<<< Date Index >>>     <<< Thread Index >>>

Re: [Mutt] #3074: Mutt hangs indefinitely



#3074: Mutt hangs indefinitely

Old description:

> On many occasions, I find that an unattended mutt (1.5.17+20080114-1+b1
> from Debian) will hang indefinitely. Never having time to debug it, I
> usually kill it and restart it, but today I let it run in strace until it
> happend. Here's what the end of the strace looks like - it seems to get
> an ERESTARTSYS while reading from the imaps socket! ERESTARTSYS is
> reserved as a signal between drivers and the kernel signal handling code
> and is, theoretically, never supposed to make it to userspace. Thus, this
> is quite odd.
>
> I'm happy to add ltraces or whatever else may be useful....
>
> write(1, "\r\33[17B\33[37m\33[40mFetching message"..., 76) = 76
> write(1, "\33[79;29H\33[37m\33[40m1834/1838 (99%"..., 54) = 54
> gettimeofday({1212696193, 854853}, NULL) = 0
> getrusage(RUSAGE_SELF, {ru_utime={0, 812050}, ru_stime={0, 236014}, ...})
> = 0
> time(NULL)                              = 1212696193
> times({tms_utime=81, tms_stime=23, tms_cutime=1, tms_cstime=2}) =
> 1734919932
> gettimeofday({1212696193, 855156}, NULL) = 0
> getrusage(RUSAGE_SELF, {ru_utime={0, 812050}, ru_stime={0, 236014}, ...})
> = 0
> time(NULL)                              = 1212696193
> times({tms_utime=81, tms_stime=23, tms_cutime=1, tms_cstime=2}) =
> 1734919932
> send(4, "\27\3\1\0\373*\341\374\221\t.p\233T*Y\354\0\223\303><\205"...,
> 256, 0) = 256
> _llseek(5, 0, [0], SEEK_SET)            = 0
> recv(4, "\27\3\1\v\33", 5, 0)           = 5
> recv(4, "x\204\315|I\203Ph`\317\205-\356\16L\315S\3551\263\302N"...,
> 2842, 0) = 1383
> recv(4, "\246\212O\2675gm\364wR!\16\31\256\361\374\373\21\352u#"...,
> 1459, 0) = 1388
> recv(4, 0x81d31d0, 71, 0)               = ? ERESTARTSYS (To be restarted)
> --- SIGWINCH (Window changed) @ 0 (0) ---
> sigreturn()                             = ? (mask now [])
> recv(4,  <unfinished ...>
>

> - Phil

New description:

 On many occasions, I find that an unattended mutt (1.5.17+20080114-1+b1
 from Debian) will hang indefinitely. Never having time to debug it, I
 usually kill it and restart it, but today I let it run in strace until it
 happend. Here's what the end of the strace looks like - it seems to get an
 ERESTARTSYS while reading from the imaps socket! ERESTARTSYS is reserved
 as a signal between drivers and the kernel signal handling code and is,
 theoretically, never supposed to make it to userspace. Thus, this is quite
 odd.

 I'm happy to add ltraces or whatever else may be useful....
 {{{
 write(1, "\r\33[17B\33[37m\33[40mFetching message"..., 76) = 76
 write(1, "\33[79;29H\33[37m\33[40m1834/1838 (99%"..., 54) = 54
 gettimeofday({1212696193, 854853}, NULL) = 0
 getrusage(RUSAGE_SELF, {ru_utime={0, 812050}, ru_stime={0, 236014}, ...})
 = 0
 time(NULL)                              = 1212696193
 times({tms_utime=81, tms_stime=23, tms_cutime=1, tms_cstime=2}) =
 1734919932
 gettimeofday({1212696193, 855156}, NULL) = 0
 getrusage(RUSAGE_SELF, {ru_utime={0, 812050}, ru_stime={0, 236014}, ...})
 = 0
 time(NULL)                              = 1212696193
 times({tms_utime=81, tms_stime=23, tms_cutime=1, tms_cstime=2}) =
 1734919932
 send(4, "\27\3\1\0\373*\341\374\221\t.p\233T*Y\354\0\223\303><\205"...,
 256, 0) = 256
 _llseek(5, 0, [0], SEEK_SET)            = 0
 recv(4, "\27\3\1\v\33", 5, 0)           = 5
 recv(4, "x\204\315|I\203Ph`\317\205-\356\16L\315S\3551\263\302N"..., 2842,
 0) = 1383
 recv(4, "\246\212O\2675gm\364wR!\16\31\256\361\374\373\21\352u#"..., 1459,
 0) = 1388
 recv(4, 0x81d31d0, 71, 0)               = ? ERESTARTSYS (To be restarted)
 --- SIGWINCH (Window changed) @ 0 (0) ---
 sigreturn()                             = ? (mask now [])
 recv(4,  <unfinished ...>
 }}}

 - Phil

-- 
Ticket URL: <http://dev.mutt.org/trac/ticket/3074#comment:5>