From patchwork Mon Nov 14 12:36:33 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Robin Murphy X-Patchwork-Id: 9427433 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id C800E6047D for ; Mon, 14 Nov 2016 12:37:34 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id C1CE3287B2 for ; Mon, 14 Nov 2016 12:37:34 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id B515D2892C; Mon, 14 Nov 2016 12:37:34 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-6.9 required=2.0 tests=BAYES_00,RCVD_IN_DNSWL_HI autolearn=ham version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id D7DBA287B2 for ; Mon, 14 Nov 2016 12:37:33 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S934051AbcKNMhM (ORCPT ); Mon, 14 Nov 2016 07:37:12 -0500 Received: from foss.arm.com ([217.140.101.70]:56666 "EHLO foss.arm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933871AbcKNMhI (ORCPT ); Mon, 14 Nov 2016 07:37:08 -0500 Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.72.51.249]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id B821CCF6; Mon, 14 Nov 2016 04:37:07 -0800 (PST) Received: from [10.1.211.71] (e104324-lin.cambridge.arm.com [10.1.211.71]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPSA id E5CC63F24D; Mon, 14 Nov 2016 04:37:04 -0800 (PST) Subject: Re: [RFC v2 3/8] iommu/dma: Allow MSI-only cookies To: Eric Auger , eric.auger.pro@gmail.com, christoffer.dall@linaro.org, marc.zyngier@arm.com, alex.williamson@redhat.com, will.deacon@arm.com, joro@8bytes.org, tglx@linutronix.de, jason@lakedaemon.net, linux-arm-kernel@lists.infradead.org References: <1478258646-3117-1-git-send-email-eric.auger@redhat.com> <1478258646-3117-4-git-send-email-eric.auger@redhat.com> Cc: kvm@vger.kernel.org, drjones@redhat.com, linux-kernel@vger.kernel.org, pranav.sawargaonkar@gmail.com, iommu@lists.linux-foundation.org, punit.agrawal@arm.com, diana.craciun@nxp.com From: Robin Murphy Message-ID: <5f93ebfd-edf4-0b5a-b54a-b96937a588b8@arm.com> Date: Mon, 14 Nov 2016 12:36:33 +0000 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Thunderbird/45.4.0 MIME-Version: 1.0 In-Reply-To: <1478258646-3117-4-git-send-email-eric.auger@redhat.com> Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP On 04/11/16 11:24, Eric Auger wrote: > From: Robin Murphy > > IOMMU domain users such as VFIO face a similar problem to DMA API ops > with regard to mapping MSI messages in systems where the MSI write is > subject to IOMMU translation. With the relevant infrastructure now in > place for managed DMA domains, it's actually really simple for other > users to piggyback off that and reap the benefits without giving up > their own IOVA management, and without having to reinvent their own > wheel in the MSI layer. > > Allow such users to opt into automatic MSI remapping by dedicating a > region of their IOVA space to a managed cookie. > > Signed-off-by: Robin Murphy > Signed-off-by: Eric Auger OK, following the discussion elsewhere I've had a go at the less stupid, but more involved, version. Thoughts? Robin. ----->8----- From: Robin Murphy Subject: [RFC PATCH] iommu/dma: Allow MSI-only cookies IOMMU domain users such as VFIO face a similar problem to DMA API ops with regard to mapping MSI messages in systems where the MSI write is subject to IOMMU translation. With the relevant infrastructure now in place for managed DMA domains, it's actually really simple for other users to piggyback off that and reap the benefits without giving up their own IOVA management, and without having to reinvent