On Wed, Feb 25, 2015 at 10:10:38 +0100, Peter Krempa wrote:
On Thu, Feb 19, 2015 at 15:59:14 +0100, Ján Tomko wrote:
> Instead of just looking at the output of fstat, call
> virStorageFileGetMetadata to get the full capacity from
> image headers.
> ---
> src/storage/storage_backend.c | 32 +++++++++++++++++++++++++++++++-
> 1 file changed, 31 insertions(+), 1 deletion(-)
>
> diff --git a/src/storage/storage_backend.c b/src/storage/storage_backend.c
> index dd33436..8471744 100644
> --- a/src/storage/storage_backend.c
> +++ b/src/storage/storage_backend.c
> @@ -1514,6 +1514,9 @@ virStorageBackendUpdateVolTargetInfo(virStorageSourcePtr
target,
> {
This function has a parameter named 'updateCapacity' ...
> int ret, fd = -1;
> struct stat sb;
> + virStorageSourcePtr meta = NULL;
> + char *buf = NULL;
> + ssize_t len = VIR_STORAGE_MAX_HEADER;
>
> if ((ret = virStorageBackendVolOpen(target->path, &sb, openflags)) <
0)
> goto cleanup;
> @@ -1523,14 +1526,41 @@ virStorageBackendUpdateVolTargetInfo(virStorageSourcePtr
target,
> updateCapacity)) < 0)
... that is obeyed when updating the capacity via the stat call ...
> goto cleanup;
>
> + if (target->type == VIR_STORAGE_VOL_FILE &&
> + target->format != VIR_STORAGE_FILE_NONE) {
... but this new code ignores it. Is there a specific reason for that?
I see. Next patch reverts all the changes regarding update capacity.
ACK after the release to this patch.
Peter