From: Fam Zheng <famz@redhat.com>
To: Eric Blake <eblake@redhat.com>
Cc: qemu-devel@nongnu.org, qemu-block@nongnu.org, kwolf@redhat.com,
stefanha@redhat.com, Max Reitz <mreitz@redhat.com>,
Ronnie Sahlberg <ronniesahlberg@gmail.com>,
Paolo Bonzini <pbonzini@redhat.com>, Peter Lieven <pl@kamp.de>
Subject: Re: [Qemu-devel] [PATCH v3 17/22] block: Switch discard length bounds to byte-based
Date: Fri, 24 Jun 2016 14:43:10 +0800 [thread overview]
Message-ID: <20160624064310.GC13266@ad.usersys.redhat.com> (raw)
In-Reply-To: <1466721446-27737-18-git-send-email-eblake@redhat.com>
On Thu, 06/23 16:37, Eric Blake wrote:
> Sector-based limits are awkward to think about; in our on-going
> quest to move to byte-based interfaces, convert max_discard and
> discard_alignment. Rename them, using 'pdiscard' as an aid to
> track which remaining discard interfaces need conversion, and so
> that the compiler will help us catch the change in semantics
> across any rebased code. The BlockLimits type is now completely
> byte-based; and in iscsi.c, sector_limits_lun2qemu() is no
> longer needed.
>
> pdiscard_alignment is made unsigned (we use power-of-2 alignments
> as bitmasks, where unsigned is easier to think about) while
> leaving max_pdiscard signed (since we still have an 'int'
> interface); this is comparable to what commit cf081fc did for
> write zeroes limits. We may later want to make everything an
> unsigned 64-bit limit - but that requires a bigger code audit.
>
> Signed-off-by: Eric Blake <eblake@redhat.com>
>
> ---
> v3: split out write_zeroes wording tweaks, improve commit message
> v2: rebase nbd and iscsi limits across earlier improvements
> ---
> include/block/block_int.h | 14 ++++++++++----
> block/io.c | 16 +++++++++-------
> block/iscsi.c | 19 ++++++-------------
> block/nbd.c | 2 +-
> qemu-img.c | 3 ++-
> 5 files changed, 28 insertions(+), 26 deletions(-)
>
> diff --git a/include/block/block_int.h b/include/block/block_int.h
> index 7a4a00f..388ef80 100644
> --- a/include/block/block_int.h
> +++ b/include/block/block_int.h
> @@ -324,11 +324,17 @@ struct BlockDriver {
> };
>
> typedef struct BlockLimits {
> - /* maximum number of sectors that can be discarded at once */
> - int max_discard;
> + /* maximum number of bytes that can be discarded at once (since it
> + * is signed, it must be < 2G, if set), should be multiple of
> + * pdiscard_alignment, but need not be power of 2. May be 0 if no
> + * inherent 32-bit limit */
> + int32_t max_pdiscard;
>
> - /* optimal alignment for discard requests in sectors */
> - int64_t discard_alignment;
> + /* optimal alignment for discard requests in bytes, must be power
> + * of 2, less than max_discard if that is set, and multiple of
s/max_discard/max_pdiscard/
> + * bs->request_alignment. May be 0 if bs->request_alignment is
> + * good enough */
> + uint32_t pdiscard_alignment;
>
> /* maximum number of bytes that can zeroized at once (since it is
> * signed, it must be < 2G, if set), should be multiple of
> diff --git a/block/io.c b/block/io.c
> index 8ca9d43..0f15d05 100644
> --- a/block/io.c
> +++ b/block/io.c
> @@ -2368,19 +2368,21 @@ int coroutine_fn bdrv_co_discard(BlockDriverState *bs, int64_t sector_num,
> goto out;
> }
>
> - max_discard = MIN_NON_ZERO(bs->bl.max_discard, BDRV_REQUEST_MAX_SECTORS);
> + max_discard = MIN_NON_ZERO(bs->bl.max_pdiscard >> BDRV_SECTOR_BITS,
> + BDRV_REQUEST_MAX_SECTORS);
> while (nb_sectors > 0) {
> int ret;
> int num = nb_sectors;
> + int discard_alignment = bs->bl.pdiscard_alignment >> BDRV_SECTOR_BITS;
>
> /* align request */
> - if (bs->bl.discard_alignment &&
> - num >= bs->bl.discard_alignment &&
> - sector_num % bs->bl.discard_alignment) {
> - if (num > bs->bl.discard_alignment) {
> - num = bs->bl.discard_alignment;
> + if (discard_alignment &&
> + num >= discard_alignment &&
> + sector_num % discard_alignment) {
> + if (num > discard_alignment) {
> + num = discard_alignment;
> }
> - num -= sector_num % bs->bl.discard_alignment;
> + num -= sector_num % discard_alignment;
Or just
num = discard_alignment - sector_num % discard_alignment;
without the if.
