From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from gate.crashing.org (gate.crashing.org [63.228.1.57]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by lists.ozlabs.org (Postfix) with ESMTPS id 3x82g36r5dzDqmC for ; Fri, 14 Jul 2017 16:22:55 +1000 (AEST) Message-ID: <1500013364.2865.73.camel@kernel.crashing.org> Subject: Re: [PATCH 3/4] powerpc/mm/radix: Avoid flushing the PWC on every flush_tlb_range From: Benjamin Herrenschmidt To: "Aneesh Kumar K.V" , linuxppc-dev@lists.ozlabs.org Cc: npiggin@gmail.com Date: Fri, 14 Jul 2017 16:22:44 +1000 In-Reply-To: <87y3rrpo6n.fsf@skywalker.in.ibm.com> References: <20170714015258.7933-1-benh@kernel.crashing.org> <20170714015258.7933-3-benh@kernel.crashing.org> <87y3rrpo6n.fsf@skywalker.in.ibm.com> Content-Type: text/plain; charset="UTF-8" Mime-Version: 1.0 List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Fri, 2017-07-14 at 11:14 +0530, Aneesh Kumar K.V wrote: > > +     pid = mm ? mm->context.id : 0; > > +     if (unlikely(pid == MMU_NO_CONTEXT)) > > +             goto no_context; > > + > > +     /* 4k page size, just blow the world */ > > +     if (PAGE_SIZE == 0x1000) { > > +             radix__flush_all_mm(mm); > > +             return; > > +     } > > What is special here ? because it is 4096 tlb flushes ? Yes, keeping in line with the existing logic in flush_tlb_pages(), more than 32, do a full flush. Cheers, Ben.