[PATCH i-g-t] tests/device_reset: Skip if no drm devices are preset

[Janusz Krzysztofik]

On Tuesday, 12 March 2024 11:54:32 CET Nirmoy Das wrote:
> 
> On 3/12/2024 11:43 AM, Janusz Krzysztofik wrote:
> > Hi Nirmoy,
> >
> > On Tuesday, 12 March 2024 11:07:57 CET Nirmoy Das wrote:
> >> Skip the test if no drm devices are preset.
> > Did you mean present? (also in the subject line)
> Right, another typo
> >
> >> v2: fix a typo.
> >>
> >> Closes:https://gitlab.freedesktop.org/drm/intel/-/issues/10087
> >> Signed-off-by: Nirmoy Das<nirmoy.das at intel.com>
> >> ---
> >>   tests/device_reset.c | 2 +-
> >>   1 file changed, 1 insertion(+), 1 deletion(-)
> >>
> >> diff --git a/tests/device_reset.c b/tests/device_reset.c
> >> index 817cbc285..73531d696 100644
> >> --- a/tests/device_reset.c
> >> +++ b/tests/device_reset.c
> >> @@ -198,7 +198,7 @@ static void init_device_fds(struct device_fds *dev)
> >>   	 * a device file descriptor open for exit handler use.
> >>   	 */
> >>   	dev->fds.dev = __drm_open_driver(DRIVER_ANY);
> >> -	igt_assert_fd(dev->fds.dev);
> >> +	igt_require(dev->fds.dev != -1);
> > What if __drm_open_driver() returns (unlikely) a negative value other than 
-1?
> 
> __drm_open_driverdoc says: "An open DRM fd or -1 on error" so that won't 
happen.

That won't happen as long as there are no bugs in __drm_open_driver().

If you don't want to skip on (fd < -1) then let's fail, but we should never 
succeed in that unlikely case, I believe.

Thanks,
Janusz

> 
> >   
> > In IGT we used to verify validity of a file descriptor by checking if its
> > value is not negative, i.e., >= 0 (see e.g. lib/igt_core:igt_assert_fd(),
> > lib/drmtest.c:drm_open_driver()).  Please follow the same convention.
> 
> I can do s/igt_assert_fd/igt_require_fd
> 
> Thanks,
> 
> Nirmoy
> 
> > Thanks,
> > Janusz
> >
> >>   	if (is_i915_device(dev->fds.dev))
> >>   		igt_require_gem(dev->fds.dev);
> >>   
> >>
> >
> >
> >