From: Yi Liu <yi.l.liu@intel.com>
To: joro@8bytes.org, alex.williamson@redhat.com, jgg@nvidia.com,
kevin.tian@intel.com, robin.murphy@arm.com
Cc: cohuck@redhat.com, eric.auger@redhat.com, nicolinc@nvidia.com,
kvm@vger.kernel.org, mjrosato@linux.ibm.com,
chao.p.peng@linux.intel.com, yi.l.liu@intel.com,
yi.y.sun@linux.intel.com, peterx@redhat.com, jasowang@redhat.com,
shameerali.kolothum.thodi@huawei.com, lulu@redhat.com,
suravee.suthikulpanit@amd.com, iommu@lists.linux.dev,
linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org,
baolu.lu@linux.intel.com
Subject: [PATCH 09/17] iommufd: Add kernel-managed hw_pagetable allocation for userspace
Date: Wed, 8 Feb 2023 20:31:45 -0800 [thread overview]
Message-ID: <20230209043153.14964-10-yi.l.liu@intel.com> (raw)
In-Reply-To: <20230209043153.14964-1-yi.l.liu@intel.com>
Today iommufd allocates the kernel-managed hw_pagetabe implicitly when
device is attached to an IOAS. This links the hw_pagetable to IOPT within
IOAS.
However, this is not the perfect way. It makes much sense to let userspace
explicitly request hw_pagtable allocation via iommufd. The reason is even
though the hw_pagetable is kernel-managed, the mappings are feed by
userspace. Also, this makes the lifecircle of kernel-managed hw_pagetable
more clear during usage. This is very important in the usage of nested
translation, in which the kernel-managed hw_pagetable would be used as the
stage-2 hw_pagetable. In such case, both stage-1 and stage-2 hw_pagetable
should be allocated by userspace to ensure the life-circle.
This adds an ioctl IOMMU_HWPT_ALLOC for the hw_pagetable allocation. For
kernel-managed hw_pagetable, userspace should provide an IOAS ID in the
allocation request.
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Signed-off-by: Yi Liu <yi.l.liu@intel.com>
---
drivers/iommu/iommufd/device.c | 11 ++-
drivers/iommu/iommufd/hw_pagetable.c | 116 ++++++++++++++++++++++++
drivers/iommu/iommufd/iommufd_private.h | 15 +++
drivers/iommu/iommufd/main.c | 3 +
include/uapi/linux/iommufd.h | 48 ++++++++++
5 files changed, 191 insertions(+), 2 deletions(-)
diff --git a/drivers/iommu/iommufd/device.c b/drivers/iommu/iommufd/device.c
index 71a8c4f1c4a9..dd7943ff02e4 100644
--- a/drivers/iommu/iommufd/device.c
+++ b/drivers/iommu/iommufd/device.c
@@ -146,8 +146,7 @@ static int iommufd_zero_fill_user(u64 ptr, int bytes)
return 0;
}
-static struct device *
-iommufd_obj_dev(struct iommufd_object *obj)
+struct device *iommufd_obj_dev(struct iommufd_object *obj)
{
struct device *dev = NULL;
@@ -160,6 +159,14 @@ iommufd_obj_dev(struct iommufd_object *obj)
return dev;
}
+/*
+ * bitmaps of supported page table data types of hardware iommu,
+ * indexed by the members defined in enum iommu_device_data_type.
+ */
+const u64 iommufd_supported_pgtbl_types[] = {
+ [IOMMU_DEVICE_DATA_INTEL_VTD] = BIT_ULL(IOMMU_PGTBL_DATA_NONE),
+};
+
int iommufd_device_get_info(struct iommufd_ucmd *ucmd)
{
struct iommu_device_info *cmd = ucmd->cmd;
diff --git a/drivers/iommu/iommufd/hw_pagetable.c b/drivers/iommu/iommufd/hw_pagetable.c
index ee97d2f3cf43..998d01490a74 100644
--- a/drivers/iommu/iommufd/hw_pagetable.c
+++ b/drivers/iommu/iommufd/hw_pagetable.c
@@ -92,3 +92,119 @@ iommufd_hw_pagetable_alloc(struct iommufd_ctx *ictx, struct iommufd_ioas *ioas,
{
return __iommufd_hw_pagetable_alloc(ictx, ioas, dev, NULL, NULL);
}
+
+/*
+ * size of page table type specific data, indexed by
+ * enum iommu_pgtbl_data_type.
