From: "J. Bruce Fields" <bfields@fieldses.org>
To: NeilBrown <neilb@suse.de>
Cc: linux-nfs@vger.kernel.org
Subject: Re: [PATCH 3/9] sunrpc: never return expired entries in sunrpc_cache_lookup
Date: Wed, 17 Mar 2010 17:33:07 -0400 [thread overview]
Message-ID: <20100317213307.GD2501@fieldses.org> (raw)
In-Reply-To: <20100203063131.12945.97779.stgit-wvvUuzkyo1EYVZTmpyfIwg@public.gmane.org>
On Wed, Feb 03, 2010 at 05:31:31PM +1100, NeilBrown wrote:
> If sunrpc_cache_lookup finds an expired entry, remove it from
> the cache and return a freshly created non-VALID entry instead.
> This ensures that we only ever get a usable entry, or an
> entry that will become usable once an update arrives.
> i.e. we will never need to repeat the lookup.
>
> This allows us to remove the 'is_expired' test from cache_check
> (i.e. from cache_is_valid). cache_check should never get an expired
> entry as 'lookup' will never return one. If it does happen - due to
> inconvenient timing - then just accept it as still valid, it won't be
> very much past it's use-by date.
Looks right to me. Thanks, applied.
By the way, if we get sunrpc_cache_update(old, new1) and
sunrpc_cache_update(old, new2) simultaneously, what happens?
More generally: should we try to ensure that a cache never contains two
entries which match the same key?
--b.
>
> Signed-off-by: NeilBrown <neilb@suse.de>
> ---
> net/sunrpc/cache.c | 17 ++++++++++++++---
> 1 files changed, 14 insertions(+), 3 deletions(-)
>
> diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c
> index 9826c5c..3e1ef8b 100644
> --- a/net/sunrpc/cache.c
> +++ b/net/sunrpc/cache.c
> @@ -59,7 +59,7 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail,
> struct cache_head *key, int hash)
> {
> struct cache_head **head, **hp;
> - struct cache_head *new = NULL;
> + struct cache_head *new = NULL, *freeme = NULL;
>
> head = &detail->hash_table[hash];
>
> @@ -68,6 +68,9 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail,
> for (hp=head; *hp != NULL ; hp = &(*hp)->next) {
> struct cache_head *tmp = *hp;
> if (detail->match(tmp, key)) {
> + if (cache_is_expired(detail, tmp))
> + /* This entry is expired, we will discard it. */
> + break;
> cache_get(tmp);
> read_unlock(&detail->hash_lock);
> return tmp;
> @@ -92,6 +95,13 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail,
> for (hp=head; *hp != NULL ; hp = &(*hp)->next) {
> struct cache_head *tmp = *hp;
> if (detail->match(tmp, key)) {
> + if (cache_is_expired(detail, tmp)) {
> + *hp = tmp->next;
> + tmp->next = NULL;
> + detail->entries --;
> + freeme = tmp;
> + break;
> + }
> cache_get(tmp);
> write_unlock(&detail->hash_lock);
> cache_put(new, detail);
> @@ -104,6 +114,8 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail,
> cache_get(new);
> write_unlock(&detail->hash_lock);
>
> + if (freeme)
> + cache_put(freeme, detail);
> return new;
> }
> EXPORT_SYMBOL_GPL(sunrpc_cache_lookup);
> @@ -189,8 +201,7 @@ static int cache_make_upcall(struct cache_detail *cd, struct cache_head *h)
>
> static inline int cache_is_valid(struct cache_detail *detail, struct cache_head *h)
> {
> - if (!test_bit(CACHE_VALID, &h->flags) ||
> - cache_is_expired(detail, h))
> + if (!test_bit(CACHE_VALID, &h->flags))
> return -EAGAIN;
> else {
> /* entry is valid */
>
>
next prev parent reply other threads:[~2010-03-17 21:31 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-02-03 6:31 [PATCH 0/9] Cache deferal improvements - try++ NeilBrown
[not found] ` <20100203060657.12945.27293.stgit-wvvUuzkyo1EYVZTmpyfIwg@public.gmane.org>
2010-02-03 6:31 ` [PATCH 9/9] sunrpc/cache: change deferred-request hash table to use hlist NeilBrown
2010-02-03 6:31 ` [PATCH 4/9] sunrpc/cache: allow threads to block while waiting for cache update NeilBrown
2010-04-15 15:55 ` J. Bruce Fields
2010-02-03 6:31 ` [PATCH 7/9] nfsd: factor out hash functions for export caches NeilBrown
[not found] ` <20100203063131.12945.38791.stgit-wvvUuzkyo1EYVZTmpyfIwg@public.gmane.org>
2010-03-17 19:35 ` J. Bruce Fields
2010-02-03 6:31 ` [PATCH 1/9] sunrpc: don't keep expired entries in the auth caches NeilBrown
[not found] ` <20100203063130.12945.29226.stgit-wvvUuzkyo1EYVZTmpyfIwg@public.gmane.org>
2010-03-15 0:58 ` J. Bruce Fields
2010-02-03 6:31 ` [PATCH 5/9] nfsd/idmap: drop special request deferal in favour of improved default NeilBrown
2010-02-03 6:31 ` [PATCH 3/9] sunrpc: never return expired entries in sunrpc_cache_lookup NeilBrown
[not found] ` <20100203063131.12945.97779.stgit-wvvUuzkyo1EYVZTmpyfIwg@public.gmane.org>
2010-03-17 21:33 ` J. Bruce Fields [this message]
2010-03-24 1:22 ` Neil Brown
2010-03-30 15:11 ` J. Bruce Fields
2010-02-03 6:31 ` [PATCH 8/9] svcauth_gss: replace a trivial 'switch' with an 'if' NeilBrown
2010-02-03 6:31 ` [PATCH 2/9] sunrpc/cache: factor out cache_is_expired NeilBrown
[not found] ` <20100203063131.12945.65321.stgit-wvvUuzkyo1EYVZTmpyfIwg@public.gmane.org>
2010-03-15 0:58 ` J. Bruce Fields
2010-02-03 6:31 ` [PATCH 6/9] sunrpc: close connection when a request is irretrievably lost NeilBrown
2010-02-03 15:43 ` Chuck Lever
2010-02-03 21:23 ` Neil Brown
2010-02-03 22:20 ` Trond Myklebust
2010-02-03 22:34 ` J. Bruce Fields
2010-02-03 22:40 ` Chuck Lever
2010-02-03 23:13 ` Trond Myklebust
2010-02-04 0:06 ` Chuck Lever
2010-02-04 0:24 ` Trond Myklebust
2010-02-03 22:34 ` Chuck Lever
2010-02-03 19:43 ` [PATCH 0/9] Cache deferal improvements - try++ J. Bruce Fields
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20100317213307.GD2501@fieldses.org \
--to=bfields@fieldses.org \
--cc=linux-nfs@vger.kernel.org \
--cc=neilb@suse.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox