On Thu, Apr 18, 2013 at 05:17:24PM +0200, Paolo Bonzini wrote: > This series includes several improvements to qcow2.py, mostly > the ability to dump the L1 and L2 tables for a full backing > chain. > > Doing this completely would require support for all image formats, > not just qcow2. To cover the common cases, I included support > for raw backing files. > > The right thing to do perhaps would be to add the read-metadata > patches from the in-place QED conversion summer of code project, > but I wanted to throw out these anyway if anyone has comments.
The patches look sane by themselves but it would be nicer to integrate with qemu-img by extending the block layer C code rather than reimplementing part of the block layer in Python for this single command. Stefan
