On Wed, Apr 15, 2009 at 11:34 AM, Alan McKinnon <[email protected]> wrote: > On Wednesday 15 April 2009 20:31:09 Mark Knecht wrote: >> Note that mythfrontend started segfaulting on an intel graphics >> machine when I went to the newer xorg. Unfortunately it did not stop >> segfaulting when I went back. I cannot explain this yet. > > When you reverted back to the prior working versions, did you also revert back > to the prior working *configuration files*? > > -- > alan dot mckinnon at gmail dot com
I believe I did but there's always a chance for mistakes. This is getting more confusing all the time. I've now gone back and forth between old and new a couple of times and there are other issues showing up with even building the older versions anymore. I cannot even get the older intel driver to build on this morning's pass with the older xorg-server, so I've gone to the new xorg again just so my wife can use the system. It definitely built last evening. 1) With the old xorg I used xorg.conf, keyboard, mouse and intel in make.conf. (I think it was intel and not i810 in make.conf.) I used i810 in xorg.conf. 2) With the new xorg I renamed xorg.conf to xorg.conf.old and used the default automatically provided by xorg and hald. In make.conf I used evdev and intel. 3) I believe hald was enabled in both configurations, but possibly it was turned off with the older xorg. Don't know any way to check for sure. With the old xorg the i810 driver got loaded from xf86-video-intel-2.1.1, at least according to my emerge.log file that was what was on the system prior to recent updates. I know for certain that the driver in memory was i810. With the new xorg the i915 driver got loaded from xf86-video-intel-2.6.3.r1. As far as I know both times the dri flag was set when the above packages were compiled. This is all pretty messed up at this point. for the day anyway I seem to be forced to use the newer xorg-server and mythfrontend still segfaults. I see a couple of messages in different log files, all relating to being unable to pin an xv buffer. The exact message was: (EE) intel(0): Failed to pin xv buffer With another line saying it had segfault. There was no backtrace of any kind. The second post in the following thread sounds similar to my results although he was able to get things working and he's using 1 version newer on xorg-server (1.6) than we are talking about here: http://forums.freebsd.org/archive/index.php/t-3222.html At this point I'm lost and with trading to do and taxes to get out today I'm buried. Maybe more tomorrow when things calm down. Thanks, Mark
