[RFA:] Fix PR optimization/15296, delayed-branch-slot bug.

[Eric Botcazou]

> Except for the occasional bugfix by me and others...

Yep, and I'm in the lot too :-)

> Hey, you just don't trust my bug report!  That hurts!

I fail to see why you say that.  I know of several bugs on the 3.3 branch 
that will not be fixed on that branch.

> To counter your statement, let me say I definitely want this
> proved and well-contained bug fixed on all open branches rather
> than remaining.  If you don't want to test the patch, then
> don't, but then I think you have no particular reason to judge
> it either.

Well, it will really be proved only after it has passed full bootstrap on 
SPARC 32-bit and 64-bit.  It looks indeed safe but <habitual rant about 
bugfixes that uncover other bugs>.

-- 
Eric Botcazou