The following patch will add a new vIOMMU allocator that will require this _iommufd_object_alloc to be sharable with IOMMU drivers (and iommufd too).

Add a new driver.c file that will be built with CONFIG_IOMMUFD_DRIVER_CORE selected by CONFIG_IOMMUFD, and put the CONFIG_DRIVER under that remaining to be selectable for drivers to build the existing iova_bitmap.c file.

Suggested-by: Jason Gunthorpe jgg@nvidia.com Reviewed-by: Jason Gunthorpe jgg@nvidia.com Signed-off-by: Nicolin Chen nicolinc@nvidia.com --- drivers/iommu/iommufd/Kconfig | 4 +++ drivers/iommu/iommufd/Makefile | 1 + drivers/iommu/iommufd/iommufd_private.h | 4 --- include/linux/iommufd.h | 13 ++++++++ drivers/iommu/iommufd/driver.c | 40 +++++++++++++++++++++++++ drivers/iommu/iommufd/main.c | 32 -------------------- 6 files changed, 58 insertions(+), 36 deletions(-) create mode 100644 drivers/iommu/iommufd/driver.c

diff --git a/drivers/iommu/iommufd/Kconfig b/drivers/iommu/iommufd/Kconfig index 76656fe0470d..0a07f9449fd9 100644 --- a/drivers/iommu/iommufd/Kconfig +++ b/drivers/iommu/iommufd/Kconfig @@ -1,4 +1,8 @@ # SPDX-License-Identifier: GPL-2.0-only +config IOMMUFD_DRIVER_CORE + tristate + default (IOMMUFD_DRIVER || IOMMUFD) if IOMMUFD!=n + config IOMMUFD tristate "IOMMU Userspace API" select INTERVAL_TREE diff --git a/drivers/iommu/iommufd/Makefile b/drivers/iommu/iommufd/Makefile index cf4605962bea..de675df52913 100644 --- a/drivers/iommu/iommufd/Makefile +++ b/drivers/iommu/iommufd/Makefile @@ -13,3 +13,4 @@ iommufd-$(CONFIG_IOMMUFD_TEST) += selftest.o

obj-$(CONFIG_IOMMUFD) += iommufd.o obj-$(CONFIG_IOMMUFD_DRIVER) += iova_bitmap.o +obj-$(CONFIG_IOMMUFD_DRIVER_CORE) += driver.o diff --git a/drivers/iommu/iommufd/iommufd_private.h b/drivers/iommu/iommufd/iommufd_private.h index 94cfcab7e9de..be347f726fda 100644 --- a/drivers/iommu/iommufd/iommufd_private.h +++ b/drivers/iommu/iommufd/iommufd_private.h @@ -206,10 +206,6 @@ iommufd_object_put_and_try_destroy(struct iommufd_ctx *ictx, iommufd_object_remove(ictx, obj, obj->id, 0); }

-struct iommufd_object *_iommufd_object_alloc(struct iommufd_ctx *ictx, - size_t size, - enum iommufd_object_type type); - #define __iommufd_object_alloc(ictx, ptr, type, obj) \ container_of(_iommufd_object_alloc( \ ictx, \ diff --git a/include/linux/iommufd.h b/include/linux/iommufd.h index 22948dd03d67..94522d4029ca 100644 --- a/include/linux/iommufd.h +++ b/include/linux/iommufd.h @@ -135,4 +135,17 @@ static inline int iommufd_vfio_compat_set_no_iommu(struct iommufd_ctx *ictx) return -EOPNOTSUPP; } #endif /* CONFIG_IOMMUFD */ + +#if IS_ENABLED(CONFIG_IOMMUFD_DRIVER_CORE) +struct iommufd_object *_iommufd_object_alloc(struct iommufd_ctx *ictx, + size_t size, + enum iommufd_object_type type); +#else /* !CONFIG_IOMMUFD_DRIVER_CORE */ +static inline struct iommufd_object * +_iommufd_object_alloc(struct iommufd_ctx *ictx, size_t size, + enum iommufd_object_type type) +{ + return ERR_PTR(-EOPNOTSUPP); +} +#endif /* CONFIG_IOMMUFD_DRIVER_CORE */ #endif diff --git a/drivers/iommu/iommufd/driver.c b/drivers/iommu/iommufd/driver.c new file mode 100644 index 000000000000..2bc47d92a0ab --- /dev/null +++ b/drivers/iommu/iommufd/driver.c @@ -0,0 +1,40 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* Copyright (c) 2024, NVIDIA CORPORATION & AFFILIATES + */ +#include "iommufd_private.h" + +struct iommufd_object *_iommufd_object_alloc(struct iommufd_ctx *ictx, + size_t size, + enum iommufd_object_type type) +{ + struct iommufd_object *obj; + int rc; + + obj = kzalloc(size, GFP_KERNEL_ACCOUNT); + if (!obj) + return ERR_PTR(-ENOMEM); + obj->type = type; + /* Starts out bias'd by 1 until it is removed from the xarray */ + refcount_set(&obj->shortterm_users, 1); + refcount_set(&obj->users, 1); + + /* + * Reserve an ID in the xarray but do not publish the pointer yet since + * the caller hasn't initialized it yet. Once the pointer is published + * in the xarray and visible to other threads we can't reliably destroy + * it anymore, so the caller must complete all errorable operations + * before calling iommufd_object_finalize(). + */ + rc = xa_alloc(&ictx->objects, &obj->id, XA_ZERO_ENTRY, xa_limit_31b, + GFP_KERNEL_ACCOUNT); + if (rc) + goto out_free; + return obj; +out_free: + kfree(obj); + return ERR_PTR(rc); +} +EXPORT_SYMBOL_NS_GPL(_iommufd_object_alloc, IOMMUFD); + +MODULE_DESCRIPTION("iommufd code shared with builtin modules"); +MODULE_LICENSE("GPL"); diff --git a/drivers/iommu/iommufd/main.c b/drivers/iommu/iommufd/main.c index 826a2b2be52f..3c32b440471b 100644 --- a/drivers/iommu/iommufd/main.c +++ b/drivers/iommu/iommufd/main.c @@ -29,38 +29,6 @@ struct iommufd_object_ops { static const struct iommufd_object_ops iommufd_object_ops[]; static struct miscdevice vfio_misc_dev;

