| vibhav | good motning | 02:24 |
|---|---|---|
| vibhav | morning, even | 02:41 |
| === sraue_ is now known as sraue | ||
| twpoland | hello | 08:34 |
| === tlyu_ is now known as tlyu | ||
| === doko_ is now known as doko | ||
| === tkamppeter__ is now known as tkamppeter | ||
| === security is now known as fire | ||
| ositake | hi, I just found a problem in Ubuntu 12.04 and I don't know if it can be consider as a bug | 18:29 |
| ositake | the problems is that if you issue strace -ttt -f -p [pid from the xorg] | 18:29 |
| ositake | (don't try | 18:29 |
| ositake | it freezes the complete Ubuntu | 18:30 |
| ositake | meaning, you need to hard reset to work | 18:30 |
| ositake | but if I do the same from the any console | 18:30 |
| ositake | like ctrl+alt+F1 | 18:30 |
| ositake | works smooth | 18:30 |
| geofft | that's disallowed by default because of the ptrace_scope stuff, right? | 18:31 |
| ositake | that's what I don't know | 18:31 |
| ositake | I guess that it is kind of stupid to strace the xorg | 18:31 |
| ositake | when you are in | 18:31 |
| ositake | but in a way, you should be able to do it | 18:32 |
| ositake | still I can't say | 18:32 |
| geofft | that too, I'd call that expected behavior. :-) | 18:32 |
| ogra_ | well, if you call it like above you will output to the xterm you run in ... meaning you strace your own command as well ... sounds like a deadly loop you force there | 18:32 |
| ositake | indeed | 18:32 |
| geofft | I bet if you write a shell script to strace xorg with output to a file, and then kill the strace process after 30 seconds | 18:32 |
| ogra_ | try redirecting to a file instead of oututting directly | 18:32 |
| geofft | it won't actually hang the system | 18:32 |
| ogra_ | *outputting | 18:32 |
| ositake | good idea | 18:32 |
| ogra_ | (since you would be stracing the strace output ... ) | 18:33 |
| ositake | I know is something that a normal user would not do | 18:35 |
| ositake | that's why I wanted to ask before submitting any bug on that topoc | 18:35 |
| ositake | topic | 18:35 |
| ositake | thanks for the tip, by forwarding to a file, it doesn't happen, | 18:39 |
| geofft | I vaguely recall that tcpdump or tshark or something is clever enough to notice if you're over ssh | 18:39 |
| geofft | and not try to dump its own output | 18:40 |
| geofft | I wonder if it's worth doing the same thing for strace | 18:40 |
| geofft | but I can't really figure out how. | 18:40 |
| ositake | that could be a good idea | 18:40 |
| infinity | It's not worth trying to save people from applying that gun to that foot, I don't think. | 18:40 |
| infinity | "Infinite loops make computers appear frozen" isn't really a bug, unless it's a loop you reach in normal use. | 18:41 |
| ositake | true infinity , that's why I wanted to ask before hand | 18:41 |
| infinity | This is all reminiscint of how, every few years, someone freaks out because you can DoS a UNIX system by forkbombing. Oooo. | 18:41 |
| infinity | The answer, of course, being twofold: "don't do that then" and "limit processes for non-root, if you care". | 18:42 |
| geofft | I thought that's no longer a problem with cgroups or something. :) | 18:42 |
| BenC | infinity: Can I get some buildd love for linux-ppc? | 18:42 |
| infinity | And in the ptrace case, we don't let regular users ptrace, and we really don't let them ptrace root-owned Xorg servers. | 18:42 |
| infinity | BenC: Maaaaybe. | 18:42 |
| infinity | BenC: adare and ross are manual, go nuts. | 18:43 |
| BenC | Thanks | 18:43 |
| geofft | I'm reminded of http://blog.michaeltang.me/yes-yes-no/ | 18:43 |
| ositake | need to leave, thanks for the clarification! | 18:44 |
| ositake | cheers | 18:44 |
| geofft | (and in particular how _two_ different users on a shared system I run decided it sounded like a good idea to run) | 18:44 |
| BenC | Now I just need some distro-manager love | 18:47 |
| infinity | BenC: Release, you mean? I can do that. :P | 18:57 |
| BenC | infinity: ah, the email says "distro manager approval"…I wasn't sure who to blame^Wask | 18:58 |
| infinity | BenC: There really was no ABI bump for PPC in that rebase? | 18:58 |
| infinity | That seems suspicious. | 18:58 |
| * ScottK hands infinity a neck tie, since he's distro manager, he should dress the part. | 19:03 | |
| infinity | ScottK: Not gonna happen. | 19:04 |
| infinity | BenC: *poke* | 19:04 |
| ScottK | ;-) | 19:04 |
| infinity | BenC: Are you sure the rebase to 3.8.4 didn't bump ABI? You don't have the abi check disabled? | 19:04 |
| BenC | infinity: I ran the build and no ABI bumps for me | 20:27 |
| BenC | infinity: debian.ppc/rules.d/powerpc.mk:skipabi = false | 20:28 |
| === mbiebl_ is now known as mbiebl | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!