Re: RFA: fix mode confusion in caller-save.c:replace_reg_with_saved_mem

[Jeff Law <law at redhat dot com>]

On 10/11/14 03:32, Joern Rennecke wrote:
> On 10 October 2014 21:13, Jeff Law <law@redhat.com> wrote:
> ...
> > ISTM it would be better to find the mode of the same class that corresponds
> > to GET_MODE_SIZE (mode) / nregs.  In your case that's obviously QImode :-)
>
> Like this?
> Or did you mean to remove the save_mode[regno] use altogether?  I can
> think of arguments for or against, but got no
> concrete examples for either.
Yea, that's basically what I was thinking.  I probably wouldn't have 
bothered with the if (hard_regno ...) check, but I can see why you might 
want that added measure of safety before slamming in a new mode.

I think you want "smode" in the mode_for_size call rather than "mode", 
right (both instances)?

If that fixes your your problem and passes the usual bootstrap and 
regression test, then it's OK with me.

I can see Richard S. getting in here one day and saying, umm, this all 
needs further refinement, but at least this hunk of code is in better shape.

jeff