On Mon, Feb 26, 2024 at 8:26 AM Vlastimil Babka <vbabka@xxxxxxx> wrote:
>
> On 2/21/24 20:40, Suren Baghdasaryan wrote:
> > Currently slab pages can store only vectors of obj_cgroup pointers in
> > page->memcg_data. Introduce slabobj_ext structure to allow more data
> > to be stored for each slab object. Wrap obj_cgroup into slabobj_ext
> > to support current functionality while allowing to extend slabobj_ext
> > in the future.
> >
> > Signed-off-by: Suren Baghdasaryan <surenb@xxxxxxxxxx>
>
> Hi, mostly good from slab perspective, just some fixups:
>
> > --- a/mm/slab.h
> > +++ b/mm/slab.h
> > -int memcg_alloc_slab_cgroups(struct slab *slab, struct kmem_cache *s,
> > - gfp_t gfp, bool new_slab);
> > -void mod_objcg_state(struct obj_cgroup *objcg, struct pglist_data *pgdat,
> > - enum node_stat_item idx, int nr);
> > -#else /* CONFIG_MEMCG_KMEM */
> > -static inline struct obj_cgroup **slab_objcgs(struct slab *slab)
> > +int alloc_slab_obj_exts(struct slab *slab, struct kmem_cache *s,
> > + gfp_t gfp, bool new_slab);
> >
>
> We could remove this declaration and make the function static in mm/slub.c.
Ack.
>
> > +#else /* CONFIG_SLAB_OBJ_EXT */
> > +
> > +static inline struct slabobj_ext *slab_obj_exts(struct slab *slab)
> > {
> > return NULL;
> > }
> >
> > -static inline int memcg_alloc_slab_cgroups(struct slab *slab,
> > - struct kmem_cache *s, gfp_t gfp,
> > - bool new_slab)
> > +static inline int alloc_slab_obj_exts(struct slab *slab,
> > + struct kmem_cache *s, gfp_t gfp,
> > + bool new_slab)
> > {
> > return 0;
> > }
>
> Ditto
Ack.
>
> > -#endif /* CONFIG_MEMCG_KMEM */
> > +
> > +static inline struct slabobj_ext *
> > +prepare_slab_obj_exts_hook(struct kmem_cache *s, gfp_t flags, void *p)
> > +{
> > + return NULL;
> > +}
>
> Same here (and the definition and usage even happens in later patch).
Ack.
>
> > +#endif /* CONFIG_SLAB_OBJ_EXT */
> > +
> > +#ifdef CONFIG_MEMCG_KMEM
> > +void mod_objcg_state(struct obj_cgroup *objcg, struct pglist_data *pgdat,
> > + enum node_stat_item idx, int nr);
> > +#endif
> >
> > size_t __ksize(const void *objp);
> >
> > diff --git a/mm/slub.c b/mm/slub.c
> > index d31b03a8d9d5..76fb600fbc80 100644
> > --- a/mm/slub.c
> > +++ b/mm/slub.c
> > @@ -683,10 +683,10 @@ static inline bool __slab_update_freelist(struct kmem_cache *s, struct slab *sla
> >
> > if (s->flags & __CMPXCHG_DOUBLE) {
> > ret = __update_freelist_fast(slab, freelist_old, counters_old,
> > - freelist_new, counters_new);
> > + freelist_new, counters_new);
> > } else {
> > ret = __update_freelist_slow(slab, freelist_old, counters_old,
> > - freelist_new, counters_new);
> > + freelist_new, counters_new);
> > }
> > if (likely(ret))
> > return true;
> > @@ -710,13 +710,13 @@ static inline bool slab_update_freelist(struct kmem_cache *s, struct slab *slab,
> >
> > if (s->flags & __CMPXCHG_DOUBLE) {
> > ret = __update_freelist_fast(slab, freelist_old, counters_old,
> > - freelist_new, counters_new);
> > + freelist_new, counters_new);
> > } else {
> > unsigned long flags;
> >
> > local_irq_save(flags);
> > ret = __update_freelist_slow(slab, freelist_old, counters_old,
> > - freelist_new, counters_new);
> > + freelist_new, counters_new);
>
> Please no drive-by fixups of whitespace in code you're not actually
> changing. I thought you agreed in v3?
Sorry, I must have misunderstood your previous comment. I thought you
were saying that the alignment I changed to was incorrect. I'll keep
them untouched.
>
> > static inline bool memcg_slab_pre_alloc_hook(struct kmem_cache *s,
> > struct list_lru *lru,
> > struct obj_cgroup **objcgp,
> > @@ -2314,7 +2364,7 @@ static __always_inline void account_slab(struct slab *slab, int order,
> > struct kmem_cache *s, gfp_t gfp)
> > {
> > if (memcg_kmem_online() && (s->flags & SLAB_ACCOUNT))
> > - memcg_alloc_slab_cgroups(slab, s, gfp, true);
> > + alloc_slab_obj_exts(slab, s, gfp, true);
>
> This is still guarded by the memcg_kmem_online() static key, which is good.
>
> >
> > mod_node_page_state(slab_pgdat(slab), cache_vmstat_idx(s),
> > PAGE_SIZE << order);
> > @@ -2323,8 +2373,7 @@ static __always_inline void account_slab(struct slab *slab, int order,
> > static __always_inline void unaccount_slab(struct slab *slab, int order,
> > struct kmem_cache *s)
> > {
> > - if (memcg_kmem_online())
> > - memcg_free_slab_cgroups(slab);
> > + free_slab_obj_exts(slab);
>
> But this no longer is, yet it still could be?
Yes, I missed that, it seems. free_slab_obj_exts() would bail out but
still checking the static key is more efficient. I'll revive this
check.
Thanks for the review!
Suren.
>
> >
> > mod_node_page_state(slab_pgdat(slab), cache_vmstat_idx(s),
> > -(PAGE_SIZE << order));
>
