[Intel-gfx] [PATCH] drm/i915: Limit the auto arming of mmio debugs on vlv/chv
Daniel Vetter
daniel at ffwll.ch
Wed Jan 20 02:45:49 PST 2016
On Wed, Jan 20, 2016 at 12:32:23PM +0200, Mika Kuoppala wrote:
> The capability to detect unclaimed register access was
> recently introduced for vlv/chv platforms. Apparently
> there are plenty of unclaimed access on these platforms,
> resulting in new dmesg warns. But as we are trying to form
> a beachhead for CI/Bat, all new warns are adding to the
> noise and thus not desirable at this point in time.
>
> Make it so that if in these platforms the automatic arming
> was responsible for mmio_debug enabling, ignore the warns.
>
> If user/dev wants to fix these, he can still do so by
> i915.mmio_debug=1234.
>
> Cc: Daniel Vetter <daniel.vetter at ffwll.ch>
> Signed-off-by: Mika Kuoppala <mika.kuoppala at intel.com>
Yeah it sucks, but otoh this will make CI more useful for everyone else
and we can keep the original test coverage even.
Reviewed-by: Daniel Vetter <daniel.vetter at ffwll.ch>
> ---
> drivers/gpu/drm/i915/intel_uncore.c | 9 +++++++++
> 1 file changed, 9 insertions(+)
>
> diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
> index c3c13dc929cb..bfa79e5c214e 100644
> --- a/drivers/gpu/drm/i915/intel_uncore.c
> +++ b/drivers/gpu/drm/i915/intel_uncore.c
> @@ -631,6 +631,15 @@ __unclaimed_reg_debug(struct drm_i915_private *dev_priv,
> const bool read,
> const bool before)
> {
> + /* XXX. We limit the auto arming traces for mmio
> + * debugs on these platforms. There are just too many
> + * revealed by these and CI/Bat suffers from the noise.
> + * Please fix and then re-enable the automatic traces.
> + */
> + if (i915.mmio_debug < 2 &&
> + (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)))
> + return;
> +
> if (WARN(check_for_unclaimed_mmio(dev_priv),
> "Unclaimed register detected %s %s register 0x%x\n",
> before ? "before" : "after",
> --
> 2.5.0
>
--
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
More information about the Intel-gfx
mailing list