I am using mencoder for video encoding with qpel enabled in xvidencopts. xvid-1.1.0 w/ qpel worked fine. xvid-1.1.0 w/o qpel worked fine. xvid-1.1.0-r1 w/ qpel fails with SIGSEGV. xvid-1.1.0-r1 w/o qpel fails with SIGSEGV. I figured out that xvid-1.1.0-textrel-2 patch, that was included in xvid-1.1.0-r1, causes the segfault. Therefore I consider the patch as _harmful_ and not to work as intended. Please fix that issue - either * recheck the patch and rewrite the part that causes qpel to SIGSEGV or * leave the patch out in portage. I don't think any further information could be helpful, but if anyone looking at this bugreport wants some evidence of the issue or any other information I of course will provide the information that is needed. Thanks alot in advance for fixing this issue.
I was just overlooking my bugreport and noticed a copy'n'paste typo. The correct table is: xvid-1.1.0 w/ qpel worked fine. xvid-1.1.0 w/o qpel worked fine. xvid-1.1.0-r1 w/ qpel fails with SIGSEGV. xvid-1.1.0-r1 w/o qpel works fine Sorry for that.
I have same observations. Tested with mencoder 1.0_pre8.
Me too. I am also getting a SIGSEGV with the xvid-1.1.0-textrel-2 patch. Removing it from the ebuild solves (i.e. works around) the problem.
This crashes mencoder on my athlon box with even without qpel. I would also request that it me removed. (In reply to comment #0) > I am using mencoder for video encoding with qpel enabled in xvidencopts. > > xvid-1.1.0 w/ qpel worked fine. > xvid-1.1.0 w/o qpel worked fine. > xvid-1.1.0-r1 w/ qpel fails with SIGSEGV. > xvid-1.1.0-r1 w/o qpel fails with SIGSEGV. > > I figured out that xvid-1.1.0-textrel-2 patch, that was included in > xvid-1.1.0-r1, causes the segfault. > > Therefore I consider the patch as _harmful_ and not to work as intended. > > Please fix that issue - either > * recheck the patch and rewrite the part that causes qpel to SIGSEGV or > * leave the patch out in portage. > > I don't think any further information could be helpful, but if anyone looking > at this bugreport wants some evidence of the issue or any other information I > of course will provide the information that is needed. > > Thanks alot in advance for fixing this issue. >
mencoder crashes with quant_type=mpeg. Removing the textrel patch from the ebuild seems to solve the problem.
*** Bug 147974 has been marked as a duplicate of this bug. ***
*** This bug has been marked as a duplicate of 129022 ***
This bug has been marked a duplicate of bug #129022 and both have been marked resolved. However, I still suffer from the textrel issue with mencoder. Removing the textrel patch solves the problem. Please, reopen this bug.
Okay, seems like there were two different problems with very different behaviour ... let's reopen this.
*** Bug 131422 has been marked as a duplicate of this bug. ***
Hardened, should I make the patch conditional pending futher examination?
I have this problem as well, causing dvdrip, mplayer and transcode problems. CPU: Athlon-XP. As long as this is not fixed the ebuild should be masked. I did so on my system ("echo =media-libs/xvid-1.1.0-r1>>/etc/portage/package.mask"), downgraded to 1.0.3 ("emerge -uDa media-libs/xvid") and have a working system again.
Me too! xvid-1.1.0-r1 does not work with transcode 1.0.2-r3 on my Athlon XP 2700+. Downgrade to xvid-1.1.0 fixed this. So xvid-1.1.0-r1 is NOT stable on x86!
it seems that this bug affects all xvid-encoding related stuff - including dvdrip-0.98.1: [export_xvid4.so] Neither './xvid4.cfg' nor '~/.transcode/xvid4.cfg' [export_xvid4.so] found. Default settings will be used instead. [decode_mpeg2.c] libmpeg2 0.4.0b loop decoder [decode_mpeg2.c] libmpeg2 acceleration: 3dnow sh: line 1: 15727 Segmentation fault execflow -n 15 transcode -H 10 -a 0 -x vob,null -i /home/antonio/rip/AOEO/vob/022/ -b 128,0,0 --a52_drc_off -f 25 -B 12,10,8 -R 1 -y xvid4,null -o /dev/null --print_status 25 Portage 2.1.1 (default-linux/x86/2006.1/desktop, gcc-4.1.1, glibc-2.4-r3, 2.6.16-gentoo-r13 i686) ================================================================= System uname: 2.6.16-gentoo-r13 i686 AMD Athlon(tm) XP 2400+ Gentoo Base System version 1.12.5 Last Sync: Sun, 08 Oct 2006 09:50:01 +0000 distcc 2.18.3 i686-pc-linux-gnu (protocols 1 and 2) (default port 3632) [disabled] app-admin/eselect-compiler: [Not Present] dev-java/java-config: 1.3.7, 2.0.30 dev-lang/python: 2.4.3-r4 dev-python/pycrypto: 2.0.1-r5 dev-util/ccache: [Not Present] dev-util/confcache: [Not Present] sys-apps/sandbox: 1.2.17 sys-devel/autoconf: 2.13, 2.59-r7 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2 sys-devel/binutils: 2.16.1-r3 sys-devel/gcc-config: 1.3.13-r4 sys-devel/libtool: 1.5.22 virtual/os-headers: 2.6.17-r1 ACCEPT_KEYWORDS="x86" AUTOCLEAN="yes" CBUILD="i686-pc-linux-gnu" CFLAGS="-march=athlon-xp -O2 -pipe" CHOST="i686-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/kde/3.5/env /usr/kde/3.5/share/config /usr/kde/3.5/shutdown /usr/share/X11/xkb /usr/share/config /usr/share/texmf/dvipdfm/config/ /usr/share/texmf/dvips/config/ /usr/share/texmf/tex/generic/config/ /usr/share/texmf/tex/platex/config/ /usr/share/texmf/xdvi/" CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/gconf /etc/java-config/vms/ /etc/revdep-rebuild /etc/terminfo" CXXFLAGS="-march=athlon-xp -O2 -pipe" DISTDIR="/usr/portage/distfiles" FEATURES="autoconfig collision-protect distlocks metadata-transfer sandbox sfperms strict" GENTOO_MIRRORS="http://gentoo.ynet.sk/pub " LANG="en_US.utf8" LC_ALL="en_US.utf8" LINGUAS="en de" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --delete-after --stats --timeout=180 --exclude='/distfiles' --exclude='/local' --exclude='/packages'" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage" SYNC="rsync://192.168.0.1/gentoo-portage" USE="x86 3dnow 3dnowext X a52 aac acpi alsa audiofile beagle berkdb bitmap-fonts bzip2 cairo cdr cli crypt css cups dbus dlloader dri dvd dvdr dvdread eds elibc_glibc emboss encode evo exif fam fbcon ffmpeg firefox flac fortran gdbm gif ginac gmp gnome gphoto2 gpm gstreamer gtk gtk2 hal howl icq input_devices_keyboard input_devices_mouse ipv6 isdnlog java javascript jpeg jpeg2k kernel_linux lcms ldap libg++ linguas_de linguas_en mad mikmod mime mmx mmxext mozsvg mp3 mpeg msn nautilus ncurses nfs nls nptl nptlonly nsplugin nvidia offensive ogg oggvorbis opengl pam pcre perl plotutils png posix ppds pppd python qt3 qt4 quicktime readline real reflection ruby sdl session sockets spell spl sqlite3 sse ssl subtitles svg tcpd tetex theora tiff truetype truetype-fonts type1-fonts udev unicode usb userland_GNU vcd video_cards_nvidia vorbis win32codecs wma xine xml xorg xv xvid zlib" Unset: CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LDFLAGS, PORTAGE_RSYNC_EXTRA_OPTS
btw, xvid-1.1.0-r2 seems to be broken too ... at least dvdrip-0.98.1 doesn't like it that much: [decode_mpeg2.c] libmpeg2 acceleration: 3dnow failed to write Y plane of framesh: line 1: 15306 Segmentation fault execflow -n 15 transcode -H 10 -a 0 -x vob,null -i /home/antonio/rip/AOEO/vob/022/ -b 128,0,0 --a52_drc_off -f 25 -Y 2,4,0,4 -Z 600x450 -R 1 -y xvid4,null -o /dev/null --print_status 25 (demuxer.c) write program stream packet: Broken pipe
*** Bug 150861 has been marked as a duplicate of this bug. ***
*** Bug 151234 has been marked as a duplicate of this bug. ***
(In reply to comment #11) > Hardened, should I make the patch conditional pending futher examination? Maybe USE="segfault" ? :P Now seriously, can we nuke the patch that's breaking lots of stuff?
I'm getting a segfault with xvid-1.1.0-r2 as well. Took a while to track it down to xvid. I thought it was mencoder choking on my firewire captured video. Going down to xvid-1.0.3 worked fine. equery -C list xvid [ Searching for package 'xvid' in all categories among: ] * installed packages [I--] [ ~] media-libs/xvid-1.1.0-r2 (1) emerge --info Portage 2.1.1 (default-linux/x86/2006.1/desktop, gcc-4.1.1, glibc-2.4-r3, 2.6.17-gentoo-r2 i686) ================================================================= System uname: 2.6.17-gentoo-r2 i686 AMD Athlon(tm) XP 1800+ Gentoo Base System version 1.12.5 Last Sync: Mon, 16 Oct 2006 06:01:01 +0000 distcc 2.18.3 i686-pc-linux-gnu (protocols 1 and 2) (default port 3632) [disabled] app-admin/eselect-compiler: [Not Present] dev-java/java-config: 1.3.7, 2.0.30 dev-lang/python: 2.3.5-r2, 2.4.3-r4 dev-python/pycrypto: 2.0.1-r5 dev-util/ccache: [Not Present] dev-util/confcache: [Not Present] sys-apps/sandbox: 1.2.17 sys-devel/autoconf: 2.13, 2.59-r7 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2 sys-devel/binutils: 2.16.1-r3 sys-devel/gcc-config: 1.3.13-r4 sys-devel/libtool: 1.5.22 virtual/os-headers: 2.6.17-r1 ACCEPT_KEYWORDS="x86" AUTOCLEAN="yes" CBUILD="i686-pc-linux-gnu" CFLAGS="-march=athlon-xp -O2 -pipe -D_FILE_OFFSET_BITS=64" CHOST="i686-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/kde/3.4/env /usr/kde/3.4/share/config /usr/kde/3.4/shutdown /usr/kde/3.5/env /usr/kde/3.5/share/config /usr/kde/3.5/shutdown /usr/share/X11/xkb /usr/share/config" CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/gconf /etc/java-config/vms/ /etc/revdep-rebuild /etc/terminfo" CXXFLAGS="-march=athlon-xp -O2 -pipe -D_FILE_OFFSET_BITS=64" DISTDIR="/usr/portage/distfiles" FEATURES="autoconfig distlocks metadata-transfer sandbox sfperms strict" GENTOO_MIRRORS="http://distfiles.gentoo.org http://distro.ibiblio.org/pub/linux/distributions/gentoo" LINGUAS="" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --delete-after --stats --timeout=180 --exclude='/distfiles' --exclude='/local' --exclude='/packages'" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage" SYNC="rsync://rsync.gentoo.org/gentoo-portage" USE="x86 3dnow 3dnowext X a52 aac alsa amr apache2 arts berkdb bitmap-fonts cairo cdr clamav cli cpudetection crypt css cups dbus debug dlloader dri dvd dvdr dvdread eds elibc_glibc emboss encode esd examples fam ffmpeg firefox fortran gdbm gif gpm gstreamer gtk gtk2 hal input_devices_keyboard input_devices_mouse ipv6 isdnlog java javascript jpeg kde kernel_linux ldap libg++ mad madwifi mikmod mmx mmxext mp3 mpeg ncurses nfs nls nntp nptl nptlonly ogg opengl oss pam pcre perl png ppds pppd python qt qt3 qt4 quicktime readline reflection samba sdl session spell spl sse sse2 ssl tcpd threadsafe truetype truetype-fonts type1-fonts udev unicode userland_GNU video_cards_radeon vorbis wifi win32codecs xinetd xml xorg xv xvid zlib" Unset: CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LANG, LC_ALL, LDFLAGS, PORTAGE_RSYNC_EXTRA_OPTS
