From nobody Mon Nov 25 05:33:03 2024 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 88BD11940B3; Tue, 29 Oct 2024 20:00:35 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1730232035; cv=none; b=cPFk+nkv/JdIi0ty9mzFyTOQ4pkzvyOUfJxnLkPRV1vxvktWrO/pd216VeVDDz+uvn/6wqn8HF9+E3bvU0gnfL3ZfDs6hxl0HnY6TUvq8QM83yrjkBZotVjB58bEDoKlAV6yOLdNORro7Ece6JyWVbK9FzOjs48Bg0w8heSGQRg= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1730232035; c=relaxed/simple; bh=Z5k6IE5EbAGovDlm8kn1eD8hO0NwIVUO7nt1+rBPFyM=; h=From:Date:Subject:MIME-Version:Content-Type:Message-Id:To:Cc; b=RqTyUGsVHtP10ky3Hp5yAIIdQwlkVfrmIzTEdGJMlDXiD03rIsJWxcRBCg4ABMQ76cYoItLRn8vnXWCOwAtXAh6Y/cghtYqn1J1sqRHZFIdrNnUyqEEz5z6Rwr6W7bFTW2EHg+MI3quMavRuWVzO3k15U0BH2+vT5f2oIUAnd88= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=EQWO6b3b; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="EQWO6b3b" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 633B5C4CECD; Tue, 29 Oct 2024 20:00:34 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1730232035; bh=Z5k6IE5EbAGovDlm8kn1eD8hO0NwIVUO7nt1+rBPFyM=; h=From:Date:Subject:To:Cc:From; b=EQWO6b3b23uFJ8yJz7kF6iDRTNE6NSAV4ffmgJXO1J7gNmVjt/Eu1gcH8hporzB1O y5fElDNqOQjt3SeWdEbuOOcgE8T8SrzX5cy1KwaFkIjeACs9fy3sX2zsYD8DOuHPW4 saYK/r/+97QgVS8riMCfOxYWhJUxHLdTVmPV9GvPvnxgHwUXPave7/XY5VamkniE2W bFDEKAvYW1vZNqwiUvZlNspiRTJxNK6q+Aat95gh+vrnZX3ZERS1bxZ3vY7FqKvefP hkITA2DKJ5+lrVDHbRtmph9eck4tJL9TwSeM2XiDmuuWQTTP5qsc00kWyW4o/rgZd8 qw5Sf3EVFkTzA== From: Jeff Layton Date: Tue, 29 Oct 2024 16:00:23 -0400 Subject: [PATCH v2] nfsd: allow for up to 32 callback session slots Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Message-Id: <20241029-bcwide-v2-1-e9010b6ef55d@kernel.org> X-B4-Tracking: v=1; b=H4sIANc+IWcC/2WMyw6CMBAAf4X0bM12KQ89+R+GA21X2GiAtAY1p P9uIR40HmcymUUE8kxBHLNFeJo58DgkwF0mbN8OHUl2iQUCagVYSGMf7EiWxlWkTZk7gyLFk6c LP7fRuUncc7iP/rV9Z7Xav8WsJEiHAHlFhYbWnq7kB7rtR9+ty09f//RAVdHWqA7OwnffxBjfI /NJOc0AAAA= X-Change-ID: 20241025-bcwide-6bd7e4b63db2 To: Chuck Lever , Neil Brown , Dai Ngo , Tom Talpey Cc: Olga Kornievskaia , linux-nfs@vger.kernel.org, linux-kernel@vger.kernel.org, Jeff Layton X-Mailer: b4 0.14.2 X-Developer-Signature: v=1; a=openpgp-sha256; l=12941; i=jlayton@kernel.org; h=from:subject:message-id; bh=Z5k6IE5EbAGovDlm8kn1eD8hO0NwIVUO7nt1+rBPFyM=; b=owEBbQKS/ZANAwAIAQAOaEEZVoIVAcsmYgBnIT7ciLNg2L6sd38JSfB3XmiIPIOrt8QRansL7 vy5HUQXL9CJAjMEAAEIAB0WIQRLwNeyRHGyoYTq9dMADmhBGVaCFQUCZyE+3AAKCRAADmhBGVaC FTMcD/9ZW5J0t0fBipWWFUe/W2Hu3O6J1/IaA5BfFieSIHBJAySi1Ts75I0/dow80CJR2njcBzt 8ZBaT/uNzvjt7uVz4yE8WXYHO9EbSavQzuUuOZ70TjnUPIv1ilt9WzcxRnh+Ud5SFIwOLxxQfaG Q161SYDqpznoiAOBh9ImdvhlSUvnJjs/tw7fUmbKFHlCSyQq8iDIJJ9aPLczig6IhOdCEsA6eBJ fD3LLDNmj1BZ+/f6sjjPDk/5pwfu5mhwUl69CSiySxxfTLXx66vdbx+N2ItijyPvP+FP2F2QzHX gWDz3RKGiHpR0wdHfZK4Ekv4zMUPK5ta7WdPLYn1AgXCAPAxFdWnTqEHyHE4ceqtSCqvXIJEp8L J29Fxmn+oSd1D4Pl0OiHPjXsBCp+W9pK8d9NKJgp9Lz0dYiHrtOTOTfNc+6oHIk0uy51XO+EcsS FN1iRcaelD8ahusYbNLPBfJTUgKIilgczQ0huxtzqIz+xPD7Mg15e/7S5EkdqPYaTqYqtduQsoJ OUaTTrE0pk5yDdWl1vWO9OczBealBRTwLcYg08o3SLUrWAdylIP08OY0s890o4Yfz++181jhEZV yH72w9FatAGZ2ksz2XUbCGXK4KKX4DEY7Ez4nxNGKCCvoQG2StcjAadJs18gs5HsGzLIfoUrP4I 