Re: [Mesa-dev] [status][help]pipe-video on Loongson3A(mipsel) platform with r600
Hi, 2011/8/5 Andy Furniss andy...@ukfsn.org: By chance it seems that your big buck bunny does not work with multithread. I don't see any difference either - probably because it's too simple, mediainfo shows it as using Main@L3.1 with no CABAC If I test with something that uses High@L5.1 and CABAC eg. http://www.andyqos.ukfsn.org/Planet_Earth_From_Pole_to_Pole_1080p_sample.mkv Then I see a 3x increase between 1 and 4 threads. Just like you said: 1 thread: 685.104s 4 threads: 219.361s -- Regards, - cee1 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [status][help]pipe-video on Loongson3A(mipsel) platform with r600
Am Donnerstag, den 04.08.2011, 13:41 +0800 schrieb fykc...@gmail.com: I guess that may be related with driver or DRM stack, after playback, my desktop corrupts http://dev.lemote.com/files/upload/software/temp/newmobcal1920-desktop-corrupt.png (Similar with -vc ffmpeg12vdpau -vo vdpau). I've seen something very similar at the very beginning of pipe-video development, back then it was related to some invalid offset calculations inside the kernel module, but that was fixed long ago. What we see here is definitely some corruption of vram wich isn't related to video playback all together, maybe a piglit run could help narrowing down the problem a bit. I don't know how close to rv670 rs780 is, but I have artifact issues with my rv670 that don't show with my rv790. Not very, if I remember correctly rs780 is an igp chipset with r7xx instruction set directly wired to the motherboard, while rv670 is an normal PCIe card with r6xx instruction set. I was able to narrow down your problems with the rv670 to somewhere in the instruction set compiler, so I don't think that these two bugs are related to each other. If you can get it to display I would interested what - http://www.andyqos.ukfsn.org/Pendulum.mpg No, this renders a corrupted playback http://dev.lemote.com/files/upload/software/temp/Pendulum-playback.png, but doesn't cause my desktop corruption. I would bet that it has something todo with the resolution/ maximum texture size. Try the tex3d-maxsize and fbo-maxsize tests in piglit, and see what the outcome of this is. Christian. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [status][help]pipe-video on Loongson3A(mipsel) platform with r600
On Thu, Aug 4, 2011 at 10:26 AM, Christian König deathsim...@vodafone.de wrote: Am Donnerstag, den 04.08.2011, 13:41 +0800 schrieb fykc...@gmail.com: I guess that may be related with driver or DRM stack, after playback, my desktop corrupts http://dev.lemote.com/files/upload/software/temp/newmobcal1920-desktop-corrupt.png (Similar with -vc ffmpeg12vdpau -vo vdpau). I've seen something very similar at the very beginning of pipe-video development, back then it was related to some invalid offset calculations inside the kernel module, but that was fixed long ago. Might try a newer kernel. What we see here is definitely some corruption of vram wich isn't related to video playback all together, maybe a piglit run could help narrowing down the problem a bit. I don't know how close to rv670 rs780 is, but I have artifact issues with my rv670 that don't show with my rv790. Not very, if I remember correctly rs780 is an igp chipset with r7xx instruction set directly wired to the motherboard, while rv670 is an normal PCIe card with r6xx instruction set. I was able to narrow down your problems with the rv670 to somewhere in the instruction set compiler, so I don't think that these two bugs are related to each other. RS780 actually has an r6xx-based (rv610) 3d core. RS880 is also r6xx (rv620) based. If you can get it to display I would interested what - http://www.andyqos.ukfsn.org/Pendulum.mpg No, this renders a corrupted playback http://dev.lemote.com/files/upload/software/temp/Pendulum-playback.png, but doesn't cause my desktop corruption. I would bet that it has something todo with the resolution/ maximum texture size. Try the tex3d-maxsize and fbo-maxsize tests in piglit, and see what the outcome of this is. Christian. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [status][help]pipe-video on Loongson3A(mipsel) platform with r600
fykc...@gmail.com wrote: Does it work now then? No, it renders a corrupted playback http://dev.lemote.com/files/upload/software/temp/newmobcal1920-playback.png I guess that may be related with driver or DRM stack, after playback, my desktop corrupts http://dev.lemote.com/files/upload/software/temp/newmobcal1920-desktop-corrupt.png (Similar with -vc ffmpeg12vdpau -vo vdpau). I don't know how close to rv670 rs780 is, but I have artifact issues with my rv670 that don't show with my rv790. If you can get it to display I would interested what - http://www.andyqos.ukfsn.org/Pendulum.mpg No, this renders a corrupted playback http://dev.lemote.com/files/upload/software/temp/Pendulum-playback.png, but doesn't cause my desktop corruption. These look far worse than I see, and I get no desktop corruption, so as Christian says it's probably not the same issue. mplayer -lavdopts threads=4 will use all of your cores. That still doesn't use all CPU power(even with thread=8) according to top. I guess a single core is maxed - press 1 (one) while top is running to see separate lines for each core. By chance it seems that your big buck bunny does not work with multithread. I don't see any difference either - probably because it's too simple, mediainfo shows it as using Main@L3.1 with no CABAC If I test with something that uses High@L5.1 and CABAC eg. http://www.andyqos.ukfsn.org/Planet_Earth_From_Pole_to_Pole_1080p_sample.mkv Then I see a 3x increase between 1 and 4 threads. With: 1. Option SwapbuffersWait off 2. xvattr -a XV_VSYNC -v 0 3. Use metacity instead of compiz 4. mplayer -nosound -benchmark -lavdopts threads=N -vo xv video-test/big_buck_bunny_720p_h264.mov threads=1 vs threads=4, the result showed little difference: * 1 thread: 1519.509s * 4 threads: 1486.825s The details http://dev.lemote.com/files/upload/software/temp/thread_1.log vs http://dev.lemote.com/files/upload/software/temp/thread_4.log Hmm, that's really slow. I guess I am spoiled by mplayer/ffmpeg using asm and mmx. Even if I force my CPUs to 800MHz I can play that stream at almost 2X speed. Maybe it would be worth researching whether there are compile tweaks you can use for your arch with ffmpeg and mplayer. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [status][help]pipe-video on Loongson3A(mipsel) platform with r600
Hi, 在 2011年8月2日 下午8:16,fykc...@gmail.com fykc...@gmail.com 写道: Depending on how it was configured mplayer needs (probably) /etc/X11/XvMCConfig containing /path/to/your/lib/libXvMCr600.so I add the path /usr/lib/libXvMCr600.so to /etc/X11/XvMCConfig, but mplayer seems not invoke it (No r600 in strace or /proc/pid/maps). Maybe I should try a newer version of mplayer. My mplayer reads /usr/lib/XvMCConfig, after I providing that file, libXvMCr600.so is loaded by mplayer. Generally r600 pipe-video still has rendering and functional issues. If you just want to test fair enough, but if you are trying to make your system play h264 that it can't currently the best thing you could do is compile svn mplayer which now has ffmpeg h264 multithread support so - mplayer -lavdopts threads=4 will use all of your cores. Tried, not make big difference at first glance. (P.S. can mplayer show fps in OSD?) That still doesn't use all CPU power(even with thread=8) according to top. -- Regards, - cee1 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [status][help]pipe-video on Loongson3A(mipsel) platform with r600