On AMD XP 2000+ and on AMD Duron 800 MHz I noticed the same Bug: the command "mencoder original.avi -o encoded.avi -ovc xvid -xvidencopts bitrate=1200:me_quality=6:vhq=4:hq_ac -oac copy" shows a segfault when using the xvid-1.1.0-r2.ebuild. My solution is to use a modified xvid-1.1.0-r1.ebuild. I deleted the following lines: # mirror://gentoo/${P}-textrel-2.patch.bz2" # epatch "${DISTDIR}/${P}-textrel-2.patch.bz2" (and made "ebuild xvid-1.1.0-r1.ebuild digest" and so on). on the AMD Duron: ------------------------------------------------ grep flags /proc/cpuinfo flags : fpu vme de pse tsc msr pae mce cx8 sep mtrr pge mca cmov pat pse36 mmx fxsr syscall mmxext 3dnowext 3dnow up mencoder MEncoder 1.0pre8-4.1.1 (C) 2000-2006 MPlayer Team CPU: AMD Duron(tm) Processor (Family: 6, Model: 3, Stepping: 1) CPUflags: Type: 6 MMX: 1 MMX2: 1 3DNow: 1 3DNow2: 1 SSE: 0 SSE2: 0 Kompiliert f
On AMD XP 2000+ and on AMD Duron 800 MHz I noticed the same Bug: the command "mencoder original.avi -o encoded.avi -ovc xvid -xvidencopts bitrate=1200:me_quality=6:vhq=4:hq_ac -oac copy" shows a segfault when using the xvid-1.1.0-r2.ebuild. My solution is to use a modified xvid-1.1.0-r1.ebuild. I deleted the following lines: # mirror://gentoo/${P}-textrel-2.patch.bz2" # epatch "${DISTDIR}/${P}-textrel-2.patch.bz2" (and made "ebuild xvid-1.1.0-r1.ebuild digest" and so on). on the AMD Duron: ------------------------------------------------ grep flags /proc/cpuinfo flags : fpu vme de pse tsc msr pae mce cx8 sep mtrr pge mca cmov pat pse36 mmx fxsr syscall mmxext 3dnowext 3dnow up mencoder MEncoder 1.0pre8-4.1.1 (C) 2000-2006 MPlayer Team CPU: AMD Duron(tm) Processor (Family: 6, Model: 3, Stepping: 1) CPUflags: Type: 6 MMX: 1 MMX2: 1 3DNow: 1 3DNow2: 1 SSE: 0 SSE2: 0 Kompiliert für x86 CPU mit folgenden Erweiterungen: MMX MMX2 3DNow 3DNowEx on the AMD XP 2000+: ------------------------------------------------ grep flags /proc/cpuinfo flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 mmx fxsr sse syscall mmxext 3dnowext 3dnow up ts mencoder MEncoder 1.0pre8-4.1.1 (C) 2000-2006 MPlayer Team CPU: AMD Athlon(tm) XP 2100+ (Family: 6, Model: 6, Stepping: 2) CPUflags: Type: 6 MMX: 1 MMX2: 1 3DNow: 1 3DNow2: 1 SSE: 1 SSE2: 0 Kompiliert für x86 CPU mit folgenden Erweiterungen: MMX MMX2 3DNow 3DNowEx SSE
well, i've just started encoding with media-libs/xvid-1.1.0-r3 USE="(-altivec) -doc" using media-video/dvdrip-0.98.1 USE="hal ogg subtitles vcd vorbis xine xvid -fping -mplayer" (see bug 149344) on x86 and got no segfaults so far (they appeared immediatly with -r2). encoding should be finished in about 6h ... if nothing fancy happens we may consider reopening bug 149344 and stable dvdrip-0.98.1 on x86 - at least in my opinion ...
arg ... wrong bug; sorry for the spam
*** Bug 153304 has been marked as a duplicate of this bug. ***
The media-libs/xvid-1.1.0-r3.ebuild looks same (same patches are included) as my working modified "-r1" (see comment #20). So it should work now on AMD XP 2000+ and so on. Thanks. [ As I see it, "textrel-2.patch" caused the segfaults. ] But now I will wait for a xvid-1.1.1.ebuild ;-)
(In reply to comment #11) > Hardened, should I make the patch conditional pending futher examination? Correct. And bug #153014 should of never happened. Vs dropping support for somethign that's supposed to fix stuff we should of CC: the author of that patch (which we failed to do) so it could be debuged and corrected.
