[reSIProcate] cross-compile errors

[Adam Roach]

The configure script now tries to be more helpful regarding path options 
when cross-compilation is used. Note that I didn't change the build 
system at all; I just made the configuration script try to make "doing 
the right thing" a bit more obvious.

/a

On 5/20/08 6:50 PM, Adam Roach wrote:
> Still working on the patch. In the meanwhile, I think you can get 
> around the problem by editing build/Makefile.conf directly and 
> changing the include directory for POPT so that it isn't /usr/include 
> anymore. You may have to also disable the use of POPT.
>
> /a
>
> On 5/20/08 5:58 PM, Adam Roach wrote:
>> Okay; I've managed to reproduce the error here. The problem is that 
>> the configure script will still try to find native headers and 
>> libraries for things like popt and bdb even when cross-compiling. I'm 
>> working on a fix for this, and will let you know when it's ready for 
>> you to try.
>>
>> /a
>>
>> On 5/20/08 5:36 PM, Michael Zhang wrote:
>>> Thanks for the quick reply, Adam
>>>
>>> I did use the crosstool to generate the toolchain. And I tried to use
>>> the mips compiler to build the code below and it worked.
>>>
>>> Regards,
>>>
>>> Michael
>>>
>>> On Tue, 2008-05-20 at 15:54, Adam Roach wrote:
>>>  
>>>> At first blush, I'd guess that there's something wrong with the 
>>>> macro for FD_ZERO in the headers you're using. It looks like you're 
>>>> using crosstool to generate your cross compiler. I'm going to try 
>>>> to replicate your toolchain environment to see whether I get the 
>>>> same error.
>>>>
>>>> In the meanwhile, I would encourage you to try compiling a minimal 
>>>> application that exercises this macro to see whether we can 
>>>> trivially reproduce the problem. Can your cross-compiler 
>>>> successfully produce a mips binary from the following code?
>>>>
>>>>
>>>> //--------------------------------------------------
>>>> #include <stdio.h>
>>>> #include <sys/select.h>
>>>>
>>>> int main (int argc, char** argv)
>>>> {
>>>>   fd_set s;
>>>>   FD_ZERO(&s);
>>>>   return 0;
>>>> }
>>>> //--------------------------------------------------
>>>>
>>>> /a
>>>>
>>>> On 5/20/08 1:55 PM, Michael Zhang wrote:
>>>>   
>>>>> Hi Adam,
>>>>>
>>>>> I am building the latest reSiprocate 1.3.1 with a mips cross 
>>>>> compiler,
>>>>> mipsel-gcc-3.3.2-glibc-2.3.2. And got some compile errors as below.
>>>>>
>>>>> make[1]: Entering directory 
>>>>> `/home/mzhang/resiprocate/resip/stack/test'
>>>>> /home/mzhang/crosstool/gcc-3.3.2-glibc-2.3.2/mipsel-unknown-linux-gnu/bin/mipsel-unknown-linux-gnu-g++ 
>>>>> -D_REENTRANT  -g -Wall -fPIC                 -I../../../build/.. 
>>>>> -I../../../build/.. -I../../../build/../contrib/ares 
>>>>> -I/usr/include 
>>>>> -I/home/mzhang/crosstool/gcc-3.3.2-glibc-2.3.2/mipsel-unknown-linux-gnu/include/c++/3.3.2/ 
>>>>> -I/home/mzhang/crosstool/gcc-3.3.2-glibc-2.3.2/mipsel-unknown-linux-gnu/include/c++/3.3.2/mipsel-unknown-linux-gnu/ 
>>>>> -DOS_MAJOR_VER=2 -DOS_MINOR_VER=4 -DOS_POINT_VER=22-1 
>>>>> -DOS_PATCH_VER=2115 -DUSE_ARES -DHAVE_POPT_H -c -o 
>>>>> obj.debug.Linux.mipsel/UAS.o UAS.cxx
>>>>> ../../../rutil/Socket.hxx: In constructor `resip::FdSet::FdSet()':
>>>>> ../../../rutil/Socket.hxx:102: error: impossible constraint in `asm'
>>>>> ../../../rutil/Socket.hxx:103: error: impossible constraint in `asm'
>>>>> ../../../rutil/Socket.hxx:104: error: impossible constraint in `asm'
>>>>> make[1]: *** [obj.debug.Linux.mipsel/UAS.o] Error 1
>>>>> make[1]: Leaving directory 
>>>>> `/home/mzhang/resiprocate/resip/stack/test'
>>>>> make: *** [tests] Error 2
>>>>>
>>>>>
>>>>> Can you please shed some light on this?
>>>>>
>>>>> Thanks,
>>>>>
>>>>> BTW - I heard about you from my colleague Mandeep Ahuja.
>>>>>
>>>>> Michael
>>>>>
>>>>>
>>>>>
>>>>>         
>>>
>>>   
>>
>> _______________________________________________
>> resiprocate-devel mailing list
>> resiprocate-devel at resiprocate.org
>> https://list.resiprocate.org/mailman/listinfo/resiprocate-devel
>
> _______________________________________________
> resiprocate-devel mailing list
> resiprocate-devel at resiprocate.org
> https://list.resiprocate.org/mailman/listinfo/resiprocate-devel