Re: Support SVM without PASID

On 8/2/2017 12:10 AM, Jean-Philippe Brucker wrote: On 01/08/17 18:38, valmiki wrote: [...] So i digged through your patches and i understood that using BIND ioctls satge-1 translations are setup in SMMU for an application. If we use VFIO_IOMMU_MAP/UNMAP_DMA ioctls they are setting up stage-2

[RFC] virtio-iommu v0.4 - Implementation notes

The following is roughly the content of topology.tex and MSI.tex --- \section{Implementation notes}\label{sec:viommu} \subsection{Virtual system topology}\label{sec:viommu / Virtual topology} \subsubsection{Example virtual topology}\label{sec:viommu / Virtual topology / Example}

[RFC] virtio-iommu v0.4 - IOMMU Device

The following is roughly the content of device-operations.tex --- \section{IOMMU device}\label{sec:Device Types / IOMMU Device} The virtio-iommu device manages Direct Memory Access (DMA) from one or more endpoints. It may act as a proxy for multiple physical IOMMUs managing devices assigned to

[RFC] virtio-iommu version 0.4

This is the continuation of my proposal for virtio-iommu, the para- virtualized IOMMU. Here is a summary of the changes since last time [1]: * The virtio-iommu document now resembles an actual specification. It is split into a formal description of the virtio device, and implementation notes.

Re: [PATCH] iommu/arm-smmu: Defer TLB flush in case of unmap op

On 03/08/17 06:35, Vivek Gautam wrote: > Hi Robin, > > > > On 08/02/2017 05:47 PM, Robin Murphy wrote: >> On 02/08/17 10:53, Vivek Gautam wrote: >>> We don't want to touch the TLB when smmu is suspended. >>> Defer it until resume. >>> >>> Signed-off-by: Vivek Gautam

Re: [PATCH v5 1/2] ACPI/IORT: Add ITS address regions reservation helper

On 01/08/17 11:49, Shameer Kolothum wrote: > On some platforms ITS address regions have to be excluded from normal > IOVA allocation in that they are detected and decoded in a HW specific > way by system components and so they cannot be considered normal IOVA > address space. > > Add an helper

[PATCH v2] iommu/of: Fix of_iommu_configure() for disabled IOMMUs

Sudeep reports that the logic got slightly broken when a PCI iommu-map entry targets an IOMMU marked as disabled in DT, since of_pci_map_rid() succeeds in following a phandle, and of_iommu_xlate() doesn't return an error value, but we miss checking whether ops was actually non-NULL. Whilst this

Re: [PATCH] iommu/of: Fix of_iommu_configure() for disabled IOMMUs

On 04/08/17 12:16, Robin Murphy wrote: > Sudeep reports that the logic is slightly broken when a PCI iommu-map > entry targets an IOMMU marked as disabled in DT, since of_pci_map_rid() > succeeds in following a phandle, and of_iommu_xlate() doesn't return an > error value, but we miss checking

Re: [PATCH v3 0/4] Add larbid init routine for mediatek's gen1 smi larb driver

On 08/04/2017 12:06 PM, Joerg Roedel wrote: On Fri, Aug 04, 2017 at 09:32:24AM +0800, honghui.zh...@mediatek.com wrote: Honghui Zhang (4): memory: mtk-smi: Use of_device_get_match_data helper memory: mtk-smi: add larbid handle routine dt-bindings: mediatek: add descriptions for

Re: [PATCH v9 12/13] iommu/amd: Clear out the GV flag when handle deferred domain attach

On Fri, Aug 04, 2017 at 09:13:20PM +0800, Baoquan He wrote: > On 08/04/17 at 02:30pm, Joerg Roedel wrote: > > On Tue, Aug 01, 2017 at 07:37:28PM +0800, Baoquan He wrote: > > > @@ -2466,11 +2472,21 @@ static struct protection_domain > > > *get_domain(struct device *dev) > > > return

Re: [PATCH v9 12/13] iommu/amd: Clear out the GV flag when handle deferred domain attach

On 08/04/17 at 02:30pm, Joerg Roedel wrote: > On Tue, Aug 01, 2017 at 07:37:28PM +0800, Baoquan He wrote: > > @@ -2466,11 +2472,21 @@ static struct protection_domain *get_domain(struct > > device *dev) > > return ERR_PTR(-EINVAL); > > > > domain = get_dev_data(dev)->domain; > >

Re: [PATCH v9 06/13] iommu/amd: copy old trans table from old kernel

On 08/04/17 at 02:21pm, Joerg Roedel wrote: > On Tue, Aug 01, 2017 at 07:37:22PM +0800, Baoquan He wrote: > > static void init_device_table_dma(void) > > { > > @@ -2137,9 +2140,49 @@ static void early_enable_iommu(struct amd_iommu > > *iommu) > > static void early_enable_iommus(void) > > { >

Re: [PATCH v9 10/13] iommu/amd: Allocate memory below 4G for dev table if translation pre-enabled

On 08/04/17 at 02:25pm, Joerg Roedel wrote: > On Tue, Aug 01, 2017 at 07:37:26PM +0800, Baoquan He wrote: > > AMD pointed out it's unsafe to update the device-table while iommu > > is enabled. It turns out that device-table pointer update is split > > up into two 32bit writes in the IOMMU

Re: [PATCH v9 05/13] iommu/amd: Add function copy_dev_tables()

On 08/04/17 at 02:51pm, Joerg Roedel wrote: > On Fri, Aug 04, 2017 at 08:30:39PM +0800, Baoquan He wrote: > > Sorry, I don't get 'this one' meaning, are you suggesting the copy for > > loop should be take out of the iommu for loop? > > > > About the temporary copy of the old device-table, you

Re: [PATCH] iommu/of: Fix of_iommu_configure() for disabled IOMMUs

On 04/08/17 12:16, Robin Murphy wrote: > Sudeep reports that the logic is slightly broken when a PCI iommu-map > entry targets an IOMMU marked as disabled in DT, since of_pci_map_rid() > succeeds in following a phandle, and of_iommu_xlate() doesn't return an > error value, but we miss checking

Re: [PATCH v9 05/13] iommu/amd: Add function copy_dev_tables()

On Fri, Aug 04, 2017 at 08:30:39PM +0800, Baoquan He wrote: > Sorry, I don't get 'this one' meaning, are you suggesting the copy for > loop should be take out of the iommu for loop? > > About the temporary copy of the old device-table, you can see in patch > 7/13, if irq sanity check failed, it

Re: [PATCH v9 05/13] iommu/amd: Add function copy_dev_tables()

On 08/04/17 at 02:09pm, Joerg Roedel wrote: > > On Tue, Aug 01, 2017 at 07:37:21PM +0800, Baoquan He wrote: > > + for_each_iommu(iommu) { .. > > + if (copied) > > + continue; > > + > > + old_devtb_phys = entry & PAGE_MASK; > > +

Re: [PATCH v9 11/13] iommu/amd: Don't copy GCR3 table root pointer

On Tue, Aug 01, 2017 at 07:37:27PM +0800, Baoquan He wrote: > @@ -908,6 +911,14 @@ static int copy_device_table(void) > old_dev_tbl_cpy[devid].data[1] > = old_devtb[devid].data[1]; > __set_bit(dom_id,

Re: [PATCH v9 12/13] iommu/amd: Clear out the GV flag when handle deferred domain attach

On Tue, Aug 01, 2017 at 07:37:28PM +0800, Baoquan He wrote: > @@ -2466,11 +2472,21 @@ static struct protection_domain *get_domain(struct > device *dev) > return ERR_PTR(-EINVAL); > > domain = get_dev_data(dev)->domain; > - if (domain == NULL &&

Re: [PATCH v9 10/13] iommu/amd: Allocate memory below 4G for dev table if translation pre-enabled

On Tue, Aug 01, 2017 at 07:37:26PM +0800, Baoquan He wrote: > AMD pointed out it's unsafe to update the device-table while iommu > is enabled. It turns out that device-table pointer update is split > up into two 32bit writes in the IOMMU hardware. So updating it while > the IOMMU is enabled could

Re: [PATCH v9 06/13] iommu/amd: copy old trans table from old kernel

On Tue, Aug 01, 2017 at 07:37:22PM +0800, Baoquan He wrote: > static void init_device_table_dma(void) > { > @@ -2137,9 +2140,49 @@ static void early_enable_iommu(struct amd_iommu *iommu) > static void early_enable_iommus(void) > { > struct amd_iommu *iommu; > + bool is_pre_disabled =

Re: [PATCH v9 05/13] iommu/amd: Add function copy_dev_tables()

Hi Joerg, Thanks for your reviewing! On 08/04/17 at 02:09pm, Joerg Roedel wrote: > Hi Baoquan, > > On Tue, Aug 01, 2017 at 07:37:21PM +0800, Baoquan He wrote: > > + for_each_iommu(iommu) { > > + /* All IOMMUs should use the same device table with the same > > size */ > > +

Re: [PATCH v9 05/13] iommu/amd: Add function copy_dev_tables()

Hi Baoquan, On Tue, Aug 01, 2017 at 07:37:21PM +0800, Baoquan He wrote: > + for_each_iommu(iommu) { > + /* All IOMMUs should use the same device table with the same > size */ > + lo = readl(iommu->mmio_base + MMIO_DEV_TABLE_OFFSET); > + hi =

[PATCH] iommu/of: Fix of_iommu_configure() for disabled IOMMUs

Sudeep reports that the logic is slightly broken when a PCI iommu-map entry targets an IOMMU marked as disabled in DT, since of_pci_map_rid() succeeds in following a phandle, and of_iommu_xlate() doesn't return an error value, but we miss checking whether ops was actually non-NULL. Whilst this

Re: [PATCH] iommu/exynos: Remove custom platform device registration code

On Fri, Aug 04, 2017 at 12:28:33PM +0200, Marek Szyprowski wrote: > Commit 09515ef5ddad ("of/acpi: Configure dma operations at probe time for > platform/amba/pci bus devices") postponed the moment of attaching IOMMU > controller to its device, so there is no need to register IOMMU controllers >

[PATCH] iommu/exynos: Remove custom platform device registration code

Commit 09515ef5ddad ("of/acpi: Configure dma operations at probe time for platform/amba/pci bus devices") postponed the moment of attaching IOMMU controller to its device, so there is no need to register IOMMU controllers very early, before all other devices in the system. This change gives us an

Re: [PATCH v3 0/4] Add larbid init routine for mediatek's gen1 smi larb driver

On Fri, Aug 04, 2017 at 09:32:24AM +0800, honghui.zh...@mediatek.com wrote: > Honghui Zhang (4): > memory: mtk-smi: Use of_device_get_match_data helper > memory: mtk-smi: add larbid handle routine > dt-bindings: mediatek: add descriptions for larbid > arm: dts: mediatek: add larbid

Re: [PATCH 0/2] iommu/omap: Rework cache functionality with DMA Streaming API

On Fri, Jul 28, 2017 at 03:49:12PM -0500, Josue Albarran wrote: > Fernando Guzman Lugo (1): > iommu/omap: Fix disabling of MMU upon a fault > > Josue Albarran (1): > iommu/omap: Use DMA-API for performing cache flushes > > drivers/iommu/omap-iommu.c | 125 >

Re: [PATCH] iommu/exynos: prevent building on big-endian kernels

On Fri, Jul 28, 2017 at 03:19:19PM +0200, Arnd Bergmann wrote: > Since we print the correct warning, an allmodconfig build is no longer > clean but always prints it, which defeats compile-testing: > > drivers/iommu/exynos-iommu.c:58:2: error: #warning "revisit driver if we can > enable

Re: [PATCH v2] memory: mtk-smi: Use of_device_get_match_data helper

On Wed, Aug 02, 2017 at 10:02:50AM +0800, Honghui Zhang wrote: > On Wed, 2017-07-26 at 20:37 +0800, honghui.zh...@mediatek.com wrote: > > From: Honghui Zhang > > > > Replace custom code with generic helper to retrieve driver data. > > Hi, Joerg, would you please take

Re: Feature Request: Ability to decode bus/dma address back into physical address

On Tue, Aug 01, 2017 at 04:42:50PM -0600, Alex Williamson wrote: > Generally that's fine, your post will get delayed until Joerg or I > approve it, but it should get there eventually. Probably just a > mis-click caused you to end up on the wrong list. I still don't see > you on the subscriber

Re: Support SVM without PASID

Hi Kevin, On 04/08/17 02:49, Tian, Kevin wrote: >> From: Jean-Philippe Brucker >> Sent: Tuesday, August 1, 2017 4:26 PM >> >> It depends what type you use when registering the IOMMU with >> VFIO_SET_IOMMU: >> >> * If the type is VFIO_TYPE1v2_IOMMU, then >> VFIO_IOMMU_MAP/UNMAP_DMA >> affects

[git pull] IOMMU Fixes for Linux v4.13-rc3

Hi Linus, The following changes since commit 520eccdfe187591a51ea9ab4c1a024ae4d0f68d9: Linux 4.13-rc2 (2017-07-23 16:15:17 -0700) are available in the git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu.git tags/iommu-fixes-v4.13-rc3 for you to fetch changes up to