Bug#179647: More info from strace
Hi, Branden!
Following up on the mail to the Xfree86 ML i posted earlier...
The strace output after the freeze has happened:
===========================================================
--- SIGALRM (Alarm clock) @ 0 (0) ---
sigreturn() = ? (mask now [])
--- SIGALRM (Alarm clock) @ 0 (0) ---
sigreturn() = ? (mask now [])
===========================================================
and so on ad infinitum...
That obviously doesn't make a whole lot of sense...
Next time I'll also capture the time-of-day of each signal. I forot that but
it would be interesting.
Here's the output while normal operation:
===========================================================
14:27:51.202001 --- SIGALRM (Alarm clock) @ 0 (0) ---
14:27:51.202168 setitimer(ITIMER_REAL, {it_interval={0, 0}, it_value={0, 0}},
NULL) = 0
14:27:51.202208 sigreturn() = ? (mask now [IO])
14:27:51.202277 gettimeofday({1044365271, 202288}, NULL) = 0
14:27:51.202309 gettimeofday({1044365271, 202319}, NULL) = 0
14:27:51.202343 select(256, [1 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
25 26 27 28 29], NULL,
NULL, {110, 560000}) = 1 (in [17], left {110, 550000})
14:27:51.212345 setitimer(ITIMER_REAL, {it_interval={0, 20000}, it_value={0,
20000}}, NULL) = 0
14:27:51.212394 read(17, "&\6\2\0;\0\0\0", 4096) = 8
14:27:51.212469 read(17, 0x4021e008, 4096) = -1 EAGAIN (Resource temporarily
unavailable)
14:27:51.212516 writev(17, [{"\1\1\r
\0\0\0\0;\0\0\0\324\t\0\1J\4\10\4J\4\10\4\f\0\0"..., 32}], 1) =
32
14:27:51.212610 gettimeofday({1044365271, 212621}, NULL) = 0
14:27:51.212645 select(256, [1 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
25 26 27 28 29], NULL,
NULL, {110, 550000}) = 1 (in [17], left {110, 550000})
14:27:51.212764 read(17, "(\6\4\0;\0\0\0p\2`\1J\4\10\4", 4096) = 16
14:27:51.212815 read(17, 0x4021e008, 4096) = -1 EAGAIN (Resource temporarily
unavailable)
14:27:51.212847 writev(17, [{"\1\1\16
\0\0\0\0\0\0\0\0&\4\6\4\0\0\0\0\10\340!@\0\0\0"..., 32}], 1) =
32
14:27:51.213589 gettimeofday({1044365271, 213610}, NULL) = 0
14:27:51.213643 select(256, [1 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
25 26 27 28 29], NULL,
NULL, {110, 549000}) = 1 (in [17], left {110, 550000})
14:27:51.213788 read(17,
"\232\6\5\0\316\4`\1\0\0\0\0\0\0\0\0\32\6\32\0;\3\5\0\251"..., 4096) = 812
14:27:51.214867 read(17, 0x4021e008, 4096) = -1 EAGAIN (Resource temporarily
unavailable)
14:27:51.214942 gettimeofday({1044365271, 214953}, NULL) = 0
14:27:51.214982 select(256, [1 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
25 26 27 28 29], NULL,
NULL, {110, 548000}) = ? ERESTARTNOHAND (To be restarted)
14:27:51.232319 --- SIGALRM (Alarm clock) @ 0 (0) ---
===========================================================
The most interesting thing is the setitimer call right after the SIGALRM.
So perhaps it's just something wrong with the sighandler...
Well, I also istalled the xserver-xfree86-dbg package but unfortunately it
doesn't work with the nvidia binary only driver. I'll try with the nv driver
and see if it also freezes with that driver.
Mika
Reply to: