Bug#351171: xserver-xorg: fbCompositeSolidMask_nx8888x8888Cmmx() is hogging the CPU
On Fri, Feb 03, 2006 at 03:25:23PM +0100, KELEMEN Peter wrote: retitle 351171 xserver-xorg: subpixel rendering hogs CPU severity 351171 normal thanks * David Nusinow ([EMAIL PROTECTED]) [20060202 21:57]: Yeah, don't run EXA since it's still largely experimental. Actually, EXA has nothing to do with it. Tried with XAA as well: samples %image nameapp name symbol name 952554 66.8796 libfb.so Xorg fbCompositeSolidMask_nxxCmmx 69217 4.8598 vmlinux Xorg delay_tsc 33075 2.3222 radeonXorg (no symbols) 24733 1.7365 opreport opreport (no symbols) 16063 1.1278 libc-2.3.5.so xterm memmove 14222 0.9985 xterm xterm (no symbols) Ok, sorry about that. It looked very much like other EXA bugs that have popped up lately. Thanks for testing with XAA too. - David Nusinow -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#351171: xserver-xorg: fbCompositeSolidMask_nx8888x8888Cmmx() is hogging the CPU
On Fri, 2006-02-03 at 00:49 +0100, KELEMEN Peter wrote: Package: xserver-xorg Version: 6.9.0.dfsg.1-4 Severity: normal Start up an xterm with some FreeType fontface: uxterm -fa Monospace -fs 12 In the terminal, start spewing lot of output lines: find / Observe the X server eating ~100% CPU. Over a 2 minutes run, oprofile reveals the following: samples %image name app name symbol name 550032 51.1805 libfb.soXorg fbCompositeSolidMask_nxxCmmx 289192 26.9093 libfb.soXorg fbCompositeSrcAdd_xmmx You seem to have set up fontconfig for subpixel anti-aliasing. Even assuming your card could actually accelerate that, it's very non-trivial and hasn't been implemented yet. You should get better results with traditional greyscale AA. -- Earthling Michel Dänzer | Debian (powerpc), X and DRI developer Libre software enthusiast| http://svcs.affero.net/rm.php?r=daenzer
Bug#351171: xserver-xorg: fbCompositeSolidMask_nx8888x8888Cmmx() is hogging the CPU
retitle 351171 xserver-xorg: subpixel rendering hogs CPU severity 351171 normal thanks * David Nusinow ([EMAIL PROTECTED]) [20060202 21:57]: Yeah, don't run EXA since it's still largely experimental. Actually, EXA has nothing to do with it. Tried with XAA as well: samples %image nameapp name symbol name 952554 66.8796 libfb.so Xorg fbCompositeSolidMask_nxxCmmx 69217 4.8598 vmlinux Xorg delay_tsc 33075 2.3222 radeonXorg (no symbols) 24733 1.7365 opreport opreport (no symbols) 16063 1.1278 libc-2.3.5.so xterm memmove 14222 0.9985 xterm xterm (no symbols) Peter -- .+'''+. .+'''+. .+'''+. .+'''+. .+'' Kelemen Péter / \ / \ /[EMAIL PROTECTED] .+' `+...+' `+...+' `+...+' `+...+'
Bug#351171: xserver-xorg: fbCompositeSolidMask_nx8888x8888Cmmx() is hogging the CPU
* Michel Dänzer ([EMAIL PROTECTED]) [20060203 11:08]: You seem to have set up fontconfig for subpixel anti-aliasing. Even assuming your card could actually accelerate that, it's very non-trivial and hasn't been implemented yet. You should get better results with traditional greyscale AA. Indeed, thanks. I've switches off subpixel rendering and I get the following profiles: EXA: samples %image name app name symbol name 104856 7.5679 libc-2.3.5.soxtermmemmove 886576.3988 xtermxterm(no symbols) 674594.8688 libc-2.3.5.soxtermwcwidth 624134.5046 radeon_drv.soXorg RadeonCompositeCP 618494.4639 libc-2.3.5.soXorg memcpy 616164.4471 vmlinux Xorg delay_tsc 553343.9937 libXft.so.2.1.2 xterm(no symbols) 496413.5828 libexa.soXorg exaComposite 441503.1865 radeon_drv.soXorg RADEONHostDataBlit 409942.9587 Xorg Xorg FindGlyphRef 359512.5947 radeon_drv.soXorg R200PrepareCompositeCP 283912.0491 libfb.so Xorg fbSolid 276021.9922 Xorg Xorg miComputeCompositeRegion 272781.9688 radeon Xorg (no symbols) 240251.7340 libexa.soXorg exaPixmapIsOffscreen 223021.6096 radeon_drv.soXorg RADEONUploadToScreenCP 200631.4480 libexa.soXorg exaGlyphs 192871.3920 radeon_drv.soXorg R200TextureSetupCP 155101.1194 libexa.soXorg exaPixmapUseScreen XAA: samples %image name app name symbol name 153097 10.9779 libfb.so Xorg fbCompositeSrcAdd_8000x8000mmx 130111 9.3297 libc-2.3.5.soxtermmemmove 113832 8.1624 xtermxterm(no symbols) 852786.1149 libc-2.3.5.soxtermwcwidth 702755.0391 libXft.so.2.1.2 xterm(no symbols) 526033.7719 Xorg Xorg FindGlyphRef 452993.2482 libfb.so Xorg fbComposite 345462.4771 libfb.so Xorg fbSolid 316522.2696 Xorg Xorg miComputeCompositeRegion 269591.9331 libxaa.soXorg XAAComposite 196281.4074 libc-2.3.5.soXorg memcpy 162541.1655 libshadowfb.so Xorg ShadowComposite 158931.1396 Xorg Xorg damageComposite 154421.1073 Xorg Xorg miValidatePicture 154271.1062 libxaa.soXorg XAADoComposite 144811.0384 Xorg Xorg CompositePicture 143031.0256 Xorg Xorg miGlyphs 142641.0228 Xorg Xorg miGlyphExtents Thanks, Peter -- .+'''+. .+'''+. .+'''+. .+'''+. .+'' Kelemen Péter / \ / \ /[EMAIL PROTECTED] .+' `+...+' `+...+' `+...+' `+...+'
Bug#351171: xserver-xorg: fbCompositeSolidMask_nx8888x8888Cmmx() is hogging the CPU
retitle 351171 xserver-xorg: EXA causes CPU hogging and/or slowdown severity 351171 minor thanks On Fri, Feb 03, 2006 at 12:49:12AM +0100, KELEMEN Peter wrote: Observe the X server eating ~100% CPU. Over a 2 minutes run, oprofile reveals the following: snip Section Device Identifier ATI Technologies, Inc. RV280 [Radeon 9200 SE] Driver radeon BusID PCI:1:0:0 Option AccelMethod EXA Option RenderAccel true # OptionAllowGLXWithComposite true Option EnablePageFliptrue Option AGPMode 4 # hangs!!! # OptionAGPFastWrite true EndSection Yeah, don't run EXA since it's still largely experimental. - David Nusinow -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#351171: xserver-xorg: fbCompositeSolidMask_nx8888x8888Cmmx() is hogging the CPU
