On 9/22/20 2:58 PM, Richard W.M. Jones wrote:
On Tue, Sep 22, 2020 at 11:15:39AM -0500, Eric Blake wrote:
> + try:
> + for c in args.command:
> + if c != '-':
> + exec(c, d, d)
> + else:
> + exec(sys.stdin.read(), d, d)
> + except nbd.Error as ex:
> + if os.environ.get("LIBNBD_DEBUG", "0") ==
"1":
Would it be better to use nbd.NBD().get_debug() == True, or
h.get_debug() == True for this test instead of duplicating the logic
inside libnbd?
I originally worried about:
nbdsh -c 'h = None'
which means we can't assume that the handle we opened prior to -c code
is still valid at the point we catch an exception. But you have a point
that opening yet another NBD handle will work. I'll push the obvious
followup.
But sure, let's do this. ACK.
Rich.
--
Eric Blake, Principal Software Engineer
Red Hat, Inc. +1-919-301-3226
Virtualization:
qemu.org |
libvirt.org