ulogd compile problems

ulogd compile problems

[Jason Frisvold]

I'm having a problem compiling ulogd 1.24.  I believe I have all of  
the appropriate development libraries installed, but I'm receiving the  
error below.  This is a CentOS 5.3 system with MySQL 5.0.67  
installed.  Can someone please shed some light on how to resolve this?

make[1]: Entering directory `/home/friz/ulogd-1.24/mysql'
gcc -I-I/usr/include/mysql -DOLD_MYSQL=1  -g -O2  -Wall - 
DULOGD_CONFIGFILE=\"/usr/local/etc/ulogd.conf\" -I/lib/modules/`uname - 
r`/build/include -I.. -I../libipulog/include -I../include -fPIC -o  
ulogd_MYSQL_sh.o -c ulogd_MYSQL.c
ld -shared -rdynamic -L/usr/lib/mysql -lmysqlclient -lz -lcrypt -lnsl - 
lm  -o ulogd_MYSQL.so ulogd_MYSQL_sh.o -lc
ld: bad -rpath option
make[1]: *** [ulogd_MYSQL.so] Error 1
make[1]: Leaving directory `/home/friz/ulogd-1.24/mysql'
make: *** [recurse] Error 1


Thanks,

-- 
Jason 'XenoPhage' Frisvold
XenoPhage0@gmail.com
http://blog.godshell.com

Re: ulogd compile problems

[Jason Frisvold]

On Sep 5, 2009, at 11:31 PM, Jason Frisvold wrote:
> I'm having a problem compiling ulogd 1.24.  I believe I have all of  
> the appropriate development libraries installed, but I'm receiving  
> the error below.  This is a CentOS 5.3 system with MySQL 5.0.67  
> installed.  Can someone please shed some light on how to resolve this?

Is this not the proper list to ask such questions?  If so, can someone  
please direct me to a more appropriate place?


-- 
Jason 'XenoPhage' Frisvold
XenoPhage0@gmail.com
http://blog.godshell.com

Re: ulogd compile problems

[Michele Petrazzo - Unipex]

Jason Frisvold wrote:
> On Sep 5, 2009, at 11:31 PM, Jason Frisvold wrote:
>> I'm having a problem compiling ulogd 1.24.  I believe I have all of
>>  the appropriate development libraries installed, but I'm receiving
>> the error below.  This is a CentOS 5.3 system with MySQL 5.0.67 
>> installed.  Can someone please shed some light on how to resolve
>> this?
> 
> Is this not the proper list to ask such questions?  If so, can
> someone please direct me to a more appropriate place?
> 


I think that it's the right one, but the developers can be busy sometime...

However, "ld: bad -rpath option " can be only a linker option so trying
to googleing a bit, I found this: http://tinyurl.com/nqewau

Modify the Makefile with that options and, if it work, try to submit a 
patch to the devels.

Michele

Re: ulogd compile problems

[Jason Frisvold]

On 09/08/2009 12:25 PM, Michele Petrazzo - Unipex wrote:
> I think that it's the right one, but the developers can be busy sometime...

Understood.  I didn't see any comments on it, so I figured I'd ask.

> However, "ld: bad -rpath option " can be only a linker option so trying
> to googleing a bit, I found this: http://tinyurl.com/nqewau

> Modify the Makefile with that options and, if it work, try to submit a
> patch to the devels.

I did modify the makefile and change the -rdynamic flag to -rpath.  I
compiled and it seems to run now, though I need to test more.

I'm a bit confused, though, because mysql_config defined -rdynamic,
which I'm inclined to believe.

Also, configure seems to be "broken" when it comes to mysql anyway
because of this :

checking for mysql_real_escape_string support... strings: invalid option
-- r
Usage: strings [option(s)] [file(s)]
 Display printable strings in [file(s)] (stdin by default)
 The options are:
  -a - --all                Scan the entire file, not just the data section
  -f --print-file-name      Print the name of the file before each string
  -n --bytes=[number]       Locate & print any NUL-terminated sequence of at
  -<number>                 least [number] characters (default 4).
  -t --radix={o,d,x}        Print the location of the string in base 8,
10 or 16
  -o                        An alias for --radix=o
  -T --target=<BFDNAME>     Specify the binary file format
  -e --encoding={s,S,b,l,B,L} Select character size and endianness:
                            s = 7-bit, S = 8-bit, {b,l} = 16-bit, {B,L}
= 32-bit
  @<file>                   Read options from <file>
  -h --help                 Display this information
  -v --version              Print the program's version number
strings: supported targets: elf32-i386 a.out-i386-linux efi-app-ia32
elf32-little elf32-big srec symbolsrec tekhex binary ihex trad-core


The source of this problem is line 1750 of configure :

   MYSQL_FUNCTION_TEST=`strings ${MYSQLLIBS}/libmysqlclient.so | grep
mysql_real_escape_string`

This would work, if MYSQLLIBS wasn't defined above as :

MYSQLLIBS=`$d/mysql_config --libs`

I'm not sure why they would try to use MYSQLLIBS as a path in the
function test as mysql_config --libs returns something similar to this:

[user@example ulogd-1.24]$ /usr/bin/mysql_config --libs
-rdynamic -L/usr/lib/mysql -lmysqlclient -lz -lcrypt -lnsl -lm

And there's that pesky -rdynamic, too.  I wonder, perhaps, if it should
be -Bdynamic instead.. that would make more sense to me..

