| Message ID | 20230321170936.478631-3-andi.shyti@linux.intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Report MMIO communication problems more clearly | expand |
> Although we now sanitycheck MMIO access during driver load to make sure > the MMIO BAR isn't returning all 0xFFFFFFFF, there have been a few cases > where (temporarily?) unreliable MMIO access has happened after GPU > resets or power events. We'll often notice this on our next GT register > access since forcewake handling will fail; let's change our handling > slightly so that when this happens we print a more meaningful message > clarifying that the problem is the MMIO access, not forcewake > specifically. > > Signed-off-by: Matt Roper <matthew.d.roper@intel.com> > Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> > Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Andi
On 21.03.2023 18:09, Andi Shyti wrote: > From: Matt Roper <matthew.d.roper@intel.com> > > Although we now sanitycheck MMIO access during driver load to make sure > the MMIO BAR isn't returning all 0xFFFFFFFF, there have been a few cases > where (temporarily?) unreliable MMIO access has happened after GPU > resets or power events. We'll often notice this on our next GT register > access since forcewake handling will fail; let's change our handling > slightly so that when this happens we print a more meaningful message > clarifying that the problem is the MMIO access, not forcewake > specifically. > > Signed-off-by: Matt Roper <matthew.d.roper@intel.com> > Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> > Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com> > --- > drivers/gpu/drm/i915/intel_uncore.c | 12 +++++++++--- > 1 file changed, 9 insertions(+), 3 deletions(-) > > diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c > index 0b69081d6d285..303a5d38c93a5 100644 > --- a/drivers/gpu/drm/i915/intel_uncore.c > +++ b/drivers/gpu/drm/i915/intel_uncore.c > @@ -178,9 +178,15 @@ static inline void > fw_domain_wait_ack_clear(const struct intel_uncore_forcewake_domain *d) > { > if (wait_ack_clear(d, FORCEWAKE_KERNEL)) { > - drm_err(&d->uncore->i915->drm, > - "%s: timed out waiting for forcewake ack to clear.\n", > - intel_uncore_forcewake_domain_to_str(d->id)); > + if (fw_ack(d) == ~0) > + drm_err(&d->uncore->i915->drm, > + "%s: MMIO unreliable (forcewake register returns 0xFFFFFFFF)!\n", > + intel_uncore_forcewake_domain_to_str(d->id)); > + else > + drm_err(&d->uncore->i915->drm, > + "%s: timed out waiting for forcewake ack to clear.\n", > + intel_uncore_forcewake_domain_to_str(d->id)); > + With: if (!wait_ack_clear(...)) return; you do not need to indent. And since drm_err have the same set of args in both cases, except fmt, it could be also simplified. Anyway I am not sure if it is worth effort, so: Reviewed-by: Andrzej Hajda <andrzej.hajda@intel.com> Regards Andrzej > add_taint_for_CI(d->uncore->i915, TAINT_WARN); /* CI now unreliable */ > } > }
diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c index 0b69081d6d285..303a5d38c93a5 100644 --- a/drivers/gpu/drm/i915/intel_uncore.c +++ b/drivers/gpu/drm/i915/intel_uncore.c @@ -178,9 +178,15 @@ static inline void fw_domain_wait_ack_clear(const struct intel_uncore_forcewake_domain *d) { if (wait_ack_clear(d, FORCEWAKE_KERNEL)) { - drm_err(&d->uncore->i915->drm, - "%s: timed out waiting for forcewake ack to clear.\n", - intel_uncore_forcewake_domain_to_str(d->id)); + if (fw_ack(d) == ~0) + drm_err(&d->uncore->i915->drm, + "%s: MMIO unreliable (forcewake register returns 0xFFFFFFFF)!\n", + intel_uncore_forcewake_domain_to_str(d->id)); + else + drm_err(&d->uncore->i915->drm, + "%s: timed out waiting for forcewake ack to clear.\n", + intel_uncore_forcewake_domain_to_str(d->id)); + add_taint_for_CI(d->uncore->i915, TAINT_WARN); /* CI now unreliable */ } }