From: Christoph Hellwig <hch@infradead.org>
To: Keith Busch <kbusch@kernel.org>
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>,
Kent Overstreet <kent.overstreet@linux.dev>,
Christoph Hellwig <hch@infradead.org>,
Jens Axboe <axboe@kernel.dk>,
linux-bcachefs@vger.kernel.org, linux-block@vger.kernel.org,
Linus Torvalds <torvalds@linux-foundation.org>
Subject: Re: [PATCH 13/14] block: Allow REQ_FUA|REQ_READ
Date: Mon, 17 Mar 2025 23:19:31 -0700 [thread overview]
Message-ID: <Z9kQc6HtB05eFQd5@infradead.org> (raw)
In-Reply-To: <Z9hcpS64HDZUJ21c@kbusch-mbp.dhcp.thefacebook.com>
On Mon, Mar 17, 2025 at 11:32:21AM -0600, Keith Busch wrote:
> I'm not even sure it makes sense to suspect the controller side, but the
> host side is considered reliable? If the media is in fact perfect, and
> if non-FUA read returns data that fails the checksum, wouldn't it be
> just as likely that the transport or host side is the problem?
From the error numbers we've seen it would in fact be much more likely
to be the host or transport side.
> Every
> NVMe SSD I've developered couldn't efficiently fill the PCIe tx-buffer
> directly from media (excluding Optane; RIP), so a read with or without
> FUA would stage the data in the very same controller-side DRAM.
Exactly. It's an writeback+invalidate+reread operation. A great
way to slow down the read, but it's not going to help you with
data recovery.
next prev parent reply other threads:[~2025-03-18 6:19 UTC|newest]
Thread overview: 63+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-11 20:15 [PATCH 00/14] better handling of checksum errors/bitrot Kent Overstreet
2025-03-11 20:15 ` [PATCH 01/14] bcachefs: Convert read path to standard error codes Kent Overstreet
2025-03-11 20:15 ` [PATCH 02/14] bcachefs: Fix BCH_ERR_data_read_csum_err_maybe_userspace in retry path Kent Overstreet
2025-03-11 20:15 ` [PATCH 03/14] bcachefs: Read error message now indicates if it was for an internal move Kent Overstreet
2025-03-11 20:15 ` [PATCH 04/14] bcachefs: BCH_ERR_data_read_buffer_too_small Kent Overstreet
2025-03-11 20:15 ` [PATCH 05/14] bcachefs: Return errors to top level bch2_rbio_retry() Kent Overstreet
2025-03-11 20:15 ` [PATCH 06/14] bcachefs: Print message on successful read retry Kent Overstreet
2025-03-11 20:15 ` [PATCH 07/14] bcachefs: Don't create bch_io_failures unless it's needed Kent Overstreet
2025-03-11 20:15 ` [PATCH 08/14] bcachefs: Checksum errors get additional retries Kent Overstreet
2025-03-11 20:15 ` [PATCH 09/14] bcachefs: __bch2_read() now takes a btree_trans Kent Overstreet
2025-03-11 20:15 ` [PATCH 10/14] bcachefs: Poison extents that can't be read due to checksum errors Kent Overstreet
2025-03-11 20:15 ` [PATCH 11/14] bcachefs: Data move can read from poisoned extents Kent Overstreet
2025-03-11 20:15 ` [PATCH 12/14] bcachefs: Debug params for data corruption injection Kent Overstreet
2025-03-11 20:15 ` [PATCH 13/14] block: Allow REQ_FUA|REQ_READ Kent Overstreet
2025-03-15 16:47 ` Jens Axboe
2025-03-15 17:01 ` Kent Overstreet
2025-03-15 17:03 ` Jens Axboe
2025-03-15 17:27 ` Kent Overstreet
2025-03-15 17:43 ` Jens Axboe
2025-03-15 18:07 ` Kent Overstreet
2025-03-15 18:32 ` Jens Axboe
2025-03-15 18:41 ` Kent Overstreet
2025-03-17 6:00 ` Christoph Hellwig
2025-03-17 12:15 ` Kent Overstreet
2025-03-17 14:13 ` Keith Busch
2025-03-17 14:49 ` Kent Overstreet
2025-03-17 15:15 ` Keith Busch
2025-03-17 15:22 ` Kent Overstreet
2025-03-17 15:30 ` Martin K. Petersen
2025-03-17 15:43 ` Kent Overstreet
2025-03-17 17:57 ` Martin K. Petersen
2025-03-17 18:21 ` Kent Overstreet
2025-03-17 19:24 ` Keith Busch
2025-03-17 19:40 ` Kent Overstreet
2025-03-17 20:39 ` Keith Busch
2025-03-17 21:13 ` Bart Van Assche
2025-03-18 1:06 ` Kent Overstreet
2025-03-18 6:16 ` Christoph Hellwig
2025-03-18 17:49 ` Bart Van Assche
2025-03-18 18:00 ` Kent Overstreet
2025-03-18 18:10 ` Keith Busch
2025-03-18 18:13 ` Kent Overstreet
2025-03-20 5:40 ` Christoph Hellwig
2025-03-20 10:28 ` Kent Overstreet
2025-03-18 0:27 ` Kent Overstreet
2025-03-18 6:11 ` Christoph Hellwig
2025-03-18 21:33 ` Kent Overstreet
2025-03-17 17:32 ` Keith Busch
2025-03-18 6:19 ` Christoph Hellwig [this message]
2025-03-18 6:01 ` Christoph Hellwig
2025-03-11 20:15 ` [PATCH 14/14] bcachefs: Read retries are after checksum errors now REQ_FUA Kent Overstreet
2025-03-17 20:55 ` [PATCH 00/14] better handling of checksum errors/bitrot John Stoffel
2025-03-17 21:12 ` errors compiling bcachefs-tools v1.20.0 on debian 12 John Stoffel
2025-03-17 21:48 ` Malte Schröder
2025-03-17 23:10 ` John Stoffel
2025-03-18 21:04 ` John Stoffel
2025-03-18 21:32 ` Malte Schröder
2025-03-19 14:16 ` John Stoffel
2025-03-24 15:25 ` Krzysztof Hajdamowicz
2025-03-26 13:45 ` John Stoffel
2025-03-18 1:15 ` [PATCH 00/14] better handling of checksum errors/bitrot Kent Overstreet
2025-03-18 14:47 ` John Stoffel
2025-03-20 17:15 ` Kent Overstreet
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=Z9kQc6HtB05eFQd5@infradead.org \
--to=hch@infradead.org \
--cc=axboe@kernel.dk \
--cc=kbusch@kernel.org \
--cc=kent.overstreet@linux.dev \
--cc=linux-bcachefs@vger.kernel.org \
--cc=linux-block@vger.kernel.org \
--cc=martin.petersen@oracle.com \
--cc=torvalds@linux-foundation.org \
/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