[PATCH v13 00/13] SMMUv3 Nested Stage Setup (VFIO part)

[PATCH v13 00/13] SMMUv3 Nested Stage Setup (VFIO part)

[Eric Auger]

This series brings the VFIO part of HW nested paging support
in the SMMUv3.

The series depends on:
[PATCH v15 00/13] SMMUv3 Nested Stage Setup (IOMMU part)

3 new IOCTLs are introduced that allow the userspace to
1) pass the guest stage 1 configuration
2) pass stage 1 MSI bindings
3) invalidate stage 1 related caches

They map onto the related new IOMMU API functions.

We introduce the capability to register specific interrupt
indexes (see [1]). A new DMA_FAULT interrupt index allows to register
an eventfd to be signaled whenever a stage 1 related fault
is detected at physical level. Also two specific regions allow to
- expose the fault records to the user space and
- inject page responses.

This latter functionality is not exercised in this series
but is provided as a POC for further vSVA activities (Shameer's input).

Best Regards

Eric

This series can be found at:
https://github.com/eauger/linux/tree/v5.12-rc6-jean-iopf-14-2stage-v15

The series series includes Tina's patch steming from
[1] "[RFC PATCH v2 1/3] vfio: Use capability chains to handle device
specific irq" plus patches originally contributed by Yi.

History:

v12 -> v13:
- Comment fix for VFIO_IOMMU_SET_PASID_TABLE (Zenghui)

v11 -> v12:
- numerous fixes following feedbacks. Many thanks to all of you
- See individual history log.

v10 -> v11:
- rebase on top of v5.10-rc4
- adapt to changes on the IOMMU API (compliant with the doc
  written by Jacob/Yi)
- addition of the page response region
- Took into account Zenghui's comments
- In this version I have kept the ioctl separate. Since
  Yi's series [2] is currently stalled, I've just rebased here.

[2] [PATCH v7 00/16] vfio: expose virtual Shared Virtual Addressing
to VMs

v9 -> v10
- rebase on top of 5.6.0-rc3 (no change versus v9)

Eric Auger (10):
  vfio: VFIO_IOMMU_SET_MSI_BINDING
  vfio/pci: Add VFIO_REGION_TYPE_NESTED region type
  vfio/pci: Register an iommu fault handler
  vfio/pci: Allow to mmap the fault queue
  vfio/pci: Add framework for custom interrupt indices
  vfio: Add new IRQ for DMA fault reporting
  vfio/pci: Register and allow DMA FAULT IRQ signaling
  vfio: Document nested stage control
  vfio/pci: Register a DMA fault response region
  vfio/pci: Inject page response upon response region fill

Liu, Yi L (2):
  vfio: VFIO_IOMMU_SET_PASID_TABLE
  vfio: VFIO_IOMMU_CACHE_INVALIDATE

Tina Zhang (1):
  vfio: Use capability chains to handle device specific irq

 Documentation/driver-api/vfio.rst   |  77 +++++
 drivers/vfio/pci/vfio_pci.c         | 447 ++++++++++++++++++++++++++--
 drivers/vfio/pci/vfio_pci_intrs.c   |  62 ++++
 drivers/vfio/pci/vfio_pci_private.h |  33 ++
 drivers/vfio/pci/vfio_pci_rdwr.c    |  84 ++++++
 drivers/vfio/vfio_iommu_type1.c     | 178 +++++++++++
 include/uapi/linux/vfio.h           | 142 ++++++++-
 7 files changed, 1003 insertions(+), 20 deletions(-)

-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 01/13] vfio: VFIO_IOMMU_SET_PASID_TABLE

[Eric Auger]

From: "Liu, Yi L" <yi.l.liu@linux.intel.com>

This patch adds an VFIO_IOMMU_SET_PASID_TABLE ioctl
which aims to pass the virtual iommu guest configuration
to the host. This latter takes the form of the so-called
PASID table.

Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Signed-off-by: Liu, Yi L <yi.l.liu@linux.intel.com>
Signed-off-by: Eric Auger <eric.auger@redhat.com>

---
v12 -> v13:
- reword the misleading doc comment of  VFIO_IOMMU_SET_PASID_TABLE

v11 -> v12:
- use iommu_uapi_set_pasid_table
- Rework the flags checks [Zenghui, Alex]
- use VFIO_BASE + 19 [Alex]
- rework the unwind in vfio_attach_pasid_table() [Alex]

v8 -> v9:
- Merge VFIO_IOMMU_ATTACH/DETACH_PASID_TABLE into a single
  VFIO_IOMMU_SET_PASID_TABLE ioctl.

v6 -> v7:
- add a comment related to VFIO_IOMMU_DETACH_PASID_TABLE

v3 -> v4:
- restore ATTACH/DETACH
- add unwind on failure

v2 -> v3:
- s/BIND_PASID_TABLE/SET_PASID_TABLE

v1 -> v2:
- s/BIND_GUEST_STAGE/BIND_PASID_TABLE
- remove the struct device arg
---
 drivers/vfio/vfio_iommu_type1.c | 58 +++++++++++++++++++++++++++++++++
 include/uapi/linux/vfio.h       | 20 ++++++++++++
 2 files changed, 78 insertions(+)

diff --git a/drivers/vfio/vfio_iommu_type1.c b/drivers/vfio/vfio_iommu_type1.c
index 45cbfd4879a5..e575ef5dd6c6 100644
--- a/drivers/vfio/vfio_iommu_type1.c
+++ b/drivers/vfio/vfio_iommu_type1.c
@@ -2834,6 +2834,39 @@ static int vfio_iommu_iova_build_caps(struct vfio_iommu *iommu,
 	return ret;
 }
 
