From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-0.8 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, MAILING_LIST_MULTI,SPF_PASS,URIBL_BLOCKED autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id D8E63ECDE5F for ; Tue, 24 Jul 2018 02:29:30 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 903E820880 for ; Tue, 24 Jul 2018 02:29:30 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 903E820880 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=linux.intel.com Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=linux-kernel-owner@vger.kernel.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S2388384AbeGXDdf (ORCPT ); Mon, 23 Jul 2018 23:33:35 -0400 Received: from mga03.intel.com ([134.134.136.65]:60379 "EHLO mga03.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S2388258AbeGXDdf (ORCPT ); Mon, 23 Jul 2018 23:33:35 -0400 X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from fmsmga001.fm.intel.com ([10.253.24.23]) by orsmga103.jf.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 23 Jul 2018 19:29:27 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.51,396,1526367600"; d="scan'208";a="74906990" Received: from blu2-desk2.ccr.corp.intel.com (HELO [10.0.2.15]) ([10.239.13.1]) by fmsmga001.fm.intel.com with ESMTP; 23 Jul 2018 19:29:18 -0700 Subject: Re: [RFC PATCH 05/10] iommu/vt-d: Setup DMA remapping for mediated devices To: "Liu, Yi L" , Joerg Roedel , David Woodhouse , Alex Williamson , Kirti Wankhede References: <1532239773-15325-1-git-send-email-baolu.lu@linux.intel.com> <1532239773-15325-6-git-send-email-baolu.lu@linux.intel.com> Cc: "Raj, Ashok" , "Kumar, Sanjay K" , "Pan, Jacob jun" , "Tian, Kevin" , "Sun, Yi Y" , "peterx@redhat.com" , "iommu@lists.linux-foundation.org" , "kvm@vger.kernel.org" , "linux-kernel@vger.kernel.org" , Jacob Pan From: Lu Baolu Message-ID: <5B568EFB.2040908@linux.intel.com> Date: Tue, 24 Jul 2018 10:29:15 +0800 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.5.1 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi, On 07/23/2018 12:44 PM, Liu, Yi L wrote: >> From: Lu Baolu [mailto:baolu.lu@linux.intel.com] >> Sent: Sunday, July 22, 2018 2:09 PM >> >> This configures the second level page table when external components request to >> allocate a domain for a mediated device. > I think it should be the time when adding a mediate device to the domain. You are right. Will correct it. > >> Cc: Ashok Raj >> Cc: Jacob Pan >> Cc: Kevin Tian >> Cc: Liu Yi L >> Signed-off-by: Sanjay Kumar >> Signed-off-by: Lu Baolu >> --- >> drivers/iommu/intel-iommu.c | 73 >> ++++++++++++++++++++++++++++++++++++++++----- >> 1 file changed, 66 insertions(+), 7 deletions(-) >> >> diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c index >> 57ccfc4..b6e9ea8 100644 >> --- a/drivers/iommu/intel-iommu.c >> +++ b/drivers/iommu/intel-iommu.c >> @@ -2569,8 +2569,9 @@ static struct dmar_domain >> *dmar_insert_one_dev_info(struct intel_iommu *iommu, >> if (dev) >> dev->archdata.iommu = info; >> >> - if (dev && dev_is_pci(dev) && sm_supported(iommu)) { >> - bool pass_through; >> + if (dev && sm_supported(iommu)) { >> + bool pass_through = hw_pass_through && >> + domain_type_is_si(domain); >> >> ret = intel_pasid_alloc_table(dev); >> if (ret) { >> @@ -2579,12 +2580,21 @@ static