Summary: | X server crashes in VMware guest when KDE started with kwin "OpenGLIsUnsafe=false" | ||
---|---|---|---|
Product: | Mesa | Reporter: | Hugh Daschbach <hdasch> |
Component: | Drivers/X11 | Assignee: | mesa-dev |
Status: | RESOLVED INVALID | QA Contact: | |
Severity: | normal | ||
Priority: | medium | ||
Version: | 7.9 | ||
Hardware: | x86-64 (AMD64) | ||
OS: | Linux (All) | ||
Whiteboard: | |||
i915 platform: | i915 features: |
Description
Hugh Daschbach
2010-12-20 08:15:27 UTC
Before finding that it's common to simply disable OpenGL in .kde4/share/config/kwinrc: OpenGLIsUnsafe=true I poked around the crash a bit. I believe the crash happens when swrastSetTexBuffer2() passes a NULL texImage->Data value down to sPriv->swrast_loader->getImage(). Here's the backtrace from gdb just before the crash: Breakpoint 1, 0xb747da06 in VMWAREGetImage () from /usr/lib/xorg/modules/drivers/vmwlegacy_drv.so (gdb) bt #0 0xb747da06 in VMWAREGetImage () from /usr/lib/xorg/modules/drivers/vmwlegacy_drv.so #1 0x081a786d in miSpriteGetImage () #2 0xb74cc14b in swrastGetImage () from /usr/lib/xorg/modules/extensions/libglx.so #3 0xaef6f81c in swrastSetTexBuffer2 (pDRICtx=0xa31e888, target=3553, texture_format=8410, dPriv=0xa0a9068) at swrast.c:97 #4 0xaef6f887 in swrastSetTexBuffer (pDRICtx=0xa31e888, target=3553, dPriv=0xa0a9068) at swrast.c:106 #5 0xb74cc0ad in __glXDRIbindTexImage () from /usr/lib/xorg/modules/extensions/libglx.so #6 0xb74c8802 in __glXDisp_BindTexImageEXT () from /usr/lib/xorg/modules/extensions/libglx.so #7 0xb74c967a in __glXDisp_VendorPrivate () from /usr/lib/xorg/modules/extensions/libglx.so #8 0xb74cb66f in __glXDispatch () from /usr/lib/xorg/modules/extensions/libglx.so #9 0x0806f187 in Dispatch () #10 0x080621b5 in main () (gdb) c Continuing. Now that I've turned on OpenGLIsSafe, this isn't a big issue for me. But I'm left wondering, is this a bug, an incomplete feature, or an unsupported configuration? Is it an itch worth scratching further? One more thing. This is Arch Linux running Mesa built from the 7.9 git tree at dc4956922dfbec2fb2f60539be97760022bd4613. I tried backporting to the 7.9 tar file and had similar results. Just wanted to note that I can reproduce this as well. As an interesting addition, the same occurs on my Lenovo Ideapad S12 netbook, too (which has a Via card, so I guess it uses the same software renderer). Ancient bug, closing. |
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.