This is the mail archive of the libc-alpha@sourceware.cygnus.com mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Bug reports for nscd



We've received the two appended bug reports which describe the same problem.
Any volunteers for checking nscd?

Andreas



Topics:
   libc/1160: nscd segfaults when attempting to cache host names with multiple IP address matches
   libc/1156: Bug in gethostbyname implementation


----------------------------------------------------------------------

Date: Sat, 12 Jun 1999 14:32:39 -0400
From: sba@srl.caltech.edu
To: bugs@gnu.org
Subject: libc/1160: nscd segfaults when attempting to cache host names with multiple IP address matches
Message-Id: <199906121832.OAA06288@delysid.gnu.org>


>Number:         1160
>Category:       libc
>Synopsis:       nscd segfaults when attempting to cache host names with multiple IP address matches
>Confidential:   no
>Severity:       critical
>Priority:       medium
>Responsible:    libc-gnats
>State:          open
>Class:          sw-bug
>Submitter-Id:   unknown
>Arrival-Date:   Sat Jun 12 14:40:01 EDT 1999
>Last-Modified:
>Originator:     sba@srl.caltech.edu
>Organization:
net
>Release:        2.1.1
>Environment:
Intel PII and PIII running RH6.0
>Description:
On both a single and dual processor Intel RH6.0 box, I have found that nscd
will segfault when a user does something as simple as ping aol.com. The
problem persists with both the RH version of glibc-2.1.1 and after compiling
and installing a clean copy of the official glibc-2.1.1 (compiled with -O
using egcs-1.1.2).

Note, the problem appears only when accessing a host name for which nslookup
returns more than one valid IP address.
>How-To-Repeat:
start nscd on a RH6.0 box
ping aol.com
nscd will segfaul
>Fix:
>Audit-Trail:
>Unformatted:


------------------------------

Date: Fri, 11 Jun 1999 21:29:31 +0200
From: Gabor Gombas <gombasg@inf.elte.hu>
To: bugs@gnu.org
Subject: libc/1156: Bug in gethostbyname implementation
Message-Id: <199906111929.VAA17364@babel.inf.elte.hu>
Content-Type: text/plain; charset=us-ascii

>Number:         1156
>Category:       libc
>Synopsis:       Bug in gethostbyname implementation
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    libc-gnats
>State:          open
>Class:          sw-bug
>Submitter-Id:   unknown
>Arrival-Date:   Fri Jun 11 16:20:02 EDT 1999
>Last-Modified:  Fri Jun 11 17:52:41 EDT 1999
>Originator:     Lord of the Files
>Organization:
 
>Release:        
>Environment:
	
Host type: i386-pc-linux-gnu
System: Linux babel 2.2.9 #1 SMP Thu Jun 3 10:31:37 CEST 1999 i686 unknown
Architecture: i686

Addons: crypt linuxthreads nss-v1

Build CC: gcc
Compiler version: egcs-2.91.66 Debian GNU/Linux (egcs-1.1.2 release)
Kernel headers: UTS_RELEASE
Symbol versioning: yes
Build static: yes
Build shared: yes
Build pic-default: no
Build profile: yes
Build omitfp: no
Build bounded: no
Build static-nss: no
Stdio: libio

Description:
	[the problem was reported by Stuart Anderson <sba@srl.caltech.edu>
	on the Linux NIS+ list. I just verified it and provided a stack
	trace.]

	There seems to be a bug in the gethostbyname function. If I issue
	'ping cnn.com' while nscd is running, nscd terminates with SIGSEGV.
	Here is a stack trace:

Program received signal SIGSEGV, Segmentation fault.
0x4009c58b in __strcasecmp (s1=0x2e6e6e63 <Address 0x2e6e6e63 out of bounds>,
    s2=0xbffff950 "cnn.com") at ../sysdeps/generic/strcasecmp.c:62
62        do
(gdb) bt
#0  0x4009c58b in __strcasecmp (
    s1=0x2e6e6e63 <Address 0x2e6e6e63 out of bounds>, s2=0xbffff950 "cnn.com")
    at ../sysdeps/generic/strcasecmp.c:62
#1  0x40145a22 in getanswer_r (answer=0xbffff2b0, anslen=468,
    qname=0xbffff8c8 "cnn.com", qtype=1, result=0xbffff950,
    buffer=0xbffff734 "", buflen=512, errnop=0x4012c2e0, h_errnop=0x4012b124)
    at nss_dns/dns-host.c:588
#2  0x40144df6 in _nss_dns_gethostbyname2_r (name=0xbffffae0 "cnn.com", af=2,
    result=0xbffff950, buffer=0xbffff734 "", buflen=512, errnop=0x4012c2e0,
    h_errnop=0x4012b124) at nss_dns/dns-host.c:174
#3  0x804db4c in __gethostbyname2_r (name=0xbffffae0 "cnn.com", af=2,
    resbuf=0xbffff950, buffer=0xbffff734 "", buflen=512, result=0xbffff94c,
    h_errnop=0x4012b124) at ../nss/getXXbyYY_r.c:182
#4  0x804d1bd in addhstbyname (db=0x8051388, fd=4, req=0xbffffc2c,
    key=0xbffffae0, caller_uid=0) at hstcache.c:415
#5  0x804a965 in handle_request (fd=4, req=0xbffffc2c, key=0xbffffae0,
    caller=0xbffffc20) at connections.c:325
#6  0x804aea7 in nscd_run (p=0x0) at connections.c:460
#7  0x804af86 in start_threads () at connections.c:496
#8  0x8049c98 in main (argc=6, argv=0xbffffce4) at nscd.c:166

	[Note: I have a somewhat modifed nscd daemon, but the problem is in
	the underlying nss code]

How-To-Repeat:
	Start nscd, and try 'ping cnn.com'
>Description:
>How-To-Repeat:
>Fix:
>Audit-Trail:

Responsible-Changed-From-To: gnats-admin->libc-gnats
Responsible-Changed-By: jaeger
Responsible-Changed-When: Fri Jun 11 17:52:29 1999
Responsible-Changed-Why:
This is a misfiled glibc report.
Andreas
>Unformatted:
Synopsis:	bug in nss layer makes nscd terminate with SIGSEGV
Priority:	medium
Category:	libc
Class:		sw-bug
Release:	libc-2.1.1


------------------------------

End of forwardK8H2aF Digest
***************************



-- 
 Andreas Jaeger   aj@arthur.rhein-neckar.de    jaeger@informatik.uni-kl.de
  for pgp-key finger ajaeger@aixd1.rhrk.uni-kl.de

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]