2f0PL8SWXcMWCVQ== X-Developer-Key: i=jlayton@kernel.org; a=openpgp; fpr=4BC0D7B24471B2A184EAF5D3000E684119568215 nfsd currently only uses a single slot in the callback channel, which is proving to be a bottleneck in some cases. Widen the callback channel to a max of 32 slots (subject to the client's target_maxreqs value). Change the cb_holds_slot boolean to an integer that tracks the current slot number (with -1 meaning "unassigned"). Move the callback slot tracking info into the session. Add a new u32 that acts as a bitmap to track which slots are in use, and a u32 to track the latest callback target_slotid that the client reports. To protect the new fields, add a new per-session spinlock (the se_lock). Fix nfsd41_cb_get_slot to always search for the lowest slotid (using ffs()), and change it to retry until there is a slot available or the rpc_task is signalled. Finally, convert the session->se_cb_seq_nr field into an array of counters and add the necessary handling to ensure that the seqids get reset at the appropriate times. Signed-off-by: Jeff Layton --- Minor update to the set I sent yesterday, mostly to address Chuck's comments. This one also adds a new per-session spinlock to protect these fields instead of using the cl_lock. Olga, you had a workload that was seeing timeouts (probably?) due to long callback queues. I'd be interested to know whether this patch helps that workload at all. --- Changes in v2: - take cl_lock when fetching fields from session to be encoded - use fls() instead of bespoke highest_unset_index() - rename variables in several functions with more descriptive names - clamp limit of for loop in update_cb_slot_table() - re-add missing rpc_wake_up_queued_task() call - fix slotid check in decode_cb_sequence4resok() - add new per-session spinlock --- fs/nfsd/nfs4callback.c | 113 +++++++++++++++++++++++++++++++++++----------= ---- fs/nfsd/nfs4state.c | 8 +++- fs/nfsd/state.h | 17 +++++--- fs/nfsd/trace.h | 2 +- 4 files changed, 98 insertions(+), 42 deletions(-) diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index e38fa834b3d91333acf1425eb14c644e5d5f2601..4516ec5f68edcf702944dd2cc30= 71a24fd350f21 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -406,6 +406,19 @@ encode_cb_getattr4args(struct xdr_stream *xdr, struct = nfs4_cb_compound_hdr *hdr, hdr->nops++; } =20 +static u32 highest_slotid(struct nfsd4_session *ses) +{ + u32 idx; + + spin_lock(&ses->se_lock); + idx =3D fls(~ses->se_cb_slot_avail); + if (idx > 0) + --idx; + idx =3D max(idx, ses->se_cb_highest_slot); + spin_unlock(&ses->se_lock); + return idx; +} + /* * CB_SEQUENCE4args * @@ -432,15 +445,35 @@ static void encode_cb_sequence4args(struct xdr_stream= *xdr, encode_sessionid4(xdr, session); =20 p =3D xdr_reserve_space(xdr, 4 + 4 + 4 + 4 + 4); - *p++ =3D cpu_to_be32(session->se_cb_seq_nr); /* csa_sequenceid */ - *p++ =3D xdr_zero; /* csa_slotid */ - *p++ =3D xdr_zero; /* csa_highest_slotid */ + *p++ =3D cpu_to_be32(session->se_cb_seq_nr[cb->cb_held_slot]); /* csa_seq= uenceid */ + *p++ =3D cpu_to_be32(cb->cb_held_slot); /* csa_slotid */ + *p++ =3D cpu_to_be32(highest_slotid(session)); /* csa_highest_slotid */ *p++ =3D xdr_zero; /* csa_cachethis */ xdr_encode_empty_array(p); /* csa_referring_call_lists */ =20 hdr->nops++; } =20 +static void update_cb_slot_table(struct nfsd4_session *ses, u32 target) +{ + /* No need to do anything if nothing changed */ + if (likely(target =3D=3D READ_ONCE(ses->se_cb_highest_slot))) + return; + + spin_lock(&ses->se_lock); + if (target > ses->se_cb_highest_slot) { + int i; + + target =3D min(target, NFSD_BC_SLOT_TABLE_MAX); + + /* Growing the slot table. Reset any new sequences to 1 */ + for (i =3D ses->se_cb_highest_slot + 1; i <=3D target; ++i) + ses->se_cb_seq_nr[i] =3D 1; + } + ses->se_cb_highest_slot =3D target; + spin_unlock(&ses->se_lock); +} + /* * CB_SEQUENCE4resok * @@ -468,7 +501,7 @@ static int decode_cb_sequence4resok(struct xdr_stream *= xdr, struct nfsd4_session *session =3D cb->cb_clp->cl_cb_session; int status =3D -ESERVERFAULT; __be32 *p; - u32 dummy; + u32 seqid, slotid, target; =20 /* * If the server returns different values for sessionID, slotID or @@ -484,21 +517,27 @@ static int decode_cb_sequence4resok(struct xdr_stream= *xdr, } p +=3D XDR_QUADLEN(NFS4_MAX_SESSIONID_LEN); =20 - dummy =3D be32_to_cpup(p++); - if (dummy !=3D session->se_cb_seq_nr) { + seqid =3D be32_to_cpup(p++); + if (seqid !=3D session->se_cb_seq_nr[cb->cb_held_slot]) { dprintk("NFS: %s Invalid sequence number\n", __func__); goto out; } =20 - dummy =3D be32_to_cpup(p++); - if (dummy !=3D 0) { + slotid =3D be32_to_cpup(p++); + if (slotid !=3D cb->cb_held_slot) { dprintk("NFS: %s Invalid slotid\n", __func__); goto out; } =20 - /* - * FIXME: process highest slotid and target highest slotid - */ + p++; // ignore current highest slot value + + target =3D be32_to_cpup(p++); + if (target =3D=3D 0) { + dprintk("NFS: %s Invalid target highest slotid\n", __func__); + goto out; + } + + update_cb_slot_table(session, target); status =3D 0; out: cb->cb_seq_status =3D status; @@ -1211,28 +1250,39 @@ void nfsd4_change_callback(struct nfs4_client *clp,= struct nfs4_cb_conn *conn) static bool nfsd41_cb_get_slot(struct nfsd4_callback *cb, struct rpc_task = *task) { struct nfs4_client *clp =3D cb->cb_clp; + struct nfsd4_session *ses =3D clp->cl_cb_session; + int idx; =20 - if (!cb->cb_holds_slot && - test_and_set_bit(0, &clp->cl_cb_slot_busy) !=3D 0) { - rpc_sleep_on(&clp->cl_cb_waitq, task, NULL); - /* Race breaker */ - if (test_and_set_bit(0, &clp->cl_cb_slot_busy) !=3D 0) { - dprintk("%s slot is busy\n", __func__); + if (cb->cb_held_slot >=3D 0) + return true; +retry: + spin_lock(&ses->se_lock); + idx =3D ffs(ses->se_cb_slot_avail) - 1; + if (idx < 0 || idx > ses->se_cb_highest_slot) { + spin_unlock(&ses->se_lock); + if (RPC_SIGNALLED(task)) return false; - } + rpc_sleep_on(&clp->cl_cb_waitq, task, NULL); rpc_wake_up_queued_task(&clp->cl_cb_waitq, task); + goto retry; } - cb->cb_holds_slot =3D true; + /* clear the bit for the slot */ + ses->se_cb_slot_avail &=3D ~BIT(idx); + spin_unlock(&ses->se_lock); + cb->cb_held_slot =3D idx; return true; } =20 static void nfsd41_cb_release_slot(struct nfsd4_callback *cb) { struct