From 6caeb33fa986151f745fc62190bc28a593b8a0d2 Mon Sep 17 00:00:00 2001 From: Lu Baolu Date: Tue, 10 Jan 2023 10:54:05 +0800 Subject: iommu: Add set_platform_dma_ops iommu ops When VFIO finishes assigning a device to user space and calls iommu_group_release_dma_owner() to return the device to kernel, the IOMMU core will attach the default domain to the device. Unfortunately, some IOMMU drivers don't support default domain, hence in the end, the core calls .detach_dev instead. This adds set_platform_dma_ops iommu ops to make it clear that what it does is returning control back to the platform DMA ops. Suggested-by: Jason Gunthorpe Reviewed-by: Jason Gunthorpe Signed-off-by: Lu Baolu Link: https://lore.kernel.org/r/20230110025408.667767-3-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel --- include/linux/iommu.h | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'include/linux') diff --git a/include/linux/iommu.h b/include/linux/iommu.h index 46e1347bfa22..7b3e3775b069 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -228,6 +228,9 @@ struct iommu_iotlb_gather { * @release_device: Remove device from iommu driver handling * @probe_finalize: Do final setup work after the device is added to an IOMMU * group and attached to the groups domain + * @set_platform_dma_ops: Returning control back to the platform DMA ops. This op + * is to support old IOMMU drivers, new drivers should use + * default domains, and the common IOMMU DMA ops. * @device_group: find iommu group for a particular device * @get_resv_regions: Request list of reserved regions for a device * @of_xlate: add OF master IDs to iommu grouping @@ -256,6 +259,7 @@ struct iommu_ops { struct iommu_device *(*probe_device)(struct device *dev); void (*release_device)(struct device *dev); void (*probe_finalize)(struct device *dev); + void (*set_platform_dma_ops)(struct device *dev); struct iommu_group *(*device_group)(struct device *dev); /* Request/Free a list of reserved regions for a device */ -- cgit v1.2.3 From dd8a25c557e109f868430bd2e3e8f394cb40eaa7 Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 10 Jan 2023 10:54:07 +0800 Subject: iommu: Remove deferred attach check from __iommu_detach_device() At the current moment, __iommu_detach_device() is only called via call chains that are after the device driver is attached - eg via explicit attach APIs called by the device driver. Commit bd421264ed30 ("iommu: Fix deferred domain attachment") has removed deferred domain attachment check from __iommu_attach_device() path, so it should just unconditionally work in the __iommu_detach_device() path. It actually looks like a bug that we were blocking detach on these paths since the attach was unconditional and the caller is going to free the (probably) UNAMANGED domain once this returns. The only place we should be testing for deferred attach is during the initial point the dma device is linked to the group, and then again during the dma api calls. Signed-off-by: Jason Gunthorpe Signed-off-by: Lu Baolu Link: https://lore.kernel.org/r/20230110025408.667767-5-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel --- include/linux/iommu.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include/linux') diff --git a/include/linux/iommu.h b/include/linux/iommu.h index 7b3e3775b069..0d10566b3cb2 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -405,6 +405,7 @@ struct iommu_fault_param { * @iommu_dev: IOMMU device this device is linked to * @priv: IOMMU Driver private data * @max_pasids: number of PASIDs this device can consume + * @attach_deferred: the dma domain attachment is deferred * * TODO: migrate other per device data pointers under iommu_dev_data, e.g. * struct iommu_group *iommu_group; @@ -417,6 +418,7 @@ struct dev_iommu { struct iommu_device *iommu_dev; void *priv; u32 max_pasids; + u32 attach_deferred:1; }; int iommu_device_register(struct iommu_device *iommu, -- cgit v1.2.3 From 8f9930fa016134ea07db4775ec596b16c3d03f05 Mon Sep 17 00:00:00 2001 From: Lu Baolu Date: Tue, 10 Jan 2023 10:54:08 +0800 Subject: iommu: Remove detach_dev callback The detach_dev callback of domain ops is not called in the IOMMU core. Remove this callback to avoid dead code. The trace event for detaching domain from device is removed accordingly. Reviewed-by: Jason Gunthorpe Signed-off-by: Lu Baolu Link: https://lore.kernel.org/r/20230110025408.667767-6-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel --- include/linux/iommu.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/iommu.h b/include/linux/iommu.h index 0d10566b3cb2..a8063f26ff69 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -299,7 +299,6 @@ struct iommu_ops { * * EBUSY - device is attached to a domain and cannot be changed * * ENODEV - device specific errors, not able to be attached * * - treated as ENODEV by the caller. Use is discouraged - * @detach_dev: detach an iommu domain from a device * @set_dev_pasid: set an iommu domain to a pasid of device * @map: map a physically contiguous memory region to an iommu domain * @map_pages: map a physically contiguous set of pages of the same size to @@ -320,7 +319,6 @@ struct iommu_ops { */ struct iommu_domain_ops { int (*attach_dev)(struct iommu_domain *domain, struct device *dev); - void (*detach_dev)(struct iommu_domain *domain, struct device *dev); int (*set_dev_pasid)(struct iommu_domain *domain, struct device *dev, ioasid_t pasid); -- cgit v1.2.3 From e251c21372c07694f547afe3c9828f7f6ef01267 Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Fri, 20 Jan 2023 18:42:48 +0100 Subject: of: Introduce of_translate_dma_region() This function is similar to of_translate_dma_address() but also reads a length in addition to an address from a device tree property. Reviewed-by: Rob Herring Signed-off-by: Thierry Reding Link: https://lore.kernel.org/r/20230120174251.4004100-2-thierry.reding@gmail.com Signed-off-by: Joerg Roedel --- include/linux/of_address.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include/linux') diff --git a/include/linux/of_address.h b/include/linux/of_address.h index 265f26eeaf6b..376671594746 100644 --- a/include/linux/of_address.h +++ b/include/linux/of_address.h @@ -38,6 +38,8 @@ struct of_pci_range { /* Translate a DMA address from device space to CPU space */ extern u64 of_translate_dma_address(struct device_node *dev, const __be32 *in_addr); +extern const __be32 *of_translate_dma_region(struct device_node *dev, const __be32 *addr, + phys_addr_t *start, size_t *length); #ifdef CONFIG_OF_ADDRESS extern u64 of_translate_address(struct device_node *np, const __be32 *addr); -- cgit v1.2.3 From a5bf3cfce8cb77d9d24613ab52d520896f83dd48 Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Fri, 20 Jan 2023 18:42:50 +0100 Subject: iommu: Implement of_iommu_get_resv_regions() This is an implementation that IOMMU drivers can use to obtain reserved memory regions from a device tree node. It uses the reserved-memory DT bindings to find the regions associated with a given device. If these regions are marked accordingly, identity mappings will be created for them in the IOMMU domain that the devices will be attached to. Cc: Frank Rowand Cc: devicetree@vger.kernel.org