https://bugs.freedesktop.org/show_bug.cgi?id=43191
Alex Deucher changed:
What|Removed |Added
Severity|normal |enhancement
Summary|AMD E-350
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #2 from Alex Deucher 2011-11-23 06:30:11 PST ---
All radeons can benefit from 2D tiling.
--
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are the ass
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #3 from Simon Farnsworth 2011-11-23
11:07:44 PST ---
Created attachment 53817
--> https://bugs.freedesktop.org/attachment.cgi?id=53817
A program to draw red/green rectangles using raw X11. Compile as gcc -o
rectangle rectangle.c -l
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #4 from Alex Deucher 2011-11-23 11:47:12 PST ---
You'll need to play with the tiling attributes (TILE_SPLIT, MACRO_TILE_ASPECT,
BANK_WIDTH, BANK_HEIGHT, NUM_BANKS) in in CB_COLORn_ATTRIB and
SQ_TEX_RESOURCE_WORD6_0, SQ_TEX_RESOURCE_WO
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #5 from Alex Deucher 2011-11-23 15:33:38 PST ---
Created attachment 53820
--> https://bugs.freedesktop.org/attachment.cgi?id=53820
possible fix
The two attached drm patches may fix the issue, or at least lay the groundwork
for a pr
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #6 from Alex Deucher 2011-11-23 15:34:05 PST ---
Created attachment 53821
--> https://bugs.freedesktop.org/attachment.cgi?id=53821
possible fix
second patch.
--
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=em
https://bugs.freedesktop.org/show_bug.cgi?id=43191
Alex Deucher changed:
What|Removed |Added
Attachment #53821|0 |1
is obsolete|
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #8 from Simon Farnsworth 2011-11-24
07:32:12 PST ---
With the two attached patches from Alex and the DDX patch, plain X11 rendering
works fine. I have some rendering issues in OpenGL to track down, though (a
rectangle at the bottom r
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #9 from Simon Farnsworth 2011-11-24
08:37:24 PST ---
More checking shows that I only see the wraparound flicker at 1920x1200, not at
1920x1080.
--
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email
--- You
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #10 from Simon Farnsworth
2011-11-24 09:45:44 PST ---
Serial console tells me that my random crashes at 1920x1200 are CP stalls:
[ 101.638081] radeon :00:01.0: GPU lockup CP stall for more than 1msec
[ 101.659447] GPU lock
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #11 from Simon Farnsworth
2011-11-24 09:54:54 PST ---
Aha - they're followed by CPU hard lockups, which is likely to be the actual
crash.
[ 123.549279] [ cut here ]
[ 123.549279] WARNING: at kernel/watchdog
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #12 from Simon Farnsworth
2011-11-28 06:00:43 PST ---
Applying some debugger knowledge to the stack traces tells me that I'm stalled
in evergreen_page_flip, in the infinite loop commented as "Wait for
update_pending to go high".
It
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #13 from Simon Farnsworth
2011-11-28 06:44:35 PST ---
Looks like the pageflip hang is a red herring; I changed the existing DRM_DEBUG
in evergreen_page_flip to DRM_ERROR, and added a new one just before the loop
to tell me what crtc_
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #14 from Simon Farnsworth
2011-11-28 09:02:14 PST ---
I've spent the day digging to understand this, so I might be completely and
utterly wrong, but, if I understand correctly, I've got the following:
drivers/gpu/drm/radeon/evergree
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #15 from Alex Deucher 2011-11-28 09:07:42 PST ---
(In reply to comment #14)
> I've spent the day digging to understand this, so I might be completely and
> utterly wrong, but, if I understand correctly, I've got the following:
>
> dr
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #16 from Simon Farnsworth
2011-11-28 09:26:32 PST ---
>From advice on IRC, I'm uncommenting the CS dumping code in evergreen_cs.c, and
I'm going to examine the CS dumps just before the failure.
