Re: [PATCH 05/17] asm-generic/tlb: Rename HAVE_RCU_TABLE_NO_INVALIDATE

[Aneesh Kumar K.V]

On 12/16/19 7:10 PM, Aneesh Kumar K.V wrote:
> On 12/16/19 6:50 PM, Peter Zijlstra wrote:
> > On Mon, Dec 16, 2019 at 06:43:53PM +0530, Aneesh Kumar K.V wrote:
> > > On 12/16/19 6:07 PM, Peter Zijlstra wrote:
> >
> > > > I'm confused, are you saing you're happy to have PowerPC eat the extra
> > > > TLB invalidates? I thought you cared about PPC performance :-)
> > >
> > > Instead can we do
> > >
> > > static inline void tlb_table_invalidate(struct mmu_gather *tlb)
> > > {
> > > #ifndef CONFIG_MMU_GATHER_RCU_TABLE_FREE
> > > ÂÂÂÂ * Invalidate page-table caches used by hardware walkers. Then we 
> > > still
> > > ÂÂÂÂ * need to RCU-sched wait while freeing the pages because software
> > > ÂÂÂÂ * walkers can still be in-flight.
> > > ÂÂÂÂ */
> > > ÂÂÂÂtlb_flush_mmu_tlbonly(tlb);
> > > #endif
> > > }
> >
> > How does that not break ARM/ARM64/s390 and x86 ?
>
> Hmm I missed that usage of RCU_TABLE_NO_INVALIDATE.
>
> Ok I guess we need to revert this change that went upstream this merge 
> window then
>
> commit 52162ec784fa05f3a4b1d8e84421279998be3773
> Author: Aneesh Kumar K.V <aneesh.kumar@xxxxxxxxxxxxx>
> Date:ÂÂ Thu Oct 24 13:28:00 2019 +0530
>
>  ÂÂÂ powerpc/mm/book3s64/radix: Use freed_tables instead of need_flush_all
>
>
>
> I will review the change closely.


So __p*_free_tlb() routines on ppc64 just mark that we need a page walk 
cache flush and the actual flush in done in tlb_flush_mmu. As per

d86564a2f085b79ec046a5cba90188e61235280 (mm/tlb, x86/mm: Support 
invalidating TLB caches for RCU_TABLE_FREE ) that is not sufficient?

-aneesh