Hej,
after latest upgrade, exim throws SIGSEGV around sometimes.
In strace this looks like:
...
open("/var/spool/exim4/input/8/1BJ6e8-0001V8-BA-D",
O_RDWR|O_APPEND|O_LARGEFILE) = 3
fcntl64(3, F_GETFD) = 0
fcntl64(3, F_SETFD, FD_CLOEXEC) = 0
fcntl64(3, F_SETLK64, {type=F_WRLCK, whence=SEEK_SET, start=0, len=19},
0xbfffeda0) = 0
fstat64(3, {st_mode=S_IFREG|0640, st_size=28934, ...}) = 0
open("/var/spool/exim4/input/8/1BJ6e8-0001V8-BA-H",
O_RDONLY|O_LARGEFILE) = 4
fstat64(4, {st_mode=S_IFREG|0640, st_size=4857, ...}) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS,
-1, 0) = 0x40018000
read(4, "1BJ6e8-0001V8-BA-H\nroot 0 0\n<>\n1"..., 4096) = 4096
time(NULL) = 1083229436
read(4, "4.inode.at with smtp (Exim 4.31)"..., 4096) = 761
read(4, "", 4096) = 0
close(4) = 0
munmap(0x40018000, 4096) = 0
open("/var/spool/exim4/input/8/1BJ6e8-0001V8-BA-J",
O_RDONLY|O_LARGEFILE) = -1 ENOENT (No such file or directory)
time(NULL) = 1083229436
open("/etc/localtime", O_RDONLY) = 4
fstat64(4, {st_mode=S_IFREG|0644, st_size=801, ...}) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS,
-1, 0) = 0x40018000
read(4, "TZif\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\6\0\0\0\6\0"...,
4096) = 801
close(4) = 0
munmap(0x40018000, 4096) = 0
--- SIGSEGV (Segmentation fault) @ 0 (0) ---
In gdb it gives following backtrace:
(gdb) run -M 1BJ6e8-0001V8-BA
...
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 16384 (LWP 24434)]
0x403c2b5f in strlen () from /lib/libc.so.6
(gdb) bt
#0 0x403c2b5f in strlen () from /lib/libc.so.6
#1 0x0805fdd6 in deliver_msglog ()
#2 0x0806891f in deliver_message ()
#3 0x0806f83b in main ()
(gdb)
Happens to about 5 out of 1 Million messages.
duh.
daniel
