linux/mm
Johannes Weiner 869712fd3d mm: memcontrol: fix network errors from failing __GFP_ATOMIC charges
While upgrading from 4.16 to 5.2, we noticed these allocation errors in
the log of the new kernel:

  SLUB: Unable to allocate memory on node -1, gfp=0xa20(GFP_ATOMIC)
    cache: tw_sock_TCPv6(960:helper-logs), object size: 232, buffer size: 240, default order: 1, min order: 0
    node 0: slabs: 5, objs: 170, free: 0

        slab_out_of_memory+1
        ___slab_alloc+969
        __slab_alloc+14
        kmem_cache_alloc+346
        inet_twsk_alloc+60
        tcp_time_wait+46
        tcp_fin+206
        tcp_data_queue+2034
        tcp_rcv_state_process+784
        tcp_v6_do_rcv+405
        __release_sock+118
        tcp_close+385
        inet_release+46
        __sock_release+55
        sock_close+17
        __fput+170
        task_work_run+127
        exit_to_usermode_loop+191
        do_syscall_64+212
        entry_SYSCALL_64_after_hwframe+68

accompanied by an increase in machines going completely radio silent
under memory pressure.

One thing that changed since 4.16 is e699e2c6a6 ("net, mm: account
sock objects to kmemcg"), which made these slab caches subject to cgroup
memory accounting and control.

The problem with that is that cgroups, unlike the page allocator, do not
maintain dedicated atomic reserves.  As a cgroup's usage hovers at its
limit, atomic allocations - such as done during network rx - can fail
consistently for extended periods of time.  The kernel is not able to
operate under these conditions.

We don't want to revert the culprit patch, because it indeed tracks a
potentially substantial amount of memory used by a cgroup.

We also don't want to implement dedicated atomic reserves for cgroups.
There is no point in keeping a fixed margin of unused bytes in the
cgroup's memory budget to accomodate a consumer that is impossible to
predict - we'd be wasting memory and get into configuration headaches,
not unlike what we have going with min_free_kbytes.  We do this for
physical mem because we have to, but cgroups are an accounting game.

Instead, account these privileged allocations to the cgroup, but let
them bypass the configured limit if they have to.  This way, we get the
benefits of accounting the consumed memory and have it exert pressure on
the rest of the cgroup, but like with the page allocator, we shift the
burden of reclaimining on behalf of atomic allocations onto the regular
allocations that can block.

Link: http://lkml.kernel.org/r/20191022233708.365764-1-hannes@cmpxchg.org
Fixes: e699e2c6a6 ("net, mm: account sock objects to kmemcg")
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Cc: Suleiman Souhlal <suleiman@google.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: <stable@vger.kernel.org>	[4.18+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-11-06 08:47:50 -08:00
..
kasan
Kconfig
Kconfig.debug
Makefile
backing-dev.c
balloon_compaction.c
cleancache.c
cma.c
cma.h
cma_debug.c
compaction.c mm, compaction: fix wrong pfn handling in __reset_isolation_pfn() 2019-10-14 15:04:01 -07:00
debug.c
debug_page_ref.c
dmapool.c
early_ioremap.c
fadvise.c
failslab.c
filemap.c mm/filemap.c: include <linux/ramfs.h> for generic_file_vm_ops definition 2019-10-19 06:32:32 -04:00
frame_vector.c
frontswap.c
gup.c mm/gup: fix a misnamed "write" argument, and a related bug 2019-10-19 06:32:32 -04:00
gup_benchmark.c
highmem.c
hmm.c
huge_memory.c mm/thp: fix node page state in split_huge_page_to_list() 2019-10-19 06:32:32 -04:00
hugetlb.c hugetlbfs: don't access uninitialized memmaps in pfn_range_valid_gigantic() 2019-10-19 06:32:32 -04:00
hugetlb_cgroup.c
hwpoison-inject.c
init-mm.c mm/init-mm.c: include <linux/mman.h> for vm_committed_as_batch 2019-10-19 06:32:32 -04:00
internal.h
interval_tree.c
khugepaged.c mm/khugepaged: fix might_sleep() warn with CONFIG_HIGHPTE=y 2019-11-06 08:47:50 -08:00
kmemleak-test.c
kmemleak.c kmemleak: Do not corrupt the object_list during clean-up 2019-10-14 08:56:16 -07:00
ksm.c
list_lru.c
maccess.c
madvise.c
memblock.c mm: memblock: do not enforce current limit for memblock_phys* family 2019-10-19 06:32:32 -04:00
memcontrol.c mm: memcontrol: fix network errors from failing __GFP_ATOMIC charges 2019-11-06 08:47:50 -08:00
memfd.c
memory-failure.c mm/memory-failure.c: don't access uninitialized memmaps in memory_failure() 2019-10-19 06:32:31 -04:00
memory.c
memory_hotplug.c mm/memory_hotplug: fix updating the node span 2019-11-06 08:47:50 -08:00
mempolicy.c
mempool.c
memremap.c mm/memunmap: don't access uninitialized memmap in memunmap_pages() 2019-10-19 06:32:32 -04:00
memtest.c
migrate.c
mincore.c
mlock.c
mm_init.c
mmap.c
mmu_context.c
mmu_gather.c
mmu_notifier.c mm/mmu_notifiers: use the right return code for WARN_ON 2019-11-06 08:47:50 -08:00
mmzone.c
mprotect.c
mremap.c
msync.c
nommu.c
oom_kill.c
page-writeback.c
page_alloc.c mm/page_alloc.c: ratelimit allocation failure warnings more aggressively 2019-11-06 08:47:50 -08:00
page_counter.c
page_ext.c mm, page_owner: fix off-by-one error in __set_page_owner_handle() 2019-10-14 15:04:00 -07:00
page_idle.c
page_io.c
page_isolation.c
page_owner.c mm/page_owner: don't access uninitialized memmaps when reading /proc/pagetypeinfo 2019-10-19 06:32:31 -04:00
page_poison.c
page_vma_mapped.c
pagewalk.c
percpu-internal.h
percpu-km.c
percpu-stats.c
percpu-vm.c
percpu.c
pgtable-generic.c
process_vm_access.c
readahead.c
rmap.c mm: include <linux/huge_mm.h> for is_vma_temporary_stack 2019-10-19 06:32:32 -04:00
rodata_test.c
shmem.c Merge branch 'work.mount3' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2019-10-10 08:16:44 -07:00
shuffle.c mm: fix -Wmissing-prototypes warnings 2019-10-07 15:47:19 -07:00
shuffle.h
slab.c mm/slab.c: fix kernel-doc warning for __ksize() 2019-10-14 15:04:01 -07:00
slab.h mm: slab: make page_cgroup_ino() to recognize non-compound slab pages properly 2019-11-06 08:47:50 -08:00
slab_common.c mm: memcg/slab: fix panic in __free_slab() caused by premature memcg pointer release 2019-10-19 06:32:32 -04:00
slob.c mm, sl[aou]b: guarantee natural alignment for kmalloc(power-of-two) 2019-10-07 15:47:20 -07:00
slub.c mm/slub.c: init_on_free=1 should wipe freelist ptr for bulk allocations 2019-10-14 15:04:01 -07:00
sparse-vmemmap.c
sparse.c mm: fix -Wmissing-prototypes warnings 2019-10-07 15:47:19 -07:00
swap.c
swap_cgroup.c
swap_slots.c
swap_state.c
swapfile.c
truncate.c mm/thp: allow dropping THP from page cache 2019-10-19 06:32:33 -04:00
usercopy.c
userfaultfd.c
util.c
vmacache.c
vmalloc.c
vmpressure.c mm/vmpressure.c: fix a signedness bug in vmpressure_register_event() 2019-10-07 15:47:19 -07:00
vmscan.c mm/vmscan.c: support removing arbitrary sized pages from mapping 2019-10-19 06:32:32 -04:00
vmstat.c mm, vmstat: reduce zone->lock holding time by /proc/pagetypeinfo 2019-11-06 08:47:50 -08:00
workingset.c
z3fold.c mm/z3fold.c: claim page in the beginning of free 2019-10-07 15:47:19 -07:00
zbud.c
zpool.c
zsmalloc.c
zswap.c