qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Fam Zheng <famz@redhat.com>
To: Markus Armbruster <armbru@redhat.com>
Cc: kwolf@redhat.com, jcody@redhat.com, hbrock@redhat.com,
	qemu-devel@nongnu.org, rjones@redhat.com, imain@redhat.com,
	stefanha@redhat.com, pbonzini@redhat.com
Subject: Re: [Qemu-devel] [PATCH v19 04/16] block: Replace in_use with operation blocker
Date: Tue, 20 May 2014 11:26:40 +0800	[thread overview]
Message-ID: <20140520032640.GC7688@T430.nay.redhat.com> (raw)
In-Reply-To: <871tvpq1ur.fsf@blackfin.pond.sub.org>

On Mon, 05/19 16:28, Markus Armbruster wrote:
> Fam Zheng <famz@redhat.com> writes:
> 
> > This drops BlockDriverState.in_use with op_blockers:
> >
> >   - Call bdrv_op_block_all in place of bdrv_set_in_use(bs, 1).
> >   - Call bdrv_op_unblock_all in place of bdrv_set_in_use(bs, 0).
> >   - Check bdrv_op_is_blocked() in place of bdrv_in_use(bs).
> >     The specific types are used, e.g. in place of starting block backup,
> >     bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_BACKUP, ...).
> >   - Check bdrv_op_blocker_is_empty() in place of assert(!bs->in_use).
> >
> > Note: there is only bdrv_op_block_all and bdrv_op_unblock_all callers at
> > this moment. So although the checks are specific to op types, this
> > changes can still be seen as identical logic with previously with
> > in_use. The difference is error message are improved because of blocker
> > error info.
> >
> > Signed-off-by: Fam Zheng <famz@redhat.com>
> > Reviewed-by: Jeff Cody <jcody@redhat.com>
> > ---
> >  block-migration.c               |  7 +++++--
> >  block.c                         | 24 +++++++-----------------
> >  blockdev.c                      | 19 +++++++++----------
> >  blockjob.c                      | 14 +++++++++-----
> >  hw/block/dataplane/virtio-blk.c | 18 ++++++++++++------
> >  include/block/block.h           |  2 --
> >  include/block/block_int.h       |  1 -
> >  include/block/blockjob.h        |  3 +++
> >  8 files changed, 45 insertions(+), 43 deletions(-)
> >
> > diff --git a/block-migration.c b/block-migration.c
> > index 56951e0..1656270 100644
> > --- a/block-migration.c
> > +++ b/block-migration.c
> > @@ -59,6 +59,7 @@ typedef struct BlkMigDevState {
> >      unsigned long *aio_bitmap;
> >      int64_t completed_sectors;
> >      BdrvDirtyBitmap *dirty_bitmap;
> > +    Error *blocker;
> >  } BlkMigDevState;
> >  
> >  typedef struct BlkMigBlock {
> > @@ -361,7 +362,8 @@ static void init_blk_migration_it(void *opaque, BlockDriverState *bs)
> >          bmds->completed_sectors = 0;
> >          bmds->shared_base = block_mig_state.shared_base;
> >          alloc_aio_bitmap(bmds);
> > -        bdrv_set_in_use(bs, 1);
> > +        error_setg(&bmds->blocker, "block device is in use by migration");
> > +        bdrv_op_block_all(bs, bmds->blocker);
> >          bdrv_ref(bs);
> >  
> >          block_mig_state.total_sector_sum += sectors;
> 
> Unlike QERR_DEVICE_IN_USE, this message doesn't include the device name.
> Did you verify it's not needed to make sense of the message?

I prefer not adding device name in the ->block error message, because we can
always get the device name from the bs pointer. In other words this should be
handled in bdrv_op_is_blocked() generally, will update patch 03/16.

<...>

> > diff --git a/blockjob.c b/blockjob.c
> > index cd4784f..60e72f5 100644
> > --- a/blockjob.c
> > +++ b/blockjob.c
> > @@ -41,14 +41,16 @@ void *block_job_create(const BlockJobDriver *driver, BlockDriverState *bs,
> >  {
> >      BlockJob *job;
> >  
> > -    if (bs->job || bdrv_in_use(bs)) {
> > +    if (bs->job || !bdrv_op_blocker_is_empty(bs)) {
> >          error_set(errp, QERR_DEVICE_IN_USE, bdrv_get_device_name(bs));
> >          return NULL;
> >      }
> 
> Elsewhere, you replace bdrv_in_use() by bdrv_op_is_blocked().  Why not
> here?

There is no suitable BLOCK_OP_TYPE_ for this because we don't know what caller
does with the created job.

Maybe add a BLOCK_OP_TYPE_ANY special type makes it clearer, but it's logically
identical to !bdrv_op_blocker_is_empty().

That said, a BLOCK_OP_TYPE_ANY is AFAICT only useful here, and even this one
will be dropped in patch 05/16.

> 
> Doesn't this contradict the commit message?

So I'd prefer to update the commit message.

Fam

  reply	other threads:[~2014-05-20  3:26 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-05-12  1:35 [Qemu-devel] [PATCH v19 00/16] Drop in_use from BlockDriverState and enable point-in-time snapshot exporting over NBD Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 01/16] vmdk: Optimize cluster allocation Fam Zheng
2014-05-14 14:00   ` Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 02/16] block: Add BlockOpType enum Fam Zheng
2014-05-19 13:55   ` Markus Armbruster
2014-05-19 15:53   ` Eric Blake
2014-05-19 16:15     ` Markus Armbruster
2014-05-20  3:09       ` Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 03/16] block: Introduce op_blockers to BlockDriverState Fam Zheng
2014-05-19 14:10   ` Markus Armbruster
2014-05-19 14:37     ` Kevin Wolf
2014-05-19 15:37       ` Jeff Cody
2014-05-20 11:43         ` Markus Armbruster
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 04/16] block: Replace in_use with operation blocker Fam Zheng
2014-05-19 14:28   ` Markus Armbruster
2014-05-20  3:26     ` Fam Zheng [this message]
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 05/16] block: Move op_blocker check from block_job_create to its caller Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 06/16] block: Add bdrv_set_backing_hd() Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 07/16] block: Add backing_blocker in BlockDriverState Fam Zheng
2014-05-19 19:35   ` Eric Blake
2014-05-19 20:23     ` Markus Armbruster
2014-05-20  3:39       ` Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 08/16] block: Parse "backing" option to reference existing BDS Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 09/16] block: Support dropping active in bdrv_drop_intermediate Fam Zheng
2014-05-19 19:38   ` Eric Blake
2014-05-20  3:53     ` Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 10/16] stream: Use bdrv_drop_intermediate and drop close_unused_images Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 11/16] commit: Use bdrv_drop_intermediate Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 12/16] qmp: Add command 'blockdev-backup' Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 13/16] block: Allow backup on referenced named BlockDriverState Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 14/16] block: Add blockdev-backup to transaction Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 15/16] qemu-iotests: Test blockdev-backup in 055 Fam Zheng
2014-05-19 19:46   ` Eric Blake
2014-05-20  3:56     ` Fam Zheng
2014-05-12  1:35 ` [Qemu-devel] [PATCH v19 16/16] qemu-iotests: Image fleecing test case 089 Fam Zheng

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=20140520032640.GC7688@T430.nay.redhat.com \
    --to=famz@redhat.com \
    --cc=armbru@redhat.com \
    --cc=hbrock@redhat.com \
    --cc=imain@redhat.com \
    --cc=jcody@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=rjones@redhat.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).