is that chongqing in sichuan? i went through there on the way to chengdu!
Oh, yes, it belongs to sichuan several years before, now it 's under directly the center govornment, chengdu is really near by chongqing and a good city for visit. If you come through it again, don't miss to visit my city, haha, I will certainly try to be a good host.
I still suffer from the problem, but now I do find it interesting if I display movie in the full screen mode, blue screen is gone, after several switch in full and windows mode, I met no blue screen any more until I play movie again. Sometime it's gone seemingly forever, but sometimes it comes back like a old bad friend. I have DRI enabled in the xorg.conf.
]]>xine was working fine with xv for me all the time. then it suddenly stopped. i did some thinking and realized i had enabled dri and glx to allow me to play tuxracer SINCE last using xine.
i just experimented:
i turned off glx and dri and xine with xvworked fine
i turned dri on and glx off and xine with xvworked fine
i turned glx on and dri off and xine with xvworked fine
i turned them both on (i.e. back to what i had before) and now xine works fine and i can use xv as root and user
then i started xine again as user and it stopped working again - might that provide some clues?
]]>Has the new version of totem changed things? I havn't upgraded and I'm not at my linux PC (face reddens, lol).
I should try making a non-module laced kernel. It is pretty bare anyway.
I really wish totem was solid and reliable.
]]>my customized kernel had almost everything turned off excecpt for stuff i needed...and stuff i needed were compiled in and not modules.
so does anyone here know the things in the kernel might affect xv?
]]>I test drove slackware to see if it was arch specific, the problem was there as well. I tried all manner of xorg.conf settings for the nvidia card, and sometimes it seemed to have been solved but came back.
It is easiest to recreate if you reboot, startx and load totem straight away then press play. I find that if you have been web browsing or doing other tasks first there is likely no problem. The problem only potentially happens on the first run of the program as well.
One plus point was I stopped using AGPGART and my GLXGEARS performance went up so it wasn't all wasted.
The nvidia card in question is a FX5200 with 128MB. I have just switched back to an old GF4 MX 64MB and the problem seems to have disappeared (for now.... evil laughter). Even stranger my glxgears score has gone up 20 points with the old dusted off card. LOL.
What cards are people using and what is their mileage with totem? I'm tempted to look for an old GF4 ti4x00. The new features of the fx series seem pointless in linux.
]]>no need for it :
http://bugs.archlinux.org/index.php?do=details&id=1090i already did it.
thx
]]>i already did it.
]]>Instead of making the symlinks, I modified the PKGBUILD from ABS to patch the configure script that came with xine-lib, so it would use libXv.so and not give up on Xv support because the .a files aren't there.
--- xine-lib-1-rc5-orig/configure 2004-06-26 12:03:25.000000000 -0400 +++ xine-lib-1-rc5/configure 2004-06-26 12:06:43.000000000 -0400 @@ -26156,7 +26156,7 @@ fi echo "$as_me:$LINENO: result: yes" >&5 echo "${ECHO_T}yes" >&6 - xv_lib="libXv.a" + xv_lib="libXv.so" lt_min_full=1.4.0 @@ -26176,7 +26176,7 @@ fi echo "$as_me:$LINENO: result: yes" >&5 echo "${ECHO_T}yes" >&6 - xvmc_lib="libXvMC.a" + xvmc_lib="libXvMC.so" lt_min_full=1.4.0
And yes, it is magnitudes faster. I couldn't run xine on my athlon 700 before this. Now it runs super.
post a feature-request @ http://bugs.archlinux.org for xine-libs for this change explaining why it is needed and i'm sure the maintainer of xine-libs will include it
]]>You can also make you own patch file using diff.
Copy the original configure to another new directory. Then edit configure...
Find the line xv_lib="libXv.a" and change it to "libXv.so"
Find the line xvmc_lib="libXvMC.a" and change it to "libXvMC.so"
Then diff -Naur orig new > xine-lib.patch
]]>patching file configure
patch unexpectedly ends in middle of line
Hunk #2 succeeded at 26176 with fuzz 1.
What's wrong? Could you someone tell me, I copy fbidwell's code and save as xine-lib.xv.patch, there is my PKGBUILD:
# $Id: PKGBUILD,v 1.20 2004/06/23 17:25:07 dorphell Exp $
# Maintainer: dorphell <dorphell@archlinux.org>
# Committer: Judd Vinet <jvinet@zeroflux.org>
pkgname=xine-lib
pkgver=1rc5
pkgrel=2
pkgdesc="A free video player for Unix"
depends=('libpng' 'libvorbis' 'sdl' 'codecs')
source=(http://dl.sourceforge.net/sourceforge/xine/$pkgname-1-rc5.tar.gz xine-lib.xv.patch)
url="http://xinehq.de/"
md5sums=('703c3e68d60524598d4d9e527fe38286' '57b9f79ea4702453b6004e9fe0853e7b')build() {
cd $startdir/src/$pkgname-1-rc5
# make xine lib to use libXv.so
patch -Np1 -i ../xine-lib.xv.patch
./configure --prefix=/usr --with-w32-path=/usr/lib/win32
--enable-alsa --disable-faad
make || return 1
make DESTDIR=$startdir/pkg install
}
~
I have compiled it under gcc3.4.1 but only with root, unable to finish compilation under normal user account .
]]>