From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-yw1-f172.google.com (mail-yw1-f172.google.com [209.85.128.172]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id D0D9441A54E for ; Wed, 15 Jul 2026 18:53:05 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.128.172 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1784141590; cv=none; b=s0Xd7ISOie5GUVU9E4b3PE58OpBm6qlsA4akgKaTBmpbfD17drzAFr91JUjue1LbLKpCPQPLU0KgVJNme/lssuV9/1uzR1wN6mQmco6cVD2PFYIjYudjM85gl6vftrQIdj1CskekReRo7ZbyHzDpNYKVskRHp06N9dvT3fvNtWs= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1784141590; c=relaxed/simple; bh=nEgYEqcTHEbRoqb9+c6GdM812W2jAXOQxH2BsJegGeU=; h=Message-ID:Subject:From:To:Cc:Date:In-Reply-To:References: Content-Type:MIME-Version; b=AejOcLxbGgWin0o/A1sChEMZVQCESBoHEuewqzGeSGGR6mYRzOy8AHjCgPN6tENNOSL3blTN1pALSeLpHHY7pRNRXQmWFwp9E+5vvRw0uGn6TvdBWDK8oreVDg7Ch4UXANXGBaX+EM9+r4D2UGZiL0LvL51b1Pkw9p0xpwAlRPQ= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=dubeyko.com; spf=pass smtp.mailfrom=dubeyko.com; dkim=pass (2048-bit key) header.d=dubeyko-com.20251104.gappssmtp.com header.i=@dubeyko-com.20251104.gappssmtp.com header.b=qKy85Gfs; arc=none smtp.client-ip=209.85.128.172 Authentication-Results: smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=dubeyko.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=dubeyko.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=dubeyko-com.20251104.gappssmtp.com header.i=@dubeyko-com.20251104.gappssmtp.com header.b="qKy85Gfs" Received: by mail-yw1-f172.google.com with SMTP id 00721157ae682-81062fdeaf5so14059087b3.0 for ; Wed, 15 Jul 2026 11:53:05 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=dubeyko-com.20251104.gappssmtp.com; s=20251104; t=1784141584; x=1784746384; darn=vger.kernel.org; h=mime-version:user-agent:content-transfer-encoding:content-type :autocrypt:references:in-reply-to:date:cc:to:from:subject:message-id :from:to:cc:subject:date:message-id:reply-to:content-type; bh=alfiDX7CYs6WFqoYlpG8kxjbkp7ZuVJOaBWU1rXmK8A=; b=qKy85Gfspib0hJEODGut4vjHMWi446NOeGCU0GIJ103KT+XqG5ujIrg134VUiFyiAV Z+m1cwBaa5aJH90rk4su+qX99Vap/h9cdYf9YaP4XofGLBNPLwSkFyyRqEXtt+VsnN5Y T5+Zq6FClKIZYiI5UDmClVrCx2tsO7Gvc6pOjPA+nPVRhIsvbCRt/k76NHSkY7qUcL3y XZY6/vjXjlVqjnmwwpIPIeDFPzSPs8jetDVnG5gYO64rtEuEBwBQXWuEm0L2Aee2r3Gm YYCZoGp2OebE3mdXx1e1cNWin0XM49yUgzKQT2NY2nZbTHMS9OJ1NYBEvaMfxRWKVBCz 