[PATCH] Fix _savef* on powerpc*-ibm-aix*

[Alan Modra]

On Wed, Jul 23, 2003 at 09:55:50AM -0400, David Edelsohn wrote:
> >>>>> Alan Modra writes:
> 
> Alan> The _savef* functions need to be called before the stack pointer is
> Alan> adjusted.  Since Jakub's 2003-07-16 change, the stack adjustment happens
> Alan> right at the start of the prologue when current_function_calls_eh_return,
> Alan> so we can't use _savef*.
> 
> Alan> I've been playing with defining FP_SAVE_INLINE for powerpc64-linux,
> Alan> which is how I ran into the problem...
> 
> 	This patch is needed in addition to Jakub's recent fix
> (ra_need_lr)?  Or do you want to replace that?

In addition.  I have to admit that I found the problem before ra_need_lr
went in my tree, but I think the two patches fix separate problems.
It's certain that this patch can't replace ra_need_lr as builtin_return
can be called in a function that doesn't muck with exceptions.  I find
it a little more difficult to prove that calls_eh_return can be set
without a builtin_return(0) being called, but I think that's the case.

-- 
Alan Modra
IBM OzLabs - Linux Technology Centre