From nobody Fri Dec 19 07:23:37 2025 Received: from mail-oa1-f97.google.com (mail-oa1-f97.google.com [209.85.160.97]) (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 E11F4288CA6 for ; Wed, 17 Dec 2025 05:35:16 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.160.97 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1765949721; cv=none; b=QN3et+cNGdcYtqIEDH+2PjUCODC1fF0MaRCnI3b42Olpy0u76GN1/+h5vI3uIOJF/JVXdfhXQy/mOiJKDSPZ+9PMpaID6pns3oQETqVJuZ3Epe98w1z34oxReU74EI6m2B0k3qXSLA+8rba6QS9OsSWrd5bccamiX2FgR73SyGY= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1765949721; c=relaxed/simple; bh=gCDcxrwsQIo9Zl4D9+H5tFsOAIyHBgcO/hwZeW+CHUU=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=CdjiY/q5+2PEz3voOFGKeB3hn1KMOl9oUEDHkOKVj07z+3z4E62qI+g9O+evlJZGrKEOWgc36XdhYJ4/WiKIEm6nFkr225EcePOkH75ON+T2kdriDjTrPQvPMTrIRGsPZCwZfRT0+oEimciyB+31fmNgeWf+O3mx6CLebm/BZHE= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=purestorage.com; spf=fail smtp.mailfrom=purestorage.com; dkim=pass (2048-bit key) header.d=purestorage.com header.i=@purestorage.com header.b=KPEueBWg; arc=none smtp.client-ip=209.85.160.97 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=purestorage.com Authentication-Results: smtp.subspace.kernel.org; spf=fail smtp.mailfrom=purestorage.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=purestorage.com header.i=@purestorage.com header.b="KPEueBWg" Received: by mail-oa1-f97.google.com with SMTP id 586e51a60fabf-3f9fb53bea4so30379fac.0 for ; Tue, 16 Dec 2025 21:35:16 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=purestorage.com; s=google2022; t=1765949714; x=1766554514; darn=vger.kernel.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=2OvJx7AxTUJ0pacBHpfR4EB7axuem5OmBv1tntyjd18=; b=KPEueBWgGwFpRBHMsmlxq+17449raT7ttquKsRKjNsssQxetX3ulLY7866k/dwShWw 5Tm9ENFk/TOFJwhRLAGay1NBoN8m+tD51thMeLoC1WbUG8U957EzPYrqQiBSmGoDbO6s nR4Og2Qz+z03peiVFERzOfjBW2bXcCjrjEvOdibugyDL1iRNQlosBPXhAGGKQa7ilDev /GX1stlQj2fxDySShZUcxHPI5bgxFRlr7QrNewYs4SPBT2857m1bhsZI6AD38KO1X5Lt 6Y+FQa1fv3IUITOKmWt9FNlGYSj83JBRTdg+j+ddPCUDG/HqlRiWk5T8gr6iHmOBobrZ vAIA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1765949714; x=1766554514; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-gg:x-gm-message-state:from :to:cc:subject:date:message-id:reply-to; bh=2OvJx7AxTUJ0pacBHpfR4EB7axuem5OmBv1tntyjd18=; b=aHbdLpyeermKJpzHTUsAdsGb3hsYKe1AQCg5Zn/aRURkjcA9sn+1+6jCrbTtk4BT+E cBHqFqbEQ8s6H4c9+PKND2zxsWizuVocR/qMhvSLdglst/bSozkSUoT3c2kit2xdt/5V 0uWsOZy6dso5qrpBHW/7f2wrp26Pf5hF6dBvJAPdk2wEtXKUuyJlxTC1lMLL4MRs6xD/ XAxmat14+TZjHcjTm0KqjFFmlAaTIRK5JrgWPhfOC3mm5PcJcoM+YNRMXsnKonz3ff8q h5qKtsdyR7bgQnrmtfYD9PDPiFPxGBz720K0myagCOIVFLPf7WY6O9iOsUR35nTgoV49 N6Kg== X-Forwarded-Encrypted: i=1; AJvYcCV3eByilXRGWqOvR08srZom7d8iT1gYWwIdj2HYOoQpGrpzWUwG5wCvsyobB3x05fiBdhsm544I1nFHMiI=@vger.kernel.org X-Gm-Message-State: AOJu0Yw39fWiDGGQ5koIna70/ua7ZVJvFKz+fupjGINNT2GYYZZTRgaB u3mAL30errei5FgutWgJ7zNYH0b6g7/ntMuCCVyt85oZ8teD5xzcR/ailqJggrfHoykCKm7K40p qJvTDaVU38vOzY7wHS3ggV3peCMaUw3mGKt9EVA28Nm5QNpgTLwNK X-Gm-Gg: AY/fxX4pIanBjkGEzS7ke8Qnoy2WeCbuRt+xAlG07J3ntqIW1khIVXZObVeoTF1YT7s jgGy0zcuf5R1n9PUUC9E2vsYtJzLe31/IZWKrJXEiYq7/O7wC7p777nCT8ypFHldsSjoc4tuV85 3qJ+zg7iwCwuhEqnkycmg5UMk3k+komUMVkdF/YDKXqmKXlWUuNBbs0YHRTLoHebyjC3QTrE8mI 3oTf1krRWEcoATHUp/hr+L5iSnQ+QH60v1ReM16PVhDt5bb+DgAi0NasMporZ/R0GVilyDwK6t1 xxcOQVVOBL6NAPnsFWlJRLoQhMX/dusTbXhW2Bobeqyl/DzaO5SWGtznnI113vy1GxODEzFXwBj xAEJas8PyvD/B6uRboS++ruAbZaE= X-Google-Smtp-Source: AGHT+IFbG6sbuUMK8Z52DIdwdRmhPJfwF/v7UIPe9gL8Y8b6oOH7po6qZZFyRaDqM7Jnd07PI/nXVcYALu2n X-Received: by 2002:a05:6870:e8e:b0:3ec:3bfe:bda7 with SMTP id 586e51a60fabf-3f5f8a3a5admr7121872fac.1.1765949713810; Tue, 16 Dec 2025 21:35:13 -0800 (PST) Received: from c7-smtp-2023.dev.purestorage.com ([2620:125:9017:12:36:3:5:0]) by smtp-relay.gmail.com with ESMTPS id 586e51a60fabf-3f614e7dab4sm1520399fac.18.2025.12.16.21.35.13 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 16 Dec 2025 21:35:13 -0800 (PST) X-Relaying-Domain: purestorage.com Received: from dev-csander.dev.purestorage.com (unknown [IPv6:2620:125:9007:640:ffff::1199]) by c7-smtp-2023.dev.purestorage.com (Postfix) with ESMTP id E028D3420F0; Tue, 16 Dec 2025 22:35:12 -0700 (MST) Received: by dev-csander.dev.purestorage.com (Postfix, from userid 1557716354) id DCFA5E41A08; Tue, 16 Dec 2025 22:35:12 -0700 (MST) From: Caleb Sander Mateos To: Ming Lei , Jens Axboe , Shuah Khan Cc: linux-block@vger.kernel.org, linux-kselftest@vger.kernel.org, linux-kernel@vger.kernel.org, Stanley Zhang , Uday Shankar , Caleb Sander Mateos Subject: [PATCH 13/20] ublk: optimize ublk_user_copy() on daemon task Date: Tue, 16 Dec 2025 22:34:47 -0700 Message-ID: <20251217053455.281509-14-csander@purestorage.com> X-Mailer: git-send-email 2.45.2 In-Reply-To: <20251217053455.281509-1-csander@purestorage.com> References: <20251217053455.281509-1-csander@purestorage.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" ublk user copy syscalls may be issued from any task, so they take a reference count on the struct ublk_io to check whether it is owned by the ublk server and prevent a concurrent UBLK_IO_COMMIT_AND_FETCH_REQ from completing the request. However, if the user copy syscall is issued on the io's daemon task, a concurrent UBLK_IO_COMMIT_AND_FETCH_REQ isn't possible, so the atomic reference count dance is unnecessary. Check for UBLK_IO_FLAG_OWNED_BY_SRV to ensure the request is dispatched to the sever and obtain the request from ublk_io's req field instead of looking it up on the tagset. Skip the reference count increment and decrement. Commit 8a8fe42d765b ("ublk: optimize UBLK_IO_REGISTER_IO_BUF on daemon task") made an analogous optimization for ublk zero copy buffer registration. Signed-off-by: Caleb Sander Mateos --- drivers/block/ublk_drv.c | 23 ++++++++++++++++++----- 1 file changed, 18 insertions(+), 5 deletions(-) diff --git a/drivers/block/ublk_drv.c b/drivers/block/ublk_drv.c index 042df4de9253..a0fbabd49feb 100644 --- a/drivers/block/ublk_drv.c +++ b/drivers/block/ublk_drv.c @@ -180,11 +180,11 @@ struct ublk_io { /* * The number of uses of this I/O by the ublk server * if user copy or zero copy are enabled: * - UBLK_REFCOUNT_INIT from dispatch to the server * until UBLK_IO_COMMIT_AND_FETCH_REQ - * - 1 for each inflight ublk_ch_{read,write}_iter() call + * - 1 for each inflight ublk_ch_{read,write}_iter() call not on task * - 1 for each io_uring registered buffer not registered on task * The I/O can only be completed once all references are dropped. * User copy and buffer registration operations are only permitted * if the reference count is nonzero. */ @@ -2644,10 +2644,11 @@ ublk_user_copy(struct kiocb *iocb, struct iov_iter = *iter, int dir) struct ublk_queue *ubq; struct request *req; struct ublk_io *io; unsigned data_len; bool is_integrity; + bool on_daemon; size_t buf_off; u16 tag, q_id; ssize_t ret; =20 if (!user_backed_iter(iter)) @@ -2670,13 +2671,24 @@ ublk_user_copy(struct kiocb *iocb, struct iov_iter = *iter, int dir) =20 if (tag >=3D ub->dev_info.queue_depth) return -EINVAL; =20 io =3D &ubq->ios[tag]; - req =3D __ublk_check_and_get_req(ub, q_id, tag, io); - if (!req) - return -EINVAL; + on_daemon =3D current =3D=3D READ_ONCE(io->task); + if (on_daemon) { + /* On daemon, io can't be completed concurrently, so skip ref */ + if (!(io->flags & UBLK_IO_FLAG_OWNED_BY_SRV)) + return -EINVAL; + + req =3D io->req; + if (!ublk_rq_has_data(req)) + return -EINVAL; + } else { + req =3D __ublk_check_and_get_req(ub, q_id, tag, io); + if (!req) + return -EINVAL; + } =20 if (is_integrity) { struct blk_integrity *bi =3D &req->q->limits.integrity; =20 data_len =3D bio_integrity_bytes(bi, blk_rq_sectors(req)); @@ -2697,11 +2709,12 @@ ublk_user_copy(struct kiocb *iocb, struct iov_iter = *iter, int dir) ret =3D ublk_copy_user_integrity(req, buf_off, iter, dir); else ret =3D ublk_copy_user_pages(req, buf_off, iter, dir); =20 out: - ublk_put_req_ref(io, req); + if (!on_daemon) + ublk_put_req_ref(io, req); return ret; } =20 static ssize_t ublk_ch_read_iter(struct kiocb *iocb, struct iov_iter *to) { --=20 2.45.2