Comment 61 for bug 1996188

@Rajat: there are still some places where image conversion is happening, but there's no image involved, for example in the vzstorage driver in _copy_volume_from_snapshot(), so at least some drivers are going to have an 'internal image' situation. I decided that since we're fixing this at the choke point in image_utils, I wouldn't modify any of the individual drivers. I think that could be done as a follow-up after the bug is public and the exploit is patched, because it really only affects log and exception messages when someone is attempting the exploit, not normal processing. What I'd like to do is propose patches to master and let the driver maintainers approve them and do the backports and further testing. But maybe I'm just being lazy; let me know what you think.