This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: PR7257-related inline flags cleanup
On Monday 29 November 2004 21:59, Richard Henderson wrote:
> On Wed, Nov 24, 2004 at 09:05:25PM +0100, Steven Bosscher wrote:
> > * common.opt (flag_no_inline): Don't set flag_no_inline to 0 when
> > * dojump.c (clear_pending_stack_adjust): Ignore flag_no_inline
> > check that only existed for the RTL inliner.
> > passed on the command line. Don't initialize it to 2 by default.
> > * c-opts.c (c_common_post_options): Don't set flag_no_inline to
> > disable the RTL inliner.
> > * flags.h (flag_really_no_inline): Remove extern declaration.
> > * opts.c (decode_options): Don't touch flag_no_inline to set or
> > clear flag_really_no_inline.
> > * toplev.c (flag_really_no_inline): Removed.
> > * c-cppbuiltin.c (c_cpp_builtins): Replace flag_really_no_inline
> > with just flag_no_inline.
> > * c-objc-common.c (c_disregard_inline_limits): Likewise.
> > (c_cannot_inline_tree_fn): Likewise.
> > * cgraph.c (cgraph_function_possibly_inlined_p): Likewise.
> > * cgraphunit.c (cgraph_analyze_function): Likewise.
> > (cgraph_decide_inlining): Likewise.
> > (cgraph_decide_inlining_incrementally): Likewise.
> > (cgraph_preserve_function_body_p): Likewise.
> > * langhooks.c (lhd_tree_inlining_cannot_inline_tree_fn): Likewise.
> >
> > ada/
> > * misc.c (gnat_post_options): Don't set flag_no_inline to disable
> > the RTL inliner.
> >
> > cp/
> > * decl.c (cxx_init_decl_processing): Don't set flag_no_inline to
> > disable the RTL inliner.
> > * tree.c (cp_cannot_inline_tree_fn): Check flag_no_inline instead
> > of flag_really_no_inline.
> >
> > fortran/
> > * options.c (gfc_post_options): Don't set flag_no_inline to disable
> > the RTL inliner.
> >
> > java/
> > * lang.c (java_post_options): Don't set flag_no_inline to disable
> > the RTL inliner.
>
> Ok.
With this patch applied, I now have one new failure that I did
not have before:
gcc.dg/debug/dwarf2/dwarf-die6.c scan-assembler-not DW_AT_inline
So I'm not going to commit the patch. I'll try to figure out what
happens, hopefully it's something simple...
Gr.
Steven