Skip to contentSkip to conversation

drm/i915: suppress atomic commit error message under gvt-g env

PATCH Superseded [email protected] 4 participants 8 messages Mar 3, 2017 – Mar 6, 2017

Subsystems (2)

v1 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 drm/i915: suppress atomic commit error message under gvt-g env

From: Bing Niu <[email protected]>

under virtualization enviroment, it is possible guest update pipe registers across vblank intervals due to overhead of mmio traps or vm schedule out. However, it is safe since those pipe update happen in virual registers and will not be committed to hardware. suppress that atomic commit error message under virtualization case to avoid confusing user.

---
 drivers/gpu/drm/i915/intel_sprite.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/gpu/drm/i915/intel_sprite.c +2 -1
index b16a295..5ce1ec6 100644
@@ -158,6 +158,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
 	int scanline_end = intel_get_crtc_scanline(crtc);
 	u32 end_vbl_count = intel_crtc_get_vblank_counter(crtc);
 	ktime_t end_vbl_time = ktime_get();
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	if (work) {
 		work->flip_queued_vblank = end_vbl_count;
@@ -184,7 +185,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
 	local_irq_enable();
 
 	if (crtc->debug.start_vbl_count &&
-	    crtc->debug.start_vbl_count != end_vbl_count) {
+	    crtc->debug.start_vbl_count != end_vbl_count && !intel_vgpu_active(dev_priv)) {
 		DRM_ERROR("Atomic update failure on pipe %c (start=%u end=%u) time %lld us, min %d, max %d, scanline start %d, end %d\n",
 			  pipe_name(pipe), crtc->debug.start_vbl_count,
 			  end_vbl_count,
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#1 lore raw

Can we directly use DRM_DEBUG_KMS() for this periodic error message?

Quoted 28 lines · On 03/03/17 19:53, [email protected] wrote:

From: Bing Niu <[email protected]>

under virtualization enviroment, it is possible guest update pipe registers across vblank intervals due to overhead of mmio traps or vm schedule out. However, it is safe since those pipe update happen in virual registers and will not be committed to hardware. suppress that atomic commit error message under virtualization case to avoid confusing user.

---
  drivers/gpu/drm/i915/intel_sprite.c | 3 ++-
  1 file changed, 2 insertions(+), 1 deletion(-)
drivers/gpu/drm/i915/intel_sprite.c +1 -0
index b16a295..5ce1ec6 100644
@@ -158,6 +158,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
  	int scanline_end = intel_get_crtc_scanline(crtc);
  	u32 end_vbl_count = intel_crtc_get_vblank_counter(crtc);
  	ktime_t end_vbl_time = ktime_get();
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
  	if (work) {
  		work->flip_queued_vblank = end_vbl_count;

drivers/gpu/drm/i915/intel_sprite.c void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work · line 191

Quoted 8 lines
drivers/gpu/drm/i915/intel_sprite.c +1 -1
@@ -184,7 +185,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
  	local_irq_enable();
 
  	if (crtc->debug.start_vbl_count &&
-	    crtc->debug.start_vbl_count != end_vbl_count) {
+	    crtc->debug.start_vbl_count != end_vbl_count && !intel_vgpu_active(dev_priv)) {
  		DRM_ERROR("Atomic update failure on pipe %c (start=%u end=%u) time %lld us, min %d, max %d, scanline start %d, end %d\n",
  			  pipe_name(pipe), crtc->debug.start_vbl_count,
  			  end_vbl_count,

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

author
#2 lore raw

that depends on whether native can accept this log level degraded. let's see if there another opinions from community~

-----Original Message-----
From: Wang, Zhi A
Sent: Friday, March 03, 2017 10:56 AM
To: Niu, Bing <[email protected]>; [email protected]
Cc: Lv, Zhiyuan <[email protected]>
Subject: Re: [Intel-gfx][PATCH] drm/i915: suppress atomic commit error message under gvt-g env

Can we directly use DRM_DEBUG_KMS() for this periodic error message?

Quoted 32 lines · On 03/03/17 19:53, [email protected] wrote:

From: Bing Niu <[email protected]>

under virtualization enviroment, it is possible guest update pipe registers across vblank intervals due to overhead of mmio traps or vm schedule out. However, it is safe since those pipe update happen in virual registers and will not be committed to hardware. suppress that atomic commit error message under virtualization case to avoid confusing user.

---
  drivers/gpu/drm/i915/intel_sprite.c | 3 ++-
  1 file changed, 2 insertions(+), 1 deletion(-)
drivers/gpu/drm/i915/intel_sprite.c +1 -0
index b16a295..5ce1ec6 100644
@@ -158,6 +158,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
  	int scanline_end = intel_get_crtc_scanline(crtc);
  	u32 end_vbl_count = intel_crtc_get_vblank_counter(crtc);
  	ktime_t end_vbl_time = ktime_get();
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
  	if (work) {
  		work->flip_queued_vblank = end_vbl_count; @@ -184,7 +185,7 @@ void 

intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
local_irq_enable();

if (crtc->debug.start_vbl_count &&

drivers/gpu/drm/i915/intel_sprite.c void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work · line 191

Quoted 6 lines
drivers/gpu/drm/i915/intel_sprite.c +2 -1
-	    crtc->debug.start_vbl_count != end_vbl_count) {
+	    crtc->debug.start_vbl_count != end_vbl_count && 
+!intel_vgpu_active(dev_priv)) {
  		DRM_ERROR("Atomic update failure on pipe %c (start=%u end=%u) time %lld us, min %d, max %d, scanline start %d, end %d\n",
  			  pipe_name(pipe), crtc->debug.start_vbl_count,
  			  end_vbl_count,

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

#3 lore raw
In reply to Zhi Wang
Quoted 1 line · Zhi Wang

Can we directly use DRM_DEBUG_KMS() for this periodic error message?

No. We want to actually know when/if this fails.

Quoted 42 lines · On 03/03/17 19:53, [email protected] wrote:
Quoted 28 lines

From: Bing Niu <[email protected]>

under virtualization enviroment, it is possible guest update pipe registers across vblank intervals due to overhead of mmio traps or vm schedule out. However, it is safe since those pipe update happen in virual registers and will not be committed to hardware. suppress that atomic commit error message under virtualization case to avoid confusing user.

---
  drivers/gpu/drm/i915/intel_sprite.c | 3 ++-
  1 file changed, 2 insertions(+), 1 deletion(-)
drivers/gpu/drm/i915/intel_sprite.c +1 -0
index b16a295..5ce1ec6 100644
@@ -158,6 +158,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
  	int scanline_end = intel_get_crtc_scanline(crtc);
  	u32 end_vbl_count = intel_crtc_get_vblank_counter(crtc);
  	ktime_t end_vbl_time = ktime_get();
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
  	if (work) {
  		work->flip_queued_vblank = end_vbl_count;

drivers/gpu/drm/i915/intel_sprite.c void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work · line 191

Quoted 8 lines
drivers/gpu/drm/i915/intel_sprite.c +1 -1
@@ -184,7 +185,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
  	local_irq_enable();
 
  	if (crtc->debug.start_vbl_count &&
-	    crtc->debug.start_vbl_count != end_vbl_count) {
+	    crtc->debug.start_vbl_count != end_vbl_count && !intel_vgpu_active(dev_priv)) {
  		DRM_ERROR("Atomic update failure on pipe %c (start=%u end=%u) time %lld us, min %d, max %d, scanline start %d, end %d\n",
  			  pipe_name(pipe), crtc->debug.start_vbl_count,
  			  end_vbl_count,

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

Signature · 6 lines
Ville Syrjälä
Intel OTC
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#4 lore raw

Hi Ville:
Thanks for the reply! :P

Is this error message for the end-user or for debug? If this is for the end-user, only warning one time should be better? I saw a case on Celeron, if this happens, usually this would happen again and agin, the error message could fill all the dmesg. If this is only for debug, should DRM_DEBUG_KMS() would be better?

Anyway, thanks for the reply! :P

Thanks,
Zhi.

Quoted 48 lines · Ville Syrjälä
Quoted 1 line · Zhi Wang

Can we directly use DRM_DEBUG_KMS() for this periodic error message?

No. We want to actually know when/if this fails.

Quoted 42 lines · On 03/03/17 19:53, [email protected] wrote:
Quoted 28 lines

From: Bing Niu <[email protected]>

under virtualization enviroment, it is possible guest update pipe registers across vblank intervals due to overhead of mmio traps or vm schedule out. However, it is safe since those pipe update happen in virual registers and will not be committed to hardware. suppress that atomic commit error message under virtualization case to avoid confusing user.

---
   drivers/gpu/drm/i915/intel_sprite.c | 3 ++-
   1 file changed, 2 insertions(+), 1 deletion(-)
drivers/gpu/drm/i915/intel_sprite.c +1 -0
index b16a295..5ce1ec6 100644
@@ -158,6 +158,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
   	int scanline_end = intel_get_crtc_scanline(crtc);
   	u32 end_vbl_count = intel_crtc_get_vblank_counter(crtc);
   	ktime_t end_vbl_time = ktime_get();
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
   	if (work) {
   		work->flip_queued_vblank = end_vbl_count;

drivers/gpu/drm/i915/intel_sprite.c void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work · line 191

Quoted 8 lines
drivers/gpu/drm/i915/intel_sprite.c +1 -1
@@ -184,7 +185,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
   	local_irq_enable();
 
   	if (crtc->debug.start_vbl_count &&
-	    crtc->debug.start_vbl_count != end_vbl_count) {
+	    crtc->debug.start_vbl_count != end_vbl_count && !intel_vgpu_active(dev_priv)) {
   		DRM_ERROR("Atomic update failure on pipe %c (start=%u end=%u) time %lld us, min %d, max %d, scanline start %d, end %d\n",
   			  pipe_name(pipe), crtc->debug.start_vbl_count,
   			  end_vbl_count,

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

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

#5 lore raw

✗ Fi.CI.BAT: warning for drm/i915: suppress atomic commit error message under gvt-g env

In reply to [email protected]
== Series Details ==

== Series Details ==

Series: drm/i915: suppress atomic commit error message under gvt-g env
URL : https://patchwork.freedesktop.org/series/20600/
State : warning

== Summary ==

Series 20600v1 drm/i915: suppress atomic commit error message under gvt-g env
https://patchwork.freedesktop.org/api/1.0/series/20600/revisions/1/mbox/

Test drv_module_reload:

        Subgroup basic-reload:
                dmesg-warn -> PASS       (fi-ilk-650)
Test gem_exec_flush:
        Subgroup basic-batch-kernel-default-uc:
                fail       -> PASS       (fi-snb-2600) fdo#100007
Test kms_force_connector_basic:
        Subgroup prune-stale-modes:
                pass       -> SKIP       (fi-snb-2600)
Test kms_pipe_crc_basic:
        Subgroup hang-read-crc-pipe-a:
                pass       -> DMESG-WARN (fi-byt-n2820)

fdo#100007 https://bugs.freedesktop.org/show_bug.cgi?id=100007

fi-bdw-5557u     total:278  pass:267  dwarn:0   dfail:0   fail:0   skip:11 
fi-bsw-n3050     total:278  pass:239  dwarn:0   dfail:0   fail:0   skip:39 
fi-bxt-j4205     total:278  pass:259  dwarn:0   dfail:0   fail:0   skip:19 
fi-bxt-t5700     total:278  pass:258  dwarn:0   dfail:0   fail:0   skip:20 
fi-byt-j1900     total:278  pass:251  dwarn:0   dfail:0   fail:0   skip:27 
fi-byt-n2820     total:278  pass:246  dwarn:1   dfail:0   fail:0   skip:31 
fi-hsw-4770      total:278  pass:262  dwarn:0   dfail:0   fail:0   skip:16 
fi-hsw-4770r     total:278  pass:262  dwarn:0   dfail:0   fail:0   skip:16 
fi-ilk-650       total:278  pass:228  dwarn:0   dfail:0   fail:0   skip:50 
fi-ivb-3520m     total:278  pass:260  dwarn:0   dfail:0   fail:0   skip:18 
fi-ivb-3770      total:278  pass:260  dwarn:0   dfail:0   fail:0   skip:18 
fi-kbl-7500u     total:278  pass:259  dwarn:1   dfail:0   fail:0   skip:18 
fi-skl-6260u     total:278  pass:268  dwarn:0   dfail:0   fail:0   skip:10 
fi-skl-6700hq    total:278  pass:261  dwarn:0   dfail:0   fail:0   skip:17 
fi-skl-6700k     total:278  pass:256  dwarn:4   dfail:0   fail:0   skip:18 
fi-skl-6770hq    total:278  pass:268  dwarn:0   dfail:0   fail:0   skip:10 
fi-snb-2520m     total:278  pass:250  dwarn:0   dfail:0   fail:0   skip:28 
fi-snb-2600      total:278  pass:248  dwarn:0   dfail:0   fail:0   skip:30 

7f38bb9ea553c613223b705326c1b8d1a8fc4a90 drm-tip: 2017y-03m-02d-22h-35m-41s UTC integration manifest bae46b8 drm/i915: suppress atomic commit error message under gvt-g env

== Logs ==

For more details see: https://intel-gfx-ci.01.org/CI/Patchwork_4047/
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#6 lore raw
In reply to [email protected]
Quoted 6 lines · On Fri, Mar 03, 2017 at 06:53:08AM -0500, [email protected] wrote:

From: Bing Niu <[email protected]>

under virtualization enviroment, it is possible guest update pipe registers across vblank intervals due to overhead of mmio traps or vm schedule out. However, it is safe since those pipe update happen in virual registers and will not be committed to hardware.

Hmm. How are these virtual registers used? Do you present some kind of virtualized display for the guest? And if this stuff fails does that mean that guest will fail in doing atomic display updates?

Quoted 22 lines

suppress that
atomic commit error message under virtualization case to avoid confusing user.

---
 drivers/gpu/drm/i915/intel_sprite.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/gpu/drm/i915/intel_sprite.c +1 -0
index b16a295..5ce1ec6 100644
@@ -158,6 +158,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
 	int scanline_end = intel_get_crtc_scanline(crtc);
 	u32 end_vbl_count = intel_crtc_get_vblank_counter(crtc);
 	ktime_t end_vbl_time = ktime_get();
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	if (work) {
 		work->flip_queued_vblank = end_vbl_count;

drivers/gpu/drm/i915/intel_sprite.c void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work · line 186

Quoted 2 lines
drivers/gpu/drm/i915/intel_sprite.c +0 -0
@@ -184,7 +185,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
 	local_irq_enable();
 

if (vgpu_active())

	return;

here might look neater, and would still work with the additional error print that Maarten is going to add [1]

[1] https://patchwork.freedesktop.org/patch/141260/

Quoted 13 lines
drivers/gpu/drm/i915/intel_sprite.c +1 -1
 	if (crtc->debug.start_vbl_count &&
-	    crtc->debug.start_vbl_count != end_vbl_count) {
+	    crtc->debug.start_vbl_count != end_vbl_count && !intel_vgpu_active(dev_priv)) {
 		DRM_ERROR("Atomic update failure on pipe %c (start=%u end=%u) time %lld us, min %d, max %d, scanline start %d, end %d\n",
 			  pipe_name(pipe), crtc->debug.start_vbl_count,
 			  end_vbl_count,
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Signature · 6 lines
Ville Syrjälä
Intel OTC
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
author
#7 lore raw

Hi ville:
thanks for the reply. :)
yeah! we present a virtual monitor to guest associating with those vregs. however, we use remote display protocol like x11vnc to display vm's desktop.

cheer~

-----Original Message-----
From: Ville Syrjälä [mailto:[email protected]] Sent: Friday, March 03, 2017 7:53 PM
To: Niu, Bing <[email protected]>
Cc: [email protected]; Lv, Zhiyuan <[email protected]>
Subject: Re: [Intel-gfx] [PATCH] drm/i915: suppress atomic commit error message under gvt-g env

Quoted 6 lines · On Fri, Mar 03, 2017 at 06:53:08AM -0500, [email protected] wrote:

From: Bing Niu <[email protected]>

under virtualization enviroment, it is possible guest update pipe registers across vblank intervals due to overhead of mmio traps or vm schedule out. However, it is safe since those pipe update happen in virual registers and will not be committed to hardware.

Hmm. How are these virtual registers used? Do you present some kind of virtualized display for the guest? And if this stuff fails does that mean that guest will fail in doing atomic display updates?

Quoted 22 lines

suppress that
atomic commit error message under virtualization case to avoid confusing user.

---
 drivers/gpu/drm/i915/intel_sprite.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/gpu/drm/i915/intel_sprite.c +1 -0
index b16a295..5ce1ec6 100644
@@ -158,6 +158,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
 	int scanline_end = intel_get_crtc_scanline(crtc);
 	u32 end_vbl_count = intel_crtc_get_vblank_counter(crtc);
 	ktime_t end_vbl_time = ktime_get();
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	if (work) {
 		work->flip_queued_vblank = end_vbl_count; @@ -184,7 +185,7 @@ void
Quoted 3 lines

intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
local_irq_enable();

if (vgpu_active())

	return;

here might look neater, and would still work with the additional error print that Maarten is going to add [1]

[1] https://patchwork.freedesktop.org/patch/141260/

Quoted 10 lines
drivers/gpu/drm/i915/intel_sprite.c +2 -2
 	if (crtc->debug.start_vbl_count &&
-	    crtc->debug.start_vbl_count != end_vbl_count) {
+	    crtc->debug.start_vbl_count != end_vbl_count &&
+!intel_vgpu_active(dev_priv)) {
 		DRM_ERROR("Atomic update failure on pipe %c (start=%u end=%u) time %lld us, min %d, max %d, scanline start %d, end %d\n",
 			  pipe_name(pipe), crtc->debug.start_vbl_count,
 			  end_vbl_count,
--

2.7.4

Quoted 4 lines

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

--
Ville Syrjälä
Intel OTC
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx