Re: [Qemu-devel] [PATCH 3/3] block: mirror - zero unallocated target sectors when zero init not present

[Paolo Bonzini]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256



On 29/09/2015 10:39, Kevin Wolf wrote:
> bdrv_has_zero_init() takes care of that, in theory. The "problem"
> here is that the target is opened with BDRV_O_NO_BACKING, so the
> block layer doesn't consider this an image with a backing file.

I think bdrv_has_zero_init() is working right. If you read the qcow2
file as it was opened (i.e. with BDRV_O_NO_BACKING), unallocated areas
will indeed read as zeroes.

Of course if the file is opened with BDRV_O_NO_BACKING but does have a
backing file, you ought not to read unallocated areas at all.

So it's not the answer (of bdrv_has_zero_init) that is wrong, but the
question that was not well-specified.

> Is there anything better than bs->backing_hd that we could check?

It's simply sync == 'full', I think.  Then the problematic case never
even reaches bdrv_has_zero_init.

Paolo
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2

iQEcBAEBCAAGBQJWCqRqAAoJEL/70l94x66DgcUH/jN8VFkGpZxXS5b+TnU8BeGV
Xmr3AjqICYS4K1mKcuu20GKZO5QSTh4Z7p/Igo2KmiGqven2kT/NIvjPRlSv4tqZ
Vov6AaamO6OIme+nA0hYbc3ANUY+b/7CqL8tDb3rKzah0FeFMSi1x7Who7aOCTQs
IjsJ37/ay+mGmPR9akDAfqjJjGPBJFL9dxz/0pgdPDUyj7IwvyolgGZ49rGNzoHE
/86Dy23ET16HQHDOz3afsrLHf9gxGZFCMsLJostqH0cuMs2sk1qnY9i9xEXYUM00
XNoaVafwCeH1ypXHNcP+GWtbbHBaMJJtmoRFB72VDRPq39XvpzWhifbKK3+c2Qw=
=/2aT
-----END PGP SIGNATURE-----