Re: [incomplete PATCH] Smoke 12088 /pro/3gl/CPAN/perl-current

On Thu, Sep 20, 2001 at 01:47:18PM +0200, H. Merijn Brand wrote:
> > Hmmm, poking around in cflags, it says we can't use -ansi because of
> > some problem with some naughty Solaris system library.  But what if
> > you had the smoke test append -ansi to the ccflags when it's gcc and
> > it's not Solaris?
> 
> Think not. perl's configuration is complicated enough. Hint files are
> maintained and you might suggest adding -ansi to the default gcc configuration
> for Perl and disable/remove it in the hints/solaris.sh
> 
> smoke should test what blead *is* and not meddle with it IMHO

Out of perverse curiousity, I'm trying configuring bleadperl with
-Wall and -ansi on.  Seems like Berkeley is guilty of a few things...

    Checking Berkeley DB version ...
    In file included from try.c:7:
    /usr/include/db.h:946: parse error before `u_long'
    I can't use Berkeley DB with your <db.h>.  I'll disable Berkeley DB.
    Removing unusable -ldb from library list

Lines 944 to 947 are
int   db_env_set_func_seek
	  __P((int (*)(int, size_t, db_pgno_t, u_int32_t, int, int)));
int   db_env_set_func_sleep __P((int (*)(u_long, u_long)));
int   db_env_set_func_unlink __P((int (*)(const char *)));


As for the make... err, we don't get very far.

$ make
`sh  cflags "optimize='-g'" libperl.a miniperlmain.o`  miniperlmain.c
          CCCMD =  cc -DPERL_CORE -c -DDEBUGGING -fno-strict-aliasing -I/usr/local/include -Wall -ansi -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -g  
In file included from miniperlmain.c:17:
perl.h:417: conflicting types for `syscall'
/usr/include/unistd.h:887: previous declaration of `syscall'
make: *** [miniperlmain.o] Error 1


Ok, scratch that idea.


-- 

Michael G. Schwern   <[email protected]>    http://www.pobox.com/~schwern/
Perl6 Quality Assurance     <[email protected]>	     Kwalitee Is Job One
That which stirs me, stirs everything.
	-- Squonk Opera, "Spoon"