On 03/25/2010 05:19 PM, Andreas Schulze wrote:
> Hello,
> 
> I have sles9, sles10 an debian lenny
> 
> # ./configure \
>         --prefix=/usr/ \
>         --libdir=/usr/lib \
>         --mandir=/usr/share/man \
>         --sysconfdir=/etc/clamav \
>         --disable-check \
>         --disable-zlib-vcheck \
>         --enable-ipv6 \
>         --disable-clamuko \
>         --disable-milter \
>         --enable-bigstack \
>         --enable-readdir_r \
>         --enable-clamdtop \
>         --with-dbdir=/var/lib/clamav
> 
> sles9 and debian compiles fine.
> make fail only on sles10:
> ...
> make[2]: Entering directory `/usr/src/packages/BUILD/clamav-0.96rc2/clamscan'
>   CC     output.o
>   CC     getopt.o
>   CC     optparser.o
>   CC     actions.o
>   CC     misc.o
>   CC     clamscan.o
>   CC     others.o
>   CC     manager.o
>   CCLD   clamscan
> ../libclamav/.libs/libclamav.so: undefined reference to 
> `__sync_val_compare_and_swap_4'
> ../libclamav/.libs/libclamav.so: undefined reference to 
> `__sync_add_and_fetch_4'
> ../libclamav/.libs/libclamav.so: undefined reference to 
> `__sync_sub_and_fetch_4'

There is a configure test that should prevent this, have a look in
libclamav/c++/llvm/config.log
"configure:12390: checking for GCC atomic builtins"
What does it say?

Also are your gcc and g++ versions the same, and your CFLAGS and
CXXFLAGS the same?

> collect2: ld returned 1 exit status
> make[2]: *** [clamscan] Fehler 1
> make[2]: Leaving directory `/usr/src/packages/BUILD/clamav-0.96rc2/clamscan'
> make[1]: *** [all-recursive] Fehler 1
> make[1]: Leaving directory `/usr/src/packages/BUILD/clamav-0.96rc2'
> make: *** [all] Fehler 2
> error: Bad exit status from /var/tmp/rpm-tmp.28193 (%build)
> 
> This problem is new in 0.96rc2. rc1 doas compile an run with these settings.
> 
> Any Hints !?

What CPU architecture, and which compiler version?

Best regards,
--Edwin
_______________________________________________
Help us build a comprehensive ClamAV guide: visit http://wiki.clamav.net
http://www.clamav.net/support/ml

Reply via email to