struct dmar_domain >> *dmar_insert_one_dev_info(struct intel_iommu *iommu, >> return NULL; >> } >> >> - /* Setup the PASID entry for requests without PASID: */ >> - pass_through = hw_pass_through && domain_type_is_si(domain); >> spin_lock(&iommu->lock); >> - intel_pasid_setup_second_level(iommu, domain, dev, >> - PASID_RID2PASID, >> - pass_through); >> + >> + /* Setup the PASID entry for requests without PASID: */ >> + if (dev_is_pci(dev)) >> + intel_pasid_setup_second_level(iommu, domain, dev, >> + PASID_RID2PASID, > can domain->default_pasid be initialized to be PASID_RID2PASID if the domain > is not a domain used by mediated device? PASID_RID2PASID is 0 as well. > >> + pass_through); >> + /* Setup the PASID entry for mediated devices: */ >> + else if (dev_is_mdev(dev)) >> + intel_pasid_setup_second_level(iommu, domain, dev, >> + domain->default_pasid, >> + false); >> + else >> + pr_err("Unsupported device %s\n", dev_name(dev)); >> + >> spin_unlock(&iommu->lock); >> } >> spin_unlock_irqrestore(&device_domain_lock, flags); @@ -4937,6 +4947,32 >> @@ static void domain_context_clear(struct intel_iommu *iommu, struct device >> *dev) >> pci_for_each_dma_alias(to_pci_dev(dev), &domain_context_clear_one_cb, >> iommu); } >> >> +static void >> +iommu_flush_ext_iotlb(struct intel_iommu *iommu, u16 did, u32 pasid, > Per latest VT-d spec, its new name is pasid-based-iotlb. Pls update the naming > accordingly. :) Okay. Best regards, Lu Baolu > > Regards, > Yi Liu > >> +u64 gran) { >> + struct qi_desc desc; >> + >> + desc.high = 0; >> + desc.low = QI_EIOTLB_PASID(pasid) | QI_EIOTLB_DID(did) | >> + QI_EIOTLB_GRAN(gran) | QI_EIOTLB_TYPE; >> + >> + qi_submit_sync(&desc, iommu); >> +} >> + >> +static void iommu_flush_pasid_dev_iotlb(struct intel_iommu *iommu, >> + struct device *dev, int sid, int pasid) { >> + struct qi_desc desc; >> + struct pci_dev *pdev = to_pci_dev(dev); >> + int qdep = pci_ats_queue_depth(pdev); >> + >> + desc.low = QI_DEV_EIOTLB_PASID(pasid) | QI_DEV_EIOTLB_SID(sid) | >> + QI_DEV_EIOTLB_QDEP(qdep) | QI_DEIOTLB_TYPE; >> + desc.high = QI_DEV_EIOTLB_ADDR(-1ULL >> 1) | QI_DEV_EIOTLB_SIZE; >> + >> + qi_submit_sync(&desc, iommu); >> +} >> + >> static void __dmar_remove_one_dev_info(struct device_domain_info *info) { >> struct intel_iommu *iommu; >> @@ -4949,6 +4985,29 @@ static void __dmar_remove_one_dev_info(struct >> device_domain_info *info) >> >> iommu = info->iommu; >> >> + if (dev_is_mdev(info->dev)) { >> + struct dmar_domain *domain = info->domain; >> + int did = domain->iommu_did[iommu->seq_id]; >> + int sid = info->bus << 8 | info->devfn; >> + struct device *dev = info->dev; >> + >> + intel_pasid_clear_entry(dev, domain->default_pasid); >> + >> + /* Flush IOTLB including PASID Cache: */ >> + iommu->flush.flush_iotlb(iommu, did, 0, 0, DMA_TLB_DSI_FLUSH); >> + >> + /* >> + * Flush EIOTLB. The only way to flush global mappings within >> + * a PASID is to use QI_GRAN_ALL_ALL. >> + */ >> + iommu_flush_ext_iotlb(iommu, did, domain->default_pasid, >> + QI_GRAN_ALL_ALL); >> + >> + /* Flush Dev TLB: */ >> + iommu_flush_pasid_dev_iotlb(iommu, dev_mdev_parent(dev), sid, >> + domain->default_pasid); >> + } >> + >> if (info->dev) { >> iommu_disable_dev_iotlb(info); >> domain_context_clear(iommu, info->dev); >> -- >> 2.7.4 >