Otherwise looks good,
Reviewed-by: Fam Zheng <famz@redhat.com>
> }
>
> /* limit request size */
> diff --git a/block/iscsi.c b/block/iscsi.c
> index 368687d..0d16c31 100644
> --- a/block/iscsi.c
> +++ b/block/iscsi.c
> @@ -1696,13 +1696,6 @@ static void iscsi_close(BlockDriverState *bs)
> memset(iscsilun, 0, sizeof(IscsiLun));
> }
>
> -static int sector_limits_lun2qemu(int64_t sector, IscsiLun *iscsilun)
> -{
> - int limit = MIN(sector_lun2qemu(sector, iscsilun), INT_MAX / 2 + 1);
> -
> - return limit < BDRV_REQUEST_MAX_SECTORS ? limit : 0;
> -}
> -
> static void iscsi_refresh_limits(BlockDriverState *bs, Error **errp)
> {
> /* We don't actually refresh here, but just return data queried in
> @@ -1722,14 +1715,14 @@ static void iscsi_refresh_limits(BlockDriverState *bs, Error **errp)
> }
>
> if (iscsilun->lbp.lbpu) {
> - if (iscsilun->bl.max_unmap < 0xffffffff) {
> - bs->bl.max_discard =
> - sector_limits_lun2qemu(iscsilun->bl.max_unmap, iscsilun);
> + if (iscsilun->bl.max_unmap < 0xffffffff / iscsilun->block_size) {
> + bs->bl.max_pdiscard =
> + iscsilun->bl.max_unmap * iscsilun->block_size;
> }
> - bs->bl.discard_alignment =
> - sector_limits_lun2qemu(iscsilun->bl.opt_unmap_gran, iscsilun);
> + bs->bl.pdiscard_alignment =
> + iscsilun->bl.opt_unmap_gran * iscsilun->block_size;
> } else {
> - bs->bl.discard_alignment = iscsilun->block_size >> BDRV_SECTOR_BITS;
> + bs->bl.pdiscard_alignment = iscsilun->block_size;
> }
>
> if (iscsilun->bl.max_ws_len < 0xffffffff / iscsilun->block_size) {
next prev parent reply other threads:[~2016-06-24 6:43 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-06-23 22:37 [Qemu-devel] [PATCH v3 00/22] Byte-based block limits Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 01/22] block: Tighter assertions on bdrv_aligned_pwritev() Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 02/22] block: Document supported flags during bdrv_aligned_preadv() Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 03/22] block: Fix harmless off-by-one in bdrv_aligned_preadv() Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 04/22] nbd: Allow larger requests Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 05/22] nbd: Advertise realistic limits to block layer Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 06/22] iscsi: " Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 07/22] scsi: Advertise limits by blocksize, not 512 Eric Blake
2016-06-24 5:22 ` Fam Zheng
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 08/22] block: Give nonzero result to blk_get_max_transfer_length() Eric Blake
2016-06-24 5:24 ` Fam Zheng
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 09/22] blkdebug: Set request_alignment during .bdrv_refresh_limits() Eric Blake
2016-06-24 5:42 ` Fam Zheng
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 10/22] iscsi: " Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 11/22] qcow2: " Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 12/22] raw-win32: " Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 13/22] block: " Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 14/22] block: Set default request_alignment during bdrv_refresh_limits() Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 15/22] block: Switch transfer length bounds to byte-based Eric Blake
2016-06-24 6:06 ` Fam Zheng
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 16/22] block: Wording tweaks to write zeroes limits Eric Blake
2016-06-24 6:12 ` Fam Zheng
2016-06-24 14:10 ` Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 17/22] block: Switch discard length bounds to byte-based Eric Blake
2016-06-24 6:43 ` Fam Zheng [this message]
2016-06-24 14:15 ` Eric Blake
2016-06-24 14:29 ` Kevin Wolf
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 18/22] block: Drop raw_refresh_limits() Eric Blake
2016-06-24 6:44 ` Fam Zheng
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 19/22] block: Split bdrv_merge_limits() from bdrv_refresh_limits() Eric Blake
2016-06-24 6:48 ` Fam Zheng
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 20/22] block: Move request_alignment into BlockLimit Eric Blake
2016-06-24 7:07 ` Fam Zheng
2016-06-24 13:45 ` Kevin Wolf
2016-06-24 14:17 ` Eric Blake
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 21/22] block: Fix error message style Eric Blake
2016-06-24 7:09 ` Fam Zheng
2016-06-23 22:37 ` [Qemu-devel] [PATCH v3 22/22] block: Use bool as appropriate for BDS members Eric Blake
2016-06-24 7:12 ` Fam Zheng
2016-06-24 14:32 ` [Qemu-devel] [PATCH v3 00/22] Byte-based block limits Kevin Wolf
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=20160624064310.GC13266@ad.usersys.redhat.com \
--to=famz@redhat.com \
--cc=eblake@redhat.com \
--cc=kwolf@redhat.com \
--cc=mreitz@redhat.com \
--cc=pbonzini@redhat.com \
--cc=pl@kamp.de \
--cc=qemu-block@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=ronniesahlberg@gmail.com \
--cc=stefanha@redhat.com \
/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;
as well as URLs for NNTP newsgroup(s).