Reviewed-by: Rob Herring Acked-by: Robin Murphy Signed-off-by: Thierry Reding Link: https://lore.kernel.org/r/20230120174251.4004100-4-thierry.reding@gmail.com Signed-off-by: Joerg Roedel --- include/linux/of_iommu.h | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'include/linux') diff --git a/include/linux/of_iommu.h b/include/linux/of_iommu.h index 55c1eb300a86..9a5e6b410dd2 100644 --- a/include/linux/of_iommu.h +++ b/include/linux/of_iommu.h @@ -12,6 +12,9 @@ extern const struct iommu_ops *of_iommu_configure(struct device *dev, struct device_node *master_np, const u32 *id); +extern void of_iommu_get_resv_regions(struct device *dev, + struct list_head *list); + #else static inline const struct iommu_ops *of_iommu_configure(struct device *dev, @@ -21,6 +24,11 @@ static inline const struct iommu_ops *of_iommu_configure(struct device *dev, return NULL; } +static inline void of_iommu_get_resv_regions(struct device *dev, + struct list_head *list) +{ +} + #endif /* CONFIG_OF_IOMMU */ #endif /* __OF_IOMMU_H */ -- cgit v1.2.3 From 1369459b2e219a6f4c861404c4f195cd81dcbb40 Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Mon, 23 Jan 2023 16:35:54 -0400 Subject: iommu: Add a gfp parameter to iommu_map() The internal mechanisms support this, but instead of exposting the gfp to the caller it wrappers it into iommu_map() and iommu_map_atomic() Fix this instead of adding more variants for GFP_KERNEL_ACCOUNT. Reviewed-by: Kevin Tian Signed-off-by: Jason Gunthorpe Reviewed-by: Mathieu Poirier Link: https://lore.kernel.org/r/1-v3-76b587fe28df+6e3-iommu_map_gfp_jgg@nvidia.com Signed-off-by: Joerg Roedel --- include/linux/iommu.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/iommu.h b/include/linux/iommu.h index 46e1347bfa22..d2020994f292 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -467,7 +467,7 @@ extern int iommu_sva_unbind_gpasid(struct iommu_domain *domain, extern struct iommu_domain *iommu_get_domain_for_dev(struct device *dev); extern struct iommu_domain *iommu_get_dma_domain(struct device *dev); extern int iommu_map(struct iommu_domain *domain, unsigned long iova, - phys_addr_t paddr, size_t size, int prot); + phys_addr_t paddr, size_t size, int prot, gfp_t gfp); extern int iommu_map_atomic(struct iommu_domain *domain, unsigned long iova, phys_addr_t paddr, size_t size, int prot); extern size_t iommu_unmap(struct iommu_domain *domain, unsigned long iova, @@ -773,7 +773,7 @@ static inline struct iommu_domain *iommu_get_domain_for_dev(struct device *dev) } static inline int iommu_map(struct iommu_domain *domain, unsigned long iova, - phys_addr_t paddr, size_t size, int prot) + phys_addr_t paddr, size_t size, int prot, gfp_t gfp) { return -ENODEV; } -- cgit v1.2.3 From 4dc6376af596d9b2a46fa9baf94c9f2fa5a3d246 Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Mon, 23 Jan 2023 16:35:55 -0400 Subject: iommu: Remove iommu_map_atomic() There is only one call site and it can now just pass the GFP_ATOMIC to the normal iommu_map(). Reviewed-by: Kevin Tian Signed-off-by: Jason Gunthorpe Link: https://lore.kernel.org/r/2-v3-76b587fe28df+6e3-iommu_map_gfp_jgg@nvidia.com Signed-off-by: Joerg Roedel --- include/linux/iommu.h | 9 --------- 1 file changed, 9 deletions(-) (limited to 'include/linux') diff --git a/include/linux/iommu.h b/include/linux/iommu.h index d2020994f292..521cd79700f4 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -468,8 +468,6 @@ extern struct iommu_domain *iommu_get_domain_for_dev(struct device *dev); extern struct iommu_domain *iommu_get_dma_domain(struct device *dev); extern int iommu_map(struct iommu_domain *domain, unsigned long iova, phys_addr_t paddr, size_t size, int prot, gfp_t gfp); -extern int iommu_map_atomic(struct iommu_domain *domain, unsigned long iova, - phys_addr_t paddr, size_t size, int prot); extern size_t iommu_unmap(struct iommu_domain *domain, unsigned long iova, size_t size); extern size_t iommu_unmap_fast(struct iommu_domain *domain, @@ -778,13 +776,6 @@ static inline int iommu_map(struct iommu_domain *domain, unsigned long iova, return -ENODEV; } -static inline int iommu_map_atomic(struct iommu_domain *domain, - unsigned long iova, phys_addr_t paddr, - size_t size, int prot) -{ - return -ENODEV; -} - static inline size_t iommu_unmap(struct iommu_domain *domain, unsigned long iova, size_t size) { -- cgit v1.2.3 From f2b2c051be6262edc3c6fce50d3d4f01b59ba228 Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Mon, 23 Jan 2023 16:35:56 -0400 Subject: iommu: Add a gfp parameter to iommu_map_sg() Follow the pattern for iommu_map() and remove iommu_map_sg_atomic(). This allows __iommu_dma_alloc_noncontiguous() to use a GFP_KERNEL allocation here, based on the provided gfp flags. Reviewed-by: Kevin Tian Signed-off-by: Jason Gunthorpe Link: https://lore.kernel.org/r/3-v3-76b587fe28df+6e3-iommu_map_gfp_jgg@nvidia.com Signed-off-by: Joerg Roedel --- include/linux/iommu.h | 18 +++++------------- 1 file changed, 5 insertions(+), 13 deletions(-) (limited to 'include/linux') diff --git a/include/linux/iommu.h b/include/linux/iommu.h index 521cd79700f4..d5c16dc33c87 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -474,10 +474,8 @@ extern size_t iommu_unmap_fast(struct iommu_domain *domain, unsigned long iova, size_t size, struct iommu_iotlb_gather *iotlb_gather); extern ssize_t iommu_map_sg(struct iommu_domain *domain, unsigned long iova, - struct scatterlist *sg, unsigned int nents, int prot); -extern ssize_t iommu_map_sg_atomic(struct iommu_domain *domain, - unsigned long iova, struct scatterlist *sg, - unsigned int nents, int prot); + struct scatterlist *sg, unsigned int nents, + int prot, gfp_t gfp); extern phys_addr_t iommu_iova_to_phys(struct iommu_domain *domain, dma_addr_t iova); extern void iommu_set_fault_handler(struct iommu_domain *domain, iommu_fault_handler_t handler, void *token); @@ -791,14 +789,7 @@ static inline size_t iommu_unmap_fast(struct iommu_domain *domain, static inline ssize_t iommu_map_sg(struct iommu_domain *domain, unsigned long iova, struct scatterlist *sg, - unsigned int nents, int prot) -{ - return -ENODEV; -} - -static inline ssize_t iommu_map_sg_atomic(struct iommu_domain *domain, - unsigned long iova, struct scatterlist *sg, - unsigned int nents, int prot) + unsigned int nents, int prot, gfp_t gfp) { return -ENODEV; } @@ -1109,7 +1100,8 @@ iommu_get_domain_for_dev_pasid(struct device *dev, ioasid_t pasid, static inline size_t iommu_map_sgtable(struct iommu_domain *domain, unsigned long iova, struct sg_table *sgt, int prot) { - return iommu_map_sg(domain, iova, sgt->sgl, sgt->orig_nents, prot); + return iommu_map_sg(domain, iova, sgt->sgl, sgt->orig_nents, prot, + GFP_KERNEL); } #ifdef CONFIG_IOMMU_DEBUGFS -- cgit v1.2.3 From d82e6ae67ac2f9d6ba51690353c477b340bba6b5 Mon Sep 17 00:00:00 2001 From: Lu Baolu Date: Tue, 31 Jan 2023 15:37:29 +0800 Subject: iommu/vt-d: Remove include/linux/intel-svm.h There's no need to have a public header for Intel SVA implementation. The device driver should interact with Intel SVA implementation via the IOMMU generic APIs. Reviewed-by: Kevin Tian Signed-off-by: Lu Baolu Link: https://lore.kernel.org/r/20230109014955.147068-2-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel --- include/linux/intel-svm.h | 16 ---------------- 1 file changed, 16 deletions(-) delete mode 100644 include/linux/intel-svm.h (limited to 'include/linux') diff --git a/include/linux/intel-svm.h b/include/linux/intel-svm.h deleted file mode 100644 index f9a0d44f6fdb..000000000000 --- a/include/linux/intel-svm.h +++ /dev/null @@ -1,16 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0-only */ -/* - * Copyright © 2015 Intel Corporation. - * - * Authors: David Woodhouse - */ - -#ifndef __INTEL_SVM_H__ -#define __INTEL_SVM_H__ - -/* Page Request Queue depth */ -#define PRQ_ORDER 4 -#define PRQ_RING_MASK ((0x1000 << PRQ_ORDER) - 0x20) -#define PRQ_DEPTH ((0x1000 << PRQ_ORDER) >> 5) - -#endif /* __INTEL_SVM_H__ */ -- cgit v1.2.3 From 4db96bfe9d7772d6ddedd62ce478895999043fd7 Mon Sep 17 00:00:00 2001 From: Kan Liang Date: Tue, 31 Jan 2023 15:37:34 +0800 Subject: iommu/vt-d: Support size of the register set in DRHD A new field, which indicates the size of the remapping hardware register set for this remapping unit, is introduced in the DMA-remapping hardware unit definition (DRHD) structure with the VT-d Spec 4.0. With this information, SW doesn't need to 'guess' the size of the register set anymore. Update the struct acpi_dmar_hardware_unit to reflect the field. Store the size of the register set in struct dmar_drhd_unit for each dmar device. The 'size' information is ResvZ for the old BIOS and platforms. Fall back to the old guessing method. There is nothing changed. Signed-off-by: Kan Liang Link: https://lore.kernel.org/r/20230128200428.1459118-2-kan.liang@linux.intel.com Signed-off-by: Lu Baolu Signed-off-by: Joerg Roedel --- include/linux/dmar.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/dmar.h b/include/linux/dmar.h index d81a51978d01..725d5e6acec0 100644 --- a/include/linux/dmar.h +++ b/include/linux/dmar.h @@ -39,6 +39,7 @@ struct dmar_drhd_unit { struct list_head list; /* list of drhd units */ struct acpi_dmar_header *hdr; /* ACPI header */ u64 reg_base_addr; /* register base address*/ + unsigned long reg_size; /* size of register set */ struct dmar_dev_scope *devices;/* target device array */ int devices_cnt; /* target device count */ u16 segment; /* PCI domain */ -- cgit v1.2.3 From 46284c6ceb5e4dfddcb00dafb7c2f3c1437fdca4 Mon Sep 17 00:00:00 2001 From: Kan Liang Date: Tue, 31 Jan 2023 15:37:38 +0800 Subject: iommu/vt-d: Support cpumask for IOMMU perfmon The perf subsystem assumes that all counters are by default per-CPU. So the user space tool reads a counter from each CPU. However, the IOMMU counters are system-wide and can be read from any CPU. Here we use a CPU mask to restrict counting to one CPU to handle the issue. (with CPU hotplug notifier to choose a different CPU if the chosen one is taken off-line). The CPU is exposed to /sys/bus/event_source/devices/dmar*/cpumask for the user space perf tool. Signed-off-by: Kan Liang Link: https://lore.kernel.org/r/20230128200428.1459118-6-kan.liang@linux.intel.com Signed-off-by: Lu Baolu Signed-off-by: Joerg Roedel --- include/linux/cpuhotplug.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/cpuhotplug.h b/include/linux/cpuhotplug.h index 6c6859bfc454..f2ea348ce3b0 100644 --- a/include/linux/cpuhotplug.h +++ b/include/linux/cpuhotplug.h @@ -221,6 +221,7 @@ enum cpuhp_state { CPUHP_AP_PERF_X86_CQM_ONLINE, CPUHP_AP_PERF_X86_CSTATE_ONLINE, CPUHP_AP_PERF_X86_IDXD_ONLINE, + CPUHP_AP_PERF_X86_IOMMU_PERF_ONLINE, CPUHP_AP_PERF_S390_CF_ONLINE, CPUHP_AP_PERF_S390_SF_ONLINE, CPUHP_AP_PERF_ARM_CCI_ONLINE, -- cgit v1.2.3