... and I'm *still* looking at it.
Firstly I wrote a test suite (see attached) which randomly generates
disk images for virt-resize to test.
This reveals a bug in the way we resized the MBR extended partition.
The new mbr_part_type function causes virt-resize to fail, but in fact
mbr_part_type is *not* to blame - it just exposes the existing bug.
Anyway, will continue looking at this tomorrow.
Rich.
--
Richard Jones, Virtualization Group, Red Hat
http://people.redhat.com/~rjones
Read my programming and virtualization blog:
http://rwmj.wordpress.com
virt-builder quickly builds VMs from scratch
http://libguestfs.org/virt-builder.1.html