0D3g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1784141584; x=1784746384; h=mime-version:user-agent:content-transfer-encoding:content-type :autocrypt:references:in-reply-to:date:cc:to:from:subject:message-id :x-gm-gg:x-gm-message-state:from:to:cc:subject:date:message-id :reply-to:content-type; bh=alfiDX7CYs6WFqoYlpG8kxjbkp7ZuVJOaBWU1rXmK8A=; b=eRrQc3tfTSskBKZpAYk0lshwWfn6iAZbEIlHIVHShWJogRQrznw4EevRc1wXinouyQ utbhh0qbIDnqLTbmnyJGmtXgcNqBVJcrQCTQZpsKPb81J5yNvnwaK18U8sanp0/HIfjC BF8bIY90SYKyqIBZfWQqTzrhxwDawBY1bG0JzHKJdHMT+QG85OJZnj0zNZ/hSoAIpfkb SQE82/6ucnX7FUPc+DlLnMp5KnDblLSgIoa6lW5JM5JkCdDOx46rZ8BI9kFqtMeKoBth bDnIdwn2FsePcd+WEisg+C+VVZQC2KVeweRVCSY2BooTnOu9V7p2SxOqEZLLyIEsIPww wMGw== X-Forwarded-Encrypted: i=1; AHgh+RqTraXmGHsRPNKs3JXXrCkjNenuDKLRgaxY6ablKGBv1z9PY8750m04aJQVU+8eLi3/9hh0HgZ6yRhm4Ng=@vger.kernel.org X-Gm-Message-State: AOJu0YxumC2z99b/HlloOIHCiFtcWUWB9L1LsntteM2gQke2PJ4xTd7p +zVHiBdChVh+0Jk4jlh6X52IyTszR8++m8ommwNkCKX9+NhO/4hpkNVy1sCkYYAXMms= X-Gm-Gg: AfdE7cmiznjODpjMRcXSHbjDennSTWkJvDQZ6463zWZHn5AwYXEW6MCdl6m6Ze4ERuK IDlURHk1vGhqcg/5VpsBpl+XKbqSFRXJqSBGGmB2EXrjp6mM3L/3Rs7yRUJI767cx1OoaHOtHV5 CZexzUjTyvhCFSwrmrJmGMhrO5XvCQLgEzhY+8ReYJXY8zIr54biIAteGCr7FkkG3C9dV4AllVU 3SNJG5vERGq4dFdJx2+rdaU0R+XrwXNOCJZ1/pscU4/Nevv3T/cMqqyeXzbWp3C05U2L7CYt1oY kngyOW8M8/aJF1HW6rFG7RUeon+gHsz7MgPdlyfNuNU8zRGQtyVQAJyBYhva/ao3WWWyBGjSAh5 u66XYjibjHNcGpS1wOvpoIC8hsmh50f2mIVWgbbSqWpSoFuyN/Kx3nQU+v1jk5jNlOEX0XiHTTd +CrjBw599He1p2tqLpExLU8x2G/XQcFcjcIntF7MaC7V+FkvGhv8Jbmx+cjB3qFWTwX5E9cuE3I 643eYR1DcxeivzaQmLQLlgk3hElpZIeG9K0TZpCUuQO X-Received: by 2002:a05:690c:3489:b0:7f0:38f7:6ca6 with SMTP id 00721157ae682-81ede3c4e3fmr5008117b3.5.1784141583555; Wed, 15 Jul 2026 11:53:03 -0700 (PDT) Received: from pop-os.attlocal.net ([2600:1700:6476:1430:393f:38b4:2407:391]) by smtp.gmail.com with ESMTPSA id 00721157ae682-81e6c1a44d6sm177708937b3.31.2026.07.15.11.53.02 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 15 Jul 2026 11:53:02 -0700 (PDT) Message-ID: Subject: Re: [PATCH v2 4/5] ceph: move mdsc->mutex into __do_request() From: Viacheslav Dubeyko To: xiubo.li@clyso.com, Ilya Dryomov , Alex Markuze Cc: ceph-devel@vger.kernel.org, linux-kernel@vger.kernel.org Date: Wed, 15 Jul 2026 11:53:01 -0700 In-Reply-To: <20260715-ceph-mdsc-mutex-optimization-v2-4-90e81b726724@clyso.com> References: <20260715-ceph-mdsc-mutex-optimization-v2-0-90e81b726724@clyso.com> <20260715-ceph-mdsc-mutex-optimization-v2-4-90e81b726724@clyso.com> Autocrypt: addr=slava@dubeyko.com; prefer-encrypt=mutual; keydata=mQINBGgaTLYBEADaJc/WqWTeunGetXyyGJ5Za7b23M/ozuDCWCp+yWUa2GqQKH40dxRIR zshgOmAue7t9RQJU9lxZ4ZHWbi1Hzz85+0omefEdAKFmxTO6+CYV0g/sapU0wPJws3sC2Pbda9/eJ ZcvScAX2n/PlhpTnzJKf3JkHh3nM1ACO3jzSe2/muSQJvqMLG2D71ccekr1RyUh8V+OZdrPtfkDam V6GOT6IvyE+d+55fzmo20nJKecvbyvdikWwZvjjCENsG9qOf3TcCJ9DDYwjyYe1To8b+mQM9nHcxp jUsUuH074BhISFwt99/htZdSgp4csiGeXr8f9BEotRB6+kjMBHaiJ6B7BIlDmlffyR4f3oR/5hxgy dvIxMocqyc03xVyM6tA4ZrshKkwDgZIFEKkx37ec22ZJczNwGywKQW2TGXUTZVbdooiG4tXbRBLxe ga/NTZ52ZdEkSxAUGw/l0y0InTtdDIWvfUT+WXtQcEPRBE6HHhoeFehLzWL/o7w5Hog+0hXhNjqte fzKpI2fWmYzoIb6ueNmE/8sP9fWXo6Av9m8B5hRvF/hVWfEysr/2LSqN+xjt9NEbg8WNRMLy/Y0MS p5fgf9pmGF78waFiBvgZIQNuQnHrM+0BmYOhR0JKoHjt7r5wLyNiKFc8b7xXndyCDYfniO3ljbr0j tXWRGxx4to6FwARAQABtCZWaWFjaGVzbGF2IER1YmV5a28gPHNsYXZhQGR1YmV5a28uY29tPokCVw QTAQoAQQIbAQUJA8JnAAULCQgHAgYVCgkICwIEFgIDAQIeAQIXgBYhBFXDC2tnzsoLQtrbBDlc2cL fhEB1BQJoGl5PAhkBAAoJEDlc2cLfhEB17DsP/jy/Dx19MtxWOniPqpQf2s65enkDZuMIQ94jSg7B F2qTKIbNR9SmsczjyjC+/J7m7WZRmcqnwFYMOyNfh12aF2WhjT7p5xEAbvfGVYwUpUrg/lcacdT0D Yk61GGc5ZB89OAWHLr0FJjI54bd7kn7E/JRQF4dqNsxU8qcPXQ0wLHxTHUPZu/w5Zu/cO+lQ3H0Pj pSEGaTAh+tBYGSvQ4YPYBcV8+qjTxzeNwkw4ARza8EjTwWKP2jWAfA/ay4VobRfqNQ2zLoo84qDtN Uxe0zPE2wobIXELWkbuW/6hoQFPpMlJWz+mbvVms57NAA1HO8F5c1SLFaJ6dN0AQbxrHi45/cQXla 9hSEOJjxcEnJG/ZmcomYHFneM9K1p1K6HcGajiY2BFWkVet9vuHygkLWXVYZ0lr1paLFR52S7T+cf 6dkxOqu1ZiRegvFoyzBUzlLh/elgp3tWUfG2VmJD3lGpB3m5ZhwQ3rFpK8A7cKzgKjwPp61Me0o9z HX53THoG+QG+o0nnIKK7M8+coToTSyznYoq9C3eKeM/J97x9+h9tbizaeUQvWzQOgG8myUJ5u5Dr4 6tv9KXrOJy0iy/dcyreMYV5lwODaFfOeA4Lbnn5vRn9OjuMg1PFhCi3yMI4lA4umXFw0V2/OI5rgW BQELhfvW6mxkihkl6KLZX8m1zcHitCpWaWFjaGVzbGF2IER1YmV5a28gPFNsYXZhLkR1YmV5a29Aa WJtLmNvbT6JAlQEEwEKAD4WIQRVwwtrZ87KC0La2wQ5XNnC34RAdQUCaBpd7AIbAQUJA8JnAAULCQ gHAgYVCgkICwIEFgIDAQIeAQIXgAAKCRA5XNnC34RAdYjFEACiWBEybMt1xjRbEgaZ3UP5i2bSway DwYDvgWW5EbRP7JcqOcZ2vkJwrK3gsqC3FKpjOPh7ecE0I4vrabH1Qobe2N8B2Y396z24mGnkTBbb 16Uz3PC93nFN1BA0wuOjlr1/oOTy5gBY563vybhnXPfSEUcXRd28jI7z8tRyzXh2tL8ZLdv1u4vQ8 E0O7lVJ55p9yGxbwgb5vXU4T2irqRKLxRvU80rZIXoEM7zLf5r7RaRxgwjTKdu6rYMUOfoyEQQZTD 4Xg9YE/X8pZzcbYFs4IlscyK6cXU0pjwr2ssjearOLLDJ7ygvfOiOuCZL+6zHRunLwq2JH/RmwuLV mWWSbgosZD6c5+wu6DxV15y7zZaR3NFPOR5ErpCFUorKzBO1nA4dwOAbNym9OGkhRgLAyxwpea0V0 