+static void
+vfio_detach_pasid_table(struct vfio_iommu *iommu)
+{
+	struct vfio_domain *d;
+
+	mutex_lock(&iommu->lock);
+	list_for_each_entry(d, &iommu->domain_list, next)
+		iommu_detach_pasid_table(d->domain);
+
+	mutex_unlock(&iommu->lock);
+}
+
+static int
+vfio_attach_pasid_table(struct vfio_iommu *iommu, unsigned long arg)
+{
+	struct vfio_domain *d;
+	int ret = 0;
+
+	mutex_lock(&iommu->lock);
+
+	list_for_each_entry(d, &iommu->domain_list, next) {
+		ret = iommu_uapi_attach_pasid_table(d->domain, (void __user *)arg);
+		if (ret) {
+			list_for_each_entry_continue_reverse(d, &iommu->domain_list, next)
+				iommu_detach_pasid_table(d->domain);
+			break;
+		}
+	}
+
+	mutex_unlock(&iommu->lock);
+	return ret;
+}
+
 static int vfio_iommu_migration_build_caps(struct vfio_iommu *iommu,
 					   struct vfio_info_cap *caps)
 {
@@ -3002,6 +3035,29 @@ static int vfio_iommu_type1_unmap_dma(struct vfio_iommu *iommu,
 			-EFAULT : 0;
 }
 
+static int vfio_iommu_type1_set_pasid_table(struct vfio_iommu *iommu,
+					    unsigned long arg)
+{
+	struct vfio_iommu_type1_set_pasid_table spt;
+	unsigned long minsz;
+
+	minsz = offsetofend(struct vfio_iommu_type1_set_pasid_table, flags);
+
+	if (copy_from_user(&spt, (void __user *)arg, minsz))
+		return -EFAULT;
+
+	if (spt.argsz < minsz)
+		return -EINVAL;
+
+	if (spt.flags == VFIO_PASID_TABLE_FLAG_SET) {
+		return vfio_attach_pasid_table(iommu, arg + minsz);
+	} else if (spt.flags == VFIO_PASID_TABLE_FLAG_UNSET) {
+		vfio_detach_pasid_table(iommu);
+		return 0;
+	}
+	return -EINVAL;
+}
+
 static int vfio_iommu_type1_dirty_pages(struct vfio_iommu *iommu,
 					unsigned long arg)
 {
@@ -3122,6 +3178,8 @@ static long vfio_iommu_type1_ioctl(void *iommu_data,
 		return vfio_iommu_type1_unmap_dma(iommu, arg);
 	case VFIO_IOMMU_DIRTY_PAGES:
 		return vfio_iommu_type1_dirty_pages(iommu, arg);
+	case VFIO_IOMMU_SET_PASID_TABLE:
+		return vfio_iommu_type1_set_pasid_table(iommu, arg);
 	default:
 		return -ENOTTY;
 	}
diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h
index 8ce36c1d53ca..c86b7b4a5064 100644
--- a/include/uapi/linux/vfio.h
+++ b/include/uapi/linux/vfio.h
@@ -14,6 +14,7 @@
 
 #include <linux/types.h>
 #include <linux/ioctl.h>
+#include <linux/iommu.h>
 
 #define VFIO_API_VERSION	0
 
@@ -1208,6 +1209,25 @@ struct vfio_iommu_type1_dirty_bitmap_get {
 
 #define VFIO_IOMMU_DIRTY_PAGES             _IO(VFIO_TYPE, VFIO_BASE + 17)
 
+/*
+ * VFIO_IOMMU_SET_PASID_TABLE - _IOWR(VFIO_TYPE, VFIO_BASE + 18,
+ *			struct vfio_iommu_type1_set_pasid_table)
+ *
+ * The SET operation passes a PASID table to the host while the
+ * UNSET operation detaches the one currently programmed. It is
+ * allowed to "SET" the table several times without unsetting as
+ * long as the table config does not stay IOMMU_PASID_CONFIG_TRANSLATE.
+ */
+struct vfio_iommu_type1_set_pasid_table {
+	__u32	argsz;
+	__u32	flags;
+#define VFIO_PASID_TABLE_FLAG_SET	(1 << 0)
+#define VFIO_PASID_TABLE_FLAG_UNSET	(1 << 1)
+	struct iommu_pasid_table_config config; /* used on SET */
+};
+
+#define VFIO_IOMMU_SET_PASID_TABLE	_IO(VFIO_TYPE, VFIO_BASE + 18)
+
 /* -------- Additional API for SPAPR TCE (Server POWERPC) IOMMU -------- */
 
 /*
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

Re: [PATCH v13 01/13] vfio: VFIO_IOMMU_SET_PASID_TABLE

[kernel test robot]

[-- Attachment #1: Type: text/plain, Size: 2997 bytes --]

Hi Eric,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on v5.12-rc6]
[also build test ERROR on next-20210409]
[cannot apply to vfio/next]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Eric-Auger/SMMUv3-Nested-Stage-Setup-VFIO-part/20210411-195216
base:    e49d033bddf5b565044e2abe4241353959bc9120
config: arm-randconfig-r003-20210411 (attached as .config)
compiler: clang version 13.0.0 (https://github.com/llvm/llvm-project 9829f5e6b1bca9b61efc629770d28bb9014dec45)
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # install arm cross compiling tool for clang build
        # apt-get install binutils-arm-linux-gnueabi
        # https://github.com/0day-ci/linux/commit/0af9db19db29eb2a707b9e1ca4ff9e1a08a1c511
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Eric-Auger/SMMUv3-Nested-Stage-Setup-VFIO-part/20210411-195216
        git checkout 0af9db19db29eb2a707b9e1ca4ff9e1a08a1c511
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=arm 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

   In file included from drivers/vfio/vfio.c:32:
   In file included from include/linux/vfio.h:16:
>> include/uapi/linux/vfio.h:1226:34: error: field has incomplete type 'struct iommu_pasid_table_config'
           struct iommu_pasid_table_config config; /* used on SET */
                                           ^
   include/uapi/linux/vfio.h:1226:9: note: forward declaration of 'struct iommu_pasid_table_config'
           struct iommu_pasid_table_config config; /* used on SET */
                  ^
   1 error generated.


