Tried with 1.2.8 with --enable-threads (we've always used this on the
1.2.6 instances, both on amd64 + i386) and got an identical backtrace.
I then recompiled without --enable-threads and got this backtrace:
Starting program: /usr/bin/memcached -m 3200 -p 11211 -u nobody -c 4096
-t 10
[Thread debugging using libthread_db enabled]
[New Thread 0xb7db76b0 (LWP 24582)]
[New Thread 0xb7d55b90 (LWP 24585)]
[New Thread 0xb7554b90 (LWP 24586)]
[New Thread 0xb6d53b90 (LWP 24587)]
[New Thread 0xb6552b90 (LWP 24588)]
[New Thread 0xb5d51b90 (LWP 24589)]
[New Thread 0xb5550b90 (LWP 24590)]
[New Thread 0xb4d4fb90 (LWP 24591)]
[New Thread 0xb454eb90 (LWP 24592)]
[New Thread 0xb3d4db90 (LWP 24593)]
[New Thread 0xb354cb90 (LWP 24594)]
calloc()
calloc()
calloc()
calloc()
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xb354cb90 (LWP 24594)]
0x0804a061 in add_iov (c=0x8b646f8, buf=0x8c19588, len=44) at
memcached.c:683
683 memcached.c: No such file or directory.
in memcached.c
(gdb) bt
#0 0x0804a061 in add_iov (c=0x8b646f8, buf=Cannot access memory at
address 0xb354c224
) at memcached.c:683
Cannot access memory at address 0xb354c23c
(gdb)
I don't understand the memory error - memcache15:~# cat /proc/meminfo
MemTotal: 3310796 kB
and the system's only using 50M with 50M cached, so enough headroom...
gdh