Actually, the -rdynamic is ok for GCC, but not LD ...  So perhaps the
Makefile is wrong?  Or outdated?

> Michele

-- 
---------------------------
Jason Frisvold
xenophage0@gmail.com
---------------------------
"I love deadlines. I like the whooshing sound they make as they fly by."
   - Douglas Adams

Re: ulogd compile problems

[Jason Frisvold]

On Sep 8, 2009, at 5:13 PM, Jason Frisvold wrote:
> I did modify the makefile and change the -rdynamic flag to -rpath.  I
> compiled and it seems to run now, though I need to test more.


Ok, so a final update to explain what I did to get this to work.  If  
the developers can come up with a patch, then awesome.  If not, then  
I've included the patch to configure that I used to make this work.

To start, I'm running this on a CentOS 5.3 system with MySQL 5.0.67  
installed via RPMs.  I'm using the rhel5 RPMS.

If we run mysql_config --libs, we get the following output :

[user@example SPECS]$ /usr/bin/mysql_config --libs
-rdynamic -L/usr/lib/mysql -lmysqlclient -lz -lcrypt -lnsl -lm

There are two problems with this.  First, -rdynamic is not a linker  
flag and just ends with an error.  I believe this can be changed to - 
Bdynamic, which is a linker flag, or it can be deleted.  I chose the  
latter.

The second problem won't show itself until you try to start ulogd with  
the MySQL extension configured.  When you try to start ulogd, you get  
this :

Starting ulogd: Thu Sep 10 22:35:20 2009 <7> ulogd.c:462 load_plugins:  
'/usr/lib/ulogd/ulogd_MYSQL.so': /usr/lib/ulogd/ulogd_MYSQL.so:  
undefined symbol: mysql_fetch_field

The problem here is that libmysqlclient exists in /usr/lib, not /usr/ 
lib/mysql.  So, we need to add -L/usr/lib to the linker options.   
Adding this allows the extension to properly build, linking it with  
the proper libraries.

Also included in the patch below is a fix for the  
mysql_real_escape_string check.  I'm not sure what the exact aim is  
here, but I do know that the original method for checking failed  
miserably.  My fix uses "locate," though, which I don't believe is  
portable.

If someone has a better patch, please let me know.  I'd love to fix  
this the correct way, but I'm not that familiar with linker and  
compiler flags.

--- ./configure	2006-01-25 06:15:22.000000000 -0500
+++ ./configure	2009-09-10 22:37:24.000000000 -0400
@@ -1728,11 +1728,11 @@
  EOF

          MYSQLINCLUDES=`$d/mysql_config --include`
-        MYSQLLIBS=`$d/mysql_config --libs`
+        MYSQLLIBS=`$d/mysql_config --libs | sed s/-rdynamic//`

  	DATABASE_DIR="${DATABASE_DIR} mysql"

-	MYSQL_LIB="${DATABASE_LIB} ${MYSQLLIBS} "
+	MYSQL_LIB="${DATABASE_LIB} -L/usr/lib ${MYSQLLIBS}"
  	# no change to DATABASE_LIB_DIR, since --libs already includes -L

  	DATABASE_DRIVERS="${DATABASE_DRIVERS} ../mysql/mysql_driver.o "
@@ -1747,7 +1747,8 @@
  	echo $ac_n "checking for mysql_real_escape_string support""...  
$ac_c" 1>&6
  echo "configure:1749: checking for mysql_real_escape_string support"  
 >&5

-	MYSQL_FUNCTION_TEST=`strings ${MYSQLLIBS}/libmysqlclient.so | grep  
mysql_real_escape_string`
+   LIBMYSQLCLIENT=`locate libmysqlclient.so | grep libmysqlclient.so$`
+   MYSQL_FUNCTION_TEST=`strings $LIBMYSQLCLIENT | grep  
mysql_real_escape_string`

  	if test "x$MYSQL_FUNCTION_TEST" = x
  	then


Thanks!

-- 
Jason 'XenoPhage' Frisvold
XenoPhage0@gmail.com
http://blog.godshell.com