On 8/1/19 3:36 PM, Daniel Henrique Barboza wrote:
On 8/1/19 11:24 AM, Daniel Henrique Barboza wrote:
>
> [...]
> So yeah. I am not sure what's different from the environments I use
> versus
> the one you're using (and probably the env of every other committer here,
> otherwise this test failure would be pointed out earlier). I'll need
> to think
> more about it.
>
I had to take a look at vsh.c to understand, but I finally got it that
it was
the absence of 'readline-devel' library in all the systems I tested.
Installing
this library made the test work as you intended in my box. It also fixed in
the Power servers as well.
Without the library, virsh interactive mode doesn't stop work per se,
but the
commands aren't echoed back to the output file (as we can see in the
code that is conditional of WITH_READLINE inside vsh.c).
Eric, feel free to change the tests as you see fit. Your idea a couple of
messages earlier (avoid using virsh interactive mode, use batch mode
instead) looks sound. It also avoid making the test dependent on the
install of readline-devel.
Okay, I've posted that alternative proposal of avoiding virsh
interactive mode (we can worry about the difference in command echoing
without readline-devel later).
--
Eric Blake, Principal Software Engineer
Red Hat, Inc. +1-919-301-3226
Virtualization:
qemu.org |
libvirt.org