Bug 690568

Summary: valgrind reports uninitialised values while running "ls -l"
Product: Red Hat Enterprise Linux 6 Reporter: Branislav Náter <bnater>
Component: coreutilsAssignee: Ondrej Vasik <ovasik>
Status: CLOSED NOTABUG QA Contact: qe-baseos-daemons
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 6.1CC: meyering, prc
Target Milestone: rc   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2011-03-24 20:14:35 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:

Description Branislav Náter 2011-03-24 16:52:54 UTC 
Description of problem:
Running ls -l within valgrind reports uninitialised values.

Version-Release number of selected component (if applicable):
coreutils-8.4-13.el6.x86_64

How reproducible:
always

Steps to Reproduce:
1. log to x86_64 machine
2. run "valgrind  ls -l"
  
Actual results:
==8525== Memcheck, a memory error detector
==8525== Copyright (C) 2002-2010, and GNU GPL'd, by Julian Seward et al.
==8525== Using Valgrind-3.6.0 and LibVEX; rerun with -h for copyright info
==8525== Command: ls -l
==8525== 
==8525== Conditional jump or move depends on uninitialised value(s)
==8525==    at 0x52D7C9B: __GI___strcasecmp_l (in /lib64/libc-2.12.so)
==8525==    by 0x5273E84: __gconv_open (in /lib64/libc-2.12.so)
==8525==    by 0x5281417: _nl_find_msg (in /lib64/libc-2.12.so)
==8525==    by 0x5281BE3: __dcigettext (in /lib64/libc-2.12.so)
==8525==    by 0x409CA3: ??? (in /bin/ls)
==8525==    by 0x5272C9C: (below main) (in /lib64/libc-2.12.so)
==8525== 
==8525== Use of uninitialised value of size 8
==8525==    at 0x52D9DD4: __GI___strcasecmp_l (in /lib64/libc-2.12.so)
==8525==    by 0x5273E84: __gconv_open (in /lib64/libc-2.12.so)
==8525==    by 0x5281417: _nl_find_msg (in /lib64/libc-2.12.so)
==8525==    by 0x5281BE3: __dcigettext (in /lib64/libc-2.12.so)
==8525==    by 0x409CA3: ??? (in /bin/ls)
==8525==    by 0x5272C9C: (below main) (in /lib64/libc-2.12.so)
==8525== 
==8525== Use of uninitialised value of size 8
==8525==    at 0x52D9DD8: __GI___strcasecmp_l (in /lib64/libc-2.12.so)
==8525==    by 0x5273E84: __gconv_open (in /lib64/libc-2.12.so)
==8525==    by 0x5281417: _nl_find_msg (in /lib64/libc-2.12.so)
==8525==    by 0x5281BE3: __dcigettext (in /lib64/libc-2.12.so)
==8525==    by 0x409CA3: ??? (in /bin/ls)
==8525==    by 0x5272C9C: (below main) (in /lib64/libc-2.12.so)
==8525== 
total 0
==8525== 
==8525== HEAP SUMMARY:
==8525==     in use at exit: 19,312 bytes in 3 blocks
==8525==   total heap usage: 137 allocs, 134 frees, 65,755 bytes allocated
==8525== 
==8525== LEAK SUMMARY:
==8525==    definitely lost: 0 bytes in 0 blocks
==8525==    indirectly lost: 0 bytes in 0 blocks
==8525==      possibly lost: 0 bytes in 0 blocks
==8525==    still reachable: 19,312 bytes in 3 blocks
==8525==         suppressed: 0 bytes in 0 blocks
==8525== Rerun with --leak-check=full to see details of leaked memory
==8525== 
==8525== For counts of detected and suppressed errors, rerun with: -v
==8525== Use --track-origins=yes to see where uninitialised values come from
==8525== ERROR SUMMARY: 3 errors from 3 contexts (suppressed: 6 from 6)

Expected results:
no uninitialised values
Comment 4 Ondrej Vasik 2011-03-25 09:10:18 UTC 
Trace with enabled debuginfo - these uninitialized value(s) reports are only on x86_64 arch.

==17674== Conditional jump or move depends on uninitialised value(s)
==17674==    at 0x3C7A285900: __GI___strcasecmp_l (strcmp.S:1998)
==17674==    by 0x3C7A21FE24: __gconv_open (gconv_open.c:70)
==17674==    by 0x3C7A22D3B7: _nl_find_msg (dcigettext.c:974)
==17674==    by 0x3C7A22DB83: __dcigettext (dcigettext.c:640)
==17674==    by 0x409CA3: main (ls.c:2063)
==17674==  Uninitialised value was created by a stack allocation
==17674==    at 0x3C7A21FD8F: __gconv_open (gconv_open.c:65)
==17674== 
==17674== Use of uninitialised value of size 8
==17674==    at 0x3C7A285D24: __GI___strcasecmp_l (strcmp.S:2257)
==17674==    by 0x3C7A21FE24: __gconv_open (gconv_open.c:70)
==17674==    by 0x3C7A22D3B7: _nl_find_msg (dcigettext.c:974)
==17674==    by 0x3C7A22DB83: __dcigettext (dcigettext.c:640)
==17674==    by 0x409CA3: main (ls.c:2063)
==17674==  Uninitialised value was created by a stack allocation
==17674==    at 0x3C7A21FD8F: __gconv_open (gconv_open.c:65)
==17674== 
==17674== Use of uninitialised value of size 8
==17674==    at 0x3C7A285D28: __GI___strcasecmp_l (strcmp.S:2258)
==17674==    by 0x3C7A21FE24: __gconv_open (gconv_open.c:70)
==17674==    by 0x3C7A22D3B7: _nl_find_msg (dcigettext.c:974)
==17674==    by 0x3C7A22DB83: __dcigettext (dcigettext.c:640)
==17674==    by 0x409CA3: main (ls.c:2063)
==17674==  Uninitialised value was created by a stack allocation
==17674==    at 0x3C7A21FD8F: __gconv_open (gconv_open.c:65)
==17674== 

So - the problematic line is

char const *locale_format = dcgettext (NULL, long_time_format[i], LC_TIME);

long_time_format array is global variable and is defined ... LC_TIME is checked above ... 

I think this is false positive caused by glibc optimalizations ... could be probably added to valgrind suppressions - if you are worried about this output.