On Tue, Aug 28, 2018 at 03:46:38PM +0200, Peter Zijlstra wrote:
> On Mon, Aug 27, 2018 at 02:44:57PM +1000, Nicholas Piggin wrote:
>
> > powerpc may be able to use the unmap granule thing to improve
> > its page size dependent flushes, but it might prefer to go
> > a different way and track start-end for different page sizes.
>
> I don't really see how tracking multiple ranges would help much with
> THP. The ranges would end up being almost the same if there is a good
> mix of page sizes.
>
> But something like:
>
> void tlb_flush_one(struct mmu_gather *tlb, unsigned long addr)
> {
> if (tlb->cleared_ptes && (addr << BITS_PER_LONG - PAGE_SHIFT))
> tblie_pte(addr);
> if (tlb->cleared_pmds && (addr << BITS_PER_LONG - PMD_SHIFT))
> tlbie_pmd(addr);
> if (tlb->cleared_puds && (addr << BITS_PER_LONG - PUD_SHIFT))
> tlbie_pud(addr);
> }
Sorry, those all should (of course) be !(addr << ...).
> void tlb_flush_range(struct mmu_gather *tlb)
> {
> unsigned long stride = 1UL << tlb_get_unmap_shift(tlb);
> unsigned long addr;
>
> for (addr = tlb->start; addr < tlb->end; addr += stride)
> tlb_flush_one(tlb, addr);
>
> ptesync();
> }
And one could; like x86 has; add a threshold above which you just kill
the complete mm.
