[OE-core] [PATCH 4/8] python: fix for libdir=/usr/lib64 case
Yu Ke
ke.yu at intel.com
Fri Jul 8 00:52:40 UTC 2011
on 2011-7-8 3:22, Richard Purdie wrote:
> On Thu, 2011-07-07 at 22:10 +0800, Yu Ke wrote:
>> python has several place hard code "lib" which breaks build
>> when libdir=/usr/lib64. SUSE has a patch to fix this issue.
>>
>> So this patch add the SUSE patch to address this issue
>>
>> Signed-off-by: Yu Ke<ke.yu at intel.com>
> [...]
>> ++AC_SUBST(LIB)
>> ++AC_MSG_CHECKING(LIB)
>> ++case $ac_sys_system in
>> ++Linux*)
>> ++ # Test if the compiler is 64bit
>> ++ echo 'int i;'> conftest.$ac_ext
>> ++ python_cv_cc_64bit_output=no
>> ++ if AC_TRY_EVAL(ac_compile); then
>> ++ case `/usr/bin/file conftest.$ac_objext` in
>> ++ *"ELF 64"*)
>> ++ python_cv_cc_64bit_output=yes
>> ++ ;;
>> ++ esac
>> ++ fi
>> ++ rm -rf conftest*
>> ++ ;;
>> ++esac
>> ++
>> ++case $ARCH:$python_cv_cc_64bit_output in
>> ++ppc64:yes | powerpc64:yes | s390x:yes | sparc64:yes | x86_64:yes)
>> ++ LIB="lib64"
>> ++ ;;
>> ++*:*)
>> ++ LIB="lib"
>> ++ ;;
>> ++esac
>> ++AC_MSG_RESULT($LIB)
>> +
>
> This patch is good apart from this bit. I think we need to be able to
> pass in our own string to use here which we derive from the libdir
> variable...
>
Yes, indeed. I will revise it to use libdir instead of hardcode lib/lib64.
Regards
Ke
More information about the Openembedded-core
mailing list