Details
-
Task
-
Resolution: Fixed
-
Minor
-
None
-
None
-
None
-
None
Description
Good day,
in radsecproxy software I found this bug:
1. Programm: radsecproxy version 1.6.8 (and earlier?)
Environment:
Sun Microsystems Sun Fire X2200 M2
Operating system: Solaris 10_x86 64bit
Compiler: Sun Studio 12.1 _and_ GCC 3.4.6 tested
I think the bug is independent of hardware, operating system and
compiler.
2. Problem/Bug:
+ With LogLevel=5 and without SourceUDP and/or SourceTLS are
NOT set in "radsecproxy.conf" (one or both variables are empty)
the binary aborted at start time:
<date time>: radsecproxy revision 1.6.8 starting
<date time>: resolvehostport<ABORT>
3. Solution (Workaround):
+ In "radsecproxy.conf" the variables SourceUDP and SourceTLS
must be set (for Loglevel 5).
And additional:
In "radsecproxy.conf" and in doc and man-gage to insert:
'## For "LogLevel 5": SourceUDP and SourceTLS must be set!'
+ Or (my workaround):
In "hostport.c" modify:
//orig (line 175):
debug(DBG_DBG, "%s: %s -> %s", __func__, hp->host,
addr2string(hp->addrinfo->ai_addr));
//korr (line 175):
debug(DBG_DBG, "%s: %s -> %s", __func__,
(hp->host == NULL ? "SourceXXX empty" : hp->host),
(hp->host == NULL ? "(SourceXXX empty)\n" :
addr2string(hp->addrinfo->ai_addr)));
Perheaps is there a better text to replace "SourceXXX empty" ?
4. For "LogLevel <= 4" the error not occurs - because the debug statement
line 175 is not executed.
Regards
Leonhard Knauff
[Created via e-mail received from: leonhard.knauff@urz.uni-halle.de]