ZlStfp0kfVaSZYo7PXd8Bbtyjali0niBjPpEVZdgtVUpBlPr97jBYZ+L5GF3hd6WJFbEYgj+5Af7C UjbX9DHweGQ/tdXWRnJHRzorxzjOS3003ddRnPtQDDN3Z/XzdAZwQAs0RqqXrTeeJrLppFUbAP+HZ TyOLVJcAAlVQROoq8PbM3ZKIaOygjj6Yw0emJi1D9OsN2UKjoe4W185vamFWX4Ba41jmCPrYJWAWH fAMjjkInIPg7RLGs8FiwxfcpkILP0YbVWHiNAabQoVmlhY2hlc2xhdiBEdWJleWtvIDx2ZHViZXlr b0BrZXJuZWwub3JnPokCVAQTAQoAPhYhBFXDC2tnzsoLQtrbBDlc2cLfhEB1BQJoVemuAhsBBQkDw mcABQsJCAcCBhUKCQgLAgQWAgMBAh4BAheAAAoJEDlc2cLfhEB1GRwP/1scX5HO9Sk7dRicLD/fxo ipwEs+UbeA0/TM8OQfdRI4C/tFBYbQCR7lD05dfq8VsYLEyrgeLqP/iRhabLky8LTaEdwoAqPDc/O 9HRffx/faJZqkKc1dZryjqS6b8NExhKOVWmDqN357+Cl/H4hT9wnvjCj1YEqXIxSd/2Pc8+yw/KRC AP7jtRzXHcc/49Lpz/NU5irScusxy2GLKa5o/13jFK3F1fWX1wsOJF8NlTx3rLtBy4GWHITwkBmu8 zI4qcJGp7eudI0l4xmIKKQWanEhVdzBm5UnfyLIa7gQ2T48UbxJlWnMhLxMPrxgtC4Kos1G3zovEy Ep+fJN7D1pwN9aR36jVKvRsX7V4leIDWGzCdfw1FGWkMUfrRwgIl6i3wgqcCP6r9YSWVQYXdmwdMu 1RFLC44iF9340S0hw9+30yGP8TWwd1mm8V/+zsdDAFAoAwisi5QLLkQnEsJSgLzJ9daAsE8KjMthv hUWHdpiUSjyCpigT+KPl9YunZhyrC1jZXERCDPCQVYgaPt+Xbhdjcem/ykv8UVIDAGVXjuk4OW8la nf8SP+uxkTTDKcPHOa5rYRaeNj7T/NClRSd4z6aV3F6pKEJnEGvv/DFMXtSHlbylhyiGKN2Amd0b4 9jg+DW85oNN7q2UYzYuPwkHsFFq5iyF1QggiwYYTpoVXsw Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable User-Agent: Evolution 3.60.1 (by Flathub.org) Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 On Wed, 2026-07-15 at 11:58 +0800, Xiubo Li via B4 Relay wrote: > From: Xiubo Li >=20 > __do_request() now acquires mdsc->mutex on entry and unlocks at > every exit point, so callers no longer need to hold it.=C2=A0 As a > result kick_requests() is now completely free of mdsc->mutex: > xa_for_each() is internally locked, list_del_init() is guarded by > wait_list_lock, and __do_request() manages its own serialization. >=20 > Remove the no-longer-needed mutex_lock/unlock around __do_request() > in ceph_mdsc_submit_request() and handle_forward().=C2=A0 The do_request(= ) > wrapper introduced in the previous commit is no longer necessary > and is dropped. Which do_request() wrapper do you mean here? Maybe, I am missing something. >=20 > Signed-off-by: Xiubo Li > --- > =C2=A0fs/ceph/mds_client.c | 48 +++++++++++++++++++++++++----------------= - > ------ > =C2=A01 file changed, 25 insertions(+), 23 deletions(-) >=20 > diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c > index a250938f32d3..be0dae919d69 100644 > --- a/fs/ceph/mds_client.c > +++ b/fs/ceph/mds_client.c > @@ -3637,9 +3637,12 @@ static void __do_request(struct > ceph_mds_client *mdsc, > =C2=A0 int err =3D 0; > =C2=A0 bool random; > =C2=A0 > + mutex_lock(&mdsc->mutex); > + > =C2=A0 if (req->r_err || test_bit(CEPH_MDS_R_GOT_RESULT, &req- > >r_req_flags)) { > =C2=A0 if (test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags)) > =C2=A0 __unregister_request(mdsc, req); > + mutex_unlock(&mdsc->mutex); > =C2=A0 return; > =C2=A0 } > =C2=A0 > @@ -3672,6 +3675,7 @@ static void __do_request(struct ceph_mds_client > *mdsc, > =C2=A0 spin_lock(&mdsc->wait_list_lock); > =C2=A0 list_add(&req->r_wait, &mdsc- > >waiting_for_map); > =C2=A0 spin_unlock(&mdsc->wait_list_lock); > + mutex_unlock(&mdsc->mutex); > =C2=A0 return; > =C2=A0 } > =C2=A0 if (!(mdsc->fsc->mount_options->flags & > @@ -3697,6 +3701,7 @@ static void __do_request(struct ceph_mds_client > *mdsc, > =C2=A0 spin_lock(&mdsc->wait_list_lock); > =C2=A0 list_add(&req->r_wait, &mdsc->waiting_for_map); > =C2=A0 spin_unlock(&mdsc->wait_list_lock); > + mutex_unlock(&mdsc->mutex); > =C2=A0 return; > =C2=A0 } > =C2=A0 > @@ -3770,6 +3775,8 @@ static void __do_request(struct ceph_mds_client > *mdsc, > =C2=A0 goto out_session; > =C2=A0 } > =C2=A0 > + mutex_unlock(&mdsc->mutex); > + > =C2=A0 /* send request */ > =C2=A0 req->r_resend_mds =3D -1;=C2=A0=C2=A0 /* forget any previous mds h= int */ > =C2=A0 > @@ -3800,6 +3807,7 @@ static void __do_request(struct ceph_mds_client > *mdsc, > =C2=A0 err =3D wait_on_bit(&di->flags, > CEPH_DENTRY_ASYNC_CREATE_BIT, > =C2=A0 =C2=A0 TASK_KILLABLE); > =C2=A0 if (err) { > + mutex_lock(&mdsc->mutex); > =C2=A0 mutex_lock(&req->r_fill_mutex); > =C2=A0 set_bit(CEPH_MDS_R_ABORTED, &req- > >r_req_flags); > =C2=A0 mutex_unlock(&req->r_fill_mutex); > @@ -3837,6 +3845,8 @@ static void __do_request(struct ceph_mds_client > *mdsc, > =C2=A0 > =C2=A0 err =3D __send_request(session, req, false); > =C2=A0 > + mutex_lock(&mdsc->mutex); > + > =C2=A0out_session: > =C2=A0 ceph_put_mds_session(session); > =C2=A0finish: > @@ -3846,12 +3856,10 @@ static void __do_request(struct > ceph_mds_client *mdsc, > =C2=A0 complete_request(mdsc, req); > =C2=A0 __unregister_request(mdsc, req); > =C2=A0 } > + mutex_unlock(&mdsc->mutex); > =C2=A0 return; > =C2=A0} > =C2=A0 > -/* > - * called under mdsc->mutex > - */ > =C2=A0static void __wake_requests(struct