Re: 3.0.1 PATCH: Avoid link failure of Objective-C testsuite on Solaris2/Intel

[Stan Shebs <shebs at apple dot com>]

Rainer Orth wrote:
> 
> Obviously, Solaris 2/Intel needs definitions of sched_get_priority_{min,
> max}, too.  Since only those two symbols are missing, I've implemented a
> proper fix, i.e. provide weak definitions for them, instead of linking with
> -lposix4 on Solaris 2.[678]/Intel as has been the hack for the Solaris
> 2/SPARC case until now.

OK, so these have a real purpose in libobjc, and in theory we might
someday write a program or testcase that needs these functions to be
defined.  Wouldn't we still need the -lposix4 library to be linked in
then?  In any case, it seems reasonable to declare them as weak symbols.

> With this patch, all Objective-C testcases succeed, with the exception of
> 
> FAIL: objc/execute/class-3.m execution,  -O
> 
> which fails with
> 
> test_instances failed

That's a new one.  I've not seen any other platform fail this one,
and the test seems pretty straightforward.  Does it fail with
other optimization levels?  (I'm working on a testsuite hack to
run ObjC tests with all the opt levels, but it's not done yet.)

Also, is formal-protocol-6.m really passing?  I've analyzed it and
it looks like it exposes a fundamental flaw in the GNU runtime
(protocols can't exist independent of classes/categories), and
it should have failed.

Stan