+ */
+static const size_t iommufd_hwpt_info_size[] = {
+ [IOMMU_PGTBL_DATA_NONE] = 0,
+};
+
+int iommufd_hwpt_alloc(struct iommufd_ucmd *ucmd)
+{
+ struct iommufd_hw_pagetable *hwpt;
+ struct iommu_hwpt_alloc *cmd = ucmd->cmd;
+ struct iommufd_ctx *ictx = ucmd->ictx;
+ struct iommufd_object *pt_obj = NULL;
+ struct iommufd_ioas *ioas = NULL;
+ struct iommufd_object *dev_obj;
+ struct device *dev;
+ const struct iommu_ops *ops;
+ void *data = NULL;
+ u32 driver_type, klen;
+ int rc;
+
+ if (cmd->__reserved || cmd->flags)
+ return -EOPNOTSUPP;
+
+ dev_obj = iommufd_get_object(ucmd->ictx, cmd->dev_id,
+ IOMMUFD_OBJ_ANY);
+ if (IS_ERR(dev_obj))
+ return PTR_ERR(dev_obj);
+
+ dev = iommufd_obj_dev(dev_obj);
+ if (!dev) {
+ rc = -EINVAL;
+ goto out_put_dev;
+ }
+
+ ops = dev_iommu_ops(dev);
+ if (!ops) {
+ rc = -EOPNOTSUPP;
+ goto out_put_dev;
+ }
+
+ driver_type = ops->driver_type;
+
+ /* data_type should be a supported type by the hardware */
+ if (!((1 << cmd->data_type) &
+ iommufd_supported_pgtbl_types[driver_type])) {
+ rc = -EINVAL;
+ goto out_put_dev;
+ }
+
+ pt_obj = iommufd_get_object(ictx, cmd->pt_id, IOMMUFD_OBJ_ANY);
+ if (IS_ERR(pt_obj)) {
+ rc = -EINVAL;
+ goto out_put_dev;
+ }
+
+ switch (pt_obj->type) {
+ case IOMMUFD_OBJ_IOAS:
+ ioas = container_of(pt_obj, struct iommufd_ioas, obj);
+ break;
+ default:
+ rc = -EINVAL;
+ goto out_put_pt;
+ }
+
+ klen = iommufd_hwpt_info_size[cmd->data_type];
+ if (klen) {
+ if (!cmd->data_len) {
+ rc = -EINVAL;
+ goto out_put_pt;
+ }
+
+ data = kzalloc(klen, GFP_KERNEL);
+ if (!data) {
+ rc = -ENOMEM;
+ goto out_put_pt;
+ }
+
+ rc = copy_struct_from_user(data, klen,
+ u64_to_user_ptr(cmd->data_uptr),
+ cmd->data_len);
+ if (rc)
+ goto out_free_data;
+ }
+
+ mutex_lock(&ioas->mutex);
+ hwpt = __iommufd_hw_pagetable_alloc(ictx, ioas, dev, NULL, data);
+ mutex_unlock(&ioas->mutex);
+ if (IS_ERR(hwpt)) {
+ rc = PTR_ERR(hwpt);
+ goto out_free_data;
+ }
+
+ cmd->out_hwpt_id = hwpt->obj.id;
+
+ rc = iommufd_ucmd_respond(ucmd, sizeof(*cmd));
+ if (rc)
+ goto out_destroy_hwpt;
+
+ kfree(data);
+ iommufd_object_finalize(ucmd->ictx, &hwpt->obj);
+ iommufd_put_object(pt_obj);
+ iommufd_put_object(dev_obj);
+ return 0;
+out_destroy_hwpt:
+ iommufd_object_abort_and_destroy(ucmd->ictx, &hwpt->obj);
+out_free_data:
+ kfree(data);
+out_put_pt:
+ iommufd_put_object(pt_obj);
+out_put_dev:
+ iommufd_put_object(dev_obj);
+ return rc;
+}
diff --git a/drivers/iommu/iommufd/iommufd_private.h b/drivers/iommu/iommufd/iommufd_private.h
index 604ad29f87b8..ee5344baf135 100644
--- a/drivers/iommu/iommufd/iommufd_private.h
+++ b/drivers/iommu/iommufd/iommufd_private.h
@@ -257,7 +257,22 @@ iommufd_hw_pagetable_alloc(struct iommufd_ctx *ictx, struct iommufd_ioas *ioas,
struct device *dev);
void iommufd_hw_pagetable_destroy(struct iommufd_object *obj);
+static inline struct iommufd_hw_pagetable *
+iommufd_get_hwpt(struct iommufd_ucmd *ucmd, u32 id)
+{
+ return container_of(iommufd_get_object(ucmd->ictx, id,
+ IOMMUFD_OBJ_HW_PAGETABLE),
