Re: [commit, 6.0] Fix fnchange ARI problems

[Daniel Jacobowitz <drow@mvista.com>]

On Mon, Aug 04, 2003 at 12:54:52PM -0400, Andrew Cagney wrote:
> [please don't edit me out of To: or Cc: lines - I'll otherwize won't see 
> it at present :-(]
> 
> >>+  static long dummy[4] = { 0xffffffff, 0xffffffff, 0xffffffff, 
> >>0xffffffff };
> >>+  static long mxcsr = 0x1f80;
> >>+  int reg;
> >>+
> >>+  for (reg = 0; reg < tdep->num_xmm_regs; reg++)
> >>+    supply_register (XMM0_REGNUM + reg, (char *) dummy);
> >>+  if (tdep->num_xmm_regs > 0)
> >>+    supply_register (MXCSR_REGNUM, (char *) &mxcsr);
> >>+}
> >
> >Won't work on a big-endian LP64 host.  You've filled them with 0
> >instead of -1 if I remember my C promotions right.
> 
> The function was cut/paste from the i386-linux-nat.c, so not me.  I'd 
> better fix it's portability problems though.

Right, not accusing.  But this is part of moving the code from -nat to
-tdep.  To be fair, I noticed it when reading your patch, but mine has
the same bug :)

> >Otherwise, I have some code to do the same thing, and it looks exactly
> >the same, which is a good sign :)  I'm too backlogged waiting for
> >review of things I've already posted to realistically flush the rest of
> >my queue, so I'm glad to see someone else save me the trouble.
> 
> Exactly how old's the patch?  You should just push these out.

Sending dozens of patches to the list has, historically, not done the
slightest bit of good.  Lately, sending single patches and repeated
reminders hasn't either.  I'm too frustrated to add another patch to
the list I have to ping every month.

When my outstanding patches are approved or at least responded to, then
I will send more.

I don't suppose you have any plans to framify MIPS?  One of my
least-ready-for-submission patches adds signal handler unwinding for
mips-linux; it's hideous, and with the new frame code it would be much
cleaner.


-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer