Re: [Qemu-block] [PATCH] vpc: use current_size field for XenServer VHD images

[Grant Wu]

Side note:  After manually patching tap\0 to qem2, Spencer and I were still unable to get qemu-img info working with the image.  Unsure if this is related to the patch or not.

Thanks,

Grant Wu

address@hidden

On Sat, Mar 19, 2016 at 3:42 PM, Stefan Hajnoczi <address@hidden> wrote:

The vpc driver has two methods of determining virtual disk size.  The
correct one to use depends on the software that generated the image
file.  Add the XenServer creator_app signature so that image size is
correctly detected for those images.

Reported-by: Grant Wu <address@hidden>
Reported-by: Spencer Baugh <address@hidden>
Signed-off-by: Stefan Hajnoczi <address@hidden>
---
 block/vpc.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/block/vpc.c b/block/vpc.c
index 8435205..7517f75 100644
--- a/block/vpc.c
+++ b/block/vpc.c
@@ -298,6 +298,7 @@ static int vpc_open(BlockDriverState *bs, QDict *options, int flags,
      *      'qem2'  :  current_size     QEMU (uses current_size)
      *      'win '  :  current_size     Hyper-V
      *      'd2v '  :  current_size     Disk2vhd
+     *      'tap\0' :  current_size     XenServer
      *
      *  The user can override the table values via drive options, however
      *  even with an override we will still use current_size for images
@@ -305,7 +306,8 @@ static int vpc_open(BlockDriverState *bs, QDict *options, int flags,
      */
     use_chs = (!!strncmp(footer->creator_app, "win ", 4) &&
                !!strncmp(footer->creator_app, "qem2", 4) &&
-               !!strncmp(footer->creator_app, "d2v ", 4)) || s->force_use_chs;
+               !!strncmp(footer->creator_app, "d2v ", 4) &&
+               !!memcmp(footer->creator_app, "tap", 4)) || s->force_use_chs;

     if (!use_chs || bs->total_sectors == VHD_MAX_GEOMETRY || s->force_use_sz) {
         bs->total_sectors = be64_to_cpu(footer->current_size) /
--
2.5.0