[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-stable] [Qemu-devel] [PATCH v4 3/3] tests: start generic qemu-
From: |
Markus Armbruster |
Subject: |
Re: [Qemu-stable] [Qemu-devel] [PATCH v4 3/3] tests: start generic qemu-qmp tests |
Date: |
Thu, 06 Oct 2016 15:26:15 +0200 |
User-agent: |
Gnus/5.13 (Gnus v5.13) Emacs/24.5 (gnu/linux) |
Marc-André Lureau <address@hidden> writes:
> Hi
>
> ----- Original Message -----
>> Marc-André Lureau <address@hidden> writes:
>>
>> > Hi
>> >
>> > ----- Original Message -----
>> > -snip-
>> >
>> >>
>> >> I'd be willing to take this as is with a suitable TODO comment
>> >> explaining where we want to go with this file. Perhaps
>> >>
>> >> /*
>> >> * This program tests QMP commands that aren't interesting enough to
>> >> * warrant their own test program.
>> >> *
>> >> * TODO The tests we got here now aren't good examples, because they
>> >> * don't really exercise the commands, but only demonstrate specific
>> >> * bugs we've fixed.
>> >> */
>> >>
>> >> What do you think?
>> >
>> > It looks like a comment that may stale. I have a few tests in some wip
>> > branch that will go naturally there, so I hope it won't remain bug-fix
>> > only checks. I can't say how long it will take to get there though, so I
>> > am fine with a comment anyway, perhaps without TODO?
>> >
>> > thanks
>>
>> Since you got more tests coming up, we have several workable options:
>>
>> (1) Delay this patch until we got more substantial tests. I'm wary of
>> rejecting the imperfect solution I can have now for a better
>> solution I might get some day, but since you already got something
>> better in the pipeline, I'd be happy to wait in this case.
>>
>> (2) Apply it now, with my TODO. Adding tests should eventually resolve
>> the TODO. If we forget to delete it then, it'll go stale. But
>> it'll be pretty obviously stale.
>>
>> (3) Apply it now, without my TODO. Until we acquire tests that would
>> resolve the TODO, the file is an unmarked bad example.
>>
>> I like (1) better than (2), because it's less churn, and I don't expect
>> to lose anything. (3) my least favourite option, because I prefer maybe
>> having an obviously stale TODO in the future over having an umarked bad
>> example now.
>
> Please go for 2, the other series is unrelated and I don't know when I am
> going to send it.
>
> thanks
Done. Also: rename to tests/qmp-test.c, cover in MAINTAINERS.