On Mon, Aug 26, 2024 at 04:29:08PM -0700, Shakeel Butt wrote: > At the moment, the slab objects are charged to the memcg at the > allocation time. However there are cases where slab objects are > allocated at the time where the right target memcg to charge it to is > not known. One such case is the network sockets for the incoming > connection which are allocated in the softirq context. > > Couple hundred thousand connections are very normal on large loaded > server and almost all of those sockets underlying those connections get > allocated in the softirq context and thus not charged to any memcg. > However later at the accept() time we know the right target memcg to > charge. Let's add new API to charge already allocated objects, so we can > have better accounting of the memory usage. > > To measure the performance impact of this change, tcp_crr is used from > the neper [1] performance suite. Basically it is a network ping pong > test with new connection for each ping pong. > > The server and the client are run inside 3 level of cgroup hierarchy > using the following commands: > > Server: > $ tcp_crr -6 > > Client: > $ tcp_crr -6 -c -H ${server_ip} > > If the client and server run on different machines with 50 GBPS NIC, > there is no visible impact of the change. > > For the same machine experiment with v6.11-rc5 as base. > > base (throughput) with-patch > tcp_crr 14545 (+- 80) 14463 (+- 56) > > It seems like the performance impact is within the noise. > > Link: https://github.com/google/neper [1] > Signed-off-by: Shakeel Butt <shakeel.butt@xxxxxxxxx> Hi Shakeel, I like the idea and performance numbers look good. However some comments on the implementation: > --- > > Changes since the RFC: > - Added check for already charged slab objects. > - Added performance results from neper's tcp_crr > > include/linux/slab.h | 1 + > mm/slub.c | 54 +++++++++++++++++++++++++++++++++ > net/ipv4/inet_connection_sock.c | 5 +-- > 3 files changed, 58 insertions(+), 2 deletions(-) > > diff --git a/include/linux/slab.h b/include/linux/slab.h > index eb2bf4629157..05cfab107c72 100644 > --- a/include/linux/slab.h > +++ b/include/linux/slab.h > @@ -547,6 +547,7 @@ void *kmem_cache_alloc_lru_noprof(struct kmem_cache *s, struct list_lru *lru, > gfp_t gfpflags) __assume_slab_alignment __malloc; > #define kmem_cache_alloc_lru(...) alloc_hooks(kmem_cache_alloc_lru_noprof(__VA_ARGS__)) > > +bool kmem_cache_charge(void *objp, gfp_t gfpflags); > void kmem_cache_free(struct kmem_cache *s, void *objp); > > kmem_buckets *kmem_buckets_create(const char *name, slab_flags_t flags, > diff --git a/mm/slub.c b/mm/slub.c > index c9d8a2497fd6..580683597b5c 100644 > --- a/mm/slub.c > +++ b/mm/slub.c > @@ -2185,6 +2185,16 @@ void memcg_slab_free_hook(struct kmem_cache *s, struct slab *slab, void **p, > > __memcg_slab_free_hook(s, slab, p, objects, obj_exts); > } > + > +static __fastpath_inline > +bool memcg_slab_post_charge(struct kmem_cache *s, void *p, gfp_t flags) > +{ > + if (likely(!memcg_kmem_online())) > + return true; We do have this check in kmem_cache_charge(), why do we need to check it again? > + > + return __memcg_slab_post_alloc_hook(s, NULL, flags, 1, &p); > +} > + > #else /* CONFIG_MEMCG */ > static inline bool memcg_slab_post_alloc_hook(struct kmem_cache *s, > struct list_lru *lru, > @@ -2198,6 +2208,13 @@ static inline void memcg_slab_free_hook(struct kmem_cache *s, struct slab *slab, > void **p, int objects) > { > } > + > +static inline bool memcg_slab_post_charge(struct kmem_cache *s, > + void *p, > + gfp_t flags) > +{ > + return true; > +} > #endif /* CONFIG_MEMCG */ > > /* > @@ -4062,6 +4079,43 @@ void *kmem_cache_alloc_lru_noprof(struct kmem_cache *s, struct list_lru *lru, > } > EXPORT_SYMBOL(kmem_cache_alloc_lru_noprof); > > +#define KMALLOC_TYPE (SLAB_KMALLOC | SLAB_CACHE_DMA | \ > + SLAB_ACCOUNT | SLAB_RECLAIM_ACCOUNT) > + > +bool kmem_cache_charge(void *objp, gfp_t gfpflags) > +{ > + struct slabobj_ext *slab_exts; > + struct kmem_cache *s; > + struct folio *folio; > + struct slab *slab; > + unsigned long off; > + > + if (!memcg_kmem_online()) > + return true; > + > + folio = virt_to_folio(objp); > + if (unlikely(!folio_test_slab(folio))) > + return false; Does it handle the case of a too-big-to-be-a-slab-object allocation? I think it's better to handle it properly. Also, why return false here? > + > + slab = folio_slab(folio); > + s = slab->slab_cache; > + > + /* Ignore KMALLOC_NORMAL cache to avoid circular dependency. */ > + if ((s->flags & KMALLOC_TYPE) == SLAB_KMALLOC) > + return true; And true here? It seems to be a bit inconsistent. Also, if we have this check here, it means your function won't handle kmallocs at all? Because !KMALLOC_NORMAL allocations won't get here. > + > + /* Ignore already charged objects. */ > + slab_exts = slab_obj_exts(slab); > + if (slab_exts) { > + off = obj_to_index(s, slab, objp); > + if (unlikely(slab_exts[off].objcg)) > + return true; > + } > + > + return memcg_slab_post_charge(s, objp, gfpflags); > +} > +EXPORT_SYMBOL(kmem_cache_charge); > + > /** > * kmem_cache_alloc_node - Allocate an object on the specified node > * @s: The cache to allocate from. > diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c > index 64d07b842e73..3c13ca8c11fb 100644 > --- a/net/ipv4/inet_connection_sock.c > +++ b/net/ipv4/inet_connection_sock.c > @@ -715,6 +715,7 @@ struct sock *inet_csk_accept(struct sock *sk, struct proto_accept_arg *arg) > release_sock(sk); > if (newsk && mem_cgroup_sockets_enabled) { > int amt = 0; > + gfp_t gfp = GFP_KERNEL | __GFP_NOFAIL; > > /* atomically get the memory usage, set and charge the > * newsk->sk_memcg. > @@ -731,8 +732,8 @@ struct sock *inet_csk_accept(struct sock *sk, struct proto_accept_arg *arg) > } > > if (amt) > - mem_cgroup_charge_skmem(newsk->sk_memcg, amt, > - GFP_KERNEL | __GFP_NOFAIL); > + mem_cgroup_charge_skmem(newsk->sk_memcg, amt, gfp); > + kmem_cache_charge(newsk, gfp); Wait, so we assume that newsk->sk_memcg === current memcg? Or we're ok with them being different? Thanks!