Re: [PATCH v2] MIPS/bpf: Enable bpf_probe_read{, str}() on MIPS again

[Thomas Bogendoerfer <tsbogend@alpha.franken.de>]

On Mon, Mar 22, 2021 at 03:12:59PM +0800, Tiezhu Yang wrote:
> On 03/22/2021 12:46 PM, Maciej W. Rozycki wrote:
> > On Thu, 18 Mar 2021, Tiezhu Yang wrote:
> > 
> > > diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
> > > index 160b3a8..4b94ec7 100644
> > > --- a/arch/mips/Kconfig
> > > +++ b/arch/mips/Kconfig
> > > @@ -6,6 +6,7 @@ config MIPS
> > >   	select ARCH_BINFMT_ELF_STATE if MIPS_FP_SUPPORT
> > >   	select ARCH_HAS_FORTIFY_SOURCE
> > >   	select ARCH_HAS_KCOV
> > > +	select ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE
> >   Hmm, documentation on ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE seems rather
> > scarce, but based on my guess shouldn't this be "if !EVA"?
> > 
> >    Maciej
> 
> I do not quite know what the effect if MIPS EVA (Enhanced Virtual
> Addressing)
> is set, I saw that ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE should be
> restricted
> to archs with non-overlapping address ranges.
> 
> I wonder whether MIPS EVA will generate overlapping address ranges?

they can overlap in EVA mode.

> If yes, it is better to make ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE depend
> on !EVA on MIPS.

Could please add the change ?

Thomas.

-- 
Crap can work. Given enough thrust pigs will fly, but it's not necessarily a
good idea.                                                [ RFC1925, 2.3 ]