I think I found the problem, seems to work now. Will keep it running for 10-20 min to confirm. On Thu, Oct 3, 2013 at 11:44 AM, Ildar Muslukhov <ildarm@xxxxxxxxxx> wrote: > Will take a look into it. > > On Thu, Oct 3, 2013 at 11:33 AM, Dave Jones <davej@xxxxxxxxxx> wrote: >> On Wed, Oct 02, 2013 at 04:46:41PM -0700, Ildar Muslukhov wrote: >> > >> > This patch removes makes the loop in child_random_syscalls aware of which >> > system calls are enabled, thus, avoiding unnecessary looping. >> > >> > Signed-off-by: Ildar Muslukhov <ildarm@xxxxxxxxxx> >> >> I knew I should have tested this before merging it.. >> >> This seems all kinds of broken: >> >> $ ./trinity -l off -q >> Trinity v1.3pre Dave Jones <davej@xxxxxxxxxx> >> Done parsing arguments. >> Marking all syscalls as enabled. >> [init] 32-bit syscalls: 351 enabled, 0 disabled. 64-bit syscalls: 314 enabled, 0 disabled. >> [init] Using pid_max = 32768 >> [init] Started watchdog process, PID is 29722 >> [watchdog] Watchdog is alive. (pid:29722) >> [main] Main thread is alive. >> [main] 375 sockets created based on info from socket cachefile. >> [main] Generating file descriptors >> [main] Added 48 filenames from /dev >> [main] Added 28849 filenames from /proc >> [main] Added 14258 filenames from /sys >> [child3:29727] uselib (134) returned ENOSYS, marking as inactive. >> [child3:29729] ni_syscall (generic) (185) returned ENOSYS, marking as inactive. >> [child3:29729] nfsservctl (180) returned ENOSYS, marking as inactive. >> [child3:29729] nfsservctl (180) returned ENOSYS, marking as inactive. >> [child3:29729] ni_syscall (generic) (182) returned ENOSYS, marking as inactive. >> [child-1:29729] ni_syscall (generic) (177) returned ENOSYS, marking as inactive. >> [child0:29724] ni_syscall (generic) (215) returned ENOSYS, marking as inactive. >> [child1:29725] ni_syscall (generic) (182) returned ENOSYS, marking as inactive. >> [child2:29726] ni_syscall (generic) (177) returned ENOSYS, marking as inactive. >> [child3:29732] ni_syscall (generic) (214) returned ENOSYS, marking as inactive. >> [child1:29739] ni_syscall (generic) (211) returned ENOSYS, marking as inactive. >> [child1:29739] ni_syscall (generic) (181) returned ENOSYS, marking as inactive. >> [child1:29739] ni_syscall (generic) (185) returned ENOSYS, marking as inactive. >> [child2:29742] ni_syscall (generic) (205) returned ENOSYS, marking as inactive. >> [child1:29750] ni_syscall (generic) (181) returned ENOSYS, marking as inactive. >> [child1:29752] ni_syscall (generic) (236) returned ENOSYS, marking as inactive. >> [child1:29754] ni_syscall (generic) (181) returned ENOSYS, marking as inactive. >> [child2:29755] ni_syscall (generic) (211) returned ENOSYS, marking as inactive. >> [child2:29755] ni_syscall (generic) (185) returned ENOSYS, marking as inactive. >> [child1:29756] ni_syscall (generic) (185) returned ENOSYS, marking as inactive. >> [child3:29757] ni_syscall (generic) (215) returned ENOSYS, marking as inactive. >> [child1:29758] BUG!: CHILD (pid:29758) GOT REPARENTED! parent pid:29723. Watchdog pid:29722 >> [child1:29758] BUG!: Last syscalls: >> [child1:29758] [0] pid:29733 call:fcntl callno:5 >> [child1:29758] [1] pid:29758 call:timer_create callno:9 >> [child1:29758] [2] pid:29755 call:fchownat callno:24 >> [child1:29758] [3] pid:29757 call:fchdir callno:10 >> [child-1:29757] child exiting. >> >> >> 1. "child-1" looks suspect >> 2. There's a lot more ni_syscall's there than there should be. >> 3. The reparent thing. >> Running with -D shows lots of segv's. >> >> >> Dave >> >> ` -- To unsubscribe from this list: send the line "unsubscribe trinity" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
