Part of the Khronos Group
OpenGL.org

The Industry's Foundation for High Performance Graphics

from games to virtual reality, mobile phones to supercomputers

Results 1 to 5 of 5

Thread: How to post in "Suggestions" forum?

  1. #1
    Junior Member Newbie
    Join Date
    Jul 2013
    Posts
    6

    How to post in "Suggestions" forum?

    Hello all.
    I'm new to this forum and I really sorry for creating topic in this section, but I did not find any rules of FAQ about how to create topics in "Suggestions for the next release of OpenGL" section.
    Can somebody explain what I need to do to gain access to it?
    Thanks!

  2. #2
    Senior Member OpenGL Pro Aleksandar's Avatar
    Join Date
    Jul 2009
    Posts
    1,161
    You are right, as a someone with no previous posts you don't have a privilege to post a topic at the "Suggestions for the next release".
    It is very unlikely that someone's first post to this forum will be a valid suggestion for the next release.
    The history has proven the previous assumption, and that's why it is forbidden.

    After a few valid posts, you'll be eligible to post also to that section. In the meantime, your suggestions for the next release could be an interesting post to the Advanced section, for example.

  3. #3
    Junior Member Newbie
    Join Date
    Jul 2013
    Posts
    6
    Ok. Thanks.

    I don't think what "Advanced" forum is right place for it.
    I'm care about unified API for GPU performance counters, I think it will help to build cross-vendor profiling/tuning tools, etc.
    Right now we have AMD_performance_monitor and INTEL_performance_query and seems to NVIDIA export counters via NVPerfKit, but it works only on Windows.
    Last edited by oh-la-la; 03-09-2014 at 06:49 AM.

  4. #4
    Senior Member OpenGL Pro Aleksandar's Avatar
    Join Date
    Jul 2009
    Posts
    1,161
    That's something I asked for several years ago, but there is no intent for the perf API to be unified in the foreseeable future.
    The only solution now is to make separate reading functions for each vendor. Another problem is in fact that hardware counters are not the same. The unification should take part also in hardware and drivers which is not likely to be done.

  5. #5
    Junior Member Newbie
    Join Date
    Jul 2013
    Posts
    6
    Sad to hear that.
    Hardware is different, right, but we can unify interface like done in ARB_texture_compression and ARB_get_program_binary.
    Some basic functions like glBeginPerformanceQuery, glEndPerformanceQuery, etc (or maybe use existing glBegin/EndQuery with new enums + some new functions)
    And query actual subset of supported enums via something like glGetIntegerv/glGetQueryiv(GL_NUM_PERFORMANCE_QUERIES, ...) in runtime.
    Serialization of integer arrays is easier when control of different subset of functions.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •