[doc] : Document -fomit-frame-pointer default change for 32bit Linux and Darwin targets.

On Sun, Aug 15, 2010 at 9:14 PM, Gerald Pfeifer <gerald@pfeifer.com> wrote:

>> Attached patch adds documentation about -fomit-frame-pointer default
>> change on selected x86 targets.
>
> Thanks for thinking of this, Uros!  Basically the same text looks good
> for the GCC 4.6 release notes (http://gcc.gnu.org/gcc-4.6/changes.html)
> I think, and I can apply it there for you.

Thanks for your kind review. I would really like to ask you to apply
the text to HTML, since my CVS skills are a bit rusty (and I don't
have a web CVS repository anymore since my last OS install).

>>       * doc/invoke.texi (-fomit-frame-pointer): Document that starting
>>       from GCC version 4.6, the default setting (when not optimizing
>
> Is it starting "from" or "with"?  Native speakers welcome! :-)

IRC says that we can live with "with". Also some non-scientific measurements:

$ grep -i "starting with gcc" doc/*.texi | wc -l
3
$ grep -i "starting from gcc" doc/*.texi | wc -l
1
$ grep -i "starting at gcc" doc/*.texi | wc -l
0

>>       for size) for 32-bit Linux x86 and 32-bit Darwin x86 targets has
>>       been changed to -fomit-frame-pointer.
>
> GNU/Linux or really Linux (that is, everything running a Linux kernel)?

It is for everything runing a Linux kernel. In fact,
-fomit-frame-pointer can be enabled for all x86_32 targets, the
subtarget can be added to line 1583 of gcc/configure.c. ATM this part
reads:

AC_ARG_ENABLE(frame-pointer,
[  --enable-frame-pointer  enable -fno-omit-frame-pointer by default
for 32bit x86], [],
[
case $target_os in
linux* | darwin[[8912]]*)
  # Enable -fomit-frame-pointer by default for Linux and Darwin with
  # DWARF2.
  enable_frame_pointer=no
  ;;
*)
  enable_frame_pointer=yes
  ;;
esac
])

>
> +Starting from GCC version 4.6, the default setting (when not optimizing for
> +size) for 32-bit Linux x86 and 32-bit Darwin x86 targets has been changed to
> +@option{-fomit-frame-pointer}.  New behavior can be reverted back to
> +@option{-fno-omit-frame-pointer} by configuring GCC with the
> +@option{--enable-frame-pointer} configure option.
>
> Same comments as for the ChangeLog apply here.
>
>
> And I suggest to slightly change the last sentence a bit to read
>
>  "The default can be reverted to @option{-fno-omit-frame-pointer} by
>  configuring..."
>
> but that's more a question of style, and thus personal preference.

I think that your version sounds better, so I took the liberty and
commit attached patch (with corrected ChangeLog).

Thanks,
Uros.