From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (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 C2939305662 for ; Mon, 3 Aug 2026 06:59:35 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1785740377; cv=none; b=lEV67yaJ9TtqNVGwgFLAaTUQ78PcGMTvOZmSDddxBTqOciPv9wzzR76zLTA9gptpsf6FKV2SPgSCv/ynKlIxA0OIyLin2vOS4oXmc8N8OZl6DVJI6wcc4ooDyThx7WaTF/ZANSjXJJEYOft4PfGL7qIG6gSqGGlQ3H+zaCduSjY= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1785740377; c=relaxed/simple; bh=1H2uacMBodTYUmv9d7HnGS4g9fXQfyBpHWEu1mHZeoM=; h=Message-ID:Date:MIME-Version:Cc:Subject:To:References:From: In-Reply-To:Content-Type; b=c6EzgM07liOlv1AGXhKaU4mh6nkiiQ3O1Qgd5chBw87oBgKu27nss01PZuHuS9aZuYBlaMctT2hKW/QHcTTGTluu+Q7salSqrQB2V9nzxCfjI36EglR7EOfXwEOCMWobtoA/zZtQ8nJjSplsfejMmiCo4hhB9QPAm4ds9FwMq4w= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=LcIFT9oE; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="LcIFT9oE" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 020311F000E9; Mon, 3 Aug 2026 06:59:33 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1785740375; bh=37UE+zEK5cc8ff/wYweX5ca4u9HVNotxR09GlvhGqwM=; h=Date:Cc:Subject:To:References:From:In-Reply-To; b=LcIFT9oEcMSU8du99sOPLL4hXjeCx7FaiI2jZIrqWCRXpBnh8yQSeoqZT6jArgNwb LwSRlkYBn2aZodGrmAFaT/HgvtHARyLkO4GC+/jyZ4vH/Qytfi4U8gy9qlLVa6NP9h cfeE/sHcAcBWE1/iPzWYN0Db0sNiadIMRr5W1MvAseA9QH0ThSR+bFOxJfJHCd1JEQ jJXisxeVcxHHNFI3agsEX5QdAP/8xzMzQgQUZs1yTgPJKS0fs0Z8qvwrmpXucgny0p suHbQAX2fdru28h9xhBVnBtJ9ggAEXKxzQjDDAuN/tOPPdxFrlLCHaAhE0CX8kk//F NS0PT6zVG7rnQ== Message-ID: <9fc81e1e-ec13-4cc0-9aed-8270bf1a4157@kernel.org> Date: Mon, 3 Aug 2026 14:59:31 +0800 Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Cc: chao@kernel.org, "linux-f2fs-devel@lists.sourceforge.net" , "linux-kernel@vger.kernel.org" , Dongjin Kim , Daejun Park Subject: Re: [PATCH] f2fs: issue multi-device flushes in parallel To: yonggil.song@samsung.com, "jaegeuk@kernel.org" References: <20260713055711epcms2p712e7add62211e42e995c54a92fd4ff0c@epcms2p7> Content-Language: en-US From: Chao Yu In-Reply-To: <20260713055711epcms2p712e7add62211e42e995c54a92fd4ff0c@epcms2p7> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit On 7/13/26 13:57, Yonggil Song wrote: > On a multi-device setup, submit_flush_wait() walked the dirty devices > in order and aborted the whole loop on the first device whose flush > failed, leaving the remaining dirty devices un-flushed. Each device > still needs its own data made durable, so a failure on one device must > not skip the others. It also waited for one device's flush to complete > before issuing the next, even though the devices have independent > flush queues and could be flushed concurrently. > > Flush every dirty device best-effort and in parallel instead: build > one PREFLUSH bio per dirty device, submit them all, then wait for > every completion, returning the first error seen (0 if all succeed). > This bounds the flush window by the slowest device rather than the sum > of all of them. No caller depends on the previous early-abort > behaviour -- fsync only checks whether the return value is zero > (fs/f2fs/file.c). The checkpoint path (f2fs_flush_device_cache) is > unaffected; this only touches the fsync flush path. > > Signed-off-by: Yonggil Song > --- > fs/f2fs/segment.c | 84 +++++++++++++++++++++++++++++++++++++++++++++++++++---- > 1 file changed, 