coredump at __correctly_grouped_prefixwc - c++

Question

the program under live env, segmentation fault some time, i try to gdb the coredump file,
but can't found the code line cause coredump.
Program terminated with signal 11, Segmentation fault.
#0 0x00000038f3a41bf5 in __correctly_grouped_prefixwc () from /lib64/libc.so.6
(gdb) bt
#0 0x00000038f3a41bf5 in __correctly_grouped_prefixwc () from /lib64/libc.so.6
#1 0x0000000000000000 in ?? ()
(gdb) info r
rax 0x1ac1b108 448901384
rbx 0x2add423b4ff0 47129787322352
rcx 0x2add48128640 47129885312576
rdx 0x0 0
rsi 0x1 1
rdi 0x2add48000020 47129884098592
rbp 0x2add3f1aef50 0x2add3f1aef50
rsp 0x2add423b4ff0 0x2add423b4ff0
r8 0x2 2
r9 0x2 2
r10 0x0 0
r11 0x0 0
r12 0x0 0
r13 0x3 3
r14 0x1000 4096
r15 0x2add3f1b0000 47129734873088
rip 0x38f3a41bf5 0x38f3a41bf5 <__correctly_grouped_prefixwc+165>
eflags 0x10246 [ PF ZF IF RF ]
cs 0x33 51
ss 0x2b 43
ds 0x0 0
es 0x0 0
fs 0x0 0
gs 0x0 0
fctrl 0x37f 895
fstat 0x0 0
ftag 0xffff 65535
fiseg 0x0 0
fioff 0xc54f06 12930822
foseg 0x2add 10973
fooff 0x423b3f00 1111179008
fop 0x0 0
mxcsr 0x1fa1 [ IE PE IM DM ZM OM UM PM ]
cat /etc/redhat-release
CentOS release 5.5 (Final)
and i want to debug glibc at the source level, run yum install yum-utils to install the debuginfo-install program.
then, run sudo debuginfo-install glibc, the result following
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* addons: centos.ustc.edu.cn
* base: mirror.bit.edu.cn
* extras: centos.ustc.edu.cn
* updates: centos.ustc.edu.cn
Checking for new repos for mirrors
Could not find debuginfo for main pkg: glibc-2.5-123.x86_64
Could not find debuginfo for main pkg: glibc-2.5-123.i686
No debuginfo packages available to install
and then i try to run yum search glibc-debuginfo
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* addons: centos.ustc.edu.cn
* base: mirrors.163.com
* extras: centos.ustc.edu.cn
* updates: centos.ustc.edu.cn
Warning: No matches found for: glibc-debuginfo
No Matches found
no matches found again.
i try to run yum search glibc
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* addons: centos.ustc.edu.cn
* base: mirrors.163.com
* extras: centos.ustc.edu.cn
* updates: centos.ustc.edu.cn
================================================================================ Matched: glibc =================================================================================
compat-glibc.i386 : Compatibility C library
compat-glibc.x86_64 : Compatibility C library
compat-glibc-headers.x86_64 : Header files for development using standard C libraries.
glibc.i686 : The GNU libc libraries.
glibc.x86_64 : The GNU libc libraries.
glibc-common.x86_64 : Common binaries and locale data for glibc
glibc-devel.i386 : Object files for development using standard C libraries.
glibc-devel.x86_64 : Object files for development using standard C libraries.
glibc-headers.x86_64 : Header files for development using standard C libraries.
glibc-utils.x86_64 : Development utilities from GNU C library
kernel-headers.x86_64 : Header files for the Linux kernel for use by glibc
nss_db.i386 : An NSS library for the Berkeley DB.
nss_db.x86_64 : An NSS library for the Berkeley DB.
yp-tools.x86_64 : NIS (or YP) client programs.
yum-protect-packages.noarch : Yum plugin to prevents Yum from removing itself and other protected packages
i try to run sudo yum install glibc-devel.x86_64, and gdb the coredump file again,
but it display the following
Reading symbols from /lib64/libc.so.6...(no debugging symbols found)...done.
Loaded symbols for /lib64/libc.so.6
how can i find the code line cause coredump? i try to google, but not found, any ideas?

Answer 1

First, although __correctly_grouped_prefixwc caused the segmentation fault, it's likely that it was passed incorrect arguments from some other piece of code, perhaps strtod or strtol or something that called them. That being said, here is how to set things up so that gdb can show the line of source code in __correctly_grouped_prefixwc that caused the segmentation fault.
To do source-level debugging, you need an executable or shared object's debug info, and its source code. Linux and Unix distributions in general do not include these by default, to conserve storage space, but they make them available as packages.
On CentOS, you just need to install the debuginfo package for each executable or library you're interested in. To do this, run
sudo yum install yum-utils
which will install the debuginfo-install program, then run
sudo debuginfo-install glibc
to download and install the glibc-debuginfo-2.5-123 package (your version number may vary). This will install, among many other files, /usr/lib/debug/lib64/libc.so.6.debug, /usr/lib/debug/lib64/libc-2.5.so.debug, and /usr/src/debug/glibc-2.5-20061008T1257/stdlib/grouping.c, which are what you need.
debuginfo-install is a short python program that enables the debuginfo repositories and downloads and installs the debuginfo package corresponding to the package you give as an argument, plus all its dependencies. As an alternative, you can download the debuginfo packages directly from http://debuginfo.centos.org (or any mirrors) and install them using rpm -i.
You mentioned that you got the error No debuginfo packages available to install. Perhaps you don't have the debuginfo repo configured. On my CentOS 5 system, the configuration is in the file /etc/yum.repos.d/CentOS-Debuginfo.repo
# All debug packages from all the various CentOS-5 releases
# are merged into a single repo, split by BaseArch
#
# Note: packages in the debuginfo repo are currently not signed
#
[base-debuginfo]
name=CentOS-5 - Debuginfo
baseurl=http://debuginfo.centos.org/5/$basearch/
gpgcheck=0
gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-CentOS-5
enabled=0
For other releases, general instructions for adding the debuginfo repo are in this CentOS wiki article.