I'm using the ATI driver compiled from GIT after commit 4ce255a9db8027f2471ba50e57724d3e4d5b16ba. The xorg-server I'm using is version 1.5.3. Today after some time of use X started to use all the CPU. The graphics didn't lock up but it did start to run slowly and was unresponsive. I then SSHed remotely into the laptop and with GDB attached to the X process. When I interupt it and show backtrace it shows one of these two: #0 0x00007f9396c16fe7 in ioctl () from /lib/libc.so.6 #1 0x00007f9395c9b91c in drmDMA () from /usr/lib/libdrm.so.2 #2 0x00007f93956143a9 in RADEONCPGetBuffer (pScrn=0x1b43de0) at radeon_accel.c:588 #3 0x00007f939561452f in RADEONCPFlushIndirect (pScrn=0x1b43de0, discard=1) at radeon_accel.c:642 #4 0x00007f9395660213 in RADEONPrepareSolidCP (pPix=0x7f20470, alu=3, pm=4294967295, fg=0) at radeon_exa_funcs.c:103 #5 0x00007f9394d98636 in exaFillRegionSolid (pDrawable=0x7f20470, pRegion=0x5274540, pixel=0, planemask=4294967295, alu=3) at exa_accel.c:1072 #6 0x00007f9394d99ee6 in exaPolyFillRect (pDrawable=0x7f20470, pGC=0x1b9f990, nrect=1, prect=0x7fffa12c18d0) at exa_accel.c:751 #7 0x000000000052c216 in damagePolyFillRect (pDrawable=0x7f20470, pGC=0x1b9f990, nRects=1, pRects=0x7fffa12c18d0) at damage.c:1337 #8 0x00007f9394d9bef8 in exaGlyphs (op=3 '\003', pSrc=0x2632800, pDst=0x6fda700, maskFormat=0x1b793b8, xSrc=0, ySrc=0, nlist=<value optimized out>, list=0x7fffa12c2270, glyphs=0x7fffa12c1a70) at exa_glyphs.c:833 #9 0x000000000052bbe1 in damageGlyphs (op=10 '\n', pSrc=0x2632800, pDst=0x6fda700, maskFormat=0x0, xSrc=0, ySrc=0, nlist=1, list=0x7fffa12c2270, glyphs=0x7fffa12c1a70) at damage.c:654 #10 0x000000000051cb64 in ProcRenderCompositeGlyphs (client=0x1d39b00) at render.c:1468 #11 0x000000000044c694 in Dispatch () at dispatch.c:454 #12 0x000000000043351d in main (argc=8, argv=0x7fffa12c2848, envp=<value optimized out>) at main.c:441 #0 0x00007f9396c16fe7 in ioctl () from /lib/libc.so.6 #1 0x00007f9395c9b91c in drmDMA () from /usr/lib/libdrm.so.2 #2 0x00007f93956143a9 in RADEONCPGetBuffer (pScrn=0x1b43de0) at radeon_accel.c:588 #3 0x00007f939561452f in RADEONCPFlushIndirect (pScrn=0x1b43de0, discard=1) at radeon_accel.c:642 #4 0x00007f9395661881 in R300PrepareCompositeCP (op=3, pSrcPicture=0x25f7a60, pMaskPicture=0x0, pDstPicture=0x25892c0, pSrc=0x9377ba0, pMask=0x0, pDst=0x263e600) at radeon_exa_render.c:1218 #5 0x00007f9394d9ea92 in exaTryDriverComposite (op=3 '\003', pSrc=0x25f7a60, pMask=0x0, pDst=0x25892c0, xSrc=0, ySrc=0, xMask=<value optimized out>, yMask=<value optimized out>, xDst=<value optimized out>, yDst=<value optimized out>, width=<value optimized out>, height=<value optimized out>) at exa_render.c:670 #6 0x00007f9394d9f45d in exaComposite (op=3 '\003', pSrc=0x25f7a60, pMask=0x0, pDst=0x25892c0, xSrc=0, ySrc=0, xMask=0, yMask=0, xDst=0, yDst=0, width=1024, height=768) at exa_render.c:935 #7 0x000000000052b8c8 in damageComposite (op=10 '\n', pSrc=0x25f7a60, pMask=0x0, pDst=0x25892c0, xSrc=0, ySrc=0, xMask=0, yMask=<value optimized out>, xDst=<value optimized out>, yDst=<value optimized out>, width=<value optimized out>, height=<value optimized out>) at damage.c:576 #8 0x0000000000513dd1 in miCompositeRects (op=<value optimized out>, pDst=0x25892c0, color=<value optimized out>, nRect=1, rects=0x11e6e284) at mirect.c:168 #9 0x000000000051b784 in ProcRenderFillRectangles (client=0x21e1ac0) at render.c:1509 #10 0x000000000044c694 in Dispatch () at dispatch.c:454 #11 0x000000000043351d in main (argc=8, argv=0x7fffa12c2848, envp=<value optimized out>) at main.c:441 The hardware I use is a laptop HP Compaq nx6325 with integrated graphics: ATI Technologies Inc RS482 [Radeon Xpress 200M]
Created attachment 21170 [details] xorg.conf
Created attachment 21171 [details] Xorg.0.log
Can you track down the problematic commit with git bisect?
I couldn't reproduce this anymore with later revisions and I guess this bug can be closed as worksforme. I'll reopen if I get this problem again sometime.
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.