-struct iommufd_object *_iommufd_object_alloc(struct iommufd_ctx *ictx, - size_t size, - enum iommufd_object_type type) -{ - struct iommufd_object *obj; - int rc; - - obj = kzalloc(size, GFP_KERNEL_ACCOUNT); - if (!obj) - return ERR_PTR(-ENOMEM); - obj->type = type; - /* Starts out bias'd by 1 until it is removed from the xarray */ - refcount_set(&obj->shortterm_users, 1); - refcount_set(&obj->users, 1); - - /* - * Reserve an ID in the xarray but do not publish the pointer yet since - * the caller hasn't initialized it yet. Once the pointer is published - * in the xarray and visible to other threads we can't reliably destroy - * it anymore, so the caller must complete all errorable operations - * before calling iommufd_object_finalize(). - */ - rc = xa_alloc(&ictx->objects, &obj->id, XA_ZERO_ENTRY, - xa_limit_31b, GFP_KERNEL_ACCOUNT); - if (rc) - goto out_free; - return obj; -out_free: - kfree(obj); - return ERR_PTR(rc); -} - /* * Allow concurrent access to the object. *

Add a new IOMMUFD_OBJ_VIOMMU with an iommufd_viommu structure to represent a slice of physical IOMMU device passed to or shared with a user space VM. This slice, now a vIOMMU object, is a group of virtualization resources of a physical IOMMU's, such as: - Security namespace for guest owned ID, e.g. guest-controlled cache tags - Non-device-affiliated event reporting, e.g. invalidation queue errors - Access to a sharable nesting parent pagetable across physical IOMMUs - Virtualization of various platforms IDs, e.g. RIDs and others - Delivery of paravirtualized invalidation - Direct assigned invalidation queues - Direct assigned interrupts

Add a new viommu_alloc op in iommu_ops, for drivers to allocate their own vIOMMU structures. And this allocation also needs a free(), so add struct iommufd_viommu_ops.

To simplify a vIOMMU allocation, provide a iommufd_viommu_alloc() helper. It's suggested that a driver should embed a core-level viommu structure in its driver-level viommu struct and call the iommufd_viommu_alloc() helper, meanwhile the driver can also implement a viommu ops: struct my_driver_viommu { struct iommufd_viommu core; /* driver-owned properties/features */ .... };

static const struct iommufd_viommu_ops my_driver_viommu_ops = { .free = my_driver_viommu_free, /* future ops for virtualization features */ .... };

static struct iommufd_viommu my_driver_viommu_alloc(...) { struct my_driver_viommu *my_viommu = iommufd_viommu_alloc(ictx, my_driver_viommu, core, my_driver_viommu_ops); /* Init my_viommu and related HW feature */ .... return &my_viommu->core; }

static struct iommu_domain_ops my_driver_domain_ops = { .... .viommu_alloc = my_driver_viommu_alloc, };

Suggested-by: Jason Gunthorpe jgg@nvidia.com Reviewed-by: Kevin Tian kevin.tian@intel.com Reviewed-by: Jason Gunthorpe jgg@nvidia.com Signed-off-by: Nicolin Chen nicolinc@nvidia.com --- include/linux/iommu.h | 14 ++++++++++++++ include/linux/iommufd.h | 40 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 54 insertions(+)

diff --git a/include/linux/iommu.h b/include/linux/iommu.h index bd722f473635..2574fc8abaf2 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -42,6 +42,8 @@ struct notifier_block; struct iommu_sva; struct iommu_dma_cookie; struct iommu_fault_param; +struct iommufd_ctx; +struct iommufd_viommu;

#define IOMMU_FAULT_PERM_READ (1 << 0) /* read */ #define IOMMU_FAULT_PERM_WRITE (1 << 1) /* write */ @@ -542,6 +544,14 @@ static inline int __iommu_copy_struct_from_user_array( * @remove_dev_pasid: Remove any translation configurations of a specific * pasid, so that any DMA transactions with this pasid * will be blocked by the hardware. + * @viommu_alloc: Allocate an iommufd_viommu on a physical IOMMU instance behind + * the @dev, as the set of virtualization resources shared/passed + * to user space IOMMU instance. And associate it with a nesting + * @parent_domain. The @viommu_type must be defined in the header + * include/uapi/linux/iommufd.h + * It is required to call iommufd_viommu_alloc() helper for + * a bundled allocation of the core and the driver structures, + * using the given @ictx pointer. * @pgsize_bitmap: bitmap of all possible supported page sizes * @owner: Driver module providing these ops * @identity_domain: An always available, always attachable identity @@ -591,6 +601,10 @@ struct iommu_ops { void (*remove_dev_pasid)(struct device *dev, ioasid_t pasid, struct iommu_domain *domain);

+ struct iommufd_viommu *(*viommu_alloc)( + struct device *dev, struct iommu_domain *parent_domain, + struct iommufd_ctx *ictx, unsigned int viommu_type); + const struct iommu_domain_ops *default_domain_ops; unsigned long pgsize_bitmap; struct module *owner; diff --git a/include/linux/iommufd.h b/include/linux/iommufd.h index 94522d4029ca..4fc2ce332f10 100644 --- a/include/linux/iommufd.h +++ b/include/linux/iommufd.h @@ -17,6 +17,7 @@ struct iommu_group; struct iommufd_access; struct iommufd_ctx; struct iommufd_device; +struct iommufd_viommu_ops; struct page;

enum iommufd_object_type { @@ -28,6 +29,7 @@ enum iommufd_object_type { IOMMUFD_OBJ_IOAS, IOMMUFD_OBJ_ACCESS, IOMMUFD_OBJ_FAULT, + IOMMUFD_OBJ_VIOMMU, #ifdef CONFIG_IOMMUFD_TEST IOMMUFD_OBJ_SELFTEST, #endif @@ -78,6 +80,26 @@ void iommufd_access_detach(struct iommufd_access *access);

void iommufd_ctx_get(struct iommufd_ctx *ictx);