Package: xserver-xorg Version: 6.9.0.dfsg.1-4 Severity: normal Start up an xterm with some FreeType fontface: uxterm -fa Monospace -fs 12 In the terminal, start spewing lot of output lines: find / Observe the X server eating ~100% CPU. Over a 2 minutes run, oprofile reveals the following: samples %image name app name symbol name 550032 51.1805 libfb.soXorg fbCompositeSolidMask_nxxCmmx 289192 26.9093 libfb.soXorg fbCompositeSrcAdd_xmmx 9791 0.9111 opreportopreport (no symbols) 0.8270 libcairo.so.2.2.3 gtik2_applet2 (no symbols) 7624 0.7094 vmlinux Xorg delay_tsc 6857 0.6380 libgobject-2.0.so.0 metacity (no symbols) 6518 0.6065 libpango-1.0.so.0 gtik2_applet2 (no symbols) 6322 0.5883 libfb.soXorg fbCopyAreammx 5833 0.5428 radeon Xorg (no symbols) 5583 0.5195 oprofiled oprofiled (no symbols) 5520 0.5136 libglib-2.0.so.0gtik2_applet2 (no symbols) Any hints? Thanks, Peter -- Package-specific info: Contents of /var/lib/xfree86/X.roster: xserver-xfree86 xserver-xorg /etc/X11/X target unchanged from checksum in /var/lib/xfree86/X.md5sum. X server symlink status: lrwxrwxrwx 1 root root 17 Jul 18 2005 /etc/X11/X - /usr/bin/X11/Xorg -rwxr-xr-x 1 root root 1878044 Jan 15 02:40 /usr/bin/X11/Xorg Contents of /var/lib/xfree86/xorg.conf.roster: xserver-xorg VGA-compatible devices on PCI bus: :01:00.0 VGA compatible controller: ATI Technologies Inc RV280 [Radeon 9200 SE] (rev 01) /etc/X11/xorg.conf does not match checksum in /var/lib/xfree86/xorg.conf.md5sum. Xorg X server configuration file status: -rw--- 1 root root 3501 Feb 3 00:21 /etc/X11/xorg.conf Contents of /etc/X11/xorg.conf: Section Files FontPathunix/:7100# local font server # if the local font server has problems, we can fall back on these FontPath/usr/lib/X11/fonts/misc FontPath/usr/lib/X11/fonts/cyrillic FontPath/usr/lib/X11/fonts/100dpi/:unscaled FontPath/usr/lib/X11/fonts/75dpi/:unscaled FontPath/usr/lib/X11/fonts/Type1 FontPath/usr/lib/X11/fonts/CID FontPath/usr/lib/X11/fonts/Speedo FontPath/usr/lib/X11/fonts/100dpi FontPath/usr/lib/X11/fonts/75dpi FontPath /var/lib/defoma/x-ttcidfont-conf.d/dirs/CID FontPath /var/lib/defoma/x-ttcidfont-conf.d/dirs/TrueType EndSection Section Module LoadGLcore Loadbitmap Loaddbe Loadddc Loaddri Loadextmod Loadfreetype Loadglx Loadint10 Loadrecord Loadspeedo Loadtype1 Loadvbe SubSection extmod Option omit XFree86-DGA EndSubSection EndSection Section InputDevice Identifier Generic Keyboard Driver keyboard Option CoreKeyboard Option XkbRules xorg Option XkbModel pc104 Option XkbLayout us EndSection Section InputDevice Identifier Configured Mouse Driver mouse Option CorePointer Option Device/dev/gpmdata Option Protocol IntelliMouse Option ZAxisMapping 4 5 EndSection Section Device Identifier ATI Technologies, Inc. RV280 [Radeon 9200 SE] Driver radeon BusID PCI:1:0:0 Option AccelMethod EXA Option RenderAccel true # OptionAllowGLXWithComposite true Option EnablePageFliptrue Option AGPMode 4 # hangs!!! # OptionAGPFastWrite true EndSection #Section Extensions # Option Composite Enable #EndSection Section Monitor Identifier Acer AL1721 HorizSync 30-83 VertRefresh 55-75 Option DPMS EndSection Section Screen Identifier Default Screen Device ATI Technologies, Inc. RV280 [Radeon 9200 SE] Monitor Acer AL1721 DefaultDepth24 SubSection Display Depth 1 Modes 1280x1024 1024x768 800x600 640x480 EndSubSection SubSection Display Depth 4 Modes 1280x1024 1024x768 800x600 640x480 EndSubSection SubSection Display Depth 8 Modes 1280x1024 1024x768 800x600 640x480