Messages in this thread Patch in this message | | | Date | Tue, 4 May 2004 13:45:02 +0100 (BST) | From | Hugh Dickins <> | Subject | Re: Slab cache seems to grow forever - 2.6.6-rc3-mm1 |
| |
On Tue, 4 May 2004, Brent Cook wrote: > > This might be related to the change in fs-writeback.c that fixed > redirtying inodes on NFS, but I'm not sure. It seems that the Slab cache > never shrinks. Could this be a memory leak? It definitely affects system > performance. Here are the numbers after running for about 12 hours with > heavy NFS traffic (this is the client).
Please apply the patch akpm posted on Saturday:
2.6.6-rc3-mm1 is totally broken in the slab-shrinking area (sorry).
--- 25/mm/vmscan.c~shrink_slab-handle-GFP_NOFS-fix 2004-05-01 14:34:25.446391008 -0700 +++ 25-akpm/mm/vmscan.c 2004-05-01 14:34:37.424570048 -0700 @@ -156,7 +156,7 @@ static int shrink_slab(unsigned long sca shrinker->nr = LONG_MAX; /* It wrapped! */ if (shrinker->nr <= SHRINK_BATCH) - break; + continue; while (shrinker->nr) { long this_scan = shrinker->nr; int shrink_ret; - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |