[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-dae...@freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #16 from kwahoo2 at wp.pl ---
Vadim, Grigori, thank you. Mainline Mesa works fine now.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL: 

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-dae...@freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

Grigori Goronzy  changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution|--- |FIXED

--- Comment #15 from Grigori Goronzy  ---
Pushed to master, should also appear in Mesa 10.0.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL: 

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-dae...@freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #14 from Fabio Pedretti  ---
Since this is a regression can it be allied before 10.0 branching:
http://lists.freedesktop.org/archives/mesa-dev/2013-October/047235.html

-- 
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL: 

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-dae...@freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

Krzysztof A. Sobiecki  changed:

   What|Removed |Added

  Attachment #84557|text/plain  |image/jpeg
  mime type||

-- 
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL: 

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-dae...@freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #13 from kwahoo2 at wp.pl ---
Grigori,
I just tested the improved patch on the vanilla Mesa (10.0.0-devel git-c325aa5)

It works fine, no artifacts in SS3.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL: 

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

Grigori Goronzy g...@chown.ath.cx changed:

   What|Removed |Added

  Attachment #85013|0   |1
is obsolete||

--- Comment #12 from Grigori Goronzy g...@chown.ath.cx ---
Created attachment 87456
  -- https://bugs.freedesktop.org/attachment.cgi?id=87456action=edit
New proposed fix

After I had a similar corrupted rendering with the WebGL water demo [1] on
TURKS, together with Vadim I tested the second patch and found that it didn't
actually work, and applied on top of the first patch, the workaround still was
always triggered.

So here is Vadim's corrected version of the patch, with an improvement to the
condition that triggers the workaround. It should hopefully only result in
minor performance degradation.

If there is no feedback, I'll just send this to mesa-dev in the coming days.

[1] http://madebyevan.com/webgl-water/

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

Vadim Girlin pt...@yandex.ru changed:

   What|Removed |Added

  Attachment #84573|0   |1
is obsolete||

--- Comment #9 from Vadim Girlin pt...@yandex.ru ---
Created attachment 85011
  -- https://bugs.freedesktop.org/attachment.cgi?id=85011action=edit
[PATCH] r600g/sb: work around hw issues with stack on eg/cm

Here is a new patch that hopefully should not affect performance, please test.

I'm not sure about correctness of workaround condition though, the description
is not very clear, so if it doesn't work as is then please find the following
line in src/gallium/drivers/r600/sb/sb_bc_finalize.cpp (should be line 713):
unsigned depth = elems; // or maybe ifs?
and replace elems with ifs:
unsigned depth = ifs;

If this doesn't help either, try 
unsigned depth = elems + 1;

and + 2, + 3. And then the same using ifs instead of elems again. One of
these variants should work.

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

Vadim Girlin pt...@yandex.ru changed:

   What|Removed |Added

  Attachment #85011|0   |1
is obsolete||

--- Comment #10 from Vadim Girlin pt...@yandex.ru ---
Created attachment 85013
  -- https://bugs.freedesktop.org/attachment.cgi?id=85013action=edit
r600g/sb: work around hw issues with stack on eg/cm

Sorry, previous patch was written on top of the first patch from this bug and
requires that patch to be applied first, this one is independent and can be
applied directly on master branch.

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #11 from kwah...@wp.pl ---
I patched after first patch (not master Mesa) and there i no artifacts neither
performance improvement. I didn't change anything in code.

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #6 from kwah...@wp.pl ---
Sorry for a delay, but I had some problems with multilib compiling. Finally
compiled Mesa today and the patch works. No more artifacts!

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #7 from Vadim Girlin pt...@yandex.ru ---
(In reply to comment #6)
 Sorry for a delay, but I had some problems with multilib compiling. Finally
 compiled Mesa today and the patch works. No more artifacts!

Is there any difference in performance with this patch? This bug is actually a
hardware issue with some chips and it needs a workaround, but possibly the
workaround will make performance a bit lower. That patch applies the workaround
always (I just wanted to make sure first that it helps with your problem), but
we'll need to restrict it to affected chips and some specific conditions to
minimize performance losses. I'll attach an improved patch for testing soon.

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #8 from kwah...@wp.pl ---
Unforunately there is no 100% reliable benchmarking mode in Serious Sam, but I
measured first 100 seconds of Summer in Cairo map intro:

Without patch:
20:59:33 INF:  - benchmark results -
20:59:33 INF:  
20:59:33 INF:   Duration: 100.0 seconds (2215 frames)
20:59:33 INF:Average: 22.2 FPS (27.3 w/o peaks)
20:59:33 INF: Spikes: 169.1 max, 1.2 min
20:59:33 INF:   Sections: AI=1%, physics=1%, sound=1%, scene=89%, shadows=6%,
misc=3%
20:59:33 INF:   Lows: 116 in 23.1 seconds (5.0 FPS)
20:59:33 INF:20 FPS: 36%
20:59:33 INF:  20-30 FPS: 10%
20:59:33 INF:  30-60 FPS:  8%
20:59:33 INF:60 FPS: 45%

Patched:
22:05:17 INF:  - benchmark results -
22:05:17 INF:  
22:05:17 INF:   Duration: 100.4 seconds (1941 frames)
22:05:17 INF:Average: 19.3 FPS (26.4 w/o peaks)
22:05:17 INF: Spikes: 148.4 max, 0.4 min
22:05:17 INF:   Sections: AI=1%, physics=0%, sound=1%, scene=89%, shadows=6%,
misc=2%
22:05:17 INF:   Lows: 46 in 28.5 seconds (1.6 FPS)
22:05:17 INF:20 FPS: 34%
22:05:17 INF:  20-30 FPS:  9%
22:05:17 INF:  30-60 FPS:  5%
22:05:17 INF:60 FPS: 52%

I noticed also small regression in Half-Life 2: Lost Coast benchmark - 83 vs 78
fps at max details.

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-dae...@freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #5 from Vadim Girlin  ---
Created attachment 84573
  --> https://bugs.freedesktop.org/attachment.cgi?id=84573=edit
patch

Does this patch help?

-- 
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL: 

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-dae...@freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #4 from kwahoo2 at wp.pl ---
Created attachment 84566
  --> https://bugs.freedesktop.org/attachment.cgi?id=84566=edit
shader no 76 enabled

Seems that shader number 76 is the culprit. 
R600_DEBUG=sb R600_SB_DSKIP_START=76 R600_SB_DSKIP_END=76 R600_SB_DSKIP_MODE=2
produces glitches.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL: 

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-dae...@freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #3 from kwahoo2 at wp.pl ---
Created attachment 84563
  --> https://bugs.freedesktop.org/attachment.cgi?id=84563=edit
output with R600_DEBUG=sb,ps,vs

-- 
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL: 

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-dae...@freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #2 from Vadim Girlin  ---
Please attach the output with R600_DEBUG=sb,ps,vs.

Could you also record the GL trace that reproduces the issue using apitrace -
http://apitrace.github.io/

You can just upload the trace if it's not too big, then it might help me to
reproduce the bug if it's not hw-specific, but also you can use recorded trace
to find the exact failing shader as described in "Regression debugging" section
here - http://people.freedesktop.org/~vadimg/r600-sb.html - you'll need to
replay it with R600_DEBUG=sb,sbstat first to figure out initial range of shader
indices used by the app, and then apply R600_SB_DSKIP_* env vars when replaying
the trace to bisect the range as described in that section. After locating
broken shader just add "ps,vs" to R600_DEBUG and attach the output. You'll have
a cmd line like this in the end (with some other shader index instead of 37):
R600_DEBUG=sb,ps,vs R600_SB_DSKIP_MODE=2 R600_SB_DSKIP_START=37
R600_SB_DSKIP_END=37  &> log_file

I can try to do everything myself if you can send me the key for SS3 (to
vadimgirlin at gmail dot com), this can save some time, but the bug may be
specific to your GPU or system configuration and in such case possibly I won't
be able to reproduce it on my hardware, then I'll still need your help with
debugging as described above.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL: 

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-dae...@freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #1 from kwahoo2 at wp.pl ---
Pics (if you cannot open the atachment): 
http://postimg.org/image/rpg5b0oz3/
http://postimg.org/image/h5acgr97n/

-- 
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL: 

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #1 from kwah...@wp.pl ---
Pics (if you cannot open the atachment): 
http://postimg.org/image/rpg5b0oz3/
http://postimg.org/image/h5acgr97n/

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #2 from Vadim Girlin pt...@yandex.ru ---
Please attach the output with R600_DEBUG=sb,ps,vs.

Could you also record the GL trace that reproduces the issue using apitrace -
http://apitrace.github.io/

You can just upload the trace if it's not too big, then it might help me to
reproduce the bug if it's not hw-specific, but also you can use recorded trace
to find the exact failing shader as described in Regression debugging section
here - http://people.freedesktop.org/~vadimg/r600-sb.html - you'll need to
replay it with R600_DEBUG=sb,sbstat first to figure out initial range of shader
indices used by the app, and then apply R600_SB_DSKIP_* env vars when replaying
the trace to bisect the range as described in that section. After locating
broken shader just add ps,vs to R600_DEBUG and attach the output. You'll have
a cmd line like this in the end (with some other shader index instead of 37):
R600_DEBUG=sb,ps,vs R600_SB_DSKIP_MODE=2 R600_SB_DSKIP_START=37
R600_SB_DSKIP_END=37 cmd to replay the trace  log_file

I can try to do everything myself if you can send me the key for SS3 (to
vadimgirlin at gmail dot com), this can save some time, but the bug may be
specific to your GPU or system configuration and in such case possibly I won't
be able to reproduce it on my hardware, then I'll still need your help with
debugging as described above.

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #3 from kwah...@wp.pl ---
Created attachment 84563
  -- https://bugs.freedesktop.org/attachment.cgi?id=84563action=edit
output with R600_DEBUG=sb,ps,vs

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #4 from kwah...@wp.pl ---
Created attachment 84566
  -- https://bugs.freedesktop.org/attachment.cgi?id=84566action=edit
shader no 76 enabled

Seems that shader number 76 is the culprit. 
R600_DEBUG=sb R600_SB_DSKIP_START=76 R600_SB_DSKIP_END=76 R600_SB_DSKIP_MODE=2
produces glitches.

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 68503] Graphical glitches in Serious Sam 3 when SB is enabled

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=68503

--- Comment #5 from Vadim Girlin pt...@yandex.ru ---
Created attachment 84573
  -- https://bugs.freedesktop.org/attachment.cgi?id=84573action=edit
patch

Does this patch help?

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel