Re: [PATCH] M386 flush_one_tlb invlpg

[Linus Torvalds]

On Tue, 27 Aug 2002, Hugh Dickins wrote:
> 
> I just sent the 2.4.20-pre4 asm-i386/pgtable.h patch to Marcelo:
> here's patch against 2.5.31 or current BK: please apply.

This test is senseless, in my opinion:

> +		if (cpu_has_pge)					\
> +			__flush_tlb_single(addr);			\

The test _should_ be for something like

	if (cpu_has_invlpg)
		__flush_tlb_single(addr);

since we want to use the invlpg instruction regardless of any PGE issues 
if it is available.

There's another issue, which is the fact that I do not believe that invlpg 
is even guaranteed to invalidate a G page at all - although obviously all 
current CPU's seem to work that way. However, I don't see that documented 
anywhere. It might make sense to mark the places that expect to invalidate 
a global page explicitly, and call that function "flush_one_global_rlb()" 
(even if it - at least for now - does the same thing as the regular single 
invalidate).

		Linus