78 insertions(+), 6 deletions(-) > > diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c > index d71ddb3ee918..51d7f76e3d1d 100644 > --- a/fs/f2fs/segment.c > +++ b/fs/f2fs/segment.c > @@ -566,24 +566,96 @@ static int __submit_flush_wait(struct f2fs_sb_info *sbi, > return ret; > } > > -static int submit_flush_wait(struct f2fs_sb_info *sbi, nid_t ino) > +static void f2fs_flush_end_io(struct bio *bio) > +{ > + complete(bio->bi_private); > +} > + > +struct f2fs_flush_bio { > + struct bio bio; > + struct completion wait; > +}; > + > +/* > + * Flush every dirty device best-effort: a failure on one device must not > + * skip the flush on the remaining dirty devices, since each device still > + * needs its own data made durable. Report the first error. > + */ > +static int submit_flush_wait_serial(struct f2fs_sb_info *sbi, nid_t ino) > { > int ret = 0; > int i; > > - if (!f2fs_is_multi_device(sbi)) > - return __submit_flush_wait(sbi, sbi->sb->s_bdev); > + for (i = 0; i < sbi->s_ndevs; i++) { > + int err; > + > + if (!f2fs_is_dirty_device(sbi, ino, i, FLUSH_INO)) > + continue; > + err = __submit_flush_wait(sbi, FDEV(i).bdev); > + if (err && !ret) > + ret = err; > + } > + return ret; > +} > + > +/* > + * Same best-effort/first-error contract as submit_flush_wait_serial(), but > + * issue every dirty device's flush before waiting for any of them, so the > + * per-device flush latencies overlap instead of adding up in series. Fall > + * back to the serial path if the bio array cannot be allocated. > + */ > +static int submit_flush_wait_parallel(struct f2fs_sb_info *sbi, nid_t ino) > +{ > + struct f2fs_flush_bio *flush_bio; > + unsigned long devices = 0; > + int ret = 0; > + int i; > + > + flush_bio = kcalloc(sbi->s_ndevs, sizeof(*flush_bio), GFP_NOFS); Can we allocate flush_bio in local stack? so that we don't need to fallback to submit_flush_wait_serial() for low memory case? > + if (!flush_bio) > + return submit_flush_wait_serial(sbi, ino); > > for (i = 0; i < sbi->s_ndevs; i++) { > if (!f2fs_is_dirty_device(sbi, ino, i, FLUSH_INO)) > continue; > - ret = __submit_flush_wait(sbi, FDEV(i).bdev); > - if (ret) > - break; > + > + bio_init(&flush_bio[i].bio, FDEV(i).bdev, NULL, 0, > + REQ_OP_WRITE | REQ_PREFLUSH); > + init_completion(&flush_bio[i].wait); > + flush_bio[i].bio.bi_private = &flush_bio[i].wait; > + flush_bio[i].bio.bi_end_io = f2fs_flush_end_io; > + submit_bio(&flush_bio[i].bio); > + devices |= BIT(i); > + } > + > + for (i = 0; i < sbi->s_ndevs; i++) { > + int err; > + > + if (!(devices & BIT(i))) > + continue; > + > + wait_for_completion(&flush_bio[i].wait); > + err = blk_status_to_errno(flush_bio[i].bio.bi_status); > + trace_f2fs_issue_flush(FDEV(i).bdev, test_opt(sbi, NOBARRIER), > + test_opt(sbi, FLUSH_MERGE), err); > + if (!err) > + f2fs_update_iostat(sbi, NULL, FS_FLUSH_IO, 0); > + else if (!ret) > + ret = err; > + bio_uninit(&flush_bio[i].bio); It may leak bio reference previously? For local bio variable, there is no such issue. Thanks, > } > + kfree(flush_bio); > return ret; > } > > +static int submit_flush_wait(struct f2fs_sb_info *sbi, nid_t ino) > +{ > + if (!f2fs_is_multi_device(sbi)) > + return __submit_flush_wait(sbi, sbi->sb->s_bdev); > + > + return submit_flush_wait_parallel(sbi, ino); > +} > + > static int issue_flush_thread(void *data) > { > struct f2fs_sb_info *sbi = data;