April 2001 - MPlayer-dev-eng - lists.mplayerhq.hu

Matrox dualhead and syncfb.
by German Gomez Garcia 02 May '01

02 May '01

Hello I'm trying to run mplayer with syncfb on the dualhead without success, in fact I'm not even able to run syncfb on first head. It fails in the ioctl() that SET_CONFIG (both when running on the first head and when running in the second head). I don't know if syncfb even supported under the second head but it should work under the first head, shouldn't it? Other thing, I'm subscribed to dev-eng only, and the traffic is really low, should I subscribe myself to users or devel ? I don't speak hungarian. Thanks, - german ------------------------------------------------------------------------- German Gomez Garcia | "This isn't right. This isn't even wrong." <german(a)piraos.com> | -- Wolfgang Pauli _______________________________________________ Mplayer-dev-eng mailing list Mplayer-dev-eng(a)lists.sourceforge.net http://lists.sourceforge.net/lists/listinfo/mplayer-dev-eng

5 25

SIGIL in fastmemcpy - Fwd: Re: MPlayer question
by Arpi 29 Apr '01

29 Apr '01

--------- Forwarded message --------- From: Kylus <pjb169(a)psu.edu> To: Arpi <arpi(a)thot.banki.hu> Subject: Re: MPlayer question > > can you try fastmemcpyebnch? > just: > cd TOOLS > ./c > ./fastmem.sh $ ./fastmem.sh k6 : v2-v1=1096167166 = 1367087us (73.148fps) 109.7MB/s k7 : v2-v1=691985437 = 863012us (115.873fps) 173.8MB/s mmx: v2-v1=695764875 = 867725us (115.244fps) 172.9MB/s sse: v2-v1=696439782 = 868568us (115.132fps) 172.7MB/s Couldn't open /dev/mga_vid ./fastmem.sh: line 12: 688 Segmentation fault (core dumped) ./fastmem2-k6 Couldn't open /dev/mga_vid ./fastmem.sh: line 14: 690 Segmentation fault (core dumped) ./fastmem2-k7 Couldn't open /dev/mga_vid ./fastmem.sh: line 16: 692 Segmentation fault (core dumped) ./fastmem2-mmx Couldn't open /dev/mga_vid ./fastmem.sh: line 18: 695 Segmentation fault (core dumped) /fastmem2-sse (I have the core file if you want it, but it's 3 MB!) > which mplayer version do you use? how long is libvo/fastmemcpy.h? I'm using the brand new 'Idegcounter' version 0.17a that was recently released. $ls -l libvo/fastmemcpy.h 8356 Apr 26 16:35 libvo/fastmemcpy.h Patrick Boyne Penn State University _______________________________________________ Mplayer-dev-eng mailing list Mplayer-dev-eng(a)lists.sourceforge.net http://lists.sourceforge.net/lists/listinfo/mplayer-dev-eng

1 0

fastmemcpy.h SIGILL on P3...
by Arpi 29 Apr '01

29 Apr '01

Some additional info. > > 0x80a9786 <fast_memcpy+166>: prefetchnta 0x140(%edx) > this should be valid on p3. > Can you please tell me kernel version (uname -a) and show content of > /proc/cpuinfo ? $ uname -a Linux C498560-A 2.2.18 #6 Sun Mar 18 20:36:12 EST 2001 i686 unknown $ cat /proc/cpuinfo processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 8 model name : Pentium III (Coppermine) stepping : 3 cpu MHz : 801.831 cache size : 256 KB fdiv_bug : no hlt_bug : no sep_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 2 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 sep mtrr pge mca cmov pat pse36 mmx fxsr xmm bogomips : 1599.07 Good luck! Patrick Boyne Penn State University _______________________________________________ Mplayer-dev-eng mailing list Mplayer-dev-eng(a)lists.sourceforge.net http://lists.sourceforge.net/lists/listinfo/mplayer-dev-eng