their own wheel in the MSI layer. Allow such users to opt into automatic MSI remapping by dedicating a region of their IOVA space to a managed cookie, and extend the mapping routine to implement a trivial linear allocator in such cases, to avoid the needless overhead of a full-blown IOVA domain. Signed-off-by: Robin Murphy --- drivers/iommu/dma-iommu.c | 118 ++++++++++++++++++++++++++++++++++++---------- include/linux/dma-iommu.h | 6 +++ 2 files changed, 100 insertions(+), 24 deletions(-) diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c index c5ab8667e6f2..33d66a8273c6 100644 --- a/drivers/iommu/dma-iommu.c +++ b/drivers/iommu/dma-iommu.c @@ -37,10 +37,19 @@ struct iommu_dma_msi_page { phys_addr_t phys; }; +enum iommu_dma_cookie_type { + IOMMU_DMA_IOVA_COOKIE, + IOMMU_DMA_MSI_COOKIE, +}; + struct iommu_dma_cookie { - struct iova_domain iovad; - struct list_head msi_page_list; - spinlock_t msi_lock; + union { + struct iova_domain iovad; + dma_addr_t msi_iova; + }; + struct list_head msi_page_list; + spinlock_t msi_lock; + enum iommu_dma_cookie_type type; }; static inline struct iova_domain *cookie_iovad(struct iommu_domain *domain) @@ -53,6 +62,19 @@ int iommu_dma_init(void) return iova_cache_get(); } +static struct iommu_dma_cookie *__cookie_alloc(enum iommu_dma_cookie_type type) +{ + struct iommu_dma_cookie *cookie; + + cookie = kzalloc(sizeof(*cookie), GFP_KERNEL); + if (cookie) { + spin_lock_init(&cookie->msi_lock); + INIT_LIST_HEAD(&cookie->msi_page_list); + cookie->type = type; + } + return cookie; +} + /** * iommu_get_dma_cookie - Acquire DMA-API resources for a domain * @domain: IOMMU domain to prepare for DMA-API usage @@ -62,25 +84,53 @@ int iommu_dma_init(void) */ int iommu_get_dma_cookie(struct iommu_domain *domain) { - struct iommu_dma_cookie *cookie; - if (domain->iova_cookie) return -EEXIST; - cookie = kzalloc(sizeof(*cookie), GFP_KERNEL); - if (!cookie) + domain->iova_cookie = __cookie_alloc(IOMMU_DMA_IOVA_COOKIE); + if (!domain->iova_cookie) return -ENOMEM; - spin_lock_init(&cookie->msi_lock); - INIT_LIST_HEAD(&cookie->msi_page_list); - domain->iova_cookie = cookie; return 0; } EXPORT_SYMBOL(iommu_get_dma_cookie); /** + * iommu_get_msi_cookie - Acquire just MSI remapping resources + * @domain: IOMMU domain to prepare + * @base: Start address of IOVA region for MSI mappings + * + * Users who manage their own IOVA allocation and do not want DMA API support, + * but would still like to take advantage of automatic MSI remapping, can use + * this to initialise their own domain appropriately. Users should reserve a + * contiguous IOVA region, starting at @base, large enough to accommodate the + * number of PAGE_SIZE mappings necessary to cover every MSI doorbell address + * used by the devices attached to @domain. + */ +int iommu_get_msi_cookie(struct iommu_domain *domain, dma_addr_t base) +{ + struct iommu_dma_cookie *cookie; + + if (domain->type != IOMMU_DOMAIN_UNMANAGED) + return -EINVAL; + + if (domain->iova_cookie) + return -EEXIST; + + cookie = __cookie_alloc(IOMMU_DMA_IOVA_COOKIE); + if (!cookie) + return -ENOMEM; + + cookie->msi_iova = base; + domain->iova_cookie = cookie; + return 0; +} +EXPORT_SYMBOL(iommu_get_msi_cookie); + +/** * iommu_put_dma_cookie - Release a domain's DMA mapping resources - * @domain: IOMMU domain previously prepared by iommu_get_dma_cookie() + * @domain: IOMMU domain previously prepared by iommu_get_dma_cookie() or + * iommu_get_msi_cookie() * * IOMMU drivers should normally call this from their domain_free callback. */ @@ -92,7 +142,7 @@ void iommu_put_dma_cookie(struct iommu_domain *domain) if (!cookie) return; - if (cookie->iovad.granule) + if (cookie->type == IOMMU_DMA_IOVA_COOKIE && cookie->iovad.granule) put_iova_domain(&cookie->iovad); list_for_each_entry_safe(msi, tmp, &cookie->msi_page_list, list) { @@ -137,11 +187,12 @@ static void iova_reserve_pci_windows(struct pci_dev *dev, int iommu_dma_init_domain(struct iommu_domain *domain, dma_addr_t base, u64 size, struct device *dev) { - struct iova_domain *iovad = cookie_iovad(domain); + struct iommu_dma_cookie *cookie = domain->iova_cookie; + struct iova_domain *iovad = &cookie->iovad; unsigned long order, base_pfn, end_pfn; - if (!iovad) - return -ENODEV; + if (!cookie || cookie->type != IOMMU_DMA_IOVA_COOKIE) + return -EINVAL; /* Use the smallest supported page size for IOVA granularity */ order = __ffs(domain->pgsize_bitmap); @@ -644,11 +695,21 @@ static struct iommu_dma_msi_page *iommu_dma_get_msi_page(struct device *dev, { struct iommu_dma_cookie *cookie = domain->iova_cookie; struct iommu_dma_msi_page *msi_page; - struct iova_domain *iovad = &cookie->iovad; + struct iova_domain *iovad; struct iova *iova; int prot = IOMMU_WRITE | IOMMU_NOEXEC | IOMMU_MMIO; + size_t size; + + if (cookie->type == IOMMU_DMA_IOVA_COOKIE) { + iovad = &cookie->iovad; + size = iovad->granule; + } else { + iovad = NULL; + size = PAGE_SIZE; + } + + msi_addr &= ~(phys_addr_t)(size - 1); - msi_addr &= ~(phys_addr_t)iova_mask(iovad); list_for_each_entry(msi_page, &cookie->msi_page_list, list) if (msi_page->phys == msi_addr) return msi_page; @@ -657,13 +718,18 @@ static struct iommu_dma_msi_page *iommu_dma_get_msi_page(struct device *dev, if (!msi_page) return NULL; - iova = __alloc_iova(domain, iovad->granule, dma_get_mask(dev)); - if (!iova) - goto out_free_page; - msi_page->phys = msi_addr; - msi_page->iova = iova_dma_addr(iovad, iova); - if (iommu_map(domain, msi_page->iova, msi_addr, iovad->granule, prot)) + if (iovad) { + iova = __alloc_iova(domain, size, dma_get_mask(dev)); + if (!iova) + goto out_free_page; + msi_page->iova = iova_dma_addr(iovad, iova); + } else { + msi_page->iova = cookie->msi_iova; + cookie->msi_iova += size; + } + + if (iommu_map(domain, msi_page->iova, msi_addr, size, prot)) goto out_free_iova; INIT_LIST_HEAD(&msi_page->list); @@ -671,7 +737,10 @@ static struct iommu_dma_msi_page *iommu_dma_get_msi_page(struct device *dev, return msi_page; out_free_iova: - __free_iova(iovad, iova); + if (iovad) + __free_iova(iovad, iova); + else + cookie->msi_iova -= size; out_free_page: kfree(msi_page); return NULL; @@ -716,3 +785,4 @@ void iommu_dma_map_msi_msg(int irq, struct msi_msg *msg) msg->address_lo += lower_32_bits(msi_page->iova); } } + diff --git a/include/linux/dma-iommu.h b/include/linux/dma-iommu.h index 32c589062bd9..d69932474576 100644 --- a/include/linux/dma-iommu.h +++ b/include/linux/dma-iommu.h @@ -27,6 +27,7 @@ int iommu_dma_init(void); /* Domain management interface for IOMMU drivers */ int iommu_get_dma_cookie(struct iommu_domain *domain); +int iommu_get_msi_cookie(struct iommu_domain *domain, dma_addr_t base); void iommu_put_dma_cookie(struct iommu_domain *domain); /* Setup call for arch DMA mapping code */ @@ -82,6 +83,11 @@ static inline int iommu_get_dma_cookie(struct iommu_domain *domain) return -ENODEV; } +static inline int iommu_get_msi_cookie(struct iommu_domain *domain, dma_addr_t base) +{ + return -ENODEV; +} + static inline void iommu_put_dma_cookie(struct iommu_domain *domain) { }