Bug #386
closedIs this a bug in ld-elf.so.2?
0%
Description
While debugging an error in evolution-2.8, I ran across this odd-
looking error message in gdb:
0x2806b445 1667 strncpy(pathname, dir, dirlen);
(gdb)
strncpy (
dst=0x2976de83
"/usr/pkg/lib/evolution/2.8:/usr/pkg/lib:/usr/pkg/lib/nss:/usr/pkg/xorg/lib:/usr/pkg/lib/nspr:/usr/lib/gcc34:/usr/lib",
src=0x1a <Error reading address 0x1a: Bad address>, n=3217027296)
at /usr/src/lib/libc_rtld/../libc/string/strncpy.c:50
The line 1667 in question is at /usr/src/libexec/rtld-elf/rtld.c:1667
and the program continues past that error as if nothing had gone
wrong. (Possibly because the return code from strncpy is never
checked? Dunno, but it worries me.)
Given a little more time and effort I may be able to find out where
the bad pointer is coming from -- but is it worth the effort? Is
this really a problem, or not?
Thanks!
Updated by corecode over 17 years ago
it not neccessarily is a stray pointer, it just means that gdb thinks that it points to 0x1a. that happens from time to time. if it doesn't segfault here, you don't have to worry.
cheers
simon