Re: Live range splitting in new allocator

[Michael Meissner <meissner at cygnus dot com>]

On Tue, Jan 30, 2001 at 09:12:30PM +1100, Fergus Henderson wrote:
> On 29-Jan-2001, Linus Torvalds <torvalds@transmeta.com> wrote:
> > 
> > But the option should NOT be called "-fundebuggable-optimizations".
> > That's like having an option that says "-fstupid-compiler". It's not
> > undebuggable - it's just a limitation of the compiler (and/or the debug
> > interface itself, of course - some debug interfaces just can't handle
> > the level of detail you need to debug the code). Make it say
> > "-fsplit-ranges", and if people want to use it with -g, let them.
> 
> I agree that `-fundebuggable-optimizations' is not a good name.
> But the point here is that there should be a single option which
> enables all such optimizations.  The user shouldn't have to know
> about the implementation details of gcc, and `-fsplit-ranges' is
> really an implementation detail.

As I tried to say in my original message, the compiler is already broken in
terms of debug support, since you could be stopped at a random instruction, and
be at a location where reload has loaded the value into a register, and if you
tell gdb to set the value, it won't affect the current computation.  LRS just
makes it more visible.  Rather than wash my hands completely of the situation,
I would rather work towards fixing it.  Unfortunately some of the debug formats
are less than flexible (COFF, stabs without GDB extensions, possibly dwarf-1)
and must be accomidated.  Also, there is going to be work needed in GDB to
support dwarf-2 to support address ranges.

-- 
Michael Meissner, Red Hat, Inc.  (GCC group)
PMB 198, 174 Littleton Road #3, Westford, Massachusetts 01886, USA
Work:	  meissner@redhat.com		phone: +1 978-486-9304
Non-work: meissner@spectacle-pond.org	fax:   +1 978-692-4482