On 2/19/19 3:21 PM, Richard W.M. Jones wrote: > On Tue, Feb 19, 2019 at 09:21:10AM -0600, Eric Blake wrote: >> Missing an unlink(). Also, does the caller properly use close(disk->fd) >> on early exits like this, or does it not matter because we're just going >> to exit()? > > Just about the early exits: I started off trying to get those paths > right but it made the code ridiculously complicated so I gave up. As > you say config_complete just calls exit, so it's not really a problem, > in fact more likely to cause bugs than fix them.
Fair enough. The unlink() is an obvious cleanup worth doing (no need to litter $TMPDIR, as the filesystem is not an automatic cleanup on process termination), but memory/fd leaks are acceptable, although it's probably also worth a comment in the code stating that we know we are leaking on failure on the assumption that the caller will exit(). -- Eric Blake, Principal Software Engineer Red Hat, Inc. +1-919-301-3226 Virtualization: qemu.org | libvirt.org _______________________________________________ Libguestfs mailing list [email protected] https://www.redhat.com/mailman/listinfo/libguestfs
