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 3 of 3

Thread: Uninitiated variables in i965_dri.so

  1. #1
    Intern Newbie
    Join Date
    Nov 2013
    Posts
    46

    Uninitiated variables in i965_dri.so

    I use the Valgrind memory leak checker (valgrind.org).
    Code :
    gcc hello_world.c -g -O0 -lglut -lGL
    valgrind --leak-check=full --track-origins=yes ./a.out
    produces an overflowing page of error messages, of which the following is an example
    ==2622== Invalid write of size 8
    ==2622== at 0x7207CD6: ??? (in /usr/lib/x86_64-linux-gnu/dri/i965_dri.so)
    ==2622== by 0x72206A0: ??? (in /usr/lib/x86_64-linux-gnu/dri/i965_dri.so)
    ==2622== by 0x72298B0: ??? (in /usr/lib/x86_64-linux-gnu/dri/i965_dri.so)
    ==2622== by 0x6E7EA28: ??? (in /usr/lib/x86_64-linux-gnu/dri/i965_dri.so)
    ==2622== by 0x4A31B41: glutBitmapCharacter (in /usr/lib/x86_64-linux-gnu/libglut.so.3.9.0)
    ==2622== by 0x15120E: toolbar (in /home/william/projects/a.out)
    ==2622== by 0x15238A: display (in /home/william/projects/a.out)
    ==2622== by 0x154333: reshape (in /home/william/projects/a.out)
    ==2622== by 0x4A39F44: glutMainLoopEvent (in /usr/lib/x86_64-linux-gnu/libglut.so.3.9.0)
    ==2622== by 0x4A3A6A4: glutMainLoop (in /usr/lib/x86_64-linux-gnu/libglut.so.3.9.0)
    ==2622== by 0x154623: main (in /home/william/projects/a.out)
    ==2622== Address 0x7f4731d1705c is not stack'd, malloc'd or (recently) free'd
    ==2622==
    ==2622== Invalid write of size 4
    ==2622== at 0x7207CDE: ??? (in /usr/lib/x86_64-linux-gnu/dri/i965_dri.so)
    ==2622== by 0x72206A0: ??? (in /usr/lib/x86_64-linux-gnu/dri/i965_dri.so)
    ==2622== by 0x72298B0: ??? (in /usr/lib/x86_64-linux-gnu/dri/i965_dri.so)
    ==2622== by 0x6E7EA28: ??? (in /usr/lib/x86_64-linux-gnu/dri/i965_dri.so)
    ==2622== by 0x4A31B41: glutBitmapCharacter (in /usr/lib/x86_64-linux-gnu/libglut.so.3.9.0)
    ==2622== by 0x15120E: toolbar (in /home/william/projects/a.out)
    ==2622== by 0x15238A: display (in /home/william/projects/a.out)
    ==2622== by 0x154333: reshape (in /home/william/projects/a.out)
    ==2622== by 0x4A39F44: glutMainLoopEvent (in /usr/lib/x86_64-linux-gnu/libglut.so.3.9.0)
    ==2622== by 0x4A3A6A4: glutMainLoop (in /usr/lib/x86_64-linux-gnu/libglut.so.3.9.0)
    ==2622== by 0x154623: main (in /home/william/projects/a.out)
    ==2622== Address 0x7f4731d17064 is not stack'd, malloc'd or (recently) free'd
    I am running 4.18.0-2-amd64 #1 SMP Debian 4.18.10-2 (2018-11-02) x86_64 GNU/Linux
    the time stamp on /usr/include/GL/gl.h and /usr/include/GL/glext.h is 10/10/18

    I installed the binary version of OpenGL via apt. If I knew how to compile it from source, I would attempt to debug the .so .


    What is the

  2. #2
    Senior Member OpenGL Guru Dark Photon's Avatar
    Join Date
    Oct 2004
    Location
    Druidia
    Posts
    4,590
    Does your program run properly, or does it crash in glutBitmapCharacter()?

    If the former, it's not unusual for graphics drivers to do things which trigger warnings deep down in the graphics driver in valgrind. If you're sure an error isn't due to an error on your part or a bug in the graphics driver, you can add a suppression to mask it in the future.

    This seems to suggest that if you want Mesa3D graphics drivers to run without valgrind errors, you nmay need to build in valgrind support.
    Last edited by Dark Photon; 01-10-2019 at 07:43 PM.

  3. #3
    Intern Newbie
    Join Date
    Nov 2013
    Posts
    46
    The Hello World trivial program does run without noticeable error. I encountered the problem on my laptop, which is an Intel machine. I ran the Hello World program under Valgrind on my desktop machine, which is an AMD machine, without any error messages. The problem is the Intel driver.
    The only issue that I have is that the error messages scroll off the page, which blankets errors that I am really concerned with.
    My reason for posting here is that I know of no better avenue for notifying concerned parties.

    Quote Originally Posted by Dark Photon View Post
    Does your program run properly, or does it crash in glutBitmapCharacter()?

    If the former, it's not unusual for graphics drivers to do things which trigger warnings deep down in the graphics driver in valgrind. If you're sure an error isn't due to an error on your part or a bug in the graphics driver, you can add a suppression to mask it in the future.

    This seems to suggest that if you want Mesa3D graphics drivers to run without valgrind errors, you nmay need to build in valgrind support.

Tags for this Thread

Posting Permissions

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