nfs4_client *clp =3D cb->cb_clp; + struct nfsd4_session *ses =3D clp->cl_cb_session; =20 - if (cb->cb_holds_slot) { - cb->cb_holds_slot =3D false; - clear_bit(0, &clp->cl_cb_slot_busy); + if (cb->cb_held_slot >=3D 0) { + spin_lock(&ses->se_lock); + ses->se_cb_slot_avail |=3D BIT(cb->cb_held_slot); + spin_unlock(&ses->se_lock); + cb->cb_held_slot =3D -1; rpc_wake_up_next(&clp->cl_cb_waitq); } } @@ -1249,8 +1299,8 @@ static void nfsd41_destroy_cb(struct nfsd4_callback *= cb) } =20 /* - * TODO: cb_sequence should support referring call lists, cachethis, multi= ple - * slots, and mark callback channel down on communication errors. + * TODO: cb_sequence should support referring call lists, cachethis, + * and mark callback channel down on communication errors. */ static void nfsd4_cb_prepare(struct rpc_task *task, void *calldata) { @@ -1292,7 +1342,7 @@ static bool nfsd4_cb_sequence_done(struct rpc_task *t= ask, struct nfsd4_callback return true; } =20 - if (!cb->cb_holds_slot) + if (cb->cb_held_slot < 0) goto need_restart; =20 /* This is the operation status code for CB_SEQUENCE */ @@ -1306,10 +1356,10 @@ static bool nfsd4_cb_sequence_done(struct rpc_task = *task, struct nfsd4_callback * If CB_SEQUENCE returns an error, then the state of the slot * (sequence ID, cached reply) MUST NOT change. */ - ++session->se_cb_seq_nr; + ++session->se_cb_seq_nr[cb->cb_held_slot]; break; case -ESERVERFAULT: - ++session->se_cb_seq_nr; + ++session->se_cb_seq_nr[cb->cb_held_slot]; nfsd4_mark_cb_fault(cb->cb_clp); ret =3D false; break; @@ -1335,17 +1385,16 @@ static bool nfsd4_cb_sequence_done(struct rpc_task = *task, struct nfsd4_callback case -NFS4ERR_BADSLOT: goto retry_nowait; case -NFS4ERR_SEQ_MISORDERED: - if (session->se_cb_seq_nr !=3D 1) { - session->se_cb_seq_nr =3D 1; + if (session->se_cb_seq_nr[cb->cb_held_slot] !=3D 1) { + session->se_cb_seq_nr[cb->cb_held_slot] =3D 1; goto retry_nowait; } break; default: nfsd4_mark_cb_fault(cb->cb_clp); } - nfsd41_cb_release_slot(cb); - trace_nfsd_cb_free_slot(task, cb); + nfsd41_cb_release_slot(cb); =20 if (RPC_SIGNALLED(task)) goto need_restart; @@ -1565,7 +1614,7 @@ void nfsd4_init_cb(struct nfsd4_callback *cb, struct = nfs4_client *clp, INIT_WORK(&cb->cb_work, nfsd4_run_cb_work); cb->cb_status =3D 0; cb->cb_need_restart =3D false; - cb->cb_holds_slot =3D false; + cb->cb_held_slot =3D -1; } =20 /** diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 5b718b349396f1aecd0ad4c63b2f43342841bbd4..2e8d2a3425f8ae28608e9e6a81e= cda49cf4b2089 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2002,6 +2002,9 @@ static struct nfsd4_session *alloc_session(struct nfs= d4_channel_attrs *fattrs, } =20 memcpy(&new->se_fchannel, fattrs, sizeof(struct nfsd4_channel_attrs)); + new->se_cb_slot_avail =3D ~0U; + new->se_cb_highest_slot =3D battrs->maxreqs - 1; + spin_lock_init(&new->se_lock); return new; out_free: while (i--) @@ -2132,7 +2135,9 @@ static void init_session(struct svc_rqst *rqstp, stru= ct nfsd4_session *new, stru =20 INIT_LIST_HEAD(&new->se_conns); =20 - new->se_cb_seq_nr =3D 1; + for (idx =3D 0; idx < NFSD_BC_SLOT_TABLE_MAX; ++idx) + new->se_cb_seq_nr[idx] =3D 1; + new->se_flags =3D cses->flags; new->se_cb_prog =3D cses->callback_prog; new->se_cb_sec =3D cses->cb_sec; @@ -3159,7 +3164,6 @@ static struct nfs4_client *create_client(struct xdr_n= etobj name, kref_init(&clp->cl_nfsdfs.cl_ref); nfsd4_init_cb(&clp->cl_cb_null, clp, NULL, NFSPROC4_CLNT_CB_NULL); clp->cl_time =3D ktime_get_boottime_seconds(); - clear_bit(0, &clp->cl_cb_slot_busy); copy_verf(clp, verf); memcpy(&clp->cl_addr, sa, sizeof(struct sockaddr_storage)); clp->cl_cb_session =3D NULL; diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h index 41cda86fea1f6166a0fd0215d3d458c93ced3e6a..665a672fe71e453809f432b8dd0= b120e92f5a36a 100644 --- a/fs/nfsd/state.h +++ b/fs/nfsd/state.h @@ -71,8 +71,8 @@ struct nfsd4_callback { struct work_struct cb_work; int cb_seq_status; int cb_status; + int cb_held_slot; bool cb_need_restart; - bool cb_holds_slot; }; =20 struct nfsd4_callback_ops { @@ -307,6 +307,9 @@ struct nfsd4_conn { unsigned char cn_flags; }; =20 +/* Max number of slots that nfsd implements for NFSv4.1+ backchannel */ +#define NFSD_BC_SLOT_TABLE_MAX (sizeof(u32) * 8) + /* * Representation of a v4.1+ session. These are refcounted in a similar fa= shion * to the nfs4_client. References are only taken when the server is active= ly @@ -314,18 +317,21 @@ struct nfsd4_conn { */ struct nfsd4_session { atomic_t se_ref; - struct list_head se_hash; /* hash by sessionid */ - struct list_head se_perclnt; /* See SESSION4_PERSIST, etc. for standard flags; this is internal-only: */ #define NFS4_SESSION_DEAD 0x010 u32 se_flags; + spinlock_t se_lock; + struct list_head se_hash; /* hash by sessionid */ + struct list_head se_perclnt; struct nfs4_client *se_client; struct nfs4_sessionid se_sessionid; struct nfsd4_channel_attrs se_fchannel; struct nfsd4_cb_sec se_cb_sec; struct list_head se_conns; u32 se_cb_prog; - u32 se_cb_seq_nr; + u32 se_cb_slot_avail; /* bitmap of available slots */ + u32 se_cb_highest_slot; /* highest slot client wants */ + u32 se_cb_seq_nr[NFSD_BC_SLOT_TABLE_MAX]; struct nfsd4_slot *se_slots[]; /* forward channel slots */ }; =20 @@ -459,9 +465,6 @@ struct nfs4_client { */ struct dentry *cl_nfsd_info_dentry; =20 - /* for nfs41 callbacks */ - /* We currently support a single back channel with a single slot */ - unsigned long cl_cb_slot_busy; struct rpc_wait_queue cl_cb_waitq; /* backchannel callers may */ /* wait here for slots */ struct net *net; diff --git a/fs/nfsd/trace.h b/fs/nfsd/trace.h index f318898cfc31614b5a84a4867e18c2b3a07122c9..a9c17186b6892f1df8d7f7b90e2= 50c2913ab23fe 100644 --- a/fs/nfsd/trace.h +++ b/fs/nfsd/trace.h @@ -1697,7 +1697,7 @@ TRACE_EVENT(nfsd_cb_free_slot, __entry->cl_id =3D sid->clientid.cl_id; __entry->seqno =3D sid->sequence; __entry->reserved =3D sid->reserved; - __entry->slot_seqno =3D session->se_cb_seq_nr; + __entry->slot_seqno =3D session->se_cb_seq_nr[cb->cb_held_slot]; ), TP_printk(SUNRPC_TRACE_TASK_SPECIFIER " sessionid=3D%08x:%08x:%08x:%08x new slot seqno=3D%u", --- base-commit: 9c9cb4242c49bbadd010e8f0a9e7daf4b392ff6b change-id: 20241025-bcwide-6bd7e4b63db2 Best regards, --=20 Jeff Layton