bug-libtool
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Bug#578996: libtool: LT_LIBM doesn't work properly on AIX


From: Török Edwin
Subject: Re: Bug#578996: libtool: LT_LIBM doesn't work properly on AIX
Date: Sun, 25 Apr 2010 15:59:36 +0300
User-agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1.9) Gecko/20100411 Icedove/3.0.4

On 04/25/2010 01:27 PM, Ralf Wildenhues wrote:
> tags +upstream
> thanks
> 
> Hello Török,
> 
> this is clearly an upstream, and not a Debian packaging bug.  Adding
> bug-libtool.

Right.

> 
> * Török Edwin wrote on Sat, Apr 24, 2010 at 09:19:54AM CEST:
>> ClamAV uses LT_LIB_M to find the math library.
>> It failed on AIX, see https://wwws.clamav.net/bugzilla/show_bug.cgi?id=1989
>>
>> configure:12861: checking for cos in -lm
>> configure:12886: gcc -o conftest -DSYSV -D_AIX -D_AIX32 -D_AIX41 -D_AIX43 
>> -D_AIX51 -D_AIX52        -D_ALL_SOURCE -DFUNCPROTO=15 -O 
>> -I/opt/freeware/include  -L/opt/freeware/lib 
>> -Wl,-blibpath:/opt/freeware/lib:/usr/lib:/lib conftest.c -lm   >&5
>> conftest.c:53: warning: conflicting types for built-in function 'cos'
>> collect2: library libm not found
>> configure:12895: result: no
>>
>> Looks like the macro looks for cos in -lm, and if it does't find it there,
>> it assumes that no library is needed for it.
>> But that is not the case here (the user didn't have libm installed).
> 
> OK, that's a bug.  For the Libtool testsuite (which I have so far
> thought to be the primary user of LT_LIB_M) the right fix would be
> to allow another failure action however, so we could skip the test.
> 
>> I think this macro should fail at configure time, and tell the user to 
>> install
>> libm, if it can't find the cos() function at all.
> 
> If you need such a macro, then you should write one yourself, and use
> a math function that you are going to need.  It's essentially a one-liner:
>   AC_SEARCH_LIBS([cos], [m], [], [AC_MSG_ERROR([need math library])])

Ok, I don't know why we used LT_LIB_M in the first place, but you are
right we can certainly change it to that instead.

Best regards,
--Edwin




reply via email to

[Prev in Thread] Current Thread [Next in Thread]