Re: [PATCH] iotests: Revert emulator selection to old behaviour

[Kevin Wolf <kwolf@redhat.com>]

Am 02.02.2021 um 15:41 hat Daniel P. Berrangé geschrieben:
> On Tue, Feb 02, 2021 at 03:28:02PM +0100, Kevin Wolf wrote:
> > If the qemu-system-{arch} binary for the host architecture can't be
> > found, the old 'check' implementation selected the alphabetically first
> > system emulator binary that it could find. The new Python implementation
> > just uses the first result of glob.iglob(), which has an undefined
> > order.
> > 
> > This is a problem that breaks CI because the iotests aren't actually
> > prepared to run on any emulator. They should be, so this is really a bug
> > in the failing test cases that should be fixed there, but as a quick
> > fix, let's revert to the old behaviour to let CI runs succeed again.
> 
> Deterministic CI is critically important.

True. I didn't mean to imply that we don't want deterministic behaviour,
but just that this hides some bugs in the test cases that we'll want to
have fixed eventually, too.

Maybe we should rely on automatic picking less and specify different
emulators explicitly in different CI jobs so that we don't only test the
same binary over and over again and others not at all.

Kevin