Re: [RFC PATCH] Drop some asm from copy_user_64.S

[Borislav Petkov <bp@alien8.de>]

On Tue, May 12, 2015 at 02:13:33PM -0700, Linus Torvalds wrote:
> On Tue, May 12, 2015 at 1:57 PM, Borislav Petkov <bp@alien8.de> wrote:
> >
> > I want to get rid of the asm glue in arch/x86/lib/copy_user_64.S which
> > prepares the copy_user* alternatives calls. And replace it with nice and
> > clean C.
> 
> Ack. I'm not a fan of the x86-64 usercopy funmctions.
> 
> That said, I think you should uninline those things, and move them
> from a header file to a C file (arch/x86/lib/uaccess.c?).

Ok.

> Move all the copy_user_generic_unrolled/string garbage there too, and
> keep the header file simple.

Those are just forward declarations for the asm functions in
arch/x86/lib/copy_user_64.S but yeah, I'll do some experimenting.

Just to make sure - I'm not getting rid of the different asm string copy
versions in copy_user_64.S - just the _copy_from_user/_copy_to_user
stubs at the beginning of that file as that gunk can be replaced with
calls with inlined copy_user_generic() workhorse.

The alternatives give us directly then

	CALL <optimal asm version>

which is as optimal as it gets.

> Because I think that we would actually be better off trying to inline
> the copy_user_generic_string() thing into the various versions (in
> that uaccess.c file) than try to inline the access_ok() check into the
> caller.

Right.

Thanks.

-- 
Regards/Gruss,
    Boris.

ECO tip #101: Trim your mails when you reply.
--