Almost description is copied from commit fb05e7a89f50
("net: don't wait for order-3 page allocation").
I saw excessive direct memory reclaim/compaction triggered by slub.
This causes performance issues and add latency. Slub uses high-order
allocation to reduce internal fragmentation and management overhead. But,
direct memory reclaim/compaction has high overhead and the benefit of
high-order allocation can't compensate the overhead of both work.
This patch makes auxiliary high-order allocation atomic. If there is
no memory pressure and memory isn't fragmented, the alloction will still
success, so we don't sacrifice high-order allocation's benefit here.
If the atomic allocation fails, direct memory reclaim/compaction will not
be triggered, allocation fallback to low-order immediately, hence
the direct memory reclaim/compaction overhead is avoided. In the
allocation failure case, kswapd is waken up and trying to make high-order
freepages, so allocation could success next time.
Following is the test to measure effect of this patch.
System: QEMU, CPU 8, 512 MB
Mem: 25% memory is allocated at random position to make fragmentation.
Memory-hogger occupies 150 MB memory.
Workload: hackbench -g 20 -l 1000
Average result by 10 runs (Base va Patched)
elapsed_time(s): 4.3468 vs 2.9838
compact_stall: 461.7 vs 73.6
pgmigrate_success: 28315.9 vs 7256.1
Signed-off-by: Joonsoo Kim <[email protected]>
---
mm/slub.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/mm/slub.c b/mm/slub.c
index 257283f..2d02a36 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -1364,6 +1364,8 @@ static struct page *allocate_slab(struct kmem_cache *s, gfp_t flags, int node)
* so we fall-back to the minimum order allocation.
*/
alloc_gfp = (flags | __GFP_NOWARN | __GFP_NORETRY) & ~__GFP_NOFAIL;
+ if ((alloc_gfp & __GFP_WAIT) && oo_order(oo) > oo_order(s->min))
+ alloc_gfp = alloc_gfp & ~__GFP_WAIT;
page = alloc_slab_page(s, alloc_gfp, node, oo);
if (unlikely(!page)) {
--
1.9.1
On Fri 31-07-15 10:09:50, Joonsoo Kim wrote:
> Almost description is copied from commit fb05e7a89f50
> ("net: don't wait for order-3 page allocation").
>
> I saw excessive direct memory reclaim/compaction triggered by slub.
> This causes performance issues and add latency. Slub uses high-order
> allocation to reduce internal fragmentation and management overhead. But,
> direct memory reclaim/compaction has high overhead and the benefit of
> high-order allocation can't compensate the overhead of both work.
>
> This patch makes auxiliary high-order allocation atomic. If there is
> no memory pressure and memory isn't fragmented, the alloction will still
> success, so we don't sacrifice high-order allocation's benefit here.
But you are also giving those allocations access to a portion of the
memory reserves which doesn't sound like an intenteded behavior here.
At least the changelog doesn't imply anything like that.
I am not oppposed to your patch but I think we should do something about
the !__GFP_WAIT behavior. This is too subtle and the mere fact the
caller doesn't want or cannot sleep doesn't make it a reserve consumer
automatically. We have __GFP_HIGH for that purpose. If this is not
desirable because of the regression risk then we might need a new gfp
flag for a best effort allocation which will fail in case we have to
dive into costly reclaim.
> If the atomic allocation fails, direct memory reclaim/compaction will not
> be triggered, allocation fallback to low-order immediately, hence
> the direct memory reclaim/compaction overhead is avoided. In the
> allocation failure case, kswapd is waken up and trying to make high-order
> freepages, so allocation could success next time.
>
> Following is the test to measure effect of this patch.
>
> System: QEMU, CPU 8, 512 MB
> Mem: 25% memory is allocated at random position to make fragmentation.
> Memory-hogger occupies 150 MB memory.
> Workload: hackbench -g 20 -l 1000
>
> Average result by 10 runs (Base va Patched)
>
> elapsed_time(s): 4.3468 vs 2.9838
> compact_stall: 461.7 vs 73.6
> pgmigrate_success: 28315.9 vs 7256.1
>
> Signed-off-by: Joonsoo Kim <[email protected]>
> ---
> mm/slub.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/mm/slub.c b/mm/slub.c
> index 257283f..2d02a36 100644
> --- a/mm/slub.c
> +++ b/mm/slub.c
> @@ -1364,6 +1364,8 @@ static struct page *allocate_slab(struct kmem_cache *s, gfp_t flags, int node)
> * so we fall-back to the minimum order allocation.
> */
> alloc_gfp = (flags | __GFP_NOWARN | __GFP_NORETRY) & ~__GFP_NOFAIL;
> + if ((alloc_gfp & __GFP_WAIT) && oo_order(oo) > oo_order(s->min))
> + alloc_gfp = alloc_gfp & ~__GFP_WAIT;
>
> page = alloc_slab_page(s, alloc_gfp, node, oo);
> if (unlikely(!page)) {
> --
> 1.9.1
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
--
Michal Hocko
SUSE Labs
On Tue, Aug 04, 2015 at 03:15:26PM +0200, Michal Hocko wrote:
> On Fri 31-07-15 10:09:50, Joonsoo Kim wrote:
> > Almost description is copied from commit fb05e7a89f50
> > ("net: don't wait for order-3 page allocation").
> >
> > I saw excessive direct memory reclaim/compaction triggered by slub.
> > This causes performance issues and add latency. Slub uses high-order
> > allocation to reduce internal fragmentation and management overhead. But,
> > direct memory reclaim/compaction has high overhead and the benefit of
> > high-order allocation can't compensate the overhead of both work.
> >
> > This patch makes auxiliary high-order allocation atomic. If there is
> > no memory pressure and memory isn't fragmented, the alloction will still
> > success, so we don't sacrifice high-order allocation's benefit here.
>
> But you are also giving those allocations access to a portion of the
> memory reserves which doesn't sound like an intenteded behavior here.
> At least the changelog doesn't imply anything like that.
>
> I am not oppposed to your patch but I think we should do something about
> the !__GFP_WAIT behavior. This is too subtle and the mere fact the
> caller doesn't want or cannot sleep doesn't make it a reserve consumer
> automatically. We have __GFP_HIGH for that purpose. If this is not
> desirable because of the regression risk then we might need a new gfp
> flag for a best effort allocation which will fail in case we have to
> dive into costly reclaim.
Hello, Michal.
We already have __GFP_NOMEMALLOC not to use emergency pool in case of
!__GFP_WAIT. Please see gfp_to_alloc_flags().
I'll send update version to use this flag.
BTW, network commit fb05e7a89f50 doesn't specify this flag. Does it
need this change, too?
Thanks.
>
> > If the atomic allocation fails, direct memory reclaim/compaction will not
> > be triggered, allocation fallback to low-order immediately, hence
> > the direct memory reclaim/compaction overhead is avoided. In the
> > allocation failure case, kswapd is waken up and trying to make high-order
> > freepages, so allocation could success next time.
> >
> > Following is the test to measure effect of this patch.
> >
> > System: QEMU, CPU 8, 512 MB
> > Mem: 25% memory is allocated at random position to make fragmentation.
> > Memory-hogger occupies 150 MB memory.
> > Workload: hackbench -g 20 -l 1000
> >
> > Average result by 10 runs (Base va Patched)
> >
> > elapsed_time(s): 4.3468 vs 2.9838
> > compact_stall: 461.7 vs 73.6
> > pgmigrate_success: 28315.9 vs 7256.1
> >
> > Signed-off-by: Joonsoo Kim <[email protected]>
> > ---
> > mm/slub.c | 2 ++
> > 1 file changed, 2 insertions(+)
> >
> > diff --git a/mm/slub.c b/mm/slub.c
> > index 257283f..2d02a36 100644
> > --- a/mm/slub.c
> > +++ b/mm/slub.c
> > @@ -1364,6 +1364,8 @@ static struct page *allocate_slab(struct kmem_cache *s, gfp_t flags, int node)
> > * so we fall-back to the minimum order allocation.
> > */
> > alloc_gfp = (flags | __GFP_NOWARN | __GFP_NORETRY) & ~__GFP_NOFAIL;
> > + if ((alloc_gfp & __GFP_WAIT) && oo_order(oo) > oo_order(s->min))
> > + alloc_gfp = alloc_gfp & ~__GFP_WAIT;
> >
> > page = alloc_slab_page(s, alloc_gfp, node, oo);
> > if (unlikely(!page)) {
> > --
> > 1.9.1
> >
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to [email protected]
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at http://www.tux.org/lkml/
>
> --
> Michal Hocko
> SUSE Labs
>
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to [email protected]. For more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=mailto:"[email protected]"> [email protected] </a>