On Fri, May 31, 2024 at 2:33 AM Vlastimil Babka <vbabka@xxxxxxx> wrote:
>
> Since commit 4f6923fbb352 ("mm: make should_failslab always available for
> fault injection") should_failslab() is unconditionally a noinline
> function. This adds visible overhead to the slab allocation hotpath,
> even if the function is empty. With CONFIG_FAILSLAB=y there's additional
> overhead when the functionality is not enabled by a boot parameter or
> debugfs.
>
> The overhead can be eliminated with a static key around the callsite.
> Fault injection and error injection frameworks can now be told that the
> this function has a static key associated, and are able to enable and
> disable it accordingly.
>
> Signed-off-by: Vlastimil Babka <vbabka@xxxxxxx>
> ---
> mm/failslab.c | 2 +-
> mm/slab.h | 3 +++
> mm/slub.c | 10 +++++++---
> 3 files changed, 11 insertions(+), 4 deletions(-)
>
> diff --git a/mm/failslab.c b/mm/failslab.c
> index ffc420c0e767..878fd08e5dac 100644
> --- a/mm/failslab.c
> +++ b/mm/failslab.c
> @@ -9,7 +9,7 @@ static struct {
> bool ignore_gfp_reclaim;
> bool cache_filter;
> } failslab = {
> - .attr = FAULT_ATTR_INITIALIZER,
> + .attr = FAULT_ATTR_INITIALIZER_KEY(&should_failslab_active.key),
> .ignore_gfp_reclaim = true,
> .cache_filter = false,
> };
> diff --git a/mm/slab.h b/mm/slab.h
> index 5f8f47c5bee0..792e19cb37b8 100644
> --- a/mm/slab.h
> +++ b/mm/slab.h
> @@ -11,6 +11,7 @@
> #include <linux/memcontrol.h>
> #include <linux/kfence.h>
> #include <linux/kasan.h>
> +#include <linux/jump_label.h>
>
> /*
> * Internal slab definitions
> @@ -160,6 +161,8 @@ static_assert(IS_ALIGNED(offsetof(struct slab, freelist), sizeof(freelist_aba_t)
> */
> #define slab_page(s) folio_page(slab_folio(s), 0)
>
> +DECLARE_STATIC_KEY_FALSE(should_failslab_active);
> +
> /*
> * If network-based swap is enabled, sl*b must keep track of whether pages
> * were allocated from pfmemalloc reserves.
> diff --git a/mm/slub.c b/mm/slub.c
> index 0809760cf789..3bb579760a37 100644
> --- a/mm/slub.c
> +++ b/mm/slub.c
> @@ -3874,13 +3874,15 @@ static __always_inline void maybe_wipe_obj_freeptr(struct kmem_cache *s,
> 0, sizeof(void *));
> }
>
> +DEFINE_STATIC_KEY_FALSE(should_failslab_active);
> +
> noinline int should_failslab(struct kmem_cache *s, gfp_t gfpflags)
> {
> if (__should_failslab(s, gfpflags))
> return -ENOMEM;
> return 0;
> }
> -ALLOW_ERROR_INJECTION(should_failslab, ERRNO);
> +ALLOW_ERROR_INJECTION_KEY(should_failslab, ERRNO, &should_failslab_active);
>
> static __fastpath_inline
> struct kmem_cache *slab_pre_alloc_hook(struct kmem_cache *s, gfp_t flags)
> @@ -3889,8 +3891,10 @@ struct kmem_cache *slab_pre_alloc_hook(struct kmem_cache *s, gfp_t flags)
>
> might_alloc(flags);
>
> - if (unlikely(should_failslab(s, flags)))
> - return NULL;
> + if (static_branch_unlikely(&should_failslab_active)) {
> + if (should_failslab(s, flags))
> + return NULL;
> + }
makes sense.
Acked-by: Alexei Starovoitov <ast@xxxxxxxxxx>
Do you have any microbenchmark numbers before/after this optimization?
