squid-1287

Version:

2.5

How it is diagnosed (reproduced or source analysis)?

We reproduced the failure. Just run “squid -NsYC’ would trigger the failure.

Bug link:

http://bugs.squid-cache.org/show_bug.cgi?id=1287

Symptom:

Segmentation Fault. When ‘squid -NsYC’

Root cause:

Null pointer dereference.

(gdb) run
Starting program: /usr/sbin/squid -NsYC
[Thread debugging using libthread_db enabled]
[New Thread -1209899488 (LWP 3808)]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread -1209899488 (LWP 3808)]
storeOffset (sio=0x0) at store_io.c:102
102         return sio->offset;
(gdb) bt
#0  storeOffset (sio=0x0) at store_io.c:102
#1  0x080b3398 in storeClientCopy2 (e=0xee60610, sc=0xee60610) at
store_client.c:255
#2  0x08065530 in clientWriteComplete (fd=67, bufnotused=0x0, size=4096,
errflag=0, data=0xf7838b0) at client_side.c:2258
#3  0x08068a8a in CommWriteStateCallbackAndFree (fd=67, code=0) at comm.c:99
#4  0x0806b887 in comm_poll (msec=10) at comm_select.c:459
#5  0x0809497a in main (argc=0, argv=0x1) at main.c:760
(gdb)

Is there Error Message?

YES. “FATAL: Received Segment Violation...dying.”

Can Errlog anticipate the error msg?

Yes. This falls into the exception signal pattern.