fykc...@gmail.com wrote: Hi, 在 2011年8月2日 下午8:16,fykc...@gmail.comfykc...@gmail.com 写道: Depending on how it was configured mplayer needs (probably) /etc/X11/XvMCConfig containing /path/to/your/lib/libXvMCr600.so I add the path /usr/lib/libXvMCr600.so to /etc/X11/XvMCConfig, but mplayer seems not invoke it (No r600 in strace or /proc/pid/maps). Maybe I should try a newer version of mplayer. My mplayer reads /usr/lib/XvMCConfig, after I providing that file, libXvMCr600.so is loaded by mplayer. Does it work now then? I don't know how close to rv670 rs780 is, but I have artifact issues with my rv670 that don't show with my rv790. If you can get it to display I would interested what - http://www.andyqos.ukfsn.org/Pendulum.mpg looks like. Generally r600 pipe-video still has rendering and functional issues. If you just want to test fair enough, but if you are trying to make your system play h264 that it can't currently the best thing you could do is compile svn mplayer which now has ffmpeg h264 multithread support so - mplayer -lavdopts threads=4 will use all of your cores. Tried, not make big difference at first glance. (P.S. can mplayer show fps in OSD?) That still doesn't use all CPU power(even with thread=8) according to top. Just to be clear, this not for pipe video - you have to decode with ffmpeg software. I guess xv would be the fastest vo to use. Personally I prefer -vo gl, but gallium is slower than classic for me (not helped by having a noisy fan requiring me to keep my card on low power when using as HTPC). I think the difference is because gallium uses blits rather than memcpy. I don't think mplayer will show fps, but it does have a benchmark option (you still have to work out your own fps). Unless you are bottlenecked on display I expect newmobcal would show a small difference with threads=4, but some high bitrate h264 would probably be better. To benchmark you need to turn off various vsyncs depending on -vo First in xorg.conf device section Option SwapbuffersWait off For gl, run mplayer with env vblank_mode=0 For xv, xvattr -a XV_VSYNC -v 0 Then try mplayer -nosound -benchmark -vo xv newmobcal1920.ts and repeat with -lavdopts threads=4 Though it shouldn't affect benchmarks too much, when actually using my PC as HTPC I've found cpufreq_ondemand to be a pain - so I set to perf. If rs780 has power saving then that's another variable to play with eg as root. echo low /sys/class/drm/card0/device/power_profile echo high /sys/class/drm/card0/device/power_profile ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [status][help]pipe-video on Loongson3A(mipsel) platform with r600
Hi, 2011/8/3 Andy Furniss andy...@ukfsn.org: My mplayer reads /usr/lib/XvMCConfig, after I providing that file, libXvMCr600.so is loaded by mplayer. Does it work now then? No, it renders a corrupted playback http://dev.lemote.com/files/upload/software/temp/newmobcal1920-playback.png I guess that may be related with driver or DRM stack, after playback, my desktop corrupts http://dev.lemote.com/files/upload/software/temp/newmobcal1920-desktop-corrupt.png (Similar with -vc ffmpeg12vdpau -vo vdpau). I don't know how close to rv670 rs780 is, but I have artifact issues with my rv670 that don't show with my rv790. If you can get it to display I would interested what - http://www.andyqos.ukfsn.org/Pendulum.mpg No, this renders a corrupted playback http://dev.lemote.com/files/upload/software/temp/Pendulum-playback.png, but doesn't cause my desktop corruption. Generally r600 pipe-video still has rendering and functional issues. If you just want to test fair enough, but if you are trying to make your system play h264 that it can't currently the best thing you could do is compile svn mplayer which now has ffmpeg h264 multithread support so - mplayer -lavdopts threads=4 will use all of your cores. Tried, not make big difference at first glance. (P.S. can mplayer show fps in OSD?) That still doesn't use all CPU power(even with thread=8) according to top. With: 1. Option SwapbuffersWait off 2. xvattr -a XV_VSYNC -v 0 3. Use metacity instead of compiz 4. mplayer -nosound -benchmark -lavdopts threads=N -vo xv video-test/big_buck_bunny_720p_h264.mov threads=1 vs threads=4, the result showed little difference: * 1 thread: 1519.509s * 4 threads: 1486.825s The details http://dev.lemote.com/files/upload/software/temp/thread_1.log vs http://dev.lemote.com/files/upload/software/temp/thread_4.log -- Regards, - cee1 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [status][help]pipe-video on Loongson3A(mipsel) platform with r600