vim +1226 include/uapi/linux/vfio.h

  1211	
  1212	/*
  1213	 * VFIO_IOMMU_SET_PASID_TABLE - _IOWR(VFIO_TYPE, VFIO_BASE + 18,
  1214	 *			struct vfio_iommu_type1_set_pasid_table)
  1215	 *
  1216	 * The SET operation passes a PASID table to the host while the
  1217	 * UNSET operation detaches the one currently programmed. It is
  1218	 * allowed to "SET" the table several times without unsetting as
  1219	 * long as the table config does not stay IOMMU_PASID_CONFIG_TRANSLATE.
  1220	 */
  1221	struct vfio_iommu_type1_set_pasid_table {
  1222		__u32	argsz;
  1223		__u32	flags;
  1224	#define VFIO_PASID_TABLE_FLAG_SET	(1 << 0)
  1225	#define VFIO_PASID_TABLE_FLAG_UNSET	(1 << 1)
> 1226		struct iommu_pasid_table_config config; /* used on SET */
  1227	};
  1228	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 21215 bytes --]

[-- Attachment #3: Type: text/plain, Size: 151 bytes --]

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

Re: [PATCH v13 01/13] vfio: VFIO_IOMMU_SET_PASID_TABLE

[kernel test robot]

[-- Attachment #1: Type: text/plain, Size: 3714 bytes --]

Hi Eric,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on v5.12-rc6]
[also build test ERROR on next-20210409]
[cannot apply to vfio/next]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Eric-Auger/SMMUv3-Nested-Stage-Setup-VFIO-part/20210411-195216
base:    e49d033bddf5b565044e2abe4241353959bc9120
config: i386-randconfig-s001-20210411 (attached as .config)
compiler: gcc-9 (Debian 9.3.0-22) 9.3.0
reproduce:
        # apt-get install sparse
        # sparse version: v0.6.3-279-g6d5d9b42-dirty
        # https://github.com/0day-ci/linux/commit/0af9db19db29eb2a707b9e1ca4ff9e1a08a1c511
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Eric-Auger/SMMUv3-Nested-Stage-Setup-VFIO-part/20210411-195216
        git checkout 0af9db19db29eb2a707b9e1ca4ff9e1a08a1c511
        # save the attached .config to linux build tree
        make W=1 C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' ARCH=i386 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

   In file included from include/linux/vfio.h:16,
                    from drivers/vfio/vfio.c:32:
>> include/uapi/linux/vfio.h:1226:34: error: field 'config' has incomplete type
    1226 |  struct iommu_pasid_table_config config; /* used on SET */
         |                                  ^~~~~~
--
   In file included from include/linux/vfio.h:16,
                    from drivers/vfio/vfio_iommu_type1.c:37:
>> include/uapi/linux/vfio.h:1226:34: error: field 'config' has incomplete type
    1226 |  struct iommu_pasid_table_config config; /* used on SET */
         |                                  ^~~~~~
   drivers/vfio/vfio_iommu_type1.c: In function 'vfio_detach_pasid_table':
>> drivers/vfio/vfio_iommu_type1.c:2844:3: error: implicit declaration of function 'iommu_detach_pasid_table'; did you mean 'vfio_detach_pasid_table'? [-Werror=implicit-function-declaration]
    2844 |   iommu_detach_pasid_table(d->domain);
         |   ^~~~~~~~~~~~~~~~~~~~~~~~
         |   vfio_detach_pasid_table
   drivers/vfio/vfio_iommu_type1.c: In function 'vfio_attach_pasid_table':
>> drivers/vfio/vfio_iommu_type1.c:2858:9: error: implicit declaration of function 'iommu_uapi_attach_pasid_table'; did you mean 'vfio_attach_pasid_table'? [-Werror=implicit-function-declaration]
    2858 |   ret = iommu_uapi_attach_pasid_table(d->domain, (void __user *)arg);
         |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
         |         vfio_attach_pasid_table
   cc1: some warnings being treated as errors


vim +/config +1226 include/uapi/linux/vfio.h

  1211	
  1212	/*
  1213	 * VFIO_IOMMU_SET_PASID_TABLE - _IOWR(VFIO_TYPE, VFIO_BASE + 18,
  1214	 *			struct vfio_iommu_type1_set_pasid_table)
  1215	 *
  1216	 * The SET operation passes a PASID table to the host while the
  1217	 * UNSET operation detaches the one currently programmed. It is
  1218	 * allowed to "SET" the table several times without unsetting as
  1219	 * long as the table config does not stay IOMMU_PASID_CONFIG_TRANSLATE.
  1220	 */
  1221	struct vfio_iommu_type1_set_pasid_table {
  1222		__u32	argsz;
  1223		__u32	flags;
  1224	#define VFIO_PASID_TABLE_FLAG_SET	(1 << 0)
  1225	#define VFIO_PASID_TABLE_FLAG_UNSET	(1 << 1)
> 1226		struct iommu_pasid_table_config config; /* used on SET */
  1227	};
  1228	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 27096 bytes --]

[-- Attachment #3: Type: text/plain, Size: 151 bytes --]

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 02/13] vfio: VFIO_IOMMU_CACHE_INVALIDATE

[Eric Auger]

From: "Liu, Yi L" <yi.l.liu@linux.intel.com>

When the guest "owns" the stage 1 translation structures,  the host
IOMMU driver has no knowledge of caching structure updates unless
the guest invalidation requests are trapped and passed down to the
host.

This patch adds the VFIO_IOMMU_CACHE_INVALIDATE ioctl with aims
at propagating guest stage1 IOMMU cache invalidations to the host.

Signed-off-by: Liu, Yi L <yi.l.liu@linux.intel.com>
Signed-off-by: Eric Auger <eric.auger@redhat.com>

---

v12 -> v13:
- resolve minor conflict due to
  898b9eaeb3fe ("vfio/type1: block on invalid vaddr")
v11 -> v12:
- share VFIO_BASE + 19 with VFIO_IOMMU_SPAPR_TCE_CREATE

v10 -> v11:
- renamed ustruct into cache_inv

v8 -> v9:
- change the ioctl ID

v6 -> v7:
- Use iommu_capsule struct
- renamed vfio_iommu_for_each_dev into vfio_iommu_lookup_dev
  due to checkpatch error related to for_each_dev suffix

v2 -> v3:
- introduce vfio_iommu_for_each_dev back in this patch

v1 -> v2:
- s/TLB/CACHE
- remove vfio_iommu_task usage
- commit message rewording
---
 drivers/vfio/vfio_iommu_type1.c | 58 +++++++++++++++++++++++++++++++++
 include/uapi/linux/vfio.h       | 13 ++++++++
 2 files changed, 71 insertions(+)

diff --git a/drivers/vfio/vfio_iommu_type1.c b/drivers/vfio/vfio_iommu_type1.c
index e575ef5dd6c6..34f8dca36ebe 100644
--- a/drivers/vfio/vfio_iommu_type1.c
+++ b/drivers/vfio/vfio_iommu_type1.c
@@ -159,6 +159,34 @@ struct vfio_regions {
 
 #define WAITED 1
 
+struct domain_capsule {
+	struct iommu_domain *domain;
+	void *data;
+};
+
+/* iommu->lock must be held */
+static int
+vfio_iommu_lookup_dev(struct vfio_iommu *iommu,
+		      int (*fn)(struct device *dev, void *data),
+		      unsigned long arg)
+{
+	struct domain_capsule dc = {.data = &arg};
+	struct vfio_domain *d;
+	struct vfio_group *g;
+	int ret = 0;
+
+	list_for_each_entry(d, &iommu->domain_list, next) {
+		dc.domain = d->domain;
+		list_for_each_entry(g, &d->group_list, next) {
+			ret = iommu_group_for_each_dev(g->iommu_group,
+						       &dc, fn);
+			if (ret)
+				break;
+		}
+	}
+	return ret;
+}
+
 static int put_pfn(unsigned long pfn, int prot);
 
 static struct vfio_group *vfio_iommu_find_iommu_group(struct vfio_iommu *iommu,
@@ -2866,6 +2894,13 @@ vfio_attach_pasid_table(struct vfio_iommu *iommu, unsigned long arg)
 	mutex_unlock(&iommu->lock);
 	return ret;
 }
+static int vfio_cache_inv_fn(struct device *dev, void *data)
+{
+	struct domain_capsule *dc = (struct domain_capsule *)data;
+	unsigned long arg = *(unsigned long *)dc->data;
+
+	return iommu_uapi_cache_invalidate(dc->domain, dev, (void __user *)arg);
+}
 
 static int vfio_iommu_migration_build_caps(struct vfio_iommu *iommu,
 					   struct vfio_info_cap *caps)
@@ -3058,6 +3093,27 @@ static int vfio_iommu_type1_set_pasid_table(struct vfio_iommu *iommu,
 	return -EINVAL;
 }
 
+static int vfio_iommu_type1_cache_invalidate(struct vfio_iommu *iommu,
+					    unsigned long arg)
+{
+	struct vfio_iommu_type1_cache_invalidate cache_inv;
+	unsigned long minsz;
+	int ret;
+
+	minsz = offsetofend(struct vfio_iommu_type1_cache_invalidate, flags);
+
+	if (copy_from_user(&cache_inv, (void __user *)arg, minsz))
+		return -EFAULT;
+
+	if (cache_inv.argsz < minsz || cache_inv.flags)
+		return -EINVAL;
+
+	mutex_lock(&iommu->lock);
+	ret = vfio_iommu_lookup_dev(iommu, vfio_cache_inv_fn, arg + minsz);
+	mutex_unlock(&iommu->lock);
+	return ret;
+}
+
 static int vfio_iommu_type1_dirty_pages(struct vfio_iommu *iommu,
 					unsigned long arg)
 {
@@ -3180,6 +3236,8 @@ static long vfio_iommu_type1_ioctl(void *iommu_data,
 		return vfio_iommu_type1_dirty_pages(iommu, arg);
 	case VFIO_IOMMU_SET_PASID_TABLE:
 		return vfio_iommu_type1_set_pasid_table(iommu, arg);
+	case VFIO_IOMMU_CACHE_INVALIDATE:
+		return vfio_iommu_type1_cache_invalidate(iommu, arg);
 	default:
 		return -ENOTTY;
 	}
diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h
index c86b7b4a5064..ada42121827a 100644
--- a/include/uapi/linux/vfio.h
+++ b/include/uapi/linux/vfio.h
@@ -1228,6 +1228,19 @@ struct vfio_iommu_type1_set_pasid_table {
 
 #define VFIO_IOMMU_SET_PASID_TABLE	_IO(VFIO_TYPE, VFIO_BASE + 18)
 
+/**
+ * VFIO_IOMMU_CACHE_INVALIDATE - _IOWR(VFIO_TYPE, VFIO_BASE + 19,
+ *			struct vfio_iommu_type1_cache_invalidate)
+ *
+ * Propagate guest IOMMU cache invalidation to the host.
+ */
+struct vfio_iommu_type1_cache_invalidate {
+	__u32   argsz;
+	__u32   flags;
+	struct iommu_cache_invalidate_info info;
+};
+#define VFIO_IOMMU_CACHE_INVALIDATE      _IO(VFIO_TYPE, VFIO_BASE + 19)
+
 /* -------- Additional API for SPAPR TCE (Server POWERPC) IOMMU -------- */
 
 /*
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 03/13] vfio: VFIO_IOMMU_SET_MSI_BINDING

[Eric Auger]

This patch adds the VFIO_IOMMU_SET_MSI_BINDING ioctl which aim
to (un)register the guest MSI binding to the host. This latter
then can use those stage 1 bindings to build a nested stage
binding targeting the physical MSIs.

Signed-off-by: Eric Auger <eric.auger@redhat.com>

---

v11 -> v12:
- Share VFIO_BASE + 20 with VFIO_IOMMU_SPAPR_TCE_REMOVE
- rework returned values

v10 -> v11:
- renamed ustruct into msi_binding
- return 0 on unbind

v8 -> v9:
- merge VFIO_IOMMU_BIND_MSI/VFIO_IOMMU_UNBIND_MSI into a single
  VFIO_IOMMU_SET_MSI_BINDING ioctl
- ioctl id changed

v6 -> v7:
- removed the dev arg

v3 -> v4:
- add UNBIND
- unwind on BIND error

v2 -> v3:
- adapt to new proto of bind_guest_msi
- directly use vfio_iommu_for_each_dev

v1 -> v2:
- s/vfio_iommu_type1_guest_msi_binding/vfio_iommu_type1_bind_guest_msi
---
 drivers/vfio/vfio_iommu_type1.c | 62 +++++++++++++++++++++++++++++++++
 include/uapi/linux/vfio.h       | 20 +++++++++++
 2 files changed, 82 insertions(+)

diff --git a/drivers/vfio/vfio_iommu_type1.c b/drivers/vfio/vfio_iommu_type1.c
index 34f8dca36ebe..5e9196ec9685 100644
--- a/drivers/vfio/vfio_iommu_type1.c
+++ b/drivers/vfio/vfio_iommu_type1.c
@@ -2902,6 +2902,41 @@ static int vfio_cache_inv_fn(struct device *dev, void *data)
 	return iommu_uapi_cache_invalidate(dc->domain, dev, (void __user *)arg);
 }
 
+static int
+vfio_bind_msi(struct vfio_iommu *iommu,
+	      dma_addr_t giova, phys_addr_t gpa, size_t size)
+{
+	struct vfio_domain *d;
+	int ret = 0;
+
+	mutex_lock(&iommu->lock);
+
+	list_for_each_entry(d, &iommu->domain_list, next) {
+		ret = iommu_bind_guest_msi(d->domain, giova, gpa, size);
+		if (ret)
+			goto unwind;
+	}
+	goto unlock;
+unwind:
+	list_for_each_entry_continue_reverse(d, &iommu->domain_list, next) {
+		iommu_unbind_guest_msi(d->domain, giova);
+	}
+unlock:
+	mutex_unlock(&iommu->lock);
+	return ret;
+}
+
+static void
+vfio_unbind_msi(struct vfio_iommu *iommu, dma_addr_t giova)
+{
+	struct vfio_domain *d;
+
+	mutex_lock(&iommu->lock);
+	list_for_each_entry(d, &iommu->domain_list, next)
+		iommu_unbind_guest_msi(d->domain, giova);
+	mutex_unlock(&iommu->lock);
+}
+
 static int vfio_iommu_migration_build_caps(struct vfio_iommu *iommu,
 					   struct vfio_info_cap *caps)
 {
@@ -3114,6 +3149,31 @@ static int vfio_iommu_type1_cache_invalidate(struct vfio_iommu *iommu,
 	return ret;
 }
 
+static int vfio_iommu_type1_set_msi_binding(struct vfio_iommu *iommu,
+					    unsigned long arg)
+{
+	struct vfio_iommu_type1_set_msi_binding msi_binding;
+	unsigned long minsz;
+
+	minsz = offsetofend(struct vfio_iommu_type1_set_msi_binding,
+			    size);
+
+	if (copy_from_user(&msi_binding, (void __user *)arg, minsz))
+		return -EFAULT;
+
+	if (msi_binding.argsz < minsz)
+		return -EINVAL;
+
+	if (msi_binding.flags == VFIO_IOMMU_UNBIND_MSI) {
+		vfio_unbind_msi(iommu, msi_binding.iova);
+		return 0;
+	} else if (msi_binding.flags == VFIO_IOMMU_BIND_MSI) {
+		return vfio_bind_msi(iommu, msi_binding.iova,
+				     msi_binding.gpa, msi_binding.size);
+	}
+	return -EINVAL;
+}
+
 static int vfio_iommu_type1_dirty_pages(struct vfio_iommu *iommu,
 					unsigned long arg)
 {
@@ -3238,6 +3298,8 @@ static long vfio_iommu_type1_ioctl(void *iommu_data,
 		return vfio_iommu_type1_set_pasid_table(iommu, arg);
 	case VFIO_IOMMU_CACHE_INVALIDATE:
 		return vfio_iommu_type1_cache_invalidate(iommu, arg);
+	case VFIO_IOMMU_SET_MSI_BINDING:
+		return vfio_iommu_type1_set_msi_binding(iommu, arg);
 	default:
 		return -ENOTTY;
 	}
diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h
index ada42121827a..afa8596c05d2 100644
--- a/include/uapi/linux/vfio.h
+++ b/include/uapi/linux/vfio.h
@@ -1241,6 +1241,26 @@ struct vfio_iommu_type1_cache_invalidate {
 };
 #define VFIO_IOMMU_CACHE_INVALIDATE      _IO(VFIO_TYPE, VFIO_BASE + 19)
 
+/**
+ * VFIO_IOMMU_SET_MSI_BINDING - _IOWR(VFIO_TYPE, VFIO_BASE + 20,
+ *			struct vfio_iommu_type1_set_msi_binding)
+ *
+ * Pass a stage 1 MSI doorbell mapping to the host so that this
+ * latter can build a nested stage2 mapping. Or conversely tear
+ * down a previously bound stage 1 MSI binding.
+ */
+struct vfio_iommu_type1_set_msi_binding {
+	__u32   argsz;
+	__u32   flags;
+#define VFIO_IOMMU_BIND_MSI	(1 << 0)
+#define VFIO_IOMMU_UNBIND_MSI	(1 << 1)
+	__u64	iova;	/* MSI guest IOVA */
+	/* Fields below are used on BIND */
+	__u64	gpa;	/* MSI guest physical address */
+	__u64	size;	/* size of stage1 mapping (bytes) */
+};
+#define VFIO_IOMMU_SET_MSI_BINDING      _IO(VFIO_TYPE, VFIO_BASE + 20)
+
 /* -------- Additional API for SPAPR TCE (Server POWERPC) IOMMU -------- */
 
