| Message ID | 165730472428.28142.17571886266553271473.stgit@klimt.1015granger.net (mailing list archive) |
|---|---|
| State | Not Applicable |
| Headers | show |
| Series | Overhaul NFSD filecache | expand |
| Context | Check | Description |
|---|---|---|
| netdev/tree_selection | success | Not a local patch |
On Fri, 2022-07-08 at 14:25 -0400, Chuck Lever wrote: > Without LRU item rotation, the shrinker visits only a few items on > the end of the LRU list, and those would always be long-term OPEN > files for NFSv4 workloads. That makes the filecache shrinker > completely ineffective. > > Adopt the same strategy as the inode LRU by using LRU_ROTATE. > > Suggested-by: Dave Chinner <david@fromorbit.com> > Signed-off-by: Chuck Lever <chuck.lever@oracle.com> > --- > fs/nfsd/filecache.c | 5 +++-- > 1 file changed, 3 insertions(+), 2 deletions(-) > > diff --git a/fs/nfsd/filecache.c b/fs/nfsd/filecache.c > index 6e9e186334ab..bd6ba63f69ae 100644 > --- a/fs/nfsd/filecache.c > +++ b/fs/nfsd/filecache.c > @@ -452,6 +452,7 @@ nfsd_file_dispose_list_delayed(struct list_head *dispose) > * > * Return values: > * %LRU_REMOVED: @item was removed from the LRU > + * %LRU_ROTATED: @item is to be moved to the LRU tail %LRU_ROTATE > * %LRU_SKIP: @item cannot be evicted > */ > static enum lru_status > @@ -490,7 +491,7 @@ nfsd_file_lru_cb(struct list_head *item, struct list_lru_one *lru, > > if (test_and_clear_bit(NFSD_FILE_REFERENCED, &nf->nf_flags)) { > trace_nfsd_file_gc_referenced(nf); > - return LRU_SKIP; > + return LRU_ROTATE; > } > > if (!test_and_clear_bit(NFSD_FILE_HASHED, &nf->nf_flags)) { > @@ -532,7 +533,7 @@ nfsd_file_gc(void) > unsigned long ret; > > ret = list_lru_walk(&nfsd_file_lru, nfsd_file_lru_cb, > - &dispose, LONG_MAX); > + &dispose, list_lru_count(&nfsd_file_lru)); > trace_nfsd_file_gc_removed(ret, list_lru_count(&nfsd_file_lru)); > nfsd_file_gc_dispose_list(&dispose); > } > >
diff --git a/fs/nfsd/filecache.c b/fs/nfsd/filecache.c index 6e9e186334ab..bd6ba63f69ae 100644 --- a/fs/nfsd/filecache.c +++ b/fs/nfsd/filecache.c @@ -452,6 +452,7 @@ nfsd_file_dispose_list_delayed(struct list_head *dispose) * * Return values: * %LRU_REMOVED: @item was removed from the LRU + * %LRU_ROTATED: @item is to be moved to the LRU tail * %LRU_SKIP: @item cannot be evicted */ static enum lru_status @@ -490,7 +491,7 @@ nfsd_file_lru_cb(struct list_head *item, struct list_lru_one *lru, if (test_and_clear_bit(NFSD_FILE_REFERENCED, &nf->nf_flags)) { trace_nfsd_file_gc_referenced(nf); - return LRU_SKIP; + return LRU_ROTATE; } if (!test_and_clear_bit(NFSD_FILE_HASHED, &nf->nf_flags)) { @@ -532,7 +533,7 @@ nfsd_file_gc(void) unsigned long ret; ret = list_lru_walk(&nfsd_file_lru, nfsd_file_lru_cb, - &dispose, LONG_MAX); + &dispose, list_lru_count(&nfsd_file_lru)); trace_nfsd_file_gc_removed(ret, list_lru_count(&nfsd_file_lru)); nfsd_file_gc_dispose_list(&dispose); }
Without LRU item rotation, the shrinker visits only a few items on the end of the LRU list, and those would always be long-term OPEN files for NFSv4 workloads. That makes the filecache shrinker completely ineffective. Adopt the same strategy as the inode LRU by using LRU_ROTATE. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Chuck Lever <chuck.lever@oracle.com> --- fs/nfsd/filecache.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)