+struct iommufd_viommu { + struct iommufd_object obj; + struct iommufd_ctx *ictx; + struct iommu_device *iommu_dev; + struct iommufd_hwpt_paging *hwpt; + + const struct iommufd_viommu_ops *ops; + + unsigned int type; +}; + +/** + * struct iommufd_viommu_ops - vIOMMU specific operations + * @destroy: Clean up all driver-specific parts of an iommufd_viommu. The memory + * of the vIOMMU will be free-ed by iommufd core after calling this op + */ +struct iommufd_viommu_ops { + void (*destroy)(struct iommufd_viommu *viommu); +}; + #if IS_ENABLED(CONFIG_IOMMUFD) struct iommufd_ctx *iommufd_ctx_from_file(struct file *file); struct iommufd_ctx *iommufd_ctx_from_fd(int fd); @@ -148,4 +170,22 @@ _iommufd_object_alloc(struct iommufd_ctx *ictx, size_t size, return ERR_PTR(-EOPNOTSUPP); } #endif /* CONFIG_IOMMUFD_DRIVER_CORE */ + +/* + * Helpers for IOMMU driver to allocate driver structures that will be freed by + * the iommufd core. The free op will be called prior to freeing the memory. + */ +#define iommufd_viommu_alloc(ictx, drv_struct, member, viommu_ops) \ + ({ \ + drv_struct *ret; \ + \ + static_assert(__same_type(struct iommufd_viommu, \ + ((drv_struct *)NULL)->member)); \ + static_assert(offsetof(drv_struct, member.obj) == 0); \ + ret = (drv_struct *)_iommufd_object_alloc( \ + ictx, sizeof(drv_struct), IOMMUFD_OBJ_VIOMMU); \ + if (!IS_ERR(ret)) \ + ret->member.ops = viommu_ops; \ + ret; \ + }) #endif

Add a new ioctl for user space to do a vIOMMU allocation. It must be based on a nesting parent HWPT, so take its refcount.

IOMMU driver wanting to support vIOMMUs must define its IOMMU_VIOMMU_TYPE_ in the uAPI header and implement a viommu_alloc op in its iommu_ops.

Reviewed-by: Jason Gunthorpe jgg@nvidia.com Reviewed-by: Kevin Tian kevin.tian@intel.com Signed-off-by: Nicolin Chen nicolinc@nvidia.com --- drivers/iommu/iommufd/Makefile | 3 +- drivers/iommu/iommufd/iommufd_private.h | 3 + include/uapi/linux/iommufd.h | 40 ++++++++++++ drivers/iommu/iommufd/main.c | 6 ++ drivers/iommu/iommufd/viommu.c | 81 +++++++++++++++++++++++++ 5 files changed, 132 insertions(+), 1 deletion(-) create mode 100644 drivers/iommu/iommufd/viommu.c

diff --git a/drivers/iommu/iommufd/Makefile b/drivers/iommu/iommufd/Makefile index de675df52913..9e321140cccd 100644 --- a/drivers/iommu/iommufd/Makefile +++ b/drivers/iommu/iommufd/Makefile @@ -7,7 +7,8 @@ iommufd-y := \ ioas.o \ main.o \ pages.o \ - vfio_compat.o + vfio_compat.o \ + viommu.o

iommufd-$(CONFIG_IOMMUFD_TEST) += selftest.o

diff --git a/drivers/iommu/iommufd/iommufd_private.h b/drivers/iommu/iommufd/iommufd_private.h index be347f726fda..a8104d9d4cef 100644 --- a/drivers/iommu/iommufd/iommufd_private.h +++ b/drivers/iommu/iommufd/iommufd_private.h @@ -506,6 +506,9 @@ static inline int iommufd_hwpt_replace_device(struct iommufd_device *idev, return iommu_group_replace_domain(idev->igroup->group, hwpt->domain); }

+int iommufd_viommu_alloc_ioctl(struct iommufd_ucmd *ucmd); +void iommufd_viommu_destroy(struct iommufd_object *obj); + #ifdef CONFIG_IOMMUFD_TEST int iommufd_test(struct iommufd_ucmd *ucmd); void iommufd_selftest_destroy(struct iommufd_object *obj); diff --git a/include/uapi/linux/iommufd.h b/include/uapi/linux/iommufd.h index 41b1a01e9293..302844136b02 100644 --- a/include/uapi/linux/iommufd.h +++ b/include/uapi/linux/iommufd.h @@ -52,6 +52,7 @@ enum { IOMMUFD_CMD_HWPT_INVALIDATE = 0x8d, IOMMUFD_CMD_FAULT_QUEUE_ALLOC = 0x8e, IOMMUFD_CMD_IOAS_MAP_FILE = 0x8f, + IOMMUFD_CMD_VIOMMU_ALLOC = 0x90, };

