[PATCH v3 6/6] powerpc/mm: Batch tlb flush when invalidating pte entries

[Aneesh Kumar K.V]

Benjamin Herrenschmidt <benh at kernel.crashing.org> writes:

> On Tue, 2016-11-22 at 00:03 +0530, Aneesh Kumar K.V wrote:
>> +               /*
>> +                * If we are trying to clear the pte, we can skip
>> +                * the below sequence and batch the tlb flush. The
>> +                * tlb flush batching is done by mmu gather code
>> +                */
>> +               if (new_pte) {
>> +                       asm volatile("ptesync" : : : "memory");
>> +                       psize = radix_get_mmu_psize(pg_sz);
>> +                       radix__flush_tlb_page_psize(mm, addr, psize);
>> +                       __radix_pte_update(ptep, 0, new_pte);
>> +               }
>>         } else
>>                 old_pte = __radix_pte_update(ptep, clr, set);
>
> Can you check the valid bit ? What if we are just setting a swap PTE
> on top of an invalid one for example ?
>
> Should the above case be limited to both old and new being valid ?

We should use set_pte_at if the old pte is not valid and that doesn't
involve a tlb flush.

-aneesh