On Wed, Aug 30, 2017 at 09:15:03PM +0300, Arkadiusz Hiler wrote: > On Wed, Aug 30, 2017 at 04:56:09PM +0300, Paul Kocialkowski wrote: > > This removes the igt_require condition on the sysfs open call used to > > write the suspend/resume delay so that it is allowed to fail. Intsead, > > the code that depends on it is put in a conditional block. > > > > This allows running test binaries as a non-privileged user for e.g. > > listing the available tests with the SuspendResumeDelay parameter set > > in igtrc configuration. Sysfs access would otherwise cause it to fail. > > > > Signed-off-by: Paul Kocialkowski <paul.kocialkowski@xxxxxxxxxxxxxxx> > Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@xxxxxxxxx> This is the wrong fix. When enumerating tests you're not supposed to touch anything hw or system related at all. Allocating a bit of memory and stuff like that is ok, but not anything with effects. The correct way to handle this is by wrapping this into an igt_fixture block. Hitting an igt_require/assert outside of an igt_fixture or igt_subtest should result in a assert. You still have igt_skip and igt_require in there, so not fixed properly yet. Aside: We need to come up with a way to have functions shared between parts of the library, without documenting and exporting them to tests. -Daniel -- Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
