Skip to contentSkip to conversation

gen6_gt_force_wake issue

PATCH Applied Nicolas Kalkhof 4 participants 7 messages Oct 5, 2011 – Nov 3, 2011

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

Hello,

current xf86-video-intel git crashex X frequently on my machine (see kernel log). Any Workarounds/Ideas?

Regards,
Nico

Oct  5 08:58:53 gonzo kernel: WARNING: at drivers/gpu/drm/i915/i915_drv.c:372 gen6_gt_force_wake_put+0x1f/0x45 [i915]()
Oct  5 08:58:53 gonzo kernel: Hardware name: 4236NGG
Oct  5 08:58:53 gonzo kernel: Modules linked in: i915 fbcon font bitblit softcursor drm_kms_helper drm fb fbdev snd_hda_codec_conexant thinkpad_acpi iwlagn cfbcopyarea nvram intel_agp intel_gtt snd_hda_intel agpgart video backlight e1000e cfbimgblt cfbfillrect snd_hda_codec Oct  5 08:58:53 gonzo kernel: Pid: 3358, comm: cat Tainted: G        W   3.1.0-rc9 #4
Oct  5 08:58:53 gonzo kernel: Call Trace:
Oct  5 08:58:53 gonzo kernel: [<ffffffff8102fc96>] ? warn_slowpath_common+0x78/0x8c
Oct  5 08:58:53 gonzo kernel: [<ffffffffa01094c0>] ? gen6_gt_force_wake_put+0x1f/0x45 [i915]
Oct  5 08:58:53 gonzo kernel: [<ffffffffa0112795>] ? i915_delayfreq_table+0x4c/0x95 [i915]
Oct  5 08:58:53 gonzo kernel: [<ffffffff810baa2d>] ? seq_read+0x16e/0x355
Oct  5 08:58:53 gonzo kernel: [<ffffffff810a1efa>] ? vfs_read+0xac/0x126
Oct  5 08:58:53 gonzo kernel: [<ffffffff810a1fb9>] ? sys_read+0x45/0x6e
Oct  5 08:58:53 gonzo kernel: [<ffffffff813b3cfb>] ? system_call_fastpath+0x16/0x1b

___________________________________________________________
SMS schreiben mit WEB.DE FreeMail - einfach, schnell und kostenguenstig. Jetzt gleich testen! http://f.web.de/?mc=021192
_______________________________________________
Intel-gfx mailing list [email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#1 lore raw

PATCH 1/1 drm/i915: forcewake warning fixes in debugfs

Some more unsafe debugfs access are fixed with this patch. I tested all reads, but didn't thoroughly test the writes.

Cc "Nicolas Kalkhof" <[email protected]>
---
 drivers/gpu/drm/i915/i915_debugfs.c |   57 +++++++++++++++++++++++++++++++----
 1 files changed, 51 insertions(+), 6 deletions(-)

---
Can you please try this patch to see if it helps?

drivers/gpu/drm/i915/i915_debugfs.c +51 -6
index 3c395a5..de233df 100644
@@ -635,11 +635,16 @@ static int i915_ringbuffer_info(struct seq_file *m, void *data)
 	struct drm_device *dev = node->minor->dev;
 	drm_i915_private_t *dev_priv = dev->dev_private;
 	struct intel_ring_buffer *ring;
+	int ret;
 
 	ring = &dev_priv->ring[(uintptr_t)node->info_ent->data];
 	if (ring->size == 0)
 		return 0;
 
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
+
 	seq_printf(m, "Ring %s:\n", ring->name);
 	seq_printf(m, "  Head :    %08x\n", I915_READ_HEAD(ring) & HEAD_ADDR);
 	seq_printf(m, "  Tail :    %08x\n", I915_READ_TAIL(ring) & TAIL_ADDR);
@@ -653,6 +658,8 @@ static int i915_ringbuffer_info(struct seq_file *m, void *data)
 	seq_printf(m, "  Control : %08x\n", I915_READ_CTL(ring));
 	seq_printf(m, "  Start :   %08x\n", I915_READ_START(ring));
 
+	mutex_unlock(&dev->struct_mutex);
+
 	return 0;
 }
 
@@ -841,7 +848,16 @@ static int i915_rstdby_delays(struct seq_file *m, void *unused)
 	struct drm_info_node *node = (struct drm_info_node *) m->private;
 	struct drm_device *dev = node->minor->dev;
 	drm_i915_private_t *dev_priv = dev->dev_private;
-	u16 crstanddelay = I915_READ16(CRSTANDVID);
+	u16 crstanddelay;
+	int ret;
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
+
+	crstanddelay = I915_READ16(CRSTANDVID);
+
+	mutex_unlock(&dev->struct_mutex);
 
 	seq_printf(m, "w/ctx: %d, w/o ctx: %d\n", (crstanddelay >> 8) & 0x3f, (crstanddelay & 0x3f));
 
@@ -939,7 +955,11 @@ static int i915_delayfreq_table(struct seq_file *m, void *unused)
 	struct drm_device *dev = node->minor->dev;
 	drm_i915_private_t *dev_priv = dev->dev_private;
 	u32 delayfreq;
-	int i;
+	int ret, i;
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
 
 	for (i = 0; i < 16; i++) {
 		delayfreq = I915_READ(PXVFREQ_BASE + i * 4);
@@ -947,6 +967,8 @@ static int i915_delayfreq_table(struct seq_file *m, void *unused)
 			   (delayfreq & PXVFREQ_PX_MASK) >> PXVFREQ_PX_SHIFT);
 	}
 
+	mutex_unlock(&dev->struct_mutex);
+
 	return 0;
 }
 
@@ -961,13 +983,19 @@ static int i915_inttoext_table(struct seq_file *m, void *unused)
 	struct drm_device *dev = node->minor->dev;
 	drm_i915_private_t *dev_priv = dev->dev_private;
 	u32 inttoext;
-	int i;
+	int ret, i;
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
 
 	for (i = 1; i <= 32; i++) {
 		inttoext = I915_READ(INTTOEXT_BASE_ILK + i * 4);
 		seq_printf(m, "INTTOEXT%02d: 0x%08x\n", i, inttoext);
 	}
 
+	mutex_unlock(&dev->struct_mutex);
+
 	return 0;
 }
 
@@ -976,9 +1004,19 @@ static int i915_drpc_info(struct seq_file *m, void *unused)
 	struct drm_info_node *node = (struct drm_info_node *) m->private;
 	struct drm_device *dev = node->minor->dev;
 	drm_i915_private_t *dev_priv = dev->dev_private;
-	u32 rgvmodectl = I915_READ(MEMMODECTL);
-	u32 rstdbyctl = I915_READ(RSTDBYCTL);
-	u16 crstandvid = I915_READ16(CRSTANDVID);
+	u32 rgvmodectl, rstdbyctl;
+	u16 crstandvid;
+	int ret;
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
+
+	rgvmodectl = I915_READ(MEMMODECTL);
+	rstdbyctl = I915_READ(RSTDBYCTL);
+	crstandvid = I915_READ16(CRSTANDVID);
+
+	mutex_unlock(&dev->struct_mutex);
 
 	seq_printf(m, "HD boost: %s\n", (rgvmodectl & MEMMODE_BOOST_EN) ?
 		   "yes" : "no");
@@ -1166,9 +1204,16 @@ static int i915_gfxec(struct seq_file *m, void *unused)
 	struct drm_info_node *node = (struct drm_info_node *) m->private;
 	struct drm_device *dev = node->minor->dev;
 	drm_i915_private_t *dev_priv = dev->dev_private;
+	int ret;
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
 
 	seq_printf(m, "GFXEC: %ld\n", (unsigned long)I915_READ(0x112f4));
 
+	mutex_unlock(&dev->struct_mutex);
+
 	return 0;
 }
 
Signature · 1 line
1.7.7
#3 lore raw

Re: [PATCH] drm/i915: forcewake warning fixes in debugfs

Hi Keith,

This patch isn't in your -next pull request. Please consider merging for 3.2.

Yours, Daniel

Quoted 12 lines · Daniel Vetter
Quoted 5 lines · Ben Widawsky

Some more unsafe debugfs access are fixed with this patch. I tested all reads, but didn't thoroughly test the writes.

Cc "Nicolas Kalkhof" <[email protected]>
Signature · 3 lines
Daniel Vetter
Mail: [email protected]
Mobile: +41 (0)79 365 57 48
Signature · 3 lines
Daniel Vetter
Mail: [email protected]
Mobile: +41 (0)79 365 57 48
#4 lore raw

Re: [PATCH] drm/i915: forcewake warning fixes in debugfs

Quoted 4 lines · Daniel Vetter

Hi Keith,

This patch isn't in your -next pull request. Please consider merging for 3.2.

I didn't ever see a reply from Nicolas that it fixed his problem; would be nice to know whether this actually worked...

Signature · 1 line
[email protected]

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#5 lore raw

Re: [PATCH] drm/i915: forcewake warning fixes in debugfs

Quoted 9 lines · On Sun, 23 Oct 2011 12:21:38 -0700 Keith Packard <[email protected]> wrote:
Quoted 4 lines · Daniel Vetter

Hi Keith,

This patch isn't in your -next pull request. Please consider merging for 3.2.

I didn't ever see a reply from Nicolas that it fixed his problem; would be nice to know whether this actually worked...

I highly doubt it fixed his problem. The problem I was trying to fix was to remove the spurious warnings we get, which was most likely the case here. This patch does that correctly.

Ben