1 0

l3codeca.acm bug in VirtualAlloc - solved?
by Arpi 29 Apr '01

29 Apr '01

Hi, Thanks to Mw3 at IRC, I think we found the problem: He got SIGILL for a file using the l3codeca.acm file (that file is uploaded to thot, robots.avi. but it doesn't matter) here is the gdb backtrace: <Mw3> Program received signal SIGILL, Illegal instruction. <Mw3> [Switching to Thread 1024 (LWP 25286)] <Mw3> 0x4a0000 in strtold () from /lib/libc.so.6 <Mw3> #0 0x4a0000 in strtold () from /lib/libc.so.6 <Mw3> #1 0x4d0918 in malloc () from /lib/libc.so.6 <Mw3> #2 0x8078f3d in VirtualAlloc (address=0x400000, size=655360, type=12288 , protection=64) at ext.c:473 <Mw3> #3 0x8082e19 in PE_LoadImage (handle=6, filename=0xbfffe8c8 "/usr/lib/win32/l3codeca.acm", version=0xbfffe8c2) at pe_image.c:569 What's happen here? libc is loaded to 0x4xxxxx and l3codeca.acm requested big VirtualAlloc to that address. So libc's code has been overwritten with data bytes, so when it executed malloc() call to libc it tried to run invalid code -> SIGILL ldd ./mplayer: <Mw3> libc.so.6 => /lib/libc.so.6 (0x00460000) Mw3's system: <Mw3> Linux pange 2.4.3-ac13-getrewted #1 Fri Apr 27 01:09:04 CEST 2001 i686 libc: 2.2.2 cpu: celeron ---------------------------------- my system: Linux arpi 2.4.3 #3 Mon Apr 2 20:08:49 CEST 2001 i686 unknown libc: 2.1.3 arpi:/usr/src/linux# gcc -v Reading specs from /usr/lib/gcc-lib/i386-slackware-linux/2.95.2/specs gcc version 2.95.2 19991024 (release) ldd ./mplayer: libc.so.6 => /lib/libc.so.6 (0x4026c000) ------------------------------------ So, for me everething is loaded above 0x40000000, while for Mw3 things are mapped from 0x00000000. maybe the -ac kernels has some patch applied to change default memory partitioning? Or it depends on libc version? Any ideas? We really has to fix mp3lib and make it default (l3codeca.acm is faster on intel cpu's but has such problems) A'rpi / Astral & ESP-team -- mailto:arpi@thot.banki.hu http://esp-team.scene.hu _______________________________________________ Mplayer-dev-eng mailing list Mplayer-dev-eng(a)lists.sourceforge.net http://lists.sourceforge.net/lists/listinfo/mplayer-dev-eng

1 0

fastmemcpy SIGILL on p3 - gdb log
by Arpi 28 Apr '01

28 Apr '01

he has P3. see the log: ----- RUNNING PROGRAM (gdb) r /usr/anime/gate_keepers_01.avi Starting program: /home/kylus/MPlayer-0.17a_IdegCounter/./mplayer /usr/anime/gate_keepers_01.avi [New Thread 1024 (LWP 4861)] MPlayer 0.17a-IdegCounter (C) 2000-2001 Arpad Gereoffy <arpi(a)thot.banki.hu> Reading config file: /etc/mplayer.conf Reading config file: /root/.mplayer/config Reading /root/.mplayer/codecs.conf: 15 audio & 26 video codecs Font /root/.mplayer/font/font.desc loaded successfully! (140 chars) Detected AVI file format! ==> Found video stream: 0 ==> Found audio stream: 1 VIDEO: [DIV3] 512x384 24bpp 23.98 fps 702.7 kbps (85.8 kbyte/s) [V] filefmt:3 fourcc:0x33564944 size:512x384 fps:23.98 ftime:=0.0417 Found audio codec: [mp3acm] drv:4 (MPEG layer-3) Loading DLL: /usr/lib/win32/l3codeca.acm OK AUDIO: samplerate=24000 channels=2 bps=2 Found video codec: [divxds] drv:4 (DivX ;-) (MS MPEG-4 v3)) Loading DLL: /usr/lib/win32/divx_c32.ax OK DShow: using YUY2 colorspace DivX setting result = 0 VO: [xv] 512x384 => 512x384 YUY2 Xvideo image format: 0x32595559 (YUY2) packed Xvideo image format: 0x32315659 (YV12) planar Xvideo image format: 0x30323449 (I420) planar Xvideo image format: 0x59565955 (UYVY) packed using Xvideo port 58 for hw scaling [xv] dcx: 1 dcy: 21 dx: 0 dy: 0 dw: 512 dh: 384 Audio buffer size: 131072 bytes, delay: 1.365s Start playing... Program received signal SIGILL, Illegal instruction. [Switching to Thread 1024 (LWP 4861)] 0x80a978d in fast_memcpy (to=0x40e23000, from=0x40cc1008, len=0) at fastmemcpy.h:149 149 __asm__ __volatile__ ( ----- BT OUTPUT (gdb) bt #0 0x80a978d in fast_memcpy (to=0x40e23000, from=0x40cc1008, len=0) at fastmemcpy.h:149 #1 0x80a95fa in draw_frame (src=0x81df538) at vo_xv.c:360 #2 0x8052567 in main (argc=2, argv=0xbffffa04, envp=0xbffffa10) at mplayer.c:1432 #3 0x4050fb5c in __libc_start_main (main=0x804dca0 <main>, argc=2, ubp_av=0xbffffa04, init=0x804bf7c <_init>, fini=0x80c7354 <_fini>, rtld_fini=0x4000d634 <_dl_fini>, stack_end=0xbffff9fc) at ../sysdeps/generic/libc-start.c:129 ----- ASSEMBLER DUMP (gdb) disass $eip-50 $eip+50 Dump of assembler code from 0x80a975b to 0x80a97bf: 0x80a975b <fast_memcpy+123>: and $0xf,%eax 0x80a975e <fast_memcpy+126>: test %eax,%eax 0x80a9760 <fast_memcpy+128>: je 0x80a97c0 <fast_memcpy+224> 0x80a9762 <fast_memcpy+130>: lea 0x0(%esi,1),%esi 0x80a9769 <fast_memcpy+137>: lea 0x0(%edi,1),%edi 0x80a9770 <fast_memcpy+144>: cmpl $0x0,0xfffffff0(%ebp) 0x80a9774 <fast_memcpy+148>: jg 0x80a9780 <fast_memcpy+160> 0x80a9776 <fast_memcpy+150>: jmp 0x80a9802 <fast_memcpy+290> 0x80a977b <fast_memcpy+155>: nop 0x80a977c <fast_memcpy+156>: lea 0x0(%esi,1),%esi 0x80a9780 <fast_memcpy+160>: mov 0xc(%ebp),%edx 0x80a9783 <fast_memcpy+163>: mov 0x8(%ebp),%ecx 0x80a9786 <fast_memcpy+166>: prefetchnta 0x140(%edx) 0x80a978d <fast_memcpy+173>: movups (%edx),%xmm0 0x80a9790 <fast_memcpy+176>: movups 0x10(%edx),%xmm1 0x80a9794 <fast_memcpy+180>: movntps %xmm0,(%ecx) 0x80a9797 <fast_memcpy+183>: movntps %xmm1,0x10(%ecx) 0x80a979b <fast_memcpy+187>: movups 0x20(%edx),%xmm0 0x80a979f <fast_memcpy+191>: movups 0x30(%edx),%xmm1 0x80a97a3 <fast_memcpy+195>: movntps %xmm0,0x20(%ecx) 0x80a97a7 <fast_memcpy+199>: movntps %xmm1,0x30(%ecx) 0x80a97ab <fast_memcpy+203>: addl $0x40,0xc(%ebp) 0x80a97af <fast_memcpy+207>: addl $0x40,0x8(%ebp) 0x80a97b3 <fast_memcpy+211>: lea 0xfffffff0(%ebp),%eax 0x80a97b6 <fast_memcpy+214>: decl (%eax) 0x80a97b8 <fast_memcpy+216>: jmp 0x80a9770 <fast_memcpy+144> 0x80a97ba <fast_memcpy+218>: lea 0x0(%esi),%esi End of assembler dump. _______________________________________________ Mplayer-dev-eng mailing list Mplayer-dev-eng(a)lists.sourceforge.net http://lists.sourceforge.net/lists/listinfo/mplayer-dev-eng

