Normally seg faults aren't a problem for me, but this is really ticking me off now. Here's the console dump: Code: Auto detecting CPU Using SSE2 Optimised binary. Enabling debug mode Server will auto-restart if there is a crash. Console initialized. Game.dll loaded for "Empires" ./srcds_run: line 352: 26997 Segmentation fault (core dumped) $HL_CMD cat: hlds.26980.pid: No such file or directory email debug.log to linux@valvesoftware.com Tue Mar 22 10:50:40 EST 2011: Server restart in 10 seconds I'm running on CentOS 5. The really weird thing is that after opening a can of -verify_all on this shit, the server worked fine...for about an hour, then it started doing this again while the server was running. I can't get the server back up now, even after uninstalling addons and downloading the OB binary files over and again.
I'm not the best at linux, but here's what I think. "cat: hlds.26980.pid: No such file or directory" now as far as I remember - in linux terms refers to a process.. It could be that it doesn't have the rights to spawn the process. "attempting to access memory the program does not own." would be unlikely but the programing related ones do seem to be more likely. "attempting to execute a program that does not compile correctly" doesn't seem like it (the compiler should show an error) nor does a buffer overflow (the programmer would see that easy on a test build). Reading into this "Segmentation fault", seems to be memory related if it's happening time and time again. Maybe your memory is corrupted? You may want to try memtest (can't remember it exactly) that comes with the Kubuntu/Ubuntu livecds and boot discs. As with the downloads, could there be a fault in the download? Maybe it's corrupted somehow and it's not downloading all the data or something? Again, I'm not a great linux guy but that's what I'm finding
In my experience with SRCDS, I've only ever had segfaults when there are problems with the binaries. If not the OB binaries, I'd be willing to assume the Empires binaries are broken. I certainly wouldn't put it past me that my box is just being stupid though; it does it a lot. That's the point of -verify_all. It makes sure that all HLDS files are up to date.
