepancies between
> v1's kmem.usage_in_bytes and the sum of the available kernel memory stats
> in v2. Adding separate memcg stats for all __GFP_ACCOUNT kernel
> allocations is an impractical maintenance burden as there a lot of those
> all over the kernel code, with more use cases likely to show up in the
> future.
>
> Therefore, add a "kernel" memcg stat that is analogous to kmem
> page counter, with added benefits such as using rstat infrastructure
> which aggregates stats more efficiently. Additionally, this provides a
> lighter alternative in case the legacy kmem is deprecated in the future
>
> Signed-off-by: Yosry Ahmed <[email protected]>
> Acked-by: Shakeel Butt <[email protected]>
> Acked-by: Johannes Weiner <[email protected]>
> Reported-by: kernel test robot <[email protected]>
Looks good, thanks Yosry!