[Bug target/30354] -Os doesn't optimize a/CONST even if it saves size.

["aldot at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=30354

--- Comment #19 from Bernhard Reutner-Fischer <aldot at gcc dot gnu.org> ---
(In reply to Denis Vlasenko from comment #17)
> Any chance of this being finally done?
> 
> I proposed a simple, working patch in 2007, it's 2016 now and all these
> years users of -Os suffer from slow divisions in important cases usch as
> "signed_int / 16" and "unsigned_int / 10".

> So.
> Any chance of this patch being accepted sometime before 2100? ;)

For GCC you need to follow https://gcc.gnu.org/contribute.html to submit
patches.

https://gcc.gnu.org/contribute.html#legal is the first obstacle unless you
happen to work for a company (like e.g. redhat) that has company-wide copyright
assignment in place. Alternatively you can put your changes in Public Domain or
go through an individual copyright assignment process for GCC with the FSF.

Then you have to create proper testcases in dejagnu form included in your patch
and have to submit that to the gcc-patches@ ML after proper bootstrapping and
regression testing. It is recommended to Cc: the maintainer(s) of the
architectures/subsystems/files you are touching in your patch submission to
gain their attention. See toplevel MAINTAINERS file.
If you want to ensure that object-size increase/regressions in your testcases
get noticed, you can add a dejagnu check for that (see
gcc/testsuite/lib/scanasm.exp:proc object-size).

For details on patch submission please refer to:

https://gcc.gnu.org/contribute.html#patches

HTH,