*** Bug 159129 has been marked as a duplicate of this bug. ***
media-libs/xvid-1.1.0-r1: xvid-1.1.0-textrel-2 patch causes SIGSEGV when using qpel we might need some help there to have a working patch for this because it still applies to 1.1.2 that I just bumped. I'm a n00b with asm so that's probably better that I don't try to fix it myself ;)
Created attachment 108302 [details, diff] fixed textrel patch for xvid-1.1.2 here's the new patch that i hope will fix the segfaults, please test it with the workload that reproduced the crashes for you and report back (if it still crashes, i'll need gdb output, stuff like 'bt', 'x/8i $pc', 'i r', etc). there's still one problem with the current handling of nasm/yasm though, the configure script will prefer the latter even on i386 therefore it will fail to build if you have it installed. see bug #90287 about the details. last but not least... this whole thing originated in bug #90287 and the problems should have been address there (that i was subscribed to already) vs. a new bug (that noone bothered to cc me, i learned about it the hard way when the textrels showed up again). please try to make a minimum effort to search bugzilla for related issues next time, these segfaults could have been resolved a long time ago.
(In reply to comment #29) > here's the new patch that i hope will fix the segfaults, please test it with > the workload that reproduced the crashes for you and report back (if it still > crashes, i'll need gdb output, stuff like 'bt', 'x/8i $pc', 'i r', etc). I'm having a segfault with that patch applied to xvid : Program received signal SIGSEGV, Segmentation fault. [Switching to Thread -1210434880 (LWP 21822)] 0xb56ba01f in enc_encode (pEnc=0x85fb340, xFrame=0xb5dfd74c, stats=0xbff2e974) at ../../src/encoder.c:1467 1467 ../../src/encoder.c: No such file or directory. in ../../src/encoder.c (gdb) bt #0 0xb56ba01f in enc_encode (pEnc=0x85fb340, xFrame=0xb5dfd74c, stats=0xbff2e974) at ../../src/encoder.c:1467 #1 0xb56bb3ea in xvid_encore (handle=0x85fb340, opt=-1521309356, param1=0xb5dfd74c, param2=0xbff2e974) at ../../src/xvid.c:852 #2 0xb5df0028 in tc_export () from /usr/lib/transcode/export_xvid4.so #3 0x0805fbef in tcv_export () #4 0x0805f231 in encoder () #5 0x080588dd in main () (gdb) bt full #0 0xb56ba01f in enc_encode (pEnc=0x85fb340, xFrame=0xb5dfd74c, stats=0xbff2e974) at ../../src/encoder.c:1467 q = <value optimized out> type = 0 bs = {bufa = 0, bufb = 0, buf = 130637824, pos = 19, tail = 0x85fb53c, start = 0x85bb720, length = 0, initpos = 0} #1 0xb56bb3ea in xvid_encore (handle=0x85fb340, opt=-1521309356, param1=0xb5dfd74c, param2=0xbff2e974) at ../../src/xvid.c:852 No locals. #2 0xb5df0028 in tc_export () from /usr/lib/transcode/export_xvid4.so No symbol table info available. #3 0x0805fbef in tcv_export () No symbol table info available. #4 0x0805f231 in encoder () No symbol table info available. #5 0x080588dd in main () No symbol table info available. (gdb) x/8i $pc 0xb56ba01f <enc_encode+2847>: add 0x44(%edx),%esi 0xb56ba022 <enc_encode+2850>: mov %edx,0xfffff838(%ebp) 0xb56ba028 <enc_encode+2856>: mov 0x158(%esi),%eax 0xb56ba02e <enc_encode+2862>: movl $0x3,0xec(%esi) 0xb56ba038 <enc_encode+2872>: movl $0x0,0x18(%esi) 0xb56ba03f <enc_encode+2879>: movl $0x0,0x10(%esi) 0xb56ba046 <enc_encode+2886>: test %eax,%eax 0xb56ba048 <enc_encode+2888>: movl $0x0,0x8(%esi) (gdb) i r eax 0x9add596c -1696769684 ecx 0x85fb340 140489536 edx 0xa552a954 -1521309356 ebx 0xb5741ff4 -1250680844 esp 0xbff29be0 0xbff29be0 ebp 0xbff2a3f8 0xbff2a3f8 esi 0xc0fc4bf0 -1057207312 edi 0xc 12 eip 0xb56ba01f 0xb56ba01f <enc_encode+2847> eflags 0x10a13 [ CF AF IF OF RF ] cs 0x73 115 ss 0x7b 123 ds 0x7b 123 es 0x7b 123 fs 0x0 0 gs 0x33 51 I hope this will be useful, I only compiled xvid with debug infos, please tell me if you need more debugger outputs or a simple way to reproduce it.
could you please post the commandline to reproduce this and your USE flags for xvid and transcode? also, could you post a /proc/pid/maps file corresponding to a segfault (and disable randomization in /proc/sys/kernel/randomize_va_space and post the same again)?
Created attachment 120299 [details] /proc/pid/maps with 1 in /proc/sys/kernel/randomize_va_space #echo 1 > /proc/sys/kernel/randomize_va_space $ transcode -x vob -i VTS_06_0.VOB -y xvid4,null -o /dev/null [...] [export_xvid4.so] Neither './xvid4.cfg' nor '~/.transcode/xvid4.cfg' [export_xvid4.so] found. Default settings will be used instead. tc_memcpy: using sse for memcpy No accelerated IMDCT transform found tc_memcpy: using sse for memcpy [decode_mpeg2.c] libmpeg2 acceleration: 3dnowSegmentation fault
Created attachment 120301 [details] /proc/pid/maps with 0 in /proc/sys/kernel/randomize_va_space # echo "0" > /proc/sys/kernel/randomize_va_space [attached maps] emerge -pv xvid transcode : [ebuild R ] media-libs/xvid-1.1.2-r1 USE="(-altivec) -doc" 0 kB [3] [ebuild R ] media-video/transcode-1.0.3 USE="3dnow X a52 dv dvdread iconv imagemagick jpeg lzo mjpeg mmx mp3 ogg quicktime sdl sse theora truetype v4l2 vorbis xml xvid (-altivec) -extrafilters -fame -gtk -mpeg -network -sse2" 0 kB [2] The vob file has been taken here : http://samples.mplayerhq.hu/MPEG-VOB/VTS_06_0.VOB Perhaps a more useful bt : (gdb) bt #0 MBCoding (frame=0x80a8fc0, pMB=0xb3015980, qcoeff=0xbfff9c40, bs=0xbfffa318, pStat=0x80a90dc) at ../../src/bitstream/bitstream.h:392 #1 0xb527214d in enc_encode (pEnc=0x80a8a40, xFrame=0xb59b574c, stats=0xbfffe8c4) at ../../src/encoder.c:1479 #2 0xb52733ea in xvid_encore (handle=0x80a8a40, opt=31468455, param1=0xb59b574c, param2=0xbfffe8c4) at ../../src/xvid.c:852 #3 0xb59a7f23 in tc_export (opt=13, para1=0xbfffe964, para2=0x80a4008) at export_xvid4.c:374 #4 0x0805fddf in tcv_export (opt=13, para1=0xbfffe964, para2=0x80a4008) at dl_loader.c:66 #5 0x0805f3f1 in encoder (vob=0x80a4008, frame_a=0, frame_b=2147483647) at encoder.c:734 #6 0x0805923e in main (argc=9, argv=0xbffff554) at transcode.c:4014 (gdb) x/8i $pc 0xb5280a17 <MBCoding+5639>: mov %eax,(%edi) 0xb5280a19 <MBCoding+5641>: add $0x4,%edi 0xb5280a1c <MBCoding+5644>: subl $0x20,0xc(%ecx) 0xb5280a20 <MBCoding+5648>: mov %edi,0xffffff68(%ebp) 0xb5280a26 <MBCoding+5654>: mov %edi,0x10(%ecx) 0xb5280a29 <MBCoding+5657>: movl $0x0,0x8(%ecx) 0xb5280a30 <MBCoding+5664>: mov 0x14(%ebp),%eax 0xb5280a33 <MBCoding+5667>: mov $0x20,%ecx
rhaaa you should stab me... on every box I had tested yasm was installed and xvid defaults to it for compiling asm... I didn't pay attention to that, even if you said that nasm is required to compile it, it doesn't fail to compile, but it seems that yasm computes wrong value for addresses... I encountered that in x264 aswell. If I remove yasm from the box and recompile xvid, it seems to work fine. Does anybody has a clue if that's a yasm bug or something else ?
most likely to be a yasm bug : http://www.tortall.net/projects/yasm/ticket/114 will test again asap
Created attachment 132669 [details, diff] textrel patch for xvid-1.1.3 just a simple forward port, no changes were needed. the transcode test from comment #32 ran fine here. can this go into portage?
(In reply to comment #36) > Created an attachment (id=132669) [edit] > textrel patch for xvid-1.1.3 > > just a simple forward port, no changes were needed. the transcode test from > comment #32 ran fine here. can this go into portage? thanks, it's fine with yasm 0.6.2 and transcode, but I still get the segfault with qpel: mencoder foo.avi -o toto.avi -ovc xvid -oac lavc -xvidencopts bitrate=1000:qpel -> seg fault mencoder foo.avi -o toto.avi -ovc xvid -oac lavc -xvidencopts bitrate=1000 -> video is fine some debugging: Program received signal SIGSEGV, Segmentation fault. 0xf7e74ff4 in ?? () from /usr/lib/libxvidcore.so.4 (gdb) bt full #0 0xf7e74ff4 in ?? () from /usr/lib/libxvidcore.so.4 No symbol table info available. #1 0xf7e25609 in xvid_me_interpolate16x16qpel (x=-134516368, y=-136816742, dir=4156896560, data=0xf7d857ba) at ../../src/motion/estimation_common.c:188 Reference = ( uint8_t *) 0xf7edf1b4 "\n+���_���\r������Zp��kS��\220\027��\220����4���:���\b��@����D���8������\220\034��07��8\037���6��P���\200���\224Z��p>� iEdgedWidth = 4158930000 rounding = 4158323644 halfpel_x = -135897344 halfpel_y = -136717808 ref1 = (const uint8_t *) 0xf7e9f104 "�S��\020?� ref2 = (const uint8_t *) 0xf7d9ec1e "U\211�WV\203�0\213E$@\203��\211E�\213E�\215\024\205" ref3 = (const uint8_t *) 0xf7edf208 "\224Z��p>� ref4 = (const uint8_t *) 0xf7e6d100 " " Backtrace stopped: previous frame inner to this frame (corrupt stack?) (gdb) x/8i $pc 0xf7e74ff4: pop %eax 0xf7e74ff5: or %ecx,(%edi) 0xf7e74ff7: add %dh,%al 0xf7e74ff9: int3 0xf7e74ffa: stc 0xf7e74ffb: divl 0x71(%eax) 0xf7e74ffe: sti 0xf7e74fff: negl 0x30f7d857(%edx) (gdb) i r eax 0xf762ef00 -144511232 ecx 0x8c6e740 147253056 edx 0x350 848 ebx 0xffa423e8 -6020120 esp 0xffa4236c 0xffa4236c ebp 0xf7e74ff4 0xf7e74ff4 esi 0x0 0 edi 0x0 0 eip 0xf7e74ff4 0xf7e74ff4 eflags 0x210246 [ PF ZF IF RF ID ] cs 0x23 35 ss 0x2b 43 ds 0x2b 43 es 0x2b 43 fs 0x0 0 gs 0x63 99 without the patch everything is fine
Created attachment 133487 [details, diff] textrel patch for xvid-1.1.3 thanks for the report, i fixed a few bugs i could reproduce.
can anyone else try it/take a look please?
I've added the patch to xvid 1.1.3-r1; big thanks for the huge work on it. Could you please try to send it upstream also ? I doubt I'll be able to port it if the asm part changes in future versions
(In reply to comment #40) > Could you please try to send it upstream also ? http://list.xvid.org/pipermail/xvid-devel/2007-December/005717.html > I doubt I'll be able to port it if the asm part changes in future versions i'll do it as time permits, no worries (maybe email me before a release as i'm not following this and other similar packages on a day-to-day basis).
*** Bug 206982 has been marked as a duplicate of this bug. ***