Hi all, I've tried mesa-7.11rc4 with pipe-video patch on a Loongson3A based notebook: * CPU: 4-core MIPS64 compatible (Little-endian, 64bit kernel with an o32 userland) * Graphics: RS780E I tried h264、mpeg2 samples with mplayer -vo vdpau and xvmc, non of them works: * For h264, mplayer produces the following error messages: [VD_FFMPEG] Trying pixfmt=0. [vdpau] Failed creating VDPAU decoder: A catch-all error, used when no other error code applies. Could not find matching colorspace - retrying with -vf scale... Opening video filter: [scale] The selected video_out device is incompatible with this codec. Try appending the scale filter to your filter list, e.g. -vf spp,scale instead of -vf spp. [h264_vdpau @ 0x2bddf1b8]decoding to PIX_FMT_NONE is not supported. [h264_vdpau @ 0x2bddf1b8]decode_slice_header error [h264_vdpau @ 0x2bddf1b8]no frame! Error while decoding frame! * For mepg4, mplayer renders a corrupted playback(http://dev.lemote.com/files/upload/software/temp/IMG_20110802_104200.jpg), and produces the following: [VD_FFMPEG] XVMC-accelerated MPEG-2. Movie-Aspect is 1.78:1 - prescaling to correct movie aspect. VO: [vdpau] 1920x1080 = 1920x1080 MPEG2 VDPAU acceleration [mpegvideo_vdpau @ 0x2cb7b1b8]warning: first frame is no keyframe V: 1.7 2/ 2 ??% ??% ??,?% 0 0 [J [mpegvideo_vdpau @ 0x2cb7b1b8]warning: first frame is no keyframe V: 1.8 3/ 3 ??% ??% ??,?% 0 0 [J I also tried the master branch(5b3c7199830b8eaac4df2f8c3f10d0e89b4bd5c5), and got the same result. The test samples can be found at: http://dev.lemote.com/files/upload/software/temp/video-test/ -- run test.sh, it will play each sample with vdpau and xvmc, recording the outputs. The detailed log message of the above test, including glxinfo can be found at: http://dev.lemote.com/files/upload/software/temp/video-test.result.log.tar.bz2 Xorg info: http://dev.lemote.com/files/upload/software/temp/Xorg.0.log Any idea? Thanks. More info about the OS environment: 1. Linux: 2.6.36 2. OS: Based on debian squeeze, and ports some ubuntu graphical packages. 2. X Server: 1.10.2.902 3. xf86-video-ati: 6.14.2 4. compiz: 0.8.6 -- Regards, - cee1 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [status][help]pipe-video on Loongson3A(mipsel) platform with r600
Hi, Am Dienstag, den 02.08.2011, 15:00 +0800 schrieb fykc...@gmail.com: I've tried mesa-7.11rc4 with pipe-video patch on a Loongson3A based notebook: * CPU: 4-core MIPS64 compatible (Little-endian, 64bit kernel with an o32 userland) * Graphics: RS780E a bit unusual hardware configuration, but lets just see what we can do. I'm not familiar with this kind of hardware so does o32 mean you run some x86 emulation, or are these native binaries? I tried h264、mpeg2 samples with mplayer -vo vdpau and xvmc, non of them works: * For h264, mplayer produces the following error messages: We currently don't have support for accelerated h264, so this test can't succeed, but your test with the newmobcal1920 (mpeg2) demo should indeed work. Any idea? Thanks. First question of all: Does glxgears work flawlessly? If yes (what I assume) then the basics seems to work fine. Please try mplayer just with the -vo vdpau option (without any -vc option), this should result in software decoding with hardware colour conversion and scaling, just the same thing that Xv does. Does this gives you a picture? Something else you should try is export VDPAU_DEBUG=4, that should give you a whole bunch of debug informations, please supply those to the list. Regards, Christian. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [status][help]pipe-video on Loongson3A(mipsel) platform with r600
