[Intel-gfx] [PATCH igt] lib: Dump /sys/kernel/debug/suspend_stats after suspend failure

[Chris Wilson]

Quoting Arkadiusz Hiler (2017-11-16 14:31:03)
> On Tue, Nov 14, 2017 at 01:19:01PM +0000, Chris Wilson wrote:
> > I noticed that dpm was storing some information about which phase of
> > suspend failed inside suspend_stats. That will be useful to help debug
> > such failures, so automatically dump it after suspend fails.
> > 
> > Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> > ---
> >  lib/igt_aux.c | 22 ++++++++++++++++++----
> >  1 file changed, 18 insertions(+), 4 deletions(-)
> > 
> > diff --git a/lib/igt_aux.c b/lib/igt_aux.c
> > index ee53559c..0bcf792c 100644
> > --- a/lib/igt_aux.c
> > +++ b/lib/igt_aux.c
> > @@ -803,10 +803,24 @@ static void suspend_via_rtcwake(enum igt_suspend_state state)
> >       snprintf(cmd, sizeof(cmd), "rtcwake -s %d -m %s ",
> >                delay, suspend_state_name[state]);
> >       ret = igt_system(cmd);
> > -     igt_assert_f(ret == 0,
> > -                  "rtcwake failed with %i\n"
> > -                  "Check dmesg for further details.\n",
> > -                  ret);
> > +     if (ret) {
> > +             const char *path = "suspend_stats";
> > +             char *info;
> > +             int dir;
> > +
> > +             igt_warn("rtcwake failed with %i\n"
> > +                      "Check dmesg for further details.\n",
> > +                      ret);
> > +
> > +             dir = open(igt_debugfs_mount(), O_RDONLY);
> > +             info = igt_sysfs_get(dir, path);
> > +             close(dir);
> > +             if (info) {
> > +                     igt_debug("%s:\n%s\n", path, info);
> > +                     free(info);
> > +             }
> > +     }
> > +     igt_assert_eq(ret, 0);
> >  }
> >  
> >  static void suspend_via_sysfs(int power_dir, enum igt_suspend_state state)
> 
> We can see it in action here:
> https://intel-gfx-ci.01.org/tree/drm-tip/IGTPW_496/shard-apl3/igt@gem_eio@in-flight-suspend.html
> 
> (gem_eio:7686) igt-aux-DEBUG: suspend_stats:
> success: 10
> fail: 1
> failed_freeze: 0
> failed_prepare: 0
> failed_suspend: 1
> failed_suspend_late: 0
> failed_suspend_noirq: 0
> failed_resume: 0
> failed_resume_early: 0
> failed_resume_noirq: 0
> failures:
>   last_failed_dev:      
>                         
>   last_failed_errno:    -16
>                         0
>   last_failed_step:     suspend
> 
> 
> Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler at intel.com>
> and pushed

Now we just need dpm to actually record the device. I about 95% certain
we did not report -EBUSY; if i915_pm_suspend returns an error it gets
printed.
-Chris