1 0

WOW!!!!!!
by TeLeNiEkO 28 Apr '01

28 Apr '01

That's nice!! Now Arpi can translate all the mails in dev-hun and send them here translated X"DDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDD viszlat! PS: Thanks for subscribing ;o) ==================================== | TeLeNiEkO | | e-Mail: telenieko(a)telenieko.com | ====================================

2 1

Re: DVD raw mode, IFO Parsing and AC3 through SPDIF
by German Gomez Garcia 27 Apr '01

27 Apr '01

Sorry for posting to the hungarian list, of course I cannot speak hungarian (I'm from Spain) so I've unsuscribed myself from mplayer-devel, and subscribe to mplayer-dev-eng. Hope everybody at the hungarian list that can read english is subscribed to this also. First of all, I would like to say that my work is just integrating other people work, I've 'stolen' code from the xine project (that in fact 'stole' the code from the livid dvd project, VIVA GPL!), although it seems that mplayer is far better at AC3 decoding that xine, because with mplayer I can play Matrix without problems while with Xine and other DVD players I get CRC problems at some scenes (Trinity fighting the cops at the beginning, Trinity and Neo at the metal detector, etc). I'm working in improving a bit the LIRC support for DVD playing, next/prev chapter, select chapter by number, change audio track, etc. Not dificult so it should be quite straightforward. Other thing I would like to see in mplayer is some 'postprocessing' capabilities, things like brightness/contrast/color correction are really usefull for some files/DVDs, as now this is beyond my knowledge, so is anybody out there? - german ------------------------------------------------------------------------- German Gomez Garcia | "This isn't right. This isn't even wrong." <german(a)piraos.com> | -- Wolfgang Pauli _______________________________________________ Mplayer-dev-eng mailing list Mplayer-dev-eng(a)lists.sourceforge.net http://lists.sourceforge.net/lists/listinfo/mplayer-dev-eng

3 3

Still sync probs audio/subtitle related
by Felix Buenemann 27 Apr '01

27 Apr '01

Hi, MPlayer still has sync probs after seeking, I tried some seeking around in TheMatrix.avi (dvdrip by me, proper sync), after some longer seeking and seeking around a-v got up to about 3secs (usually 0.000 and ct 0.015 to 0.030), after seeking audio sync was bad bud corrected itself (a-v slowly went down to 0.000 again, ct was at about 2,5sec then), but now audio was synced again but subtitles were out of sync (microdvd format). -- Best Regards, Felix _______________________________________________ Mplayer-dev-eng mailing list Mplayer-dev-eng(a)lists.sourceforge.net http://lists.sourceforge.net/lists/listinfo/mplayer-dev-eng

1 0