[ccache] Re: ccache segfault

[Lapo Luchini]

Martin Pool <mbp <at> sourcefrog.net> writes:

> If you can reproduce it, would you please get a backtrace from the core
> file and post it here?  Something like this:
> 
>   % gdb `which ccache` core
>   gdb> backtrace full

(gdb) bt full
#0  0x00000008008025d6 in memcpy () from /lib/libc.so.6
No symbol table info available.
#1  0x0000000000403f9c in x_realloc (ptr=0x514800, size=2056) at util.c:195
        p2 = (void *) 0x513000
#2  0x00000000004044c2 in args_add (args=0x512040, 
    s=0x7fffffffe1c4 "p12_key.So") at args.c:40
No locals.
#3  0x0000000000404546 in args_init (init_argc=200, init_args=0x7fffffffcdd0)
    at args.c:32
        args = (ARGS *) 0x512040
        i = 200
#4  0x00000000004029c4 in main (argc=455, argv=0x7fffffffc5d0) at ccache.c:564
        mask = 2
        p = 0x2 <Address 0x2 out of bounds>
(gdb)

PS: I can reproduce it anytime and am willing to help with any more tests.
PPS: I see those functions use sizeof(char*) and not 4.. so AFAIK should be
already 64bit-aware, aren't they?