Quoting Tvrtko Ursulin (2019-06-17 15:21:54) > > On 17/06/2019 11:09, Chris Wilson wrote: > > If the user is clearing the log buffer too slowly, we overflow. As this > > is an expected condition, and the driver tries to handle it, reduce the > > error message down to a notice. > > > > Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=110817 > > Signed-off-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> > > Cc: Michal Wajdeczko <michal.wajdeczko@xxxxxxxxx> > > Cc: Tvrtko Ursulin <tvrtko.ursulin@xxxxxxxxx> > > --- > > drivers/gpu/drm/i915/intel_guc_log.c | 4 +++- > > 1 file changed, 3 insertions(+), 1 deletion(-) > > > > diff --git a/drivers/gpu/drm/i915/intel_guc_log.c b/drivers/gpu/drm/i915/intel_guc_log.c > > index bf1446629703..e3b83ecb90b5 100644 > > --- a/drivers/gpu/drm/i915/intel_guc_log.c > > +++ b/drivers/gpu/drm/i915/intel_guc_log.c > > @@ -208,7 +208,9 @@ static bool guc_check_log_buf_overflow(struct intel_guc_log *log, > > /* buffer_full_cnt is a 4 bit counter */ > > log->stats[type].sampled_overflow += 16; > > } > > - DRM_ERROR_RATELIMITED("GuC log buffer overflow\n"); > > + > > + dev_notice_ratelimited(guc_to_i915(log_to_guc(log))->drm.dev, > > + "GuC log buffer overflow\n"); > > } > > > > return overflow; > > > > Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@xxxxxxxxx> > > Maybe we could even go debug, although I not sure if userspace is > expected to look at dmesg to detect overflows or if the thing has some > marker it writes into the file in that case. I don't see it so that's > probably a slight flaw. It wasn't completely clear to me how well userspace did handle this (I haven't run the debug tools myself) so settled for leaving a hint visible in dmesg. Ideally, if we are expecting userspace to be handling the ringbuf interface, it should be coping with the overflows then this is just noise and can be removed. -Chris _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
