On Fri, Oct 03, 2008 at 05:44:31AM -0700, Arjan van de Ven wrote:
quoted text
> On Fri, 3 Oct 2008 12:25:52 +0300
> "Kirill A. Shutemov" <kirill@shutemov.name> wrote:
>=20
> > On Fri, Oct 03, 2008 at 10:02:44AM +0200, Ingo Molnar wrote:
> > >=20
> > > * Kirill A. Shutemov <kirill@shutemov.name> wrote:
> > >=20
> > > > -	/* for MAP_32BIT mappings we force the legact mmap base
> > > > */
> > > > -	if (!test_thread_flag(TIF_IA32) && (flags & MAP_32BIT))
> > > > +	/* for MAP_32BIT mappings and ADDR_LIMIT_32BIT
> > > > personality we force the
> > > > +	 * legact mmap base
> > > > +	 */
> > >=20
> > > please use the customary multi-line comment style:
> > >=20
> > >   /*
> > >    * Comment .....
> > >    * ...... goes here:
> > >    */
> > >=20
> > > and you might use the opportunity to fix the s/legact/legacy typo
> > > as well.
> >=20
> > Ok, I'll fix it.
> >=20
> > >=20
> > > but more generally, we already have ADDR_LIMIT_3GB support on x86.
> >=20
> > Does ADDR_LIMIT_3GB really work?
>=20
> if it's broken we should fix it.... not invent a new one.
> Also, traditionally often personalities only start at exec() time iirc.
> (but I could be wrong on that)

What is difference beetween ADDR_LIMIT_3GB and ADDR_LIMIT_32BIT? Probably,
I implement ADDR_LIMIT_3GB, not ADDR_LIMIT_32BIT...

--=20
Regards,  Kirill A. Shutemov
 + Belarus, Minsk
 + ALT Linux Team, http://www.altlinux.com/