linux-kernel - Re: [Intel-gfx] Excessive WARN()s in Intel 915 driver

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Tue, 28 Jan 2014 20:42:14 +0100
From:	Daniel Vetter <daniel@...ll.ch>
To:	Ville Syrjälä <ville.syrjala@...ux.intel.com>
Cc:	Alan Stern <stern@...land.harvard.edu>,
	intel-gfx <intel-gfx@...ts.freedesktop.org>,
	Kernel development list <linux-kernel@...r.kernel.org>
Subject: Re: [Intel-gfx] Excessive WARN()s in Intel 915 driver

On Tue, Jan 28, 2014 at 06:17:00PM +0100, Daniel Vetter wrote:
> On Wed, Jan 8, 2014 at 7:43 PM, Ville Syrjälä
> <ville.syrjala@...ux.intel.com> wrote:
> >
> > The log looks fairly clear to me:
> >
> > 1. initially panel fitter is enabled on pipe B, and pipe B is outputting
> >    to LVDS and VGA. Border is enabled.
> > 2. pipe A gets enabled outputting to LVDS. This will overwrite the
> >    LVDS border bits
> > 3. pipe B is still active so we do the state check, but as the LVDS
> >    border bits have been clobbered earlier, the state checker gets
> >    angry
> 
> Meh, I've been fairly dense the entire time. This is indeed the
> root-cause, with the twist that we're allowing the impossible:
> Essentially we take away the panel fitter from pipe B to pipe A while
> it is strictly still in use by pipe B for VGA. Currently no idea how
> to properly fix this in a not too intrusive way.
> 
> The other issue is that the encoders connected to pipe B change in the
> first modeset, but that's not reflected in the pipe masks.

I think I need a bit more debug output first. Can you please apply the
below patch to drm-intel-nightly and then grab a drm.debug=0xe dmesg from
boot?

Thanks, Daniel
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 122f87155b8e..554c30e89308 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -9144,6 +9144,11 @@ intel_modeset_affected_pipes(struct drm_crtc *crtc, unsigned *modeset_pipes,
 		if (connector->new_encoder)
 			*prepare_pipes |=
 				1 << connector->new_encoder->new_crtc->pipe;
+
+		DRM_DEBUG_KMS("[CONNECTOR:%d:%s]: prepare_pipes %u\n",
+			connector->base.base.id,
+			drm_get_connector_name(&connector->base),
+			*prepare_pipes);
 	}
 
 	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
@@ -9159,6 +9164,11 @@ intel_modeset_affected_pipes(struct drm_crtc *crtc, unsigned *modeset_pipes,
 
 		if (encoder->new_crtc)
 			*prepare_pipes |= 1 << encoder->new_crtc->pipe;
+
+		DRM_DEBUG_KMS("[ENCODER:%d:%s]: prepare_pipes %u\n",
+			encoder->base.base.id,
+			drm_get_encoder_name(&encoder->base),
+			*prepare_pipes);
 	}
 
 	/* Check for pipes that will be enabled/disabled ... */
@@ -9171,6 +9181,11 @@ intel_modeset_affected_pipes(struct drm_crtc *crtc, unsigned *modeset_pipes,
 			*disable_pipes |= 1 << intel_crtc->pipe;
 		else
 			*prepare_pipes |= 1 << intel_crtc->pipe;
+
+		DRM_DEBUG_KMS("[CRTC:%d:%s]: prepare_pipes %u\n",
+			intel_crtc->base.base.id,
+			pipe_name(intel_crtc->pipe),
+			*prepare_pipes);
 	}
 
 
-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/