From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx0a-001b2d01.pphosted.com (mx0a-001b2d01.pphosted.com [148.163.156.1]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by lists.ozlabs.org (Postfix) with ESMTPS id 3wJhjw0qsjzDqBJ for ; Fri, 5 May 2017 03:25:07 +1000 (AEST) Received: from pps.filterd (m0098396.ppops.net [127.0.0.1]) by mx0a-001b2d01.pphosted.com (8.16.0.20/8.16.0.20) with SMTP id v44HNwSp108355 for ; Thu, 4 May 2017 13:24:58 -0400 Received: from e06smtp12.uk.ibm.com (e06smtp12.uk.ibm.com [195.75.94.108]) by mx0a-001b2d01.pphosted.com with ESMTP id 2a86ge6r5a-1 (version=TLSv1.2 cipher=AES256-SHA bits=256 verify=NOT) for ; Thu, 04 May 2017 13:24:57 -0400 Received: from localhost by e06smtp12.uk.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Thu, 4 May 2017 18:24:55 +0100 Subject: Re: [RFC 1/2] powerpc/mm: Add marker for contexts requiring global TLB invalidations To: "Aneesh Kumar K.V" , alistair@popple.id.au, bsingharora@gmail.com, linuxppc-dev@lists.ozlabs.org, mpe@ellerman.id.au References: <20170503142906.29220-1-fbarrat@linux.vnet.ibm.com> <20170503142906.29220-2-fbarrat@linux.vnet.ibm.com> <87inlhp1iq.fsf@skywalker.in.ibm.com> From: Frederic Barrat Date: Thu, 4 May 2017 19:24:51 +0200 MIME-Version: 1.0 In-Reply-To: <87inlhp1iq.fsf@skywalker.in.ibm.com> Content-Type: text/plain; charset=windows-1252; format=flowed Message-Id: <848dceb7-1814-6123-df67-008db71ace0d@linux.vnet.ibm.com> List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Le 04/05/2017 à 08:41, Aneesh Kumar K.V a écrit : > Frederic Barrat writes: > >> Introduce a new 'flags' attribute per context and define its first bit >> to be a marker requiring all TLBIs for that context to be broadcasted >> globally. Once that marker is set on a context, it cannot be removed. >> >> Such a marker is useful for memory contexts used by devices behind the >> NPU and CAPP/PSL. The NPU and the PSL keep their own >> translation cache so they need to see all the TLBIs for those >> contexts. > > Can we also switch existing cxl_ctx_in_use() to this ? That was my initial intent. But in the hash code, when calling the tlbie, it seems that we no longer have the related context handy. Or did I miss it? so it would require quite a bit of changes. So I've just focused on fixing the pb for radix for the time being. That being said, we'll have to update what we do for hash if we ever want to support opencapi with hash/powervm (which seems like a strong possibility for next year), as we could have more than one opencapi drivers. Fred