--
Configure bugmail: https://bugs.fr
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #17 from Alex Deucher 2011-11-28 09:35:53 PST ---
Created attachment 53906
--> https://bugs.freedesktop.org/attachment.cgi?id=53906
avoid infinite loops in pageflip code
This patch should avoid the infinite loops in the pageflip co
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #18 from Simon Farnsworth
2011-11-28 10:44:16 PST ---
Created attachment 53912
--> https://bugs.freedesktop.org/attachment.cgi?id=53912
CS dumping (from the kernel) for the lockup case
I'm about to go home for the day, and will co
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #19 from Jerome Glisse 2011-11-28 16:18:20
PST ---
What you need to print is what is the last fence the hw signaled each time you
emit a fence. Once you know that you know which cs is guilty, then it could be
any instruction inside t
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #20 from Simon Farnsworth
2011-11-29 06:27:37 PST ---
I cannot reproduce the lockup now that I've added the "avoid infinite loops in
pageflip code" patch; I'm guessing that the udelay(1) in there keeps the chip
happy.
I have, howeve
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #21 from Simon Farnsworth
2011-12-01 09:50:56 PST ---
Created attachment 54028
--> https://bugs.freedesktop.org/attachment.cgi?id=54028
Program to demonstrate misrendering issue
Thanks to Jerome's hint, I took some time to try and
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #22 from Simon Farnsworth
2011-12-01 10:01:28 PST ---
A bit of checking by drawing rectangles in different places shows that it's the
top-left of the backbuffer that overwrites the front buffer.
--
Configure bugmail: https://bugs.f
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #23 from Simon Farnsworth
2011-12-05 10:32:59 PST ---
I had the kernel print the size and gpu_offset of each BO assigned to
CB_COLOR0_BASE:
$ git diff
diff --git a/drivers/gpu/drm/radeon/evergreen.c
b/drivers/gpu/drm/radeon/evergree
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #24 from Michel Dänzer 2011-12-06 08:17:51 PST
---
(In reply to comment #23)
> I now have no idea what's going on - visually, it looks like the rendering
> engine writes to parts of the color buffer *before* the gpu_offset set in
> C
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #25 from Simon Farnsworth
2011-12-07 06:49:01 PST ---
Alex Deucher has given me data on Evergreen tiling, which has allowed me to get
to the bottom of this:
Evergreen doesn't have one fixed 2D tiling pattern; it has configurable
per
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #26 from Simon Farnsworth
2011-12-13 07:18:25 PST ---
Created attachment 54396
--> https://bugs.freedesktop.org/attachment.cgi?id=54396
Kernel patch setting tile shape
This is the patch I sent to dri-devel, setting the kernel's ti
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #27 from Simon Farnsworth
2011-12-13 07:22:40 PST ---
Created attachment 54397
--> https://bugs.freedesktop.org/attachment.cgi?id=54397
DDX patch sorting out alignment
The DDX patch I sent to xorg-driver-ati, making alignment matc
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #28 from Simon Farnsworth
2011-12-13 07:52:17 PST ---
Created attachment 54398
--> https://bugs.freedesktop.org/attachment.cgi?id=54398
Mesa patch sorting out alignment
And the Mesa patch, for future reference.
--
Configure bugm
https://bugs.freedesktop.org/show_bug.cgi?id=43191
Michal Suchanek changed:
What|Removed |Added
CC||hramr...@gmail.com
--
Configure bugma
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #29 from Florian Mickler 2012-01-12 14:14:55
PST ---
A patch referencing this bug report has been merged in Linux v3.2-rc5:
commit f64964796dedca340608fb1075ab6baad5625851
Author: Alex Deucher
Date: Mon Nov 28 14:49:26 2011 -0500
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #30 from Michal Suchanek 2012-01-18 05:06:42
PST ---
That's not really much related, that fixes an issue that you might hit when the
tiling is wrong and the gpu locks up.
The current patches at http://people.freedesktop.org/~glisse/
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #31 from Florian Mickler 2012-01-21 08:47:57
PST ---
A patch referencing this bug report has been merged in Linux v3.2-rc5:
commit 392e37229f0d6358dcc7b43641df776e9f62a6e6
Author: Alex Deucher
Date: Mon Nov 28 14:49:27 2011 -0500
https://bugs.freedesktop.org/show_bug.cgi?id=43191
--- Comment #32 from Florian Mickler 2012-01-21 08:50:33
PST ---
A patch referencing this bug report has been merged in Linux v3.2-rc5:
commit f3a71df05082c84d1408129084736c5f742a6165
Author: Alex Deucher
Date: Mon Nov 28 14:49:28 2011 -0500
https://bugs.freedesktop.org/show_bug.cgi?id=43191
Alex Deucher changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|
34 matches
Mail list logo
