From: Kanchan Joshi <[email protected]>
To: John Garry <[email protected]>,
[email protected], [email protected], [email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected], [email protected],
[email protected]
Cc: [email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected],
Alan Adamson <[email protected]>
Subject: Re: [PATCH v8 10/10] nvme: Atomic write support
Date: Mon, 17 Jun 2024 22:54:00 +0530 [thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>
On 6/10/2024 4:13 PM, John Garry wrote:
> +static bool nvme_valid_atomic_write(struct request *req)
> +{
> + struct request_queue *q = req->q;
> + u32 boundary_bytes = queue_atomic_write_boundary_bytes(q);
> +
> + if (blk_rq_bytes(req) > queue_atomic_write_unit_max_bytes(q))
> + return false;
> +
> + if (boundary_bytes) {
> + u64 mask = boundary_bytes - 1, imask = ~mask;
> + u64 start = blk_rq_pos(req) << SECTOR_SHIFT;
> + u64 end = start + blk_rq_bytes(req) - 1;
> +
> + /* If greater then must be crossing a boundary */
> + if (blk_rq_bytes(req) > boundary_bytes)
> + return false;
Nit: I'd cache blk_rq_bytes(req), since that is repeating and this
function is called for each atomic IO.
> +
> + if ((start & imask) != (end & imask))
> + return false;
> + }
> +
> + return true;
> +}
> +
> static inline blk_status_t nvme_setup_rw(struct nvme_ns *ns,
> struct request *req, struct nvme_command *cmnd,
> enum nvme_opcode op)
> @@ -941,6 +965,12 @@ static inline blk_status_t nvme_setup_rw(struct nvme_ns *ns,
>
> if (req->cmd_flags & REQ_RAHEAD)
> dsmgmt |= NVME_RW_DSM_FREQ_PREFETCH;
> + /*
> + * Ensure that nothing has been sent which cannot be executed
> + * atomically.
> + */
> + if (req->cmd_flags & REQ_ATOMIC && !nvme_valid_atomic_write(req))
> + return BLK_STS_INVAL;
>
Is this validity check specific to NVMe or should this be moved up to
block layer as it also knows the limits?
next prev parent reply other threads:[~2024-06-17 17:24 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-10 10:43 [PATCH v8 00/10] block atomic writes John Garry
2024-06-10 10:43 ` [PATCH v8 01/10] block: Pass blk_queue_get_max_sectors() a request pointer John Garry
2024-06-10 10:43 ` [PATCH v8 02/10] block: Generalize chunk_sectors support as boundary support John Garry
2024-06-10 10:43 ` [PATCH v8 03/10] fs: Initial atomic write support John Garry
2024-06-12 20:51 ` Darrick J. Wong
2024-06-10 10:43 ` [PATCH v8 04/10] fs: Add initial atomic write support info to statx John Garry
2024-06-12 20:54 ` Darrick J. Wong
2024-06-13 7:25 ` John Garry
2024-06-10 10:43 ` [PATCH v8 05/10] block: Add core atomic write support John Garry
2024-06-17 18:56 ` Keith Busch
2024-06-18 6:51 ` Christoph Hellwig
2024-06-18 7:46 ` John Garry
2024-06-18 17:25 ` Keith Busch
2024-06-19 7:59 ` John Garry
2024-06-19 8:02 ` Christoph Hellwig
2024-06-19 10:42 ` John Garry
2024-06-19 16:07 ` Martin K. Petersen
2024-06-10 10:43 ` [PATCH v8 06/10] block: Add atomic write support for statx John Garry
2024-06-10 10:43 ` [PATCH v8 07/10] block: Add fops atomic write support John Garry
2024-06-10 10:43 ` [PATCH v8 08/10] scsi: sd: Atomic " John Garry
2024-06-10 10:43 ` [PATCH v8 09/10] scsi: scsi_debug: " John Garry
2024-06-10 10:43 ` [PATCH v8 10/10] nvme: " John Garry
2024-06-17 17:24 ` Kanchan Joshi [this message]
2024-06-17 18:04 ` John Garry
2024-06-18 6:49 ` Christoph Hellwig
2024-06-18 7:22 ` John Garry
2024-06-14 2:01 ` [PATCH v8 00/10] block atomic writes Martin K. Petersen
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 \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
/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