+ struct iommufd_hw_pagetable, obj);
+}
+
+int iommufd_hwpt_alloc(struct iommufd_ucmd *ucmd);
+
+struct device *iommufd_obj_dev(struct iommufd_object *obj);
+
void iommufd_device_destroy(struct iommufd_object *obj);
+
+extern const u64 iommufd_supported_pgtbl_types[];
+
int iommufd_device_get_info(struct iommufd_ucmd *ucmd);
struct iommufd_access {
diff --git a/drivers/iommu/iommufd/main.c b/drivers/iommu/iommufd/main.c
index 59aa30ad1090..831303d64abe 100644
--- a/drivers/iommu/iommufd/main.c
+++ b/drivers/iommu/iommufd/main.c
@@ -251,6 +251,7 @@ static int iommufd_option(struct iommufd_ucmd *ucmd)
union ucmd_buffer {
struct iommu_destroy destroy;
struct iommu_device_info info;
+ struct iommu_hwpt_alloc hwpt;
struct iommu_ioas_alloc alloc;
struct iommu_ioas_allow_iovas allow_iovas;
struct iommu_ioas_copy ioas_copy;
@@ -284,6 +285,8 @@ static const struct iommufd_ioctl_op iommufd_ioctl_ops[] = {
IOCTL_OP(IOMMU_DESTROY, iommufd_destroy, struct iommu_destroy, id),
IOCTL_OP(IOMMU_DEVICE_GET_INFO, iommufd_device_get_info, struct iommu_device_info,
__reserved),
+ IOCTL_OP(IOMMU_HWPT_ALLOC, iommufd_hwpt_alloc, struct iommu_hwpt_alloc,
+ __reserved),
IOCTL_OP(IOMMU_IOAS_ALLOC, iommufd_ioas_alloc_ioctl,
struct iommu_ioas_alloc, out_ioas_id),
IOCTL_OP(IOMMU_IOAS_ALLOW_IOVAS, iommufd_ioas_allow_iovas,
diff --git a/include/uapi/linux/iommufd.h b/include/uapi/linux/iommufd.h
index bbffb63d2513..f501add5ffe9 100644
--- a/include/uapi/linux/iommufd.h
+++ b/include/uapi/linux/iommufd.h
@@ -46,6 +46,7 @@ enum {
IOMMUFD_CMD_OPTION,
IOMMUFD_CMD_VFIO_IOAS,
IOMMUFD_CMD_DEVICE_GET_INFO,
+ IOMMUFD_CMD_HWPT_ALLOC,
};
/**
@@ -373,6 +374,14 @@ struct iommu_device_info_vtd {
__aligned_u64 ecap_reg;
};
+/**
+ * enum iommu_pgtbl_data_type - IOMMU Page Table User Data type
+ * @IOMMU_PGTBL_DATA_NONE: no user data
+ */
+enum iommu_pgtbl_data_type {
+ IOMMU_PGTBL_DATA_NONE,
+};
+
/**
* struct iommu_device_info - ioctl(IOMMU_DEVICE_GET_INFO)
* @size: sizeof(struct iommu_device_info)
@@ -461,6 +470,45 @@ struct iommu_hwpt_intel_vtd {
__u32 __reserved;
};
+/**
+ * struct iommu_hwpt_alloc - ioctl(IOMMU_HWPT_ALLOC)
+ * @size: sizeof(struct iommu_hwpt_alloc)
+ * @flags: Must be 0
+ * @dev_id: The device to allocate this HWPT for
+ * @pt_id: The parent of this HWPT (IOAS or HWPT)
+ * @data_type: One of enum iommu_pgtbl_data_type
+ * @data_len: Length of the type specific data
+ * @data_uptr: User pointer to the type specific data
+ * @out_hwpt_id: Output HWPT ID for the allocated object
+ * @__reserved: Must be 0
+ *
+ * Allocate hw_pagetable for managing page tables in userspace. Such page
+ * tables can be user-managed or kernel-managed. @pt_id is needed for either
+ * case. While the @data_type, @data_len and @data_uptr are optional. For
+ * the user-managed page tables, userspace should provide the data_type, the
+ * data_len and the type speficific data. While for the kernel-managed page
+ * tables, use the IOMMU_PGTBL_DATA_NONE data_type, @data_len and @data_uptr
+ * will be ignored.
