Bug 30143

Summary:	[KMS] mesa demo lodbias broken
Product:	Mesa	Reporter:	Andrew Randrianasulu <randrik>
Component:	Drivers/DRI/r200	Assignee:	Default DRI bug account <dri-devel>
Status:	RESOLVED FIXED	QA Contact:
Severity:	normal
Priority:	medium
Version:	git
Hardware:	x86 (IA32)
OS:	Linux (All)
Whiteboard:
i915 platform:		i915 features:
Attachments:	Screenshot of corruption X log Piglit data (bz2 compressed) RADEON_DEBUG=tex ./lodbias

Description Andrew Randrianasulu 2010-09-11 22:30:31 UTC

Running software:

kernel 2.6.36-rc3

mesa - commit d8a36620089e72d431ae853ec168f193f3376782
Author: Ian Romanick <ian.d.romanick@intel.com>
Date:   Fri Sep 10 13:10:26 2010 -0700

    mesa: Remove unused Emit flags from gl_shader_compiler_options

libdrm - commit b04515c5d6c95f573457a94267b855cceb639105
Author: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Date:   Fri Aug 6 13:55:11 2010 +1000

    libdrm: Fix PCI domain domain support

xf86-video-ati - commit bdd41fecdb19c83c6c7b793016b61d38065dcd13
Author: Alex Deucher <alexdeucher@gmail.com>
Date:   Fri Aug 20 01:17:13 2010 -0400

    evergreen: set encoder type to DVI for HDMI

Xserver: 1.9.0

self-build with  gcc (GCC) 4.4.4


Hw:

01:00.0 VGA compatible controller: ATI Technologies Inc RV280 [Radeon 9200 SE] (rev 01) (prog-if 00 [VGA controller])
        Subsystem: Hightech Information System Ltd. Excalibur 9200SE VIVO 128M
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
        Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 64 (2000ns min), Cache Line Size: 32 bytes
        Interrupt: pin A routed to IRQ 16
        Region 0: Memory at d0000000 (32-bit, prefetchable) [size=128M]
        Region 1: I/O ports at c800 [size=256]
        Region 2: Memory at dfef0000 (32-bit, non-prefetchable) [size=64K]
        Expansion ROM at dfec0000 [disabled] [size=128K]
        Capabilities: [58] AGP version 2.0
                Status: RQ=80 Iso- ArqSz=0 Cal=0 SBA+ ITACoh- GART64- HTrans- 64bit- FW+ AGP3- Rate=x1,x2,x4
                Command: RQ=32 ArqSz=0 Cal=0 SBA+ AGP+ GART64- 64bit- FW- Rate=x4


I'll add logs and screenshot. For me it looks like only very negative bias resulted in image displayed, everything else just looks like  heavily corrupted mipmaps of texture. Probably bug in _mesa_meta_GenerateMipmap(), may be due to driver setting something wrong

Comment 1 Andrew Randrianasulu 2010-09-11 22:32:46 UTC

Created attachment 38636 [details]
Screenshot of corruption

Comment 2 Andrew Randrianasulu 2010-09-11 22:34:06 UTC

Created attachment 38637 [details]
X log

Comment 3 Andrew Randrianasulu 2010-09-11 22:37:34 UTC

Created attachment 38638 [details]
Piglit data (bz2 compressed)

Comment 4 Andrew Randrianasulu 2010-09-11 22:41:48 UTC

From quick look at piglit's summary:

bugs  	13/15

--------
fdo20701  	fail

Returncode: 1

Errors:
Mesa: Mesa 7.9-devel DEBUG build Sep 11 2010 06:16:40


Output:
init:104: framebuffer status = 0x8cdd
------

then

fdo25614-genmipmap  	fail

-------
Returncode: 1

Errors:
Mesa: Mesa 7.9-devel DEBUG build Sep 11 2010 06:16:40


Output:
Probe at (0,0)
  Expected: 0.000000 1.000000 0.000000 1.000000
  Observed: 0.501961 0.501961 0.501961 0.501961
Probe at (0,0)
  Expected: 0.000000 1.000000 0.000000 1.000000
  Observed: 0.000000 0.000000 0.000000 0.000000
Probe at (0,0)
  Expected: 0.000000 1.000000 0.000000 1.000000
  Observed: 0.000000 0.000000 0.000000 0.000000
Probe at (0,0)
  Expected: 0.000000 1.000000 0.000000 1.000000
  Observed: 0.000000 0.000000 0.000000 0.000000
Probe at (0,0)
  Expected: 0.000000 1.000000 0.000000 1.000000
  Observed: 0.000000 0.000000 0.000000 0.000000

Comment 5 Andrew Randrianasulu 2010-09-13 11:42:02 UTC

Created attachment 38677 [details]
RADEON_DEBUG=tex ./lodbias

Comment 6 Andrew Randrianasulu 2010-09-24 06:09:05 UTC

Still broken, as of mesa version up to:


commit 4e27e935ca670d0cb1748e40a46a2357d4331a6f
Author: Dave Airlie <airlied@redhat.com>
Date:   Fri Sep 24 12:36:08 2010 +1000
    r600g: drop index_offset parameter to index buffer translate.
    r600 doesn't need this as we always have working index bias

using kernel 2.6.36-rc5.

Comment 7 smoki 2014-02-25 17:10:41 UTC

 This works OK today :). Just checked with my radeon 9250 and git mesa 10.2.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.