Re: Problem with std_expand_builtin_va_start on targets where sizetype and ptr_type_node don't have the same mode

7 msg	Crossbuild problem x86_64-pc-mingw32 related to...
1 msg	gcc-4.1-20070709 is now available
2 msg	Reply Immediately
28 msg	AMD64 ABI compatibility
5 msg	Uncatchable exceptions on some systems.
3 msg	Bootstrap failure on x86
2 msg	gcc-4.3-20070707 is now available
8 msg	abs insn with QI and HI mode
10 msg	Re: RFC: GIMPLE tuples. Design and implementati...
1 msg	gcc-4.3-20070706 is now available
10 msg	Q: middle-end problem when variadic builtins pr...
14 msg	Ongoing bootstrap failures on ppc64 since 2007-...
3 msg	X86_64 bit shifts
4 msg	RFC: Enabling DFP/libbid on dditional target (m...
2 msg	[MIPS] One test failed

Problem with std_expand_builtin_va_start on tar...
\ Zack Weinberg (5 Jul 2007)
. \ Andrew Pinski (5 Jul 2007)
. . \ Andrew Pinski (5 Jul 2007)
. . \ Zack Weinberg (5 Jul 2007)
. . . \ Zack Weinberg (6 Jul 2007)
. \ Ian Lance Taylor (6 Jul 2007)
. . \ Zack Weinberg (6 Jul 2007)
. . . \ Paolo Bonzini (6 Jul 2007)
. . . . \ Zack Weinberg (6 Jul 2007)
. . \ Zack Weinberg (6 Jul 2007)
. . . \ Ian Lance Taylor (6 Jul 2007)

7 msg	Fixing jumps reachability after block reordering
1 msg	Re: Rainer Orth appointed IRIX and Tru64 UNIX m...
3 msg	Bootstrap is broken on x86_64, comparison betwe...
11 msg	Re: PATCH: Add myself as libbid maintainer

Subject:	Re: Problem with std_expand_builtin_va_start on targets where sizetype and ptr_type_node don't have the same mode
Group:	Gcc
From:	Ian Lance Taylor
Date:	6 Jul 2007

"Zack Weinberg" <zackw> writes:

> On 05 Jul 2007 18:22:50 -0700, Ian Lance Taylor <iant> wrote:
> > "Zack Weinberg" <zackw> writes:
> >
> > [std_expand_builtin_va_start]
> >
> > > rtx va_r = expand_normal (valist);
> > > emit_move_insn (va_r, nextarg);
> > >
> > > but this is a part of the compiler I am not at all familiar with...
> >
> > I think you would want something more like
> > rtx va_r = expand_expr (valist, NULL_RTX, VOIDmode, EXPAND_WRITE);
> > convert_move (va_r, nextarg, 0);
>
> No regressions on arm-elf (simulator, default and -mthumb); in fact,
> execute/20040709-1.c stops ICEing at -O3 -fomit-frame-pointer, but I
> think this is spurious, as that test has no variadic functions in it.
> OK to commit?

Yes, this is OK. Thanks for testing it.

Ian