This is the mail archive of the gdb-patches@sources.redhat.com mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [RFA] sh-tdep.c: Fix float handling


On Fri, Oct 10, 2003 at 04:25:42PM -0400, Elena Zannoni wrote:
> Corinna Vinschen writes:
>  > >  > First, in contrast to non-FPU CPUs, arguments are never split between
>  > >  > registers and stack.  If an argument doesn't fit in the remaining registers
>  > >  > it's always pushed entirely on the stack.  My testing led me to the wrong
>  > >  > assumption that just types > 16 bytes are pushed entirely on the stack.
>  > 
>  > I have a problem with this first part of the comment.  This one does
>  > not fit well as a comment to the helper function.  I would suggest to
>  > add this (carefully rephrased) to sh_push_dummy_call_fpu(), substituting
>  > the very short
>  > 
>  >   /* No data is passed partially in registers.  */
>  > 
> 
> Where is that? I think you can add it before the function
> sh_push_dummy_call_fpu. Or where it makes more sense. I don't care, as
> long as the whole thing is as clear as possible. It is wasy to
> understand the code now, but I bet a few months from now we will all
> have forgotten the details.

I've applied this with the comment at the point I think it fit most well.
I agree to what you're saying, therefore I'm always trying to put some
additional comments in.

The above comment is now inline, beginnig at line 890 in
sh_push_dummy_call_fpu(), exactly above the code which cares for this
very situation.

Corinna

-- 
Corinna Vinschen
Cygwin Developer
Red Hat, Inc.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]