Re: extensions module and Solaris

[Eric Blake]

Eric Blake <ebb9 <at> byu.net> writes:

> But in my (limited) testing, I wasn't able to get that to 
> happen with either gcc 3.3 or /usr/ucb/cc as the compiler and -
D_POSIX_SOURCE=1 
> on the command line.  Any ideas on how to select the standards-compliant 
system
> (2) implementation on Solaris?

After reading 'man standards' again, I figured out that this works to expose 
the difference in implementation (using the particular system call that tripped 
up the M4 testsuite):

$ cat foo.c
#include <stdlib.h>
#include <sys/wait.h>

int main()
{
   exit (WEXITSTATUS (system ("echo hi >&3")));
}
$ gcc -o foo -Wall foo.c
$ ./foo
hi
$ gcc -o foo -Wall foo.c /usr/ccs/lib/values-xpg4.o
$ ./foo
sh: 3: bad file unit number
$ nm /usr/ccs/lib/values-xpg4.o  
00000000 D __xpg4

So maybe the question is better worded:

Should we figure out a way to make inclusion of the extensions module cause 
compilation on Solaris to favor the standards-compliant behavior of various 
functions?  Should we try to directly add values-xpg4.o to the link line, or 
are we better off providing a .c file which provides the same definition of 
__xpg4 as what values-xpg4.o provides?

-- 
Eric Blake