Uploaded image for project: 'Minecraft: Java Edition'
  1. Minecraft: Java Edition
  2. MC-166846

Corruption, Crash, & Even Destabilization of Other Programs Caused by Resource Pack Reload for Single-Channel RAM Systems

    XMLWordPrintable

Details

    • Bug
    • Status: Resolved
    • Resolution: Awaiting Response
    • 1.15 Pre-release 3, 1.15 Pre-release 7, 1.15, 1.15.1, 1.15.1 Pre-release 1
    • None
    • None
    • Unconfirmed
    • Resource Packs

    Description

      Memory leak? Definitely.

      Crashing and artifacting stops after installing a second DIMM.

      Upon loading or reloading a resource pack (F3+T and/or from the menu in-game or on the main menu) in 1.15, the game will artifact significantly (sometimes with a big of scattered lag) and proceed to crash between 5-30 seconds later. No crash report is generated (sometimes hs_pid logs appear), and other programs may be destabilized by this as well.

      Minecraft is doing this independently of any software.

       

      How to reproduce:

      -Enter into a Minecraft world

      -EITHER Switch between resource packs, reload resource packs (either using F3+T or menu) until something goes wrong. It takes me around 5-10 reloads. 128x128 is the size I'm using.

      -OR Focus out of game using the Windows key until it refuses/doesn't 'hear' one time. The next time you attempt to reload packs, the game will crash possibly taking another program with it.

       

      *This does not occur in 1.14.4

      *RAM has been tested using MemTest. There are no issues there.

      *Changing Java preference or modifying memory allocation does not change outcome

      *Exiting and re-entering a world does not change outcome

      *Crashes can be averted (regarding other programs) if you are fast enough to tab out at the sign of corruption (if there is one) and End Task, trigger a manual crash, or ALT-F4

      *Seemingly, the larger the OS Page File, the longer it takes for Minecraft to destabilize everything. MC-167219 begins to come into play progressively.

      ========================

       CRASH REPORT: The crash report below was manually triggered after a failed reload (specifically, the artifact where all of the textures are white and tinted if biome-colored.

      HS_PID.log / REPLAY_PID.log Crashes: No typical crash reports generate. Game may run into a couple of different issues after a pack reload (missing textures, extreme artifacting, loading bar on-screen will remain until crash while player can still look around, white screen). Crashes imminently and harms other programs upon this crash.

      LAUNCHER_LOG.txt Crashes: This crash is occurs relative to Minecraft only (other programs are unaffected) and is immediate- however, despite the launcher giving me an option to see a crash report, none are generated. No hs_pid nor replay_pid logs for this one- a couple of launcher_logs had interesting bits:

      (1: repeats an insane amount of times)

      "[Info: 2019-12-14 04:50:27.6678835: GameCallbacks.cpp(142)] Game/cxg (Render thread) Info OpenGL debug message, id=0, source=API, type=ERROR, severity=HIGH, message=Unknown internal debug message. The NVIDIA OpenGL driver has encountered an out of memory error. This application might behave inconsistently and fail.
      (pid=10984 javaw.exe 64bit)"

      [Info: 2019-12-14 04:50:27.7448504: GameCallbacks.cpp(142)] Game/cxg (Render thread) Info OpenGL debug message, id=1281, source=API, type=ERROR, severity=HIGH, message=GL_INVALID_VALUE error generated. Invalid texture format.

       

      (2: repeats for lots of textures, alternates between 'corrupt jpeg' and 'out of memory')

      [Info: 2019-12-14 09:20:13.9488145: GameCallbacks.cpp(142)] Game/eaf (Server-Worker-17) Error Using missing texture, unable to load minecraft:textures/block/diorite.png
      java.io.IOException: Could not load image: Corrupt JPEG

      [Info: 2019-12-14 09:20:13.9472044: GameCallbacks.cpp(142)] Game/eaf (Server-Worker-16) Error Using missing texture, unable to load minecraft:textures/block/destroy_stage_7.png
      java.io.IOException: Could not load image: Out of memory

       

      Attachments

        1. 2019-12-09_18.32.54.png
          2019-12-09_18.32.54.png
          1.58 MB
        2. 2019-12-09_18.33.01.png
          2019-12-09_18.33.01.png
          1.69 MB
        3. 2019-12-09_18.33.39.png
          2019-12-09_18.33.39.png
          2.13 MB
        4. 2019-12-09_18.33.49.png
          2019-12-09_18.33.49.png
          1.08 MB
        5. 2019-12-19_20.58.42.png
          2019-12-19_20.58.42.png
          951 kB
        6. BeforeCrash.jpg
          BeforeCrash.jpg
          214 kB
        7. crash-2019-12-15_04.38.37-client.txt
          3 kB
        8. CrashGimp.jpg
          CrashGimp.jpg
          233 kB
        9. CrashMC.jpg
          CrashMC.jpg
          252 kB
        10. error_message.PNG
          error_message.PNG
          25 kB
        11. hs_err_pid12372.log
          40 kB
        12. hs_err_pid7076.log
          581 kB
        13. launcher_log.txt
          217 kB
        14. launcher_log.txt
          489 kB
        15. MemoryLeak.png
          MemoryLeak.png
          1.26 MB
        16. replay_pid12372.log
          604 kB
        17. replay_pid7076.log
          653 kB
        18. RM_1_ThisDoesntLookRight.png
          RM_1_ThisDoesntLookRight.png
          204 kB
        19. RM_2_ThisDoesntLookRight.png
          RM_2_ThisDoesntLookRight.png
          214 kB
        20. RM_3_WhenMCCrashed.png
          RM_3_WhenMCCrashed.png
          212 kB

        Issue Links

          Activity

            People

              Unassigned Unassigned
              Niknokinater Niknokinater
              Votes:
              1 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: