Bug 459 - Shader compilation fails with caps.setStencilBits(8) set
Summary: Shader compilation fails with caps.setStencilBits(8) set
Status: VERIFIED FIXED
Alias: None
Product: Jogl
Classification: JogAmp
Component: opengl (show other bugs)
Version: 2
Hardware: pc_x86_64 windows
: --- normal
Assignee: Sven Gothel
URL:
Depends on:
Blocks:
 
Reported: 2011-01-16 21:20 CET by Wade Walker
Modified: 2011-02-05 04:04 CET (History)
0 users

See Also:
Type: ---
SCM Refs:
8adc04788a6d9dd44de5a4636b46d14dbb70b799 e2040ada5a94b466ba95e84a497290c069d377bb
Workaround: ---


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Wade Walker 2011-01-16 21:20:50 CET
On Windows, shader compilation fails if we set caps.setStencilBits(8). Setting caps.setSampleBuffers(true) fixes the problem, as does removing the stencil bit setting.

Created unit test, confirmed failure on Windows and pass on Linux. Submitting pull request for unit test soon, after preliminary debugging.
Comment 1 Wade Walker 2011-01-16 22:49:00 CET
Submitted pull request for failing JUnit test case at  https://github.com/sgothel/jogl/pull/19. The original report of this bug was on the forum at http://jogamp.762907.n3.nabble.com/JOGL2-GLCapabilities-Windows-td2153961.html.

The problem seems to be in the capabilities choosing code, so I'll wait for your next checkin before digging in too deep.
Comment 2 Sven Gothel 2011-02-02 21:30:25 CET
merged test, fixed by GLCapabilities change 8adc04788a6d9dd44de5a4636b46d14dbb70b799
test passed in 280
Comment 3 Sven Gothel 2011-02-05 02:58:39 CET
still exists on NV/Win7 .. fixing now
Comment 4 Sven Gothel 2011-02-05 04:04:31 CET
commit e2040ada5a94b466ba95e84a497290c069d377bb
Author: Sven Gothel <sgothel@jausoft.com>
Date:   Sat Feb 5 04:03:51 2011 +0100

    Fix bug #459 NV/Win7 PDF/caps selection
    
    Scenario
      - NV / Win7 driver version 266.58's
      - Caps: on-scr, rgba 8/8/8/0, accum-rgba 0/0/0/0, dp/st/ms: 16/8/0, dbl, mono
    
    The above 'wglChoosePixelFormatARB' impl returns an array of pixelformats,
    where the 1st entry is not hardware accelerated!
    This should be considered a bug in NV's driver, since the array should return
    a list ordered from 'best' to 'worst'.
    
    Workaround trying explicit hw acceleration 1st, then generic, then software.