Version: 2.1.0-2 Hi,
Niels Thykier wrote (23 Jun 2016 21:30:52 GMT) : > intrigeri: >> syscall numbers are architecture dependent, so: what architecture did >> you experience this on? > amd64. Thanks. So, that would be the keyctl syscall: https://sources.debian.net/src/linux/3.16.7-ckt11-1%2Bdeb8u6~bpo70%2B1/arch/x86/syscalls/syscall_64.tbl/#L259 I don't see it allowed in torsocks source tree. >> (I can't reproduce this here with torsocks 2.1.0-2 on sid.) > True, the particular system call seem to have been fixed. Cool :) Now, I see no relevant difference before 2.1.0-1 and 2.1.0-2, so I don't understand why this problem has disappeared in the meantime. Do you mind if I close this bug report anyway, until someone reproduces it again, maybe, some day? > However, I can reproduce it with syscall "204" (sched_getaffinity). A > particularly > exciting case would be "torsocks mplayer" because mplayer apparently > does not give up making it an infinite fprintf loop of warnings. That's https://bugs.debian.org/805741, which has been fixed in upstream Git yesterday, so it'll be fixed in 2.2 final. Cheers, -- intrigeri