 /*
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

Re: [PATCH v13 03/13] vfio: VFIO_IOMMU_SET_MSI_BINDING

[kernel test robot]

[-- Attachment #1: Type: text/plain, Size: 3936 bytes --]

Hi Eric,

I love your patch! Yet something to improve:

[auto build test ERROR on v5.12-rc6]
[also build test ERROR on next-20210409]
[cannot apply to vfio/next]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Eric-Auger/SMMUv3-Nested-Stage-Setup-VFIO-part/20210411-195216
base:    e49d033bddf5b565044e2abe4241353959bc9120
config: i386-randconfig-s001-20210411 (attached as .config)
compiler: gcc-9 (Debian 9.3.0-22) 9.3.0
reproduce:
        # apt-get install sparse
        # sparse version: v0.6.3-279-g6d5d9b42-dirty
        # https://github.com/0day-ci/linux/commit/8a9991f0409cae6af9a387b8b90dbb7ad004f590
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Eric-Auger/SMMUv3-Nested-Stage-Setup-VFIO-part/20210411-195216
        git checkout 8a9991f0409cae6af9a387b8b90dbb7ad004f590
        # save the attached .config to linux build tree
        make W=1 C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' ARCH=i386 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

   In file included from include/linux/vfio.h:16,
                    from drivers/vfio/vfio_iommu_type1.c:37:
   include/uapi/linux/vfio.h:1226:34: error: field 'config' has incomplete type
    1226 |  struct iommu_pasid_table_config config; /* used on SET */
         |                                  ^~~~~~
   drivers/vfio/vfio_iommu_type1.c: In function 'vfio_detach_pasid_table':
   drivers/vfio/vfio_iommu_type1.c:2872:3: error: implicit declaration of function 'iommu_detach_pasid_table'; did you mean 'vfio_detach_pasid_table'? [-Werror=implicit-function-declaration]
    2872 |   iommu_detach_pasid_table(d->domain);
         |   ^~~~~~~~~~~~~~~~~~~~~~~~
         |   vfio_detach_pasid_table
   drivers/vfio/vfio_iommu_type1.c: In function 'vfio_attach_pasid_table':
   drivers/vfio/vfio_iommu_type1.c:2886:9: error: implicit declaration of function 'iommu_uapi_attach_pasid_table'; did you mean 'vfio_attach_pasid_table'? [-Werror=implicit-function-declaration]
    2886 |   ret = iommu_uapi_attach_pasid_table(d->domain, (void __user *)arg);
         |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
         |         vfio_attach_pasid_table
   drivers/vfio/vfio_iommu_type1.c: In function 'vfio_bind_msi':
>> drivers/vfio/vfio_iommu_type1.c:2915:9: error: implicit declaration of function 'iommu_bind_guest_msi' [-Werror=implicit-function-declaration]
    2915 |   ret = iommu_bind_guest_msi(d->domain, giova, gpa, size);
         |         ^~~~~~~~~~~~~~~~~~~~
>> drivers/vfio/vfio_iommu_type1.c:2922:3: error: implicit declaration of function 'iommu_unbind_guest_msi' [-Werror=implicit-function-declaration]
    2922 |   iommu_unbind_guest_msi(d->domain, giova);
         |   ^~~~~~~~~~~~~~~~~~~~~~
   cc1: some warnings being treated as errors


vim +/iommu_bind_guest_msi +2915 drivers/vfio/vfio_iommu_type1.c

  2904	
  2905	static int
  2906	vfio_bind_msi(struct vfio_iommu *iommu,
  2907		      dma_addr_t giova, phys_addr_t gpa, size_t size)
  2908	{
  2909		struct vfio_domain *d;
  2910		int ret = 0;
  2911	
  2912		mutex_lock(&iommu->lock);
  2913	
  2914		list_for_each_entry(d, &iommu->domain_list, next) {
> 2915			ret = iommu_bind_guest_msi(d->domain, giova, gpa, size);
  2916			if (ret)
  2917				goto unwind;
  2918		}
  2919		goto unlock;
  2920	unwind:
  2921		list_for_each_entry_continue_reverse(d, &iommu->domain_list, next) {
> 2922			iommu_unbind_guest_msi(d->domain, giova);
  2923		}
  2924	unlock:
  2925		mutex_unlock(&iommu->lock);
  2926		return ret;
  2927	}
  2928	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 27096 bytes --]

[-- Attachment #3: Type: text/plain, Size: 151 bytes --]

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 04/13] vfio/pci: Add VFIO_REGION_TYPE_NESTED region type

[Eric Auger]

Add a new specific DMA_FAULT region aiming to exposed nested mode
translation faults. This region only is exposed if the device
is attached to a nested domain.

The region has a ring buffer that contains the actual fault
records plus a header allowing to handle it (tail/head indices,
max capacity, entry size). At the moment the region is dimensionned
for 512 fault records.

