Re: Re-organize initial optimization passes

["Joseph S. Myers" <joseph at codesourcery dot com>]

On Tue, 10 May 2005, Diego Novillo wrote:

> The bad news is that our weak uninitialized warning harness
> gave me more than a few headaches.  It either misfires causing
> bootstrap failures or breaks some uninit tests in the testsuite.
> I finally settled for one that breaks one of the uninit tests,
> fixes another one and allows us to bootstrap.

If uninit-1.c is fixed by this patch, then the associated bug (21005) 
should be closed.

> 	* gcc.dg/tree-ssa/loop-4.c: Fix expected pattern.

If this is fixing bug 21341, then that bug should be closed.

In general, FAILing testcases should have bugs and those bugs should be 
(noted in the ChangeLog entry / log message and) checked for and closed 
when fixing those FAILs.

(The uninit-1.c FAIL appears to have become a PASS today and the loop-4.c 
FAIL appears to have disappeared, but the results from my mainline builds 
today are sufficiently bad with in particular many coverage tests failing 
with the relevant generated files not found that it's difficult to tell 
which changes are meaningful and which are noise.)

-- 
Joseph S. Myers               http://www.srcf.ucam.org/~jsm28/gcc/
    jsm@polyomino.org.uk (personal mail)
    joseph@codesourcery.com (CodeSourcery mail)
    jsm28@gcc.gnu.org (Bugzilla assignments and CCs)