qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Kevin Wolf <kwolf@redhat.com>
To: Max Reitz <mreitz@redhat.com>
Cc: qemu-devel@nongnu.org, Stefan Hajnoczi <stefanha@redhat.com>
Subject: Re: [Qemu-devel] [PATCH v5 07/26] qcow2: Respect error in qcow2_alloc_bytes()
Date: Wed, 4 Feb 2015 12:40:15 +0100	[thread overview]
Message-ID: <20150204114015.GA5641@noname.redhat.com> (raw)
In-Reply-To: <1418647857-3589-8-git-send-email-mreitz@redhat.com>

Am 15.12.2014 um 13:50 hat Max Reitz geschrieben:
> qcow2_update_cluster_refcount() may fail, and qcow2_alloc_bytes() should
> mind that case.
> 
> Signed-off-by: Max Reitz <mreitz@redhat.com>
> Reviewed-by: Eric Blake <eblake@redhat.com>
> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
>  block/qcow2-refcount.c | 33 +++++++++++++++++++++------------
>  1 file changed, 21 insertions(+), 12 deletions(-)
> 
> diff --git a/block/qcow2-refcount.c b/block/qcow2-refcount.c
> index 0308a7e..db81647 100644
> --- a/block/qcow2-refcount.c
> +++ b/block/qcow2-refcount.c
> @@ -778,8 +778,8 @@ int qcow2_alloc_clusters_at(BlockDriverState *bs, uint64_t offset,
>  int64_t qcow2_alloc_bytes(BlockDriverState *bs, int size)
>  {
>      BDRVQcowState *s = bs->opaque;
> -    int64_t offset, cluster_offset;
> -    int free_in_cluster;
> +    int64_t offset, cluster_offset, new_cluster;
> +    int free_in_cluster, ret;
>  
>      BLKDBG_EVENT(bs->file, BLKDBG_CLUSTER_ALLOC_BYTES);
>      assert(size > 0 && size <= s->cluster_size);
> @@ -800,23 +800,32 @@ int64_t qcow2_alloc_bytes(BlockDriverState *bs, int size)
>          free_in_cluster -= size;
>          if (free_in_cluster == 0)
>              s->free_byte_offset = 0;
> -        if (offset_into_cluster(s, offset) != 0)
> -            qcow2_update_cluster_refcount(bs, offset >> s->cluster_bits, 1,
> -                                          false, QCOW2_DISCARD_NEVER);
> +        if (offset_into_cluster(s, offset) != 0) {
> +            ret = qcow2_update_cluster_refcount(bs, offset >> s->cluster_bits,
> +                                                1, false, QCOW2_DISCARD_NEVER);
> +            if (ret < 0) {
> +                return ret;

Not sure how relevant it is, but s->free_byte_offset has already been
increased, so we're leaving sub-cluster space unused. (It's not really
leaking as freeing all references still frees the cluster.)

> +            }
> +        }
>      } else {
> -        offset = qcow2_alloc_clusters(bs, s->cluster_size);
> -        if (offset < 0) {
> -            return offset;
> +        new_cluster = qcow2_alloc_clusters(bs, s->cluster_size);
> +        if (new_cluster < 0) {
> +            return new_cluster;
>          }

offset is the return value of this function, and now there are cases
where it isn't set to new_cluster any more (I wonder why gcc doesn't
warn).

Why can't we keep offset where it was used and only assign new_cluster
additionally so we can do the cleanup?

>          cluster_offset = start_of_cluster(s, s->free_byte_offset);
> -        if ((cluster_offset + s->cluster_size) == offset) {
> +        if ((cluster_offset + s->cluster_size) == new_cluster) {
>              /* we are lucky: contiguous data */
>              offset = s->free_byte_offset;
> -            qcow2_update_cluster_refcount(bs, offset >> s->cluster_bits, 1,
> -                                          false, QCOW2_DISCARD_NEVER);
> +            ret = qcow2_update_cluster_refcount(bs, offset >> s->cluster_bits,
> +                                                1, false, QCOW2_DISCARD_NEVER);
> +            if (ret < 0) {
> +                qcow2_free_clusters(bs, new_cluster, s->cluster_size,
> +                                    QCOW2_DISCARD_NEVER);
> +                return ret;
> +            }
>              s->free_byte_offset += size;
>          } else {
> -            s->free_byte_offset = offset;
> +            s->free_byte_offset = new_cluster;
>              goto redo;
>          }
>      }

Kevin

  reply	other threads:[~2015-02-04 11:40 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-12-15 12:50 [Qemu-devel] [PATCH v5 00/26] qcow2: Support refcount orders != 4 Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 01/26] qcow2: Add two new fields to BDRVQcowState Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 02/26] qcow2: Add refcount_bits to format-specific info Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 03/26] qcow2: Do not return new value after refcount update Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 04/26] qcow2: Only return status from qcow2_get_refcount Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 05/26] qcow2: Use unsigned addend for update_refcount() Max Reitz
2015-01-22 15:33   ` Eric Blake
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 06/26] qcow2: Use 64 bits for refcount values Max Reitz
2015-01-22 15:35   ` Eric Blake
2015-02-03 19:26   ` Kevin Wolf
2015-02-03 19:48     ` Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 07/26] qcow2: Respect error in qcow2_alloc_bytes() Max Reitz
2015-02-04 11:40   ` Kevin Wolf [this message]
2015-02-04 15:04     ` Max Reitz
2015-02-04 15:12       ` Kevin Wolf
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 08/26] qcow2: Refcount overflow and qcow2_alloc_bytes() Max Reitz
2015-02-04 11:55   ` Kevin Wolf
2015-02-04 15:33     ` Max Reitz
2015-02-04 16:10       ` Kevin Wolf
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 09/26] qcow2: Helper for refcount array reallocation Max Reitz
2015-02-04 13:21   ` Kevin Wolf
2015-02-04 15:57     ` Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 10/26] qcow2: Helper function for refcount modification Max Reitz
2015-02-04 16:06   ` Kevin Wolf
2015-02-04 17:12     ` Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 11/26] qcow2: More helpers " Max Reitz
2015-02-04 13:53   ` Kevin Wolf
2015-02-04 15:59     ` Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 12/26] qcow2: Open images with refcount order != 4 Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 13/26] qcow2: refcount_order parameter for qcow2_create2 Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 14/26] qcow2: Use symbolic macros in qcow2_amend_options Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 15/26] iotests: Prepare for refcount_bits option Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 16/26] qcow2: Allow creation with refcount order != 4 Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 17/26] progress: Allow regressing progress Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 18/26] block: Add opaque value to the amend CB Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 19/26] qcow2: Use error_report() in qcow2_amend_options() Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 20/26] qcow2: Use abort() instead of assert(false) Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 21/26] qcow2: Split upgrade/downgrade paths for amend Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 22/26] qcow2: Use intermediate helper CB " Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 23/26] qcow2: Add function for refcount order amendment Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 24/26] qcow2: Invoke refcount order amendment function Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 25/26] qcow2: Point to amend function in check Max Reitz
2014-12-15 12:50 ` [Qemu-devel] [PATCH v5 26/26] iotests: Add test for different refcount widths Max Reitz
2015-01-20 22:48 ` [Qemu-devel] [PATCH v5 00/26] qcow2: Support refcount orders != 4 Max Reitz

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=20150204114015.GA5641@noname.redhat.com \
    --to=kwolf@redhat.com \
    --cc=mreitz@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --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).