/** @@ -822,4 +823,43 @@ struct iommu_fault_alloc { __u32 out_fault_fd; }; #define IOMMU_FAULT_QUEUE_ALLOC _IO(IOMMUFD_TYPE, IOMMUFD_CMD_FAULT_QUEUE_ALLOC) + +/** + * enum iommu_viommu_type - Virtual IOMMU Type + * @IOMMU_VIOMMU_TYPE_DEFAULT: Reserved for future use + */ +enum iommu_viommu_type { + IOMMU_VIOMMU_TYPE_DEFAULT = 0, +}; + +/** + * struct iommu_viommu_alloc - ioctl(IOMMU_VIOMMU_ALLOC) + * @size: sizeof(struct iommu_viommu_alloc) + * @flags: Must be 0 + * @type: Type of the virtual IOMMU. Must be defined in enum iommu_viommu_type + * @dev_id: The device's physical IOMMU will be used to back the virtual IOMMU + * @hwpt_id: ID of a nesting parent HWPT to associate to + * @out_viommu_id: Output virtual IOMMU ID for the allocated object + * + * Allocate a virtual IOMMU object, representing the underlying physical IOMMU's + * virtualization support that is a security-isolated slice of the real IOMMU HW + * that is unique to a specific VM. Operations global to the IOMMU are connected + * to the vIOMMU, such as: + * - Security namespace for guest owned ID, e.g. guest-controlled cache tags + * - Non-device-affiliated event reporting, e.g. invalidation queue errors + * - Access to a sharable nesting parent pagetable across physical IOMMUs + * - Virtualization of various platforms IDs, e.g. RIDs and others + * - Delivery of paravirtualized invalidation + * - Direct assigned invalidation queues + * - Direct assigned interrupts + */ +struct iommu_viommu_alloc { + __u32 size; + __u32 flags; + __u32 type; + __u32 dev_id; + __u32 hwpt_id; + __u32 out_viommu_id; +}; +#define IOMMU_VIOMMU_ALLOC _IO(IOMMUFD_TYPE, IOMMUFD_CMD_VIOMMU_ALLOC) #endif diff --git a/drivers/iommu/iommufd/main.c b/drivers/iommu/iommufd/main.c index 30e6c2af3b45..cc514f9bc3e6 100644 --- a/drivers/iommu/iommufd/main.c +++ b/drivers/iommu/iommufd/main.c @@ -307,6 +307,7 @@ union ucmd_buffer { struct iommu_ioas_unmap unmap; struct iommu_option option; struct iommu_vfio_ioas vfio_ioas; + struct iommu_viommu_alloc viommu; #ifdef CONFIG_IOMMUFD_TEST struct iommu_test_cmd test; #endif @@ -360,6 +361,8 @@ static const struct iommufd_ioctl_op iommufd_ioctl_ops[] = { val64), IOCTL_OP(IOMMU_VFIO_IOAS, iommufd_vfio_ioas, struct iommu_vfio_ioas, __reserved), + IOCTL_OP(IOMMU_VIOMMU_ALLOC, iommufd_viommu_alloc_ioctl, + struct iommu_viommu_alloc, out_viommu_id), #ifdef CONFIG_IOMMUFD_TEST IOCTL_OP(IOMMU_TEST_CMD, iommufd_test, struct iommu_test_cmd, last), #endif @@ -495,6 +498,9 @@ static const struct iommufd_object_ops iommufd_object_ops[] = { [IOMMUFD_OBJ_FAULT] = { .destroy = iommufd_fault_destroy, }, + [IOMMUFD_OBJ_VIOMMU] = { + .destroy = iommufd_viommu_destroy, + }, #ifdef CONFIG_IOMMUFD_TEST [IOMMUFD_OBJ_SELFTEST] = { .destroy = iommufd_selftest_destroy, diff --git a/drivers/iommu/iommufd/viommu.c b/drivers/iommu/iommufd/viommu.c new file mode 100644 index 000000000000..888239b78667 --- /dev/null +++ b/drivers/iommu/iommufd/viommu.c @@ -0,0 +1,81 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* Copyright (c) 2024, NVIDIA CORPORATION & AFFILIATES + */ +#include "iommufd_private.h" + +void iommufd_viommu_destroy(struct iommufd_object *obj) +{ + struct iommufd_viommu *viommu = + container_of(obj, struct iommufd_viommu, obj); + + if (viommu->ops && viommu->ops->destroy) + viommu->ops->destroy(viommu); + refcount_dec(&viommu->hwpt->common.obj.users); +} + +int iommufd_viommu_alloc_ioctl(struct iommufd_ucmd *ucmd) +{ + struct iommu_viommu_alloc *cmd = ucmd->cmd; + struct iommufd_hwpt_paging *hwpt_paging; + struct iommufd_viommu *viommu; + struct iommufd_device *idev; + const struct iommu_ops *ops; + int rc; + + if (cmd->flags || cmd->type == IOMMU_VIOMMU_TYPE_DEFAULT) + return -EOPNOTSUPP; + + idev = iommufd_get_device(ucmd, cmd->dev_id); + if (IS_ERR(idev)) + return PTR_ERR(idev); + + ops = dev_iommu_ops(idev->dev); + if (!ops->viommu_alloc) { + rc = -EOPNOTSUPP; + goto out_put_idev; + } + + hwpt_paging = iommufd_get_hwpt_paging(ucmd, cmd->hwpt_id); + if (IS_ERR(hwpt_paging)) { + rc = PTR_ERR(hwpt_paging); + goto out_put_idev; + } + + if (!hwpt_paging->nest_parent) { + rc = -EINVAL; + goto out_put_hwpt; + } + + viommu = ops->viommu_alloc(idev->dev, hwpt_paging->common.domain, + ucmd->ictx, cmd->type); + if (IS_ERR(viommu)) { + rc = PTR_ERR(viommu); + goto out_put_hwpt; + } + + viommu->type = cmd->type; + viommu->ictx = ucmd->ictx; + viommu->hwpt = hwpt_paging; + refcount_inc(&viommu->hwpt->common.obj.users); + /* + * It is the most likely case that a physical IOMMU is unpluggable. A + * pluggable IOMMU instance (if exists) is responsible for refcounting + * on its own. + */ + viommu->iommu_dev = __iommu_get_iommu_dev(idev->dev); + + cmd->out_viommu_id = viommu->obj.id; + rc = iommufd_ucmd_respond(ucmd, sizeof(*cmd)); + if (rc) + goto out_abort; + iommufd_object_finalize(ucmd->ictx, &viommu->obj); + goto out_put_hwpt; + +out_abort: + iommufd_object_abort_and_destroy(ucmd->ictx, &viommu->obj); +out_put_hwpt: + iommufd_put_object(ucmd->ictx, &hwpt_paging->common.obj); +out_put_idev: + iommufd_put_object(ucmd->ictx, &idev->obj); + return rc; +}

Now a vIOMMU holds a shareable nesting parent HWPT. So, it can act like that nesting parent HWPT to allocate a nested HWPT.

Support that in the IOMMU_HWPT_ALLOC ioctl handler, and update its kdoc.

Also, add an iommufd_viommu_alloc_hwpt_nested helper to allocate a nested HWPT for a vIOMMU object. Since a vIOMMU object holds the parent hwpt's refcount already, increase the refcount of the vIOMMU only.

Reviewed-by: Kevin Tian kevin.tian@intel.com Signed-off-by: Nicolin Chen nicolinc@nvidia.com --- drivers/iommu/iommufd/iommufd_private.h | 1 + include/uapi/linux/iommufd.h | 14 +++-- drivers/iommu/iommufd/hw_pagetable.c | 73 ++++++++++++++++++++++++- 3 files changed, 81 insertions(+), 7 deletions(-)

diff --git a/drivers/iommu/iommufd/iommufd_private.h b/drivers/iommu/iommufd/iommufd_private.h index a8104d9d4cef..e8f5ef550cc9 100644 --- a/drivers/iommu/iommufd/iommufd_private.h +++ b/drivers/iommu/iommufd/iommufd_private.h @@ -290,6 +290,7 @@ struct iommufd_hwpt_paging { struct iommufd_hwpt_nested { struct iommufd_hw_pagetable common; struct iommufd_hwpt_paging *parent; + struct iommufd_viommu *viommu; };

static inline bool hwpt_is_paging(struct iommufd_hw_pagetable *hwpt) diff --git a/include/uapi/linux/iommufd.h b/include/uapi/linux/iommufd.h index 302844136b02..a498d4838f9a 100644 --- a/include/uapi/linux/iommufd.h +++ b/include/uapi/linux/iommufd.h @@ -435,7 +435,7 @@ enum iommu_hwpt_data_type { * @size: sizeof(struct iommu_hwpt_alloc) * @flags: Combination of enum iommufd_hwpt_alloc_flags * @dev_id: The device to allocate this HWPT for - * @pt_id: The IOAS or HWPT to connect this HWPT to + * @pt_id: The IOAS or HWPT or vIOMMU to connect this HWPT to * @out_hwpt_id: The ID of the new HWPT * @__reserved: Must be 0 * @data_type: One of enum iommu_hwpt_data_type @@ -454,11 +454,13 @@ enum iommu_hwpt_data_type { * IOMMU_HWPT_DATA_NONE. The HWPT can be allocated as a parent HWPT for a * nesting configuration by passing IOMMU_HWPT_ALLOC_NEST_PARENT via @flags. * - * A user-managed nested HWPT will be created from a given parent HWPT via - * @pt_id, in which the parent HWPT must be allocated previously via the - * same ioctl from a given IOAS (@pt_id). In this case, the @data_type - * must be set to a pre-defined type corresponding to an I/O page table - * type supported by the underlying IOMMU hardware. + * A user-managed nested HWPT will be created from a given vIOMMU (wrapping a + * parent HWPT) or a parent HWPT via @pt_id, in which the parent HWPT must be + * allocated previously via the same ioctl from a given IOAS (@pt_id). In this + * case, the @data_type must be set to a pre-defined type corresponding to an + * I/O page table type supported by the underlying IOMMU hardware. The device + * via @dev_id and the vIOMMU via @pt_id must be associated to the same IOMMU + * instance. * * If the @data_type is set to IOMMU_HWPT_DATA_NONE, @data_len and * @data_uptr should be zero. Otherwise, both @data_len and @data_uptr diff --git a/drivers/iommu/iommufd/hw_pagetable.c b/drivers/iommu/iommufd/hw_pagetable.c index d06bf6e6c19f..982bf4a35a2b 100644 --- a/drivers/iommu/iommufd/hw_pagetable.c +++ b/drivers/iommu/iommufd/hw_pagetable.c @@ -57,7 +57,10 @@ void iommufd_hwpt_nested_destroy(struct iommufd_object *obj) container_of(obj, struct iommufd_hwpt_nested, common.obj);

__iommufd_hwpt_destroy(&hwpt_nested->common); - refcount_dec(&hwpt_nested->parent->common.obj.users); + if (hwpt_nested->viommu) + refcount_dec(&hwpt_nested->viommu->obj.users); + else + refcount_dec(&hwpt_nested->parent->common.obj.users); }

void iommufd_hwpt_nested_abort(struct iommufd_object *obj) @@ -260,6 +263,58 @@ iommufd_hwpt_nested_alloc(struct iommufd_ctx *ictx, return ERR_PTR(rc); }

+/** + * iommufd_viommu_alloc_hwpt_nested() - Get a hwpt_nested for a vIOMMU + * @viommu: vIOMMU ojbect to associate the hwpt_nested/domain with + * @flags: Flags from userspace + * @user_data: user_data pointer. Must be valid + * + * Allocate a new IOMMU_DOMAIN_NESTED for a vIOMMU and return it as a NESTED + * hw_pagetable. + */ +static struct iommufd_hwpt_nested * +iommufd_viommu_alloc_hwpt_nested(struct iommufd_viommu *viommu, u32 flags, + const struct iommu_user_data *user_data) +{ + struct iommufd_hwpt_nested *hwpt_nested; + struct iommufd_hw_pagetable *hwpt; + int rc; + + if (!user_data->len) + return ERR_PTR(-EOPNOTSUPP); + if (!viommu->ops || !viommu->ops->alloc_domain_nested) + return ERR_PTR(-EOPNOTSUPP); + + hwpt_nested = __iommufd_object_alloc( + viommu->ictx, hwpt_nested, IOMMUFD_OBJ_HWPT_NESTED, common.obj); + if (IS_ERR(hwpt_nested)) + return ERR_CAST(hwpt_nested); + hwpt = &hwpt_nested->common; + + hwpt_nested->viommu = viommu; + refcount_inc(&viommu->obj.users); + hwpt_nested->parent = viommu->hwpt; + + hwpt->domain = + viommu->ops->alloc_domain_nested(viommu, flags, user_data); + if (IS_ERR(hwpt->domain)) { + rc = PTR_ERR(hwpt->domain); + hwpt->domain = NULL; + goto out_abort; + } + hwpt->domain->owner = viommu->iommu_dev->ops; + + if (WARN_ON_ONCE(hwpt->domain->type != IOMMU_DOMAIN_NESTED)) { + rc = -EINVAL; + goto out_abort; + } + return hwpt_nested; + +out_abort: + iommufd_object_abort_and_destroy(viommu->ictx, &hwpt->obj); + return ERR_PTR(rc); +} + int iommufd_hwpt_alloc(struct iommufd_ucmd *ucmd) { struct iommu_hwpt_alloc *cmd = ucmd->cmd; @@ -316,6 +371,22 @@ int iommufd_hwpt_alloc(struct iommufd_ucmd *ucmd) goto out_unlock; } hwpt = &hwpt_nested->common; + } else if (pt_obj->type == IOMMUFD_OBJ_VIOMMU) { + struct iommufd_hwpt_nested *hwpt_nested; + struct iommufd_viommu *viommu; + + viommu = container_of(pt_obj, struct iommufd_viommu, obj); + if (viommu->iommu_dev != __iommu_get_iommu_dev(idev->dev)) { + rc = -EINVAL; + goto out_unlock; + } + hwpt_nested = iommufd_viommu_alloc_hwpt_nested( + viommu, cmd->flags, &user_data); + if (IS_ERR(hwpt_nested)) { + rc = PTR_ERR(hwpt_nested); + goto out_unlock; + } + hwpt = &hwpt_nested->common; } else { rc = -EINVAL; goto out_put_pt;

A nested domain now can be allocated for a parent domain or for a vIOMMU object. Rework the existing allocators to prepare for the latter case.

Reviewed-by: Kevin Tian kevin.tian@intel.com Reviewed-by: Jason Gunthorpe jgg@nvidia.com Signed-off-by: Nicolin Chen nicolinc@nvidia.com --- drivers/iommu/iommufd/selftest.c | 89 ++++++++++++++++++-------------- 1 file changed, 50 insertions(+), 39 deletions(-)

diff --git a/drivers/iommu/iommufd/selftest.c b/drivers/iommu/iommufd/selftest.c index 322e57ff3605..92d753985640 100644 --- a/drivers/iommu/iommufd/selftest.c +++ b/drivers/iommu/iommufd/selftest.c @@ -318,55 +318,39 @@ static struct iommu_domain *mock_domain_alloc_paging(struct device *dev) return &mock->domain; }

-static struct iommu_domain * -__mock_domain_alloc_nested(struct mock_iommu_domain *mock_parent, - const struct iommu_hwpt_selftest *user_cfg) +static struct mock_iommu_domain_nested * +__mock_domain_alloc_nested(const struct iommu_user_data *user_data) { struct mock_iommu_domain_nested *mock_nested; - int i; + struct iommu_hwpt_selftest user_cfg; + int rc, i; + + if (user_data->type != IOMMU_HWPT_DATA_SELFTEST) + return ERR_PTR(-EOPNOTSUPP); + + rc = iommu_copy_struct_from_user(&user_cfg, user_data, + IOMMU_HWPT_DATA_SELFTEST, iotlb); + if (rc) + return ERR_PTR(rc);

mock_nested = kzalloc(sizeof(*mock_nested), GFP_KERNEL); if (!mock_nested) return ERR_PTR(-ENOMEM); - mock_nested->parent = mock_parent; mock_nested->domain.ops = &domain_nested_ops; mock_nested->domain.type = IOMMU_DOMAIN_NESTED; for (i = 0; i < MOCK_NESTED_DOMAIN_IOTLB_NUM; i++) - mock_nested->iotlb[i] = user_cfg->iotlb; - return &mock_nested->domain; + mock_nested->iotlb[i] = user_cfg.iotlb; + return mock_nested; }

static struct iommu_domain * -mock_domain_alloc_user(struct device *dev, u32 flags, - struct iommu_domain *parent, - const struct iommu_user_data *user_data) +mock_domain_alloc_nested(struct iommu_domain *parent, u32 flags, + const struct iommu_user_data *user_data) { + struct mock_iommu_domain_nested *mock_nested; struct mock_iommu_domain *mock_parent; - struct iommu_hwpt_selftest user_cfg; - int rc; - - /* must be mock_domain */ - if (!parent) { - struct mock_dev *mdev = to_mock_dev(dev); - bool has_dirty_flag = flags & IOMMU_HWPT_ALLOC_DIRTY_TRACKING; - bool no_dirty_ops = mdev->flags & MOCK_FLAGS_DEVICE_NO_DIRTY; - struct iommu_domain *domain; - - if (flags & (~(IOMMU_HWPT_ALLOC_NEST_PARENT | - IOMMU_HWPT_ALLOC_DIRTY_TRACKING))) - return ERR_PTR(-EOPNOTSUPP); - if (user_data || (has_dirty_flag && no_dirty_ops)) - return ERR_PTR(-EOPNOTSUPP); - domain = mock_domain_alloc_paging(dev); - if (!domain) - return ERR_PTR(-ENOMEM); - if (has_dirty_flag) - domain->dirty_ops = &dirty_ops; - return domain; - }

- /* must be mock_domain_nested */ - if (user_data->type != IOMMU_HWPT_DATA_SELFTEST || flags) + if (flags) return ERR_PTR(-EOPNOTSUPP); if (!parent || parent->ops != mock_ops.default_domain_ops) return ERR_PTR(-EINVAL); @@ -375,12 +359,39 @@ mock_domain_alloc_user(struct device *dev, u32 flags, if (!mock_parent) return ERR_PTR(-EINVAL);

- rc = iommu_copy_struct_from_user(&user_cfg, user_data, - IOMMU_HWPT_DATA_SELFTEST, iotlb); - if (rc) - return ERR_PTR(rc); + mock_nested = __mock_domain_alloc_nested(user_data); + if (IS_ERR(mock_nested)) + return ERR_CAST(mock_nested); + mock_nested->parent = mock_parent; + return &mock_nested->domain; +} + +static struct iommu_domain * +mock_domain_alloc_user(struct device *dev, u32 flags, + struct iommu_domain *parent, + const struct iommu_user_data *user_data) +{ + bool has_dirty_flag = flags & IOMMU_HWPT_ALLOC_DIRTY_TRACKING; + const u32 PAGING_FLAGS = IOMMU_HWPT_ALLOC_DIRTY_TRACKING | + IOMMU_HWPT_ALLOC_NEST_PARENT; + bool no_dirty_ops = to_mock_dev(dev)->flags & + MOCK_FLAGS_DEVICE_NO_DIRTY; + struct iommu_domain *domain; + + if (parent) + return mock_domain_alloc_nested(parent, flags, user_data);

- return __mock_domain_alloc_nested(mock_parent, &user_cfg); + if (user_data) + return ERR_PTR(-EOPNOTSUPP); + if ((flags & ~PAGING_FLAGS) || (has_dirty_flag && no_dirty_ops)) + return ERR_PTR(-EOPNOTSUPP); + + domain = mock_domain_alloc_paging(dev); + if (!domain) + return ERR_PTR(-ENOMEM); + if (has_dirty_flag) + domain->dirty_ops = &dirty_ops; + return domain; }

static void mock_domain_free(struct iommu_domain *domain)

Implement the viommu alloc/free functions to increase/reduce refcount of its dependent mock iommu device. User space can verify this loop via the IOMMU_VIOMMU_TYPE_SELFTEST.

Reviewed-by: Kevin Tian kevin.tian@intel.com Reviewed-by: Jason Gunthorpe jgg@nvidia.com Signed-off-by: Nicolin Chen nicolinc@nvidia.com --- drivers/iommu/iommufd/iommufd_test.h | 2 + drivers/iommu/iommufd/selftest.c | 67 ++++++++++++++++++++++++++++ 2 files changed, 69 insertions(+)

diff --git a/drivers/iommu/iommufd/iommufd_test.h b/drivers/iommu/iommufd/iommufd_test.h index f4bc23a92f9a..edced4ac7cd3 100644 --- a/drivers/iommu/iommufd/iommufd_test.h +++ b/drivers/iommu/iommufd/iommufd_test.h @@ -180,4 +180,6 @@ struct iommu_hwpt_invalidate_selftest { __u32 iotlb_id; };

+#define IOMMU_VIOMMU_TYPE_SELFTEST 0xdeadbeef + #endif diff --git a/drivers/iommu/iommufd/selftest.c b/drivers/iommu/iommufd/selftest.c index 4f67a83f667a..31c8f78a3a66 100644 --- a/drivers/iommu/iommufd/selftest.c +++ b/drivers/iommu/iommufd/selftest.c @@ -134,6 +134,7 @@ to_mock_domain(struct iommu_domain *domain)

struct mock_iommu_domain_nested { struct iommu_domain domain; + struct mock_viommu *mock_viommu; struct mock_iommu_domain *parent; u32 iotlb[MOCK_NESTED_DOMAIN_IOTLB_NUM]; }; @@ -144,6 +145,16 @@ to_mock_nested(struct iommu_domain *domain) return container_of(domain, struct mock_iommu_domain_nested, domain); }

+struct mock_viommu { + struct iommufd_viommu core; + struct mock_iommu_domain *s2_parent; +}; + +static inline struct mock_viommu *to_mock_viommu(struct iommufd_viommu *viommu) +{ + return container_of(viommu, struct mock_viommu, core); +} + enum selftest_obj_type { TYPE_IDEV, }; @@ -569,6 +580,61 @@ static int mock_dev_disable_feat(struct device *dev, enum iommu_dev_features fea return 0; }

+static void mock_viommu_destroy(struct iommufd_viommu *viommu) +{ + struct mock_iommu_device *mock_iommu = container_of( + viommu->iommu_dev, struct mock_iommu_device, iommu_dev); + + if (refcount_dec_and_test(&mock_iommu->users)) + complete(&mock_iommu->complete); + + /* iommufd core frees mock_viommu and viommu */ +} + +static struct iommu_domain * +mock_viommu_alloc_domain_nested(struct iommufd_viommu *viommu, u32 flags, + const struct iommu_user_data *user_data) +{ + struct mock_viommu *mock_viommu = to_mock_viommu(viommu); + struct mock_iommu_domain_nested *mock_nested; + + if (flags & ~IOMMU_HWPT_FAULT_ID_VALID) + return ERR_PTR(-EOPNOTSUPP); + + mock_nested = __mock_domain_alloc_nested(user_data); + if (IS_ERR(mock_nested)) + return ERR_CAST(mock_nested); + mock_nested->mock_viommu = mock_viommu; + mock_nested->parent = mock_viommu->s2_parent; + return &mock_nested->domain; +} + +static struct iommufd_viommu_ops mock_viommu_ops = { + .destroy = mock_viommu_destroy, + .alloc_domain_nested = mock_viommu_alloc_domain_nested, +}; + +static struct iommufd_viommu *mock_viommu_alloc(struct device *dev, + struct iommu_domain *domain, + struct iommufd_ctx *ictx, + unsigned int viommu_type) +{ + struct mock_iommu_device *mock_iommu = + iommu_get_iommu_dev(dev, struct mock_iommu_device, iommu_dev); + struct mock_viommu *mock_viommu; + + if (viommu_type != IOMMU_VIOMMU_TYPE_SELFTEST) + return ERR_PTR(-EOPNOTSUPP); + + mock_viommu = iommufd_viommu_alloc(ictx, struct mock_viommu, core, + &mock_viommu_ops); + if (IS_ERR(mock_viommu)) + return ERR_CAST(mock_viommu); + + refcount_inc(&mock_iommu->users); + return &mock_viommu->core; +} + static const struct iommu_ops mock_ops = { /* * IOMMU_DOMAIN_BLOCKED cannot be returned from def_domain_type() @@ -588,6 +654,7 @@ static const struct iommu_ops mock_ops = { .dev_enable_feat = mock_dev_enable_feat, .dev_disable_feat = mock_dev_disable_feat, .user_pasid_table = true, + .viommu_alloc = mock_viommu_alloc, .default_domain_ops = &(struct iommu_domain_ops){ .free = mock_domain_free,

With the introduction of the new object and its infrastructure, update the doc to reflect that and add a new graph.

Reviewed-by: Jason Gunthorpe jgg@nvidia.com Reviewed-by: Kevin Tian kevin.tian@intel.com Signed-off-by: Nicolin Chen nicolinc@nvidia.com --- Documentation/userspace-api/iommufd.rst | 69 ++++++++++++++++++++++++- 1 file changed, 68 insertions(+), 1 deletion(-)

diff --git a/Documentation/userspace-api/iommufd.rst b/Documentation/userspace-api/iommufd.rst index 2deba93bf159..a8b7766c2849 100644 --- a/Documentation/userspace-api/iommufd.rst +++ b/Documentation/userspace-api/iommufd.rst @@ -63,6 +63,37 @@ Following IOMMUFD objects are exposed to userspace: space usually has mappings from guest-level I/O virtual addresses to guest- level physical addresses.

+- IOMMUFD_OBJ_VIOMMU, representing a slice of the physical IOMMU instance, + passed to or shared with a VM. It may be some HW-accelerated virtualization + features and some SW resources used by the VM. For examples: + * Security namespace for guest owned ID, e.g. guest-controlled cache tags + * Non-device-affiliated event reporting, e.g. invalidation queue errors + * Access to a sharable nesting parent pagetable across physical IOMMUs + * Virtualization of various platforms IDs, e.g. RIDs and others + * Delivery of paravirtualized invalidation + * Direct assigned invalidation queues + * Direct assigned interrupts + Such a vIOMMU object generally has the access to a nesting parent pagetable + to support some HW-accelerated virtualization features. So, a vIOMMU object + must be created given a nesting parent HWPT_PAGING object, and then it would + encapsulate that HWPT_PAGING object. Therefore, a vIOMMU object can be used + to allocate an HWPT_NESTED object in place of the encapsulated HWPT_PAGING. + + .. note:: + + The name "vIOMMU" isn't necessarily identical to a virtualized IOMMU in a + VM. A VM can have one giant virtualized IOMMU running on a machine having + multiple physical IOMMUs, in which case the VMM will dispatch the requests + or configurations from this single virtualized IOMMU instance to multiple + vIOMMU objects created for individual slices of different physical IOMMUs. + In other words, a vIOMMU object is always a representation of one physical + IOMMU, not necessarily of a virtualized IOMMU. For VMMs that want the full + virtualization features from physical IOMMUs, it is suggested to build the + same number of virtualized IOMMUs as the number of physical IOMMUs, so the + passed-through devices would be connected to their own virtualized IOMMUs + backed by corresponding vIOMMU objects, in which case a guest OS would do + the "dispatch" naturally instead of VMM trappings. + All user-visible objects are destroyed via the IOMMU_DESTROY uAPI.

The diagrams below show relationships between user-visible objects and kernel @@ -101,6 +132,28 @@ creating the objects and links:: |------------>|iommu_domain|<----|iommu_domain|<----|device| |____________| |____________| |______|

+ _______________________________________________________________________ + | iommufd (with vIOMMU) | + | | + | [5] | + | _____________ | + | | | | + | |----------------| vIOMMU | | + | | | | | + | | | | | + | | [1] | | [4] [2] | + | | ______ | | _____________ ________ | + | | | | | [3] | | | | | | + | | | IOAS |<---|(HWPT_PAGING)|<---| HWPT_NESTED |<--| DEVICE | | + | | |______| |_____________| |_____________| |________| | + | | | | | | | + |______|________|______________|__________________|_______________|_____| + | | | | | + ______v_____ | ______v_____ ______v_____ ___v__ + | struct | | PFN | (paging) | | (nested) | |struct| + |iommu_device| |------>|iommu_domain|<----|iommu_domain|<----|device| + |____________| storage|____________| |____________| |______| + 1. IOMMUFD_OBJ_IOAS is created via the IOMMU_IOAS_ALLOC uAPI. An iommufd can hold multiple IOAS objects. IOAS is the most generic object and does not expose interfaces that are specific to single IOMMU drivers. All operations @@ -132,7 +185,8 @@ creating the objects and links:: flag is set.

4. IOMMUFD_OBJ_HWPT_NESTED can be only manually created via the IOMMU_HWPT_ALLOC - uAPI, provided an hwpt_id via @pt_id to associate the new HWPT_NESTED object + uAPI, provided an hwpt_id or a viommu_id of a vIOMMU object encapsulating a + nesting parent HWPT_PAGING via @pt_id to associate the new HWPT_NESTED object to the corresponding HWPT_PAGING object. The associating HWPT_PAGING object must be a nesting parent manually allocated via the same uAPI previously with an IOMMU_HWPT_ALLOC_NEST_PARENT flag, otherwise the allocation will fail. The @@ -149,6 +203,18 @@ creating the objects and links:: created via the same IOMMU_HWPT_ALLOC uAPI. The difference is at the type of the object passed in via the @pt_id field of struct iommufd_hwpt_alloc.

+5. IOMMUFD_OBJ_VIOMMU can be only manually created via the IOMMU_VIOMMU_ALLOC + uAPI, provided a dev_id (for the device's physical IOMMU to back the vIOMMU) + and an hwpt_id (to associate the vIOMMU to a nesting parent HWPT_PAGING). The + iommufd core will link the vIOMMU object to the struct iommu_device that the + struct device is behind. And an IOMMU driver can implement a viommu_alloc op + to allocate its own vIOMMU data structure embedding the core-level structure + iommufd_viommu and some driver-specific data. If necessary, the driver can + also configure its HW virtualization feature for that vIOMMU (and thus for + the VM). Successful completion of this operation sets up the linkages between + the vIOMMU object and the HWPT_PAGING, then this vIOMMU object can be used + as a nesting parent object to allocate an HWPT_NESTED object described above. + A device can only bind to an iommufd due to DMA ownership claim and attach to at most one IOAS object (no support of PASID yet).

@@ -161,6 +227,7 @@ User visible objects are backed by following datastructures: - iommufd_device for IOMMUFD_OBJ_DEVICE. - iommufd_hwpt_paging for IOMMUFD_OBJ_HWPT_PAGING. - iommufd_hwpt_nested for IOMMUFD_OBJ_HWPT_NESTED. +- iommufd_viommu for IOMMUFD_OBJ_VIOMMU.

Several terminologies when looking at these datastructures: