Re: [RFC PATCH v4 2/2] arm64: tlb: Use the TLBI RANGE feature in arm64

[Robin Murphy]

On 2020-07-07 18:46, Catalin Marinas wrote:
> On Tue, Jul 07, 2020 at 06:43:35PM +0100, Marc Zyngier wrote:
> > On 2020-07-07 18:36, Catalin Marinas wrote:
> > > On Mon, Jun 01, 2020 at 10:47:13PM +0800, Zhenyu Ye wrote:
> > > > @@ -59,6 +69,47 @@
> > > >   		__ta;						\
> > > >   	})
> > > >
> > > > +/*
> > > > + * __TG defines translation granule of the system, which is decided
> > > > by
> > > > + * PAGE_SHIFT.  Used by TTL.
> > > > + *  - 4KB	: 1
> > > > + *  - 16KB	: 2
> > > > + *  - 64KB	: 3
> > > > + */
> > > > +#define __TG	((PAGE_SHIFT - 12) / 2 + 1)
> > >
> > > Nitpick: maybe something like __TLBI_TG to avoid clashes in case someone
> > > else defines a __TG macro.
> >
> > I have commented on this in the past, and still maintain that this
> > would be better served by a switch statement similar to what is used
> > for TTL already (I don't think this magic formula exists in the
> > ARM ARM).
>
> Good point, it would be cleaner indeed.

FWIW, we use the somewhat obtuse "(shift - 10) / 2" in the SMMUv3 
driver, but that's because we support multiple different granules at 
runtime and want to generate efficient code. Anything based on 
PAGE_SHIFT that resolves to a compile-time constant has no excuse for 
not being written in a clear and obvious manner ;)

Robin.