[Intel-gfx] i915: hotplug events gone wild

Andrew Lutomirski luto at mit.edu
Sun Jan 31 12:49:09 CET 2010


On Sat, Jan 30, 2010 at 10:02 PM, Andrew Lutomirski <luto at mit.edu> wrote:
> [I posted this bug earlier with a terrible description as "resume
> lagginess and other problems."  Here it is again with a better bug
> report.]
>
> I'm running 2.6.33-rc5 (plus some wireless-testing stuff, but I've
> seen this problem on a variety of 2.6.33-rc? kernels).  Every now and
> then, X starts to lag badly on my GM45 laptop.  When this happens, I
> usually see a bunch of events in udevmonitor.  Running with
> drm.debug=0x02 (and the patch below to keep the log under control), I
> see tons of messages like this:

I triggered it again.  This time, the messages looked like
(drm.debug=3 from a different VT to avoid all the hotplug stuff
running off the screen, and running a different debugging hack -- see
all the way at the bottom):

[ 1324.285057] [drm:i915_driver_irq_handler], hotplug event received,
stat 0x28200000, mask 0x38000800
[ 1324.285086] [drm:i915_hotplug_work_func], i915_hotplug_work_func
[ 1324.285093] [drm:drm_sysfs_hotplug_event], generating hotplug event
[ 1324.290356] [drm:i915_driver_irq_handler], hotplug event received,
stat 0x28200000, mask 0x38000800
[ 1324.290388] [drm:i915_hotplug_work_func], i915_hotplug_work_func
[ 1324.290395] [drm:drm_sysfs_hotplug_event], generating hotplug event
[ 1324.296512] [drm:i915_driver_irq_handler], hotplug event received,
stat 0x28200000, mask 0x38000800
[ 1324.296536] [drm:i915_hotplug_work_func], i915_hotplug_work_func
[ 1324.296543] [drm:drm_sysfs_hotplug_event], generating hotplug event
[ 1324.307078] [drm:i915_driver_irq_handler], hotplug event received,
stat 0x20200000, mask 0x38000800
[ 1324.307104] [drm:i915_hotplug_work_func], i915_hotplug_work_func
[ 1324.307111] [drm:drm_sysfs_hotplug_event], generating hotplug event

>From X, I got:

[ 1174.177474] [drm:drm_ioctl], pid=1889, cmd=0x40046460, nr=0x60, dev
0xe200, auth=1
[ 1174.177480] [drm:drm_ioctl], pid=1889, cmd=0x400c645f, nr=0x5f, dev
0xe200, auth=1
[ 1174.177485] [drm:drm_ioctl], pid=1889, cmd=0x400c645f, nr=0x5f, dev
0xe200, auth=1
[ 1174.177492] [drm:drm_ioctl], pid=1889, cmd=0x40046460, nr=0x60, dev
0xe200, auth=1
[ 1174.177497] [drm:drm_ioctl], pid=1889, cmd=0x40046460, nr=0x60, dev
0xe200, auth=1
[ 1174.177522] [drm:drm_ioctl], pid=1889, cmd=0xc0086457, nr=0x57, dev
0xe200, auth=1
[ 1174.177528] [drm:drm_ioctl], pid=1889, cmd=0xc00c6466, nr=0x66, dev
0xe200, auth=1

Using intel_reg_write to set PORT_HOTPLUG_EN to 0 silenced udevmonitor
but did *not* make X stop lagging.

Restarting X helped for a few minutes, but then the bug triggered again.

--Andy

The patch was:

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 0c67924..d0977ea 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1618,7 +1618,7 @@ i915_add_request(struct drm_device *dev, struct
drm_file *file_priv,
 	OUT_RING(MI_USER_INTERRUPT);
 	ADVANCE_LP_RING();

-	DRM_DEBUG_DRIVER("%d\n", seqno);
+	//DRM_DEBUG_DRIVER("%d\n", seqno);

 	request->seqno = seqno;
 	request->emitted_jiffies = jiffies;
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 89a071a..0cda630 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -266,6 +266,7 @@ static void i915_hotplug_work_func(struct work_struct *work)
 		}
 	}
 	/* Just fire off a uevent and let userspace tell us what to do */
+	DRM_DEBUG_DRIVER("i915_hotplug_work_func\n");
 	drm_sysfs_hotplug_event(dev);
 }

@@ -616,8 +617,8 @@ irqreturn_t i915_driver_irq_handler(DRM_IRQ_ARGS)
 		    (iir & I915_DISPLAY_PORT_INTERRUPT)) {
 			u32 hotplug_status = I915_READ(PORT_HOTPLUG_STAT);

-			DRM_DEBUG_DRIVER("hotplug event received, stat 0x%08x\n",
-				  hotplug_status);
+			DRM_DEBUG_DRIVER("hotplug event received, stat 0x%08x, mask 0x%08x\n",
+				  hotplug_status, dev_priv->hotplug_supported_mask);
 			if (hotplug_status & dev_priv->hotplug_supported_mask)
 				queue_work(dev_priv->wq,
 					   &dev_priv->hotplug_work);
diff --git a/drivers/gpu/drm/i915/intel_display.c
b/drivers/gpu/drm/i915/intel_display.c
index 45da78e..f2040bf 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -3759,7 +3759,7 @@ static void intel_gpu_idle_timer(unsigned long arg)
 	struct drm_device *dev = (struct drm_device *)arg;
 	drm_i915_private_t *dev_priv = dev->dev_private;

-	DRM_DEBUG_DRIVER("idle timer fired, downclocking\n");
+	//DRM_DEBUG_DRIVER("idle timer fired, downclocking\n");

 	dev_priv->busy = false;

@@ -3774,7 +3774,7 @@ static void intel_crtc_idle_timer(unsigned long arg)
 	struct drm_crtc *crtc = &intel_crtc->base;
 	drm_i915_private_t *dev_priv = crtc->dev->dev_private;

-	DRM_DEBUG_DRIVER("idle timer fired, downclocking\n");
+	//DRM_DEBUG_DRIVER("idle timer fired, downclocking\n");

 	intel_crtc->busy = false;



More information about the Intel-gfx mailing list