From: Jeff Layton <jlayton@kernel.org>
To: NeilBrown <neilb@suse.de>, Chuck Lever <chuck.lever@oracle.com>
Cc: linux-nfs@vger.kernel.org,
Olga Kornievskaia <okorniev@redhat.com>,
Dai Ngo <Dai.Ngo@oracle.com>, Tom Talpey <tom@talpey.com>
Subject: Re: [PATCH 3/4] nfsd: free unused session-DRC slots
Date: Wed, 13 Nov 2024 11:52:38 -0500 [thread overview]
Message-ID: <1fe0918a2cb4a13613acf2a47ec5c1bd222dec10.camel@kernel.org> (raw)
In-Reply-To: <20241113055345.494856-4-neilb@suse.de>
On Wed, 2024-11-13 at 16:38 +1100, NeilBrown wrote:
> When the client confirms that some currently allocated DRC slots are not
> used, free some. We currently keep 6 (NFSD_MAX_UNUSED_SLOTS) unused
> slots to support bursts of client activity. This could be tuned with a
> shrinker.
>
> When we free a slot we store the seqid in the slot pointer so that it can
> be restored when we reactivate the slot. The RFC requires the seqid for
> each slot to increment on each request and does not permit it ever to be
> reset.
>
I think we need to get some clarification on the above.
IMO, once you free a slot, it's gone and its seqid should also be
forgotten. A reactivated slot, IOW, is effectively "new" and its seqid
should start again at 1. I don't think the spec intended to require
that both ends to remember seqids for defunct slots.
> We decoding sa_highest_slotid into maxslots we need to add 1 - this
> matches how it is encoded for the reply.
>
> Signed-off-by: NeilBrown <neilb@suse.de>
> ---
> fs/nfsd/nfs4state.c | 25 +++++++++++++++++++------
> fs/nfsd/nfs4xdr.c | 3 ++-
> 2 files changed, 21 insertions(+), 7 deletions(-)
>
> diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
> index 683bb908039b..15de62416243 100644
> --- a/fs/nfsd/nfs4state.c
> +++ b/fs/nfsd/nfs4state.c
> @@ -1910,14 +1910,21 @@ gen_sessionid(struct nfsd4_session *ses)
> #define NFSD_MIN_HDR_SEQ_SZ (24 + 12 + 44)
>
> static void
> -free_session_slots(struct nfsd4_session *ses)
> +free_session_slots(struct nfsd4_session *ses, int from)
> {
> int i;
>
> - for (i = 0; i < ses->se_fchannel.maxreqs; i++) {
> + if (from >= ses->se_fchannel.maxreqs)
> + return;
> +
> + for (i = from; i < ses->se_fchannel.maxreqs; i++) {
> + uintptr_t seqid = ses->se_slots[i]->sl_seqid;
> free_svc_cred(&ses->se_slots[i]->sl_cred);
> kfree(ses->se_slots[i]);
> + /* Save the seqid in case we reactivate this slot */
> + ses->se_slots[i] = (void *)seqid;
> }
> + ses->se_fchannel.maxreqs = from;
> }
>
> /*
> @@ -2069,7 +2076,7 @@ static void nfsd4_del_conns(struct nfsd4_session *s)
>
> static void __free_session(struct nfsd4_session *ses)
> {
> - free_session_slots(ses);
> + free_session_slots(ses, 0);
> kfree(ses);
> }
>
> @@ -4214,6 +4221,9 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
> if (status)
> goto out_put_session;
>
> + /* If there are lots of unused slots, free some */
> + free_session_slots(session, seq->maxslots + NFSD_MAX_UNUSED_SLOTS);
> +
> buflen = (seq->cachethis) ?
> session->se_fchannel.maxresp_cached :
> session->se_fchannel.maxresp_sz;
> @@ -4244,10 +4254,13 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
> session->se_fchannel.maxreqs < NFSD_MAX_SLOTS_PER_SESSION) {
> int s = session->se_fchannel.maxreqs;
>
> - session->se_slots[s] = kzalloc(slot_bytes(&session->se_fchannel),
> - GFP_NOWAIT);
> - if (session->se_slots[s])
> + struct nfsd4_slot *slot = kzalloc(slot_bytes(&session->se_fchannel),
> + GFP_NOWAIT);
> + if (slot) {
> + slot->sl_seqid = (uintptr_t)session->se_slots[s];
> + session->se_slots[s] = slot;
> session->se_fchannel.maxreqs += 1;
> + }
> }
> seq->maxslots = session->se_fchannel.maxreqs;
>
> diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
> index f118921250c3..846ed52fdaf5 100644
> --- a/fs/nfsd/nfs4xdr.c
> +++ b/fs/nfsd/nfs4xdr.c
> @@ -1884,7 +1884,8 @@ nfsd4_decode_sequence(struct nfsd4_compoundargs *argp,
> return nfserr_bad_xdr;
> seq->seqid = be32_to_cpup(p++);
> seq->slotid = be32_to_cpup(p++);
> - seq->maxslots = be32_to_cpup(p++);
> + /* sa_highest_slotid counts from 0 but maxslots counts from 1 ... */
> + seq->maxslots = be32_to_cpup(p++) + 1;
> seq->cachethis = be32_to_cpup(p);
>
> seq->status_flags = 0;
--
Jeff Layton <jlayton@kernel.org>
next prev parent reply other threads:[~2024-11-13 16:52 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-11-13 5:38 [PATCH 0/4 RFC] nfsd: allocate/free session-based DRC slots on demand NeilBrown
2024-11-13 5:38 ` [PATCH 1/4] nfsd: remove artificial limits on the session-based DRC NeilBrown
2024-11-13 16:48 ` Jeff Layton
2024-11-15 4:01 ` NeilBrown
2024-11-15 14:07 ` Jeff Layton
2024-11-13 5:38 ` [PATCH 2/4] nfsd: allocate new session-based DRC slots on demand NeilBrown
2024-11-13 5:38 ` [PATCH 3/4] nfsd: free unused session-DRC slots NeilBrown
2024-11-13 16:52 ` Jeff Layton [this message]
2024-11-13 5:38 ` [PATCH 4/4] nfsd: add shrinker to reduce number of slots allocated per session NeilBrown
2024-11-13 11:23 ` [PATCH 0/4 RFC] nfsd: allocate/free session-based DRC slots on demand Daire Byrne
2024-11-15 4:56 ` NeilBrown
2024-11-15 14:18 ` Chuck Lever
2024-11-18 15:43 ` Daire Byrne
2024-11-13 14:48 ` Chuck Lever III
2024-11-15 9:03 ` NeilBrown
2024-11-15 14:27 ` Chuck Lever
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=1fe0918a2cb4a13613acf2a47ec5c1bd222dec10.camel@kernel.org \
--to=jlayton@kernel.org \
--cc=Dai.Ngo@oracle.com \
--cc=chuck.lever@oracle.com \
--cc=linux-nfs@vger.kernel.org \
--cc=neilb@suse.de \
--cc=okorniev@redhat.com \
--cc=tom@talpey.com \
/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