+ *
+ * +==============================+=====================================+
+ * | @data_type | Data structure in @data_uptr |
+ * +------------------------------+-------------------------------------+
+ * | IOMMU_PGTBL_DATA_NONE | N/A |
+ * +------------------------------+-------------------------------------+
+ */
+struct iommu_hwpt_alloc {
+ __u32 size;
+ __u32 flags;
+ __u32 dev_id;
+ __u32 pt_id;
+ __u32 data_type;
+ __u32 data_len;
+ __aligned_u64 data_uptr;
+ __u32 out_hwpt_id;
+ __u32 __reserved;
+};
+#define IOMMU_HWPT_ALLOC _IO(IOMMUFD_TYPE, IOMMUFD_CMD_HWPT_ALLOC)
+
/**
* enum iommu_vtd_qi_granularity - Intel VT-d specific granularity of
* queued invalidation
--
2.34.1
next prev parent reply other threads:[~2023-02-09 4:35 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-09 4:31 [PATCH 00/17] Add Intel VT-d nested translation Yi Liu
2023-02-09 4:31 ` [PATCH 01/17] iommu: Add new iommu op to create domains owned by userspace Yi Liu
2023-02-10 8:24 ` Tian, Kevin
2023-02-11 3:16 ` Baolu Lu
2023-02-09 4:31 ` [PATCH 02/17] iommu: Add nested domain support Yi Liu
2023-02-14 18:47 ` Nicolin Chen
2023-02-09 4:31 ` [PATCH 03/17] iommu/vt-d: Extend dmar_domain to support nested domain Yi Liu
2023-02-09 4:31 ` [PATCH 04/17] iommu/vt-d: Add helper to setup pasid nested translation Yi Liu
2023-02-09 4:31 ` [PATCH 05/17] iommu/vt-d: Add nested domain support Yi Liu
2023-02-09 4:31 ` [PATCH 06/17] iommufd/hw_pagetable: Use domain_alloc_user op for domain allocation Yi Liu
2023-02-09 17:59 ` Matthew Rosato
2023-02-09 18:36 ` Jason Gunthorpe
2023-02-09 19:51 ` Nicolin Chen
2023-02-09 20:39 ` Jason Gunthorpe
2023-02-09 22:22 ` Nicolin Chen
2023-02-09 23:59 ` Jason Gunthorpe
2023-02-10 10:50 ` Liu, Yi L
2023-02-09 4:31 ` [PATCH 07/17] iommufd: Add/del hwpt to IOAS at alloc/destroy() Yi Liu
2023-02-09 4:31 ` [PATCH 08/17] iommufd: Split iommufd_hw_pagetable_alloc() Yi Liu
2023-02-09 4:31 ` Yi Liu [this message]
2023-02-09 20:45 ` [PATCH 09/17] iommufd: Add kernel-managed hw_pagetable allocation for userspace Jason Gunthorpe
2023-02-10 10:52 ` Liu, Yi L
2023-02-09 4:31 ` [PATCH 10/17] iommufd/device: Move IOAS attaching and detaching operations into helpers Yi Liu
2023-02-09 4:31 ` [PATCH 11/17] iommufd: Add infrastructure for user-managed hw_pagetable allocation Yi Liu
2023-02-09 4:31 ` [PATCH 12/17] iommufd: Add " Yi Liu
2023-02-09 4:31 ` [PATCH 13/17] iommufd/device: Report supported stage-1 page table types Yi Liu
2023-02-09 4:31 ` [PATCH 14/17] iommufd/selftest: Add IOMMU_TEST_OP_MOCK_DOMAIN_REPLACE test op Yi Liu
2023-02-14 18:36 ` Nicolin Chen
2023-02-09 4:31 ` [PATCH 15/17] iommufd/selftest: Add coverage for IOMMU_HWPT_ALLOC ioctl Yi Liu
2023-02-09 4:31 ` [PATCH 16/17] iommufd/selftest: Add IOMMU_TEST_OP_MD_CHECK_IOTLB test op Yi Liu
2023-02-09 4:31 ` [PATCH 17/17] iommufd/selftest: Add coverage for IOMMU_HWPT_INVALIDATE ioctl Yi Liu
2023-02-09 10:11 ` [PATCH 00/17] Add Intel VT-d nested translation Shameerali Kolothum Thodi
2023-02-09 16:10 ` Nicolin Chen
2023-02-09 16:16 ` Shameerali Kolothum Thodi
2023-02-17 18:20 ` Nicolin Chen
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20230209043153.14964-10-yi.l.liu@intel.com \
--to=yi.l.liu@intel.com \
--cc=alex.williamson@redhat.com \
--cc=baolu.lu@linux.intel.com \
--cc=chao.p.peng@linux.intel.com \
--cc=cohuck@redhat.com \
--cc=eric.auger@redhat.com \
--cc=iommu@lists.linux.dev \
--cc=jasowang@redhat.com \
--cc=jgg@nvidia.com \
--cc=joro@8bytes.org \
--cc=kevin.tian@intel.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=lulu@redhat.com \
--cc=mjrosato@linux.ibm.com \
--cc=nicolinc@nvidia.com \
--cc=peterx@redhat.com \
--cc=robin.murphy@arm.com \
--cc=shameerali.kolothum.thodi@huawei.com \
--cc=suravee.suthikulpanit@amd.com \
--cc=yi.y.sun@linux.intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).