Hi Christian, Thanks for your reply. 2011/8/2 Christian König deathsim...@vodafone.de: Hi, Am Dienstag, den 02.08.2011, 15:00 +0800 schrieb fykc...@gmail.com: I've tried mesa-7.11rc4 with pipe-video patch on a Loongson3A based notebook: * CPU: 4-core MIPS64 compatible (Little-endian, 64bit kernel with an o32 userland) * Graphics: RS780E a bit unusual hardware configuration, but lets just see what we can do. I'm not familiar with this kind of hardware so does o32 mean you run some x86 emulation, or are these native binaries? These are native binaries, o32 is an ABI for mips, which means 32bit address space + assumes all general purpose registers are 32bit. I tried h264、mpeg2 samples with mplayer -vo vdpau and xvmc, non of them works: * For h264, mplayer produces the following error messages: We currently don't have support for accelerated h264, so this test can't succeed, but your test with the newmobcal1920 (mpeg2) demo should indeed work. Any idea? Thanks. First question of all: Does glxgears work flawlessly? If yes (what I assume) then the basics seems to work fine. Yes, it works fine. Please try mplayer just with the -vo vdpau option (without any -vc option), this should result in software decoding with hardware colour conversion and scaling, just the same thing that Xv does. Does this gives you a picture? Yes, it works. Something else you should try is export VDPAU_DEBUG=4, that should give you a whole bunch of debug informations, please supply those to the list. Attachment is the log. -- Regards, - cee1 VDPAU_DEBUG=4 VDPAU_DRIVER=r600 mplayer -vc ffmpeg12vdpau -vo vdpau newmobcal1920.ts MPlayer 1.0rc4-4.4.5 (C) 2000-2010 MPlayer Team mplayer: could not connect to socket mplayer: No such file or directory Failed to open LIRC support. You will not be able to use your remote control. Playing newmobcal1920.ts. TS file format detected. VIDEO MPEG2(pid=49) NO AUDIO! NO SUBS (yet)! PROGRAM N. 1 VIDEO: MPEG2 1920x1080 (aspect 3) 29.970 fps 22496.0 kbps (2812.0 kbyte/s) [VDPAU] Device created succesfully [VDPAU] Got proc adress 0x2ecd06f0 for id 0 [VDPAU] Got proc adress 0x2ecd0a04 for id 5 [VDPAU] Got proc adress 0x2ecd3584 for id 9 [VDPAU] Got proc adress 0x2ecd3428 for id 10 [VDPAU] Got proc adress 0x2ecd3164 for id 13 [VDPAU] Got proc adress 0x2ecd1848 for id 22 [VDPAU] Got proc adress 0x2ecd1b5c for id 18 [VDPAU] Got proc adress 0x2ecd1e6c for id 19 [VDPAU] Got proc adress 0x2ecd16b8 for id 46 [VDPAU] Got proc adress 0x2ecd160c for id 53 [VDPAU] Got proc adress 0x2ecd1508 for id 54 [VDPAU] Got proc adress 0x2ecd146c for id 47 [VDPAU] Got proc adress 0x2ecd1408 for id 48 [VDPAU] Got proc adress 0x2ecd0b0c for id 55 [VDPAU] Got proc adress 0x2ecd2574 for id 56 [VDPAU] Got proc adress 0x2ecd24b0 for id 57 [VDPAU] Got proc adress 0x2ecd214c for id 63 [VDPAU] Got proc adress 0x2ecd2050 for id 64 [VDPAU] Got proc adress 0x2ecd0bb8 for id 4096 [VDPAU] Got proc adress 0x2ecd1860 for id 33 [VDPAU] Got proc adress 0x2ecd1850 for id 23 [VDPAU] Got proc adress 0x2ecd0478 for id 37 [VDPAU] Got proc adress 0x2eccfe80 for id 40 [VDPAU] Got proc adress 0x2ecd037c for id 38 [VDPAU] Got proc adress 0x2eccfd48 for id 26 [VDPAU] Got proc adress 0x2eccfc10 for id 27 [VDPAU] Got proc adress 0x2eccfc50 for id 29 [VDPAU] Got proc adress 0x2ecd1868 for id 34 [VDPAU] Got proc adress 0x2ecd1390 for id 6 [VDPAU] Got proc adress 0x2ecd2018 for id 66 == Forced video codec: ffmpeg12vdpau Opening video decoder: [ffmpeg] FFmpeg's libavcodec codec family Selected video codec: [ffmpeg12vdpau] vfm: ffmpeg (FFmpeg MPEG-1/2 (VDPAU)) == Audio: no sound Starting playback... [VD_FFMPEG] XVMC-accelerated MPEG-2. [VDPAU] Creating decoder [VDPAU] Decoder created succesfully [VDPAU] Destroying decoder [VDPAU] Creating decoder [VDPAU] Decoder created succesfully Movie-Aspect is 1.78:1 - prescaling to correct movie aspect. VO: [vdpau] 1920x1080 = 1920x1080 MPEG2 VDPAU acceleration [VDPAU] Destroying decoder [VDPAU] Creating decoder [VDPAU] Decoder created succesfully [VDPAU] Destroying decoder [VDPAU] Creating decoder [VDPAU] Decoder created succesfully [VDPAU] Creating PresentationQueueTarget [VDPAU] Creating PresentationQueue [VDPAU] Creating VideoMixer [VDPAU] Setting VideoMixer features [VDPAU] Generating CSCMatrix [VDPAU] Creating output surface [VDPAU] Creating output surface [VDPAU] Creating output surface [VDPAU] Creating output surface [VDPAU] Creating a surface [VDPAU] Decoding [VDPAU] Decoding MPEG2 [VDPAU] Creating a surface [mpegvideo_vdpau @ 0x2c83b1b8]warning: first frame is no keyframe [VDPAU] Creating a surface [VDPAU] Decoding [VDPAU] Decoding MPEG2 V: 1.7 2/ 2 ??% ??% ??,?% 0 0 [J [mpegvideo_vdpau @ 0x2c83b1b8]warning: first frame is no keyframe [VDPAU] Creating a surface [VDPAU] Decoding [VDPAU] Decoding MPEG2 V: 1.8 3/ 3 ??%
Re: [Mesa-dev] [status][help]pipe-video on Loongson3A(mipsel) platform with r600
Hi, Thanks for your reply. 2011/8/2 Andy Furniss andy...@ukfsn.org: fykc...@gmail.com wrote: Hi all, I've tried mesa-7.11rc4 with pipe-video patch on a Loongson3A based notebook: * CPU: 4-core MIPS64 compatible (Little-endian, 64bit kernel with an o32 userland) * Graphics: RS780E I tried h264、mpeg2 samples with mplayer -vo vdpau and xvmc, non of them works: * For h264, mplayer produces the following error messages: Only mpeg 1 and 2 are supported. [VD_FFMPEG] Trying pixfmt=0. [vdpau] Failed creating VDPAU decoder: A catch-all error, used when no other error code applies. Does VDPAU_DRIVER=r600 vdpauinfo find your libs - if not you may need to copy them to where ever vdpau is looking. Yes. I specified a non-exists driver, mplayer quit early, complains not find the non-exists so. Any idea? Thanks. Compiz may complicate things - I haven't tested with it. I've did the test without compiz, it seems not make any difference. Depending on how it was configured mplayer needs (probably) /etc/X11/XvMCConfig containing /path/to/your/lib/libXvMCr600.so I add the path /usr/lib/libXvMCr600.so to /etc/X11/XvMCConfig, but mplayer seems not invoke it (No r600 in strace or /proc/pid/maps). Maybe I should try a newer version of mplayer. Your mplayer is old - maybe svn would be better if you can build it. mplayer -v will tell you more about where it's looking for things like xvmc libs - but with newmobcal you will also get swamped with transport stream info. vanilla xf86-video-ati: 6.14.2 does not have xvmc support, but I see from your Xorg log that you must be using a patched/other version. Yes, some r600-related patches are backported from the master branch. Generally r600 pipe-video still has rendering and functional issues. If you just want to test fair enough, but if you are trying to make your system play h264 that it can't currently the best thing you could do is compile svn mplayer which now has ffmpeg h264 multithread support so - mplayer -lavdopts threads=4 will use all of your cores. Got it, thanks. -- Regards, - cee1 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [status][help]pipe-video on Loongson3A(mipsel) platform with r600
fykc...@gmail.com wrote: Depending on how it was configured mplayer needs (probably) /etc/X11/XvMCConfig containing /path/to/your/lib/libXvMCr600.so I add the path /usr/lib/libXvMCr600.so to /etc/X11/XvMCConfig, but mplayer seems not invoke it (No r600 in strace or /proc/pid/maps). Maybe I should try a newer version of mplayer. mplayer -v may help, just in case it's looking somewhere else. For me mplayer xvmc is not built by default, I have to --enable-xvmc when configuring. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev