Attachment 208237 Details for Bug 241048 – proposed patch

[patch] proposed patch

vm_page.c.diff (text/plain), 2.03 KB, created by Mark Johnston on 2019-10-11 03:01:32 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Mark Johnston

Created: 2019-10-11 03:01:32 UTC

Size: 2.03 KB

patch

obsolete

>commit e27ff405d661e94b79289203f09f53c54d7c1892
>Author: Mark Johnston <markjdb@gmail.com>
>Date:   Thu Oct 10 13:43:24 2019 -0400
>
>    Further constrain the use of per-CPU page caches.
>    
>    The calculation which determines whether or not to use UMA cache zones
>    is wrong in that it does not account for the fact that UMA maintains two
>    buckets per CPU, not one.  In the worst case, both the alloc and free
>    buckets on a CPU may be full of unused pages.
>    
>    Also increase the amount of required RAM per CPU to enable the per-CPU
>    caches.  With this change, on amd64 we require ~2.5GB or more RAM
>    per CPU to enable the caches.
>    
>    I have seen a couple of reports of unexpected OOM kills and memory
>    allocation failures which I believe are caused by a large number of
>    cached pages in systems with relatively small amounts of RAM.  In
>    particular, uma_reclaim() does not do anything to reclaim items from
>    cache zones, so cached pages are effectively unreclaimable right now.
>    This is a bug.
>    
>    We could also use uma_zone_set_max() to set an upper limit on the number
>    of items cached in the zone, i.e., use a smaller maximum bucket size
>    instead of 256.
>    
>    However, let's first adjust the threshold since that helps alleviate the
>    problem and is not too risky for 12.1.
>
>diff --git a/sys/vm/vm_page.c b/sys/vm/vm_page.c
>index c361a731db9e..cbbdabd816a3 100644
>--- a/sys/vm/vm_page.c
>+++ b/sys/vm/vm_page.c
>@@ -220,10 +220,12 @@ vm_page_init_cache_zones(void *dummy __unused)
> 		vmd = VM_DOMAIN(domain);
> 
> 		/*
>-		 * Don't allow the page caches to take up more than .25% of
>-		 * memory.
>+		 * Don't allow the page caches to take up more than .1875% of
>+		 * memory.  A UMA bucket contains at most 256 free pages, and we
>+		 * have two buckets per CPU per free pool.
> 		 */
>-		if (vmd->vmd_page_count / 400 < 256 * mp_ncpus * VM_NFREEPOOL)
>+		if (vmd->vmd_page_count / 600 < 2 * 256 * mp_ncpus *
>+		    VM_NFREEPOOL)
> 			continue;
> 		for (pool = 0; pool < VM_NFREEPOOL; pool++) {
> 			pgcache = &vmd->vmd_pgcache[pool];

Actions: View | Diff

Attachments on bug 241048: 208104 | 208105 | 208118 | 208237 | 208276 | 208284