ceph_mds_client *mdsc, > =C2=A0 =C2=A0=C2=A0=C2=A0 struct list_head *head) > =C2=A0{ > @@ -3969,11 +3977,12 @@ int ceph_mdsc_submit_request(struct > ceph_mds_client *mdsc, struct inode *dir, > =C2=A0 doutc(cl, "submit_request on %p for inode %p\n", req, dir); > =C2=A0 mutex_lock(&mdsc->mutex); > =C2=A0 __register_request(mdsc, req, dir); > + mutex_unlock(&mdsc->mutex); > + > =C2=A0 trace_ceph_mdsc_submit_request(mdsc, req); > =C2=A0 if (!req->r_err) > =C2=A0 __do_request(mdsc, req); > =C2=A0 err =3D req->r_err; > - mutex_unlock(&mdsc->mutex); > =C2=A0 return err; > =C2=A0} > =C2=A0 > @@ -4346,13 +4355,14 @@ static void handle_forward(struct > ceph_mds_client *mdsc, > =C2=A0 req->r_num_fwd =3D fwd_seq; > =C2=A0 req->r_resend_mds =3D next_mds; > =C2=A0 put_request_session(req); > - __do_request(mdsc, req); > =C2=A0 } > =C2=A0 mutex_unlock(&mdsc->mutex); > =C2=A0 > =C2=A0 /* kick calling process */ > =C2=A0 if (aborted) > =C2=A0 complete_request(mdsc, req); > + else if (!test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags)) > + __do_request(mdsc, req); > =C2=A0 ceph_mdsc_put_request(req); > =C2=A0 return; > =C2=A0 > @@ -4676,11 +4686,9 @@ static void handle_session(struct > ceph_mds_session *session, > =C2=A0 > =C2=A0 mutex_unlock(&session->s_mutex); > =C2=A0 if (wake) { > - mutex_lock(&mdsc->mutex); > =C2=A0 __wake_requests(mdsc, &session->s_waiting); > =C2=A0 if (wake =3D=3D 2) > =C2=A0 kick_requests(mdsc, mds); > - mutex_unlock(&mdsc->mutex); > =C2=A0 } > =C2=A0 if (op =3D=3D CEPH_SESSION_CLOSE) > =C2=A0 ceph_put_mds_session(session); > @@ -5326,9 +5334,7 @@ static int send_mds_reconnect(struct > ceph_mds_client *mdsc, > =C2=A0 > =C2=A0 mutex_unlock(&session->s_mutex); > =C2=A0 > - mutex_lock(&mdsc->mutex); > =C2=A0 __wake_requests(mdsc, &session->s_waiting); > - mutex_unlock(&mdsc->mutex); > =C2=A0 > =C2=A0 up_read(&mdsc->snap_rwsem); > =C2=A0 ceph_pagelist_release(recon_state.pagelist); > @@ -5773,8 +5779,8 @@ static void ceph_mdsc_reset_workfn(struct > work_struct *work) > =C2=A0 } > =C2=A0 sessions[i]->s_state =3D CEPH_MDS_SESSION_CLOSED; > =C2=A0 __unregister_session(mdsc, sessions[i]); > - __wake_requests(mdsc, &sessions[i]->s_waiting); > =C2=A0 mutex_unlock(&mdsc->mutex); > + __wake_requests(mdsc, &sessions[i]->s_waiting); > =C2=A0 > =C2=A0 mutex_lock(&sessions[i]->s_mutex); > =C2=A0 cleanup_session_requests(mdsc, sessions[i]); > @@ -5785,9 +5791,7 @@ static void ceph_mdsc_reset_workfn(struct > work_struct *work) > =C2=A0 > =C2=A0 ceph_put_mds_session(sessions[i]); > =C2=A0 > - mutex_lock(&mdsc->mutex); > =C2=A0 kick_requests(mdsc, mds); > - mutex_unlock(&mdsc->mutex); > =C2=A0 > =C2=A0 torn_down++; > =C2=A0 pr_info_client(cl, "mds%d session reset complete\n", > mds); > @@ -5903,8 +5907,8 @@ static void check_new_map(struct > ceph_mds_client *mdsc, > =C2=A0 /* force close session for stopped mds */ > =C2=A0 ceph_get_mds_session(s); > =C2=A0 __unregister_session(mdsc, s); > - __wake_requests(mdsc, &s->s_waiting); > =C2=A0 mutex_unlock(&mdsc->mutex); > + __wake_requests(mdsc, &s->s_waiting); > =C2=A0 > =C2=A0 mutex_lock(&s->s_mutex); > =C2=A0 cleanup_session_requests(mdsc, s); > @@ -5913,8 +5917,8 @@ static void check_new_map(struct > ceph_mds_client *mdsc, > =C2=A0 > =C2=A0 ceph_put_mds_session(s); > =C2=A0 > - mutex_lock(&mdsc->mutex); > =C2=A0 kick_requests(mdsc, i); > + mutex_lock(&mdsc->mutex); > =C2=A0 continue; > =C2=A0 } > =C2=A0 > @@ -5958,8 +5962,8 @@ static void check_new_map(struct > ceph_mds_client *mdsc, > =C2=A0 =C2=A0=C2=A0=C2=A0 oldstate !=3D CEPH_MDS_STATE_STARTING) > =C2=A0 pr_info_client(cl, "mds%d recovery > completed\n", > =C2=A0 =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 s->s_mds); > - kick_requests(mdsc, i); > =C2=A0 mutex_unlock(&mdsc->mutex); > + kick_requests(mdsc, i); > =C2=A0 mutex_lock(&s->s_mutex); > =C2=A0 mutex_lock(&mdsc->mutex); > =C2=A0 ceph_kick_flushing_caps(mdsc, s); > @@ -6931,8 +6935,8 @@ void ceph_mdsc_force_umount(struct > ceph_mds_client *mdsc) > =C2=A0 > =C2=A0 if (session->s_state =3D=3D CEPH_MDS_SESSION_REJECTED) > =C2=A0 __unregister_session(mdsc, session); > - __wake_requests(mdsc, &session->s_waiting); > =C2=A0 mutex_unlock(&mdsc->mutex); > + __wake_requests(mdsc, &session->s_waiting); > =C2=A0 > =C2=A0 mutex_lock(&session->s_mutex); > =C2=A0 __close_session(mdsc, session); > @@ -6943,11 +6947,11 @@ void ceph_mdsc_force_umount(struct > ceph_mds_client *mdsc) > =C2=A0 mutex_unlock(&session->s_mutex); > =C2=A0 ceph_put_mds_session(session); > =C2=A0 > - mutex_lock(&mdsc->mutex); > =C2=A0 kick_requests(mdsc, mds); > + mutex_lock(&mdsc->mutex); > =C2=A0 } > - __wake_requests(mdsc, &mdsc->waiting_for_map); > =C2=A0 mutex_unlock(&mdsc->mutex); > + __wake_requests(mdsc, &mdsc->waiting_for_map); > =C2=A0} > =C2=A0 > =C2=A0static void ceph_mdsc_stop(struct ceph_mds_client *mdsc) > @@ -7091,8 +7095,8 @@ void ceph_mdsc_handle_fsmap(struct > ceph_mds_client *mdsc, struct ceph_msg *msg) > =C2=A0err_out: > =C2=A0 mutex_lock(&mdsc->mutex); > =C2=A0 mdsc->mdsmap_err =3D err; > - __wake_requests(mdsc, &mdsc->waiting_for_map); > =C2=A0 mutex_unlock(&mdsc->mutex); > + __wake_requests(mdsc, &mdsc->waiting_for_map); > =C2=A0} > =C2=A0 > =C2=A0/* > @@ -7143,11 +7147,11 @@ void ceph_mdsc_handle_mdsmap(struct > ceph_mds_client *mdsc, struct ceph_msg *msg) > =C2=A0 mdsc->fsc->max_file_size =3D min((loff_t)mdsc->mdsmap- > >m_max_file_size, > =C2=A0 MAX_LFS_FILESIZE); > =C2=A0 > + mutex_unlock(&mdsc->mutex); > =C2=A0 __wake_requests(mdsc, &mdsc->waiting_for_map); > =C2=A0 ceph_monc_got_map(&mdsc->fsc->client->monc, CEPH_SUB_MDSMAP, > =C2=A0 =C2=A0 mdsc->mdsmap->m_epoch); > =C2=A0 > - mutex_unlock(&mdsc->mutex); > =C2=A0 schedule_delayed(mdsc, 0); > =C2=A0 return; > =C2=A0 > @@ -7245,8 +7249,8 @@ static void mds_peer_reset(struct > ceph_connection *con) > =C2=A0 ceph_get_mds_session(s); > =C2=A0 s->s_state =3D CEPH_MDS_SESSION_CLOSED; > =C2=A0 __unregister_session(mdsc, s); > - __wake_requests(mdsc, &s->s_waiting); > =C2=A0 mutex_unlock(&mdsc->mutex); > + __wake_requests(mdsc, &s->s_waiting); > =C2=A0 > =C2=A0 mutex_lock(&s->s_mutex); > =C2=A0 cleanup_session_requests(mdsc, s); > @@ -7255,9 +7259,7 @@ static void mds_peer_reset(struct > ceph_connection *con) > =C2=A0 > =C2=A0 wake_up_all(&mdsc->session_close_wq); > =C2=A0 > - mutex_lock(&mdsc->mutex); > =C2=A0 kick_requests(mdsc, s->s_mds); > - mutex_unlock(&mdsc->mutex); > =C2=A0 > =C2=A0 ceph_put_mds_session(s); > =C2=A0 break; Maybe, I am missing something. But kick_requests() reads req- >r_attempts [1] and req->r_session [2] with no lock at all. While __do_request()/__prepare_send_request() can write those same fields [3,4] from another thread without holding mdsc->mutex at the write point either. Could it be the issue? Thanks, Slava. [1] https://elixir.bootlin.com/linux/v7.2-rc3/source/fs/ceph/mds_client.c#L3832 [2] https://elixir.bootlin.com/linux/v7.2-rc3/source/fs/ceph/mds_client.c#L3834 [3] https://elixir.bootlin.com/linux/v7.2-rc3/source/fs/ceph/mds_client.c#L3658 [4] https://elixir.bootlin.com/linux/v7.2-rc3/source/fs/ceph/mds_client.c#L3475