Re: [Qemu-devel] [PATCH for-1.5 0/3] qcow2: Catch some L1 table index overflows

[Richard W.M. Jones]

On Mon, May 13, 2013 at 03:42:53PM +0200, Kevin Wolf wrote:
[...]

I tested the first patch (didn't try the tests) with qemu-img
and libguestfs and it works.

New qemu-img fails with:
qemu-img: The image size is too large for file format 'qcow2'

If you use the old qemu-img to create the file, and then try
to add it with the new qemu, you get:

-drive file=/tmp/huge.qcow2,cache=none,id=hd0,if=none: could not open disk 
image /tmp/huge.qcow2: File too large

(instead of a segfault).

So:

ACK.

Tested-by: Richard W.M. Jones <address@hidden>

My only gripe is it would be nice if the qemu-img error message
mentioned that you can use the -o cluster_size=XXX option.

> Suggestions for a better test case are welcome. But now that creating a large
> image file fails, and if you somehow manage to create it anyway (qcow2.py)
> opening it fails, it's hard to test the actual read/write cases.

It'd be nice if the test included a test of the huge case using -o
cluster_size=2M and a few reads and writes at the end of the disk,
just to make sure this doesn't break in future.

Also nice to have would be to be able to specify disk sizes using 'P'
and 'E' :-)

Rich.

-- 
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
virt-df lists disk usage of guests without needing to install any
software inside the virtual machine.  Supports Linux and Windows.
http://people.redhat.com/~rjones/virt-df/