PR2076
Dale Johannesen
dalej@apple.com
Wed Apr 25 14:07:00 GMT 2001
On Wednesday, April 25, 2001, at 12:26 PM, Geoff Keating wrote:
>
> This sounds very much like the bug that Dale Johannesen
> <dalej@apple.com>
> was trying to fix for Darwin. The problem is that
> STRICT_ARGUMENT_NAMING
> needs to be switched on, but it needs to be done carefully so as not to
> change the ABI, and I haven't had time to work out how to get it done.
I have in fact fixed it for Darwin, and believe that fix would be
applicable to AIX.
I attempted to preserve the old code in rs6000.c for non-Darwin targets,
so as
not to break anything (further). This introduced a lot of conditionals
into the
code which will be unnecessary when it's fixed for all ABIs, and Geoff
Keating
rejected the patch on this basis. It is ugly, at that. I'll rewrite it
on the assumption
that STRICT_ARGUMENT_NAMING is on for all ABIs, and post this later
today,
assuming the regression testing passes.
More information about the Gcc
mailing list