[Bug target/67492] insn-attrtab and insn-automata for arm target neede huge amount of memory to generate/compile

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

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=67492

ktkachov at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
                 CC|                            |ktkachov at gcc dot gnu.org
         Resolution|---                         |WONTFIX

--- Comment #5 from ktkachov at gcc dot gnu.org ---
(In reply to Michael Weiser from comment #4)
> Okay, I can see this is headed for a WONTFIX. Can you please answer my
> questions before we chuck it?
> 
> 1. How hard would it be or is there even some effort/patch to make gcc only
> compile in support for a selected number of CPUs of a given target,
> indirectly reducing size of insn-attrtab/insn-automata?

There's no effort to do that that I know of. There are ways to reduce the
automaton size, but that involves modifying the pipeline descriptions
(<core-name>.md files) and as such would require a lot of
benchmarking/validation to make sure it doesn't regress code quality.

> 
> 2. Incidentally: Is my patch likely to produce a seriously broken compiler?

Well, it's definitely not appropriate for submission upstream. Just removing
cores is not the way to go.

If that's a patch you want to use privately for your arm board and you only
ever expect to compile for a single -mcpu then I suppose it may work, but don't
expect any support for it ;)

> 
> Thanks!