Signed-off-by: Eric Auger <eric.auger@redhat.com>

---
v11 -> v12:
- set fault_pages to NULL after free
- check new_tail >= header->nb_entries (Zenghui)
- update value of VFIO_REGION_TYPE_NESTED
- handle the case where the domain is NULL
- pass an int pointer to iommu_domain_get_attr [Shenming]

v10 -> v11:
- rename vfio_pci_init_dma_fault_region into
  vfio_pci_dma_fault_init
- free fault_pages in vfio_pci_dma_fault_release
- only register the region if the device is attached
  to a nested domain

v8 -> v9:
- Use a single region instead of a prod/cons region

v4 -> v5
- check cons is not null in vfio_pci_check_cons_fault

v3 -> v4:
- use 2 separate regions, respectively in read and write modes
- add the version capability
---
 drivers/vfio/pci/vfio_pci.c         | 79 +++++++++++++++++++++++++++++
 drivers/vfio/pci/vfio_pci_private.h |  6 +++
 drivers/vfio/pci/vfio_pci_rdwr.c    | 44 ++++++++++++++++
 include/uapi/linux/vfio.h           | 35 +++++++++++++
 4 files changed, 164 insertions(+)

diff --git a/drivers/vfio/pci/vfio_pci.c b/drivers/vfio/pci/vfio_pci.c
index 65e7e6b44578..f37cabddf1d6 100644
--- a/drivers/vfio/pci/vfio_pci.c
+++ b/drivers/vfio/pci/vfio_pci.c
@@ -309,6 +309,81 @@ int vfio_pci_set_power_state(struct vfio_pci_device *vdev, pci_power_t state)
 	return ret;
 }
 
+static void vfio_pci_dma_fault_release(struct vfio_pci_device *vdev,
+				       struct vfio_pci_region *region)
+{
+	kfree(vdev->fault_pages);
+}
+
+static int vfio_pci_dma_fault_add_capability(struct vfio_pci_device *vdev,
+					     struct vfio_pci_region *region,
+					     struct vfio_info_cap *caps)
+{
+	struct vfio_region_info_cap_fault cap = {
+		.header.id = VFIO_REGION_INFO_CAP_DMA_FAULT,
+		.header.version = 1,
+		.version = 1,
+	};
+	return vfio_info_add_capability(caps, &cap.header, sizeof(cap));
+}
+
+static const struct vfio_pci_regops vfio_pci_dma_fault_regops = {
+	.rw		= vfio_pci_dma_fault_rw,
+	.release	= vfio_pci_dma_fault_release,
+	.add_capability = vfio_pci_dma_fault_add_capability,
+};
+
+#define DMA_FAULT_RING_LENGTH 512
+
+static int vfio_pci_dma_fault_init(struct vfio_pci_device *vdev)
+{
+	struct vfio_region_dma_fault *header;
+	struct iommu_domain *domain;
+	size_t size;
+	int nested;
+	int ret;
+
+	domain = iommu_get_domain_for_dev(&vdev->pdev->dev);
+	if (!domain)
+		return 0;
+
+	ret = iommu_domain_get_attr(domain, DOMAIN_ATTR_NESTING, &nested);
+	if (ret || !nested)
+		return ret;
+
+	mutex_init(&vdev->fault_queue_lock);
+
+	/*
+	 * We provision 1 page for the header and space for
+	 * DMA_FAULT_RING_LENGTH fault records in the ring buffer.
+	 */
+	size = ALIGN(sizeof(struct iommu_fault) *
+		     DMA_FAULT_RING_LENGTH, PAGE_SIZE) + PAGE_SIZE;
+
+	vdev->fault_pages = kzalloc(size, GFP_KERNEL);
+	if (!vdev->fault_pages)
+		return -ENOMEM;
+
+	ret = vfio_pci_register_dev_region(vdev,
+		VFIO_REGION_TYPE_NESTED,
+		VFIO_REGION_SUBTYPE_NESTED_DMA_FAULT,
+		&vfio_pci_dma_fault_regops, size,
+		VFIO_REGION_INFO_FLAG_READ | VFIO_REGION_INFO_FLAG_WRITE,
+		vdev->fault_pages);
+	if (ret)
+		goto out;
+
+	header = (struct vfio_region_dma_fault *)vdev->fault_pages;
+	header->entry_size = sizeof(struct iommu_fault);
+	header->nb_entries = DMA_FAULT_RING_LENGTH;
+	header->offset = sizeof(struct vfio_region_dma_fault);
+	return 0;
+out:
+	kfree(vdev->fault_pages);
+	vdev->fault_pages = NULL;
+	return ret;
+}
+
 static int vfio_pci_enable(struct vfio_pci_device *vdev)
 {
 	struct pci_dev *pdev = vdev->pdev;
@@ -407,6 +482,10 @@ static int vfio_pci_enable(struct vfio_pci_device *vdev)
 		}
 	}
 
+	ret = vfio_pci_dma_fault_init(vdev);
+	if (ret)
+		goto disable_exit;
+
 	vfio_pci_probe_mmaps(vdev);
 
 	return 0;
diff --git a/drivers/vfio/pci/vfio_pci_private.h b/drivers/vfio/pci/vfio_pci_private.h
index 9cd1882a05af..76cb3bd11375 100644
--- a/drivers/vfio/pci/vfio_pci_private.h
+++ b/drivers/vfio/pci/vfio_pci_private.h
@@ -134,6 +134,8 @@ struct vfio_pci_device {
 	int			ioeventfds_nr;
 	struct eventfd_ctx	*err_trigger;
 	struct eventfd_ctx	*req_trigger;
+	u8			*fault_pages;
+	struct mutex		fault_queue_lock;
 	struct list_head	dummy_resources_list;
 	struct mutex		ioeventfds_lock;
 	struct list_head	ioeventfds_list;
@@ -170,6 +172,10 @@ extern ssize_t vfio_pci_vga_rw(struct vfio_pci_device *vdev, char __user *buf,
 extern long vfio_pci_ioeventfd(struct vfio_pci_device *vdev, loff_t offset,
 			       uint64_t data, int count, int fd);
 
+extern size_t vfio_pci_dma_fault_rw(struct vfio_pci_device *vdev,
+				    char __user *buf, size_t count,
+				    loff_t *ppos, bool iswrite);
+
 extern int vfio_pci_init_perm_bits(void);
 extern void vfio_pci_uninit_perm_bits(void);
 
diff --git a/drivers/vfio/pci/vfio_pci_rdwr.c b/drivers/vfio/pci/vfio_pci_rdwr.c
index a0b5fc8e46f4..164120607469 100644
--- a/drivers/vfio/pci/vfio_pci_rdwr.c
+++ b/drivers/vfio/pci/vfio_pci_rdwr.c
@@ -356,6 +356,50 @@ ssize_t vfio_pci_vga_rw(struct vfio_pci_device *vdev, char __user *buf,
 	return done;
 }
 
+size_t vfio_pci_dma_fault_rw(struct vfio_pci_device *vdev, char __user *buf,
+			     size_t count, loff_t *ppos, bool iswrite)
+{
+	unsigned int i = VFIO_PCI_OFFSET_TO_INDEX(*ppos) - VFIO_PCI_NUM_REGIONS;
+	loff_t pos = *ppos & VFIO_PCI_OFFSET_MASK;
+	void *base = vdev->region[i].data;
+	int ret = -EFAULT;
+
+	if (pos >= vdev->region[i].size)
+		return -EINVAL;
+
+	count = min(count, (size_t)(vdev->region[i].size - pos));
+
+	mutex_lock(&vdev->fault_queue_lock);
+
+	if (iswrite) {
+		struct vfio_region_dma_fault *header =
+			(struct vfio_region_dma_fault *)base;
+		u32 new_tail;
+
+		if (pos != 0 || count != 4) {
+			ret = -EINVAL;
+			goto unlock;
+		}
+
+		if (copy_from_user((void *)&new_tail, buf, count))
+			goto unlock;
+
+		if (new_tail >= header->nb_entries) {
+			ret = -EINVAL;
+			goto unlock;
+		}
+		header->tail = new_tail;
+	} else {
+		if (copy_to_user(buf, base + pos, count))
+			goto unlock;
+	}
+	*ppos += count;
+	ret = count;
+unlock:
+	mutex_unlock(&vdev->fault_queue_lock);
+	return ret;
+}
+
 static void vfio_pci_ioeventfd_do_write(struct vfio_pci_ioeventfd *ioeventfd,
 					bool test_mem)
 {
diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h
index afa8596c05d2..ca8cc796e254 100644
--- a/include/uapi/linux/vfio.h
+++ b/include/uapi/linux/vfio.h
@@ -325,6 +325,7 @@ struct vfio_region_info_cap_type {
 #define VFIO_REGION_TYPE_GFX                    (1)
 #define VFIO_REGION_TYPE_CCW			(2)
 #define VFIO_REGION_TYPE_MIGRATION              (3)
+#define VFIO_REGION_TYPE_NESTED			(4)
 
 /* sub-types for VFIO_REGION_TYPE_PCI_* */
 
@@ -349,6 +350,9 @@ struct vfio_region_info_cap_type {
 /* sub-types for VFIO_REGION_TYPE_GFX */
 #define VFIO_REGION_SUBTYPE_GFX_EDID            (1)
 
+/* sub-types for VFIO_REGION_TYPE_NESTED */
+#define VFIO_REGION_SUBTYPE_NESTED_DMA_FAULT	(1)
+
 /**
  * struct vfio_region_gfx_edid - EDID region layout.
  *
@@ -995,6 +999,37 @@ struct vfio_device_feature {
  */
 #define VFIO_DEVICE_FEATURE_PCI_VF_TOKEN	(0)
 
+/*
+ * Capability exposed by the DMA fault region
+ * @version: ABI version
+ */
+#define VFIO_REGION_INFO_CAP_DMA_FAULT	6
+
+struct vfio_region_info_cap_fault {
+	struct vfio_info_cap_header header;
+	__u32 version;
+};
+
+/*
+ * DMA Fault Region Layout
+ * @tail: index relative to the start of the ring buffer at which the
+ *        consumer finds the next item in the buffer
+ * @entry_size: fault ring buffer entry size in bytes
+ * @nb_entries: max capacity of the fault ring buffer
+ * @offset: ring buffer offset relative to the start of the region
+ * @head: index relative to the start of the ring buffer at which the
+ *        producer (kernel) inserts items into the buffers
+ */
+struct vfio_region_dma_fault {
+	/* Write-Only */
+	__u32   tail;
+	/* Read-Only */
+	__u32   entry_size;
+	__u32	nb_entries;
+	__u32	offset;
+	__u32   head;
+};
+
 /* -------- API for Type1 VFIO IOMMU -------- */
 
 /**
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 05/13] vfio/pci: Register an iommu fault handler

[Eric Auger]

Register an IOMMU fault handler which records faults in
the DMA FAULT region ring buffer. In a subsequent patch, we
will add the signaling of a specific eventfd to allow the
userspace to be notified whenever a new fault has shown up.

Signed-off-by: Eric Auger <eric.auger@redhat.com>

---
v11 -> v12:
- take the fault_queue_lock before reading header (Zenghui)
- also record recoverable errors
- only WARN_ON if the unregistration returns -EBUSY
- make vfio_pci_iommu_dev_fault_handler static

v10 -> v11:
- move iommu_unregister_device_fault_handler into
  vfio_pci_disable
- check fault_pages != 0

v8 -> v9:
- handler now takes an iommu_fault handle
- eventfd signaling moved to a subsequent patch
- check the fault type and return an error if != UNRECOV
- still the fault handler registration can fail. We need to
  reach an agreement about how to deal with the situation

v3 -> v4:
- move iommu_unregister_device_fault_handler to vfio_pci_release
---
 drivers/vfio/pci/vfio_pci.c | 48 ++++++++++++++++++++++++++++++++++++-
 1 file changed, 47 insertions(+), 1 deletion(-)

diff --git a/drivers/vfio/pci/vfio_pci.c b/drivers/vfio/pci/vfio_pci.c
index f37cabddf1d6..92840e0f46bf 100644
--- a/drivers/vfio/pci/vfio_pci.c
+++ b/drivers/vfio/pci/vfio_pci.c
@@ -27,6 +27,7 @@
 #include <linux/vgaarb.h>
 #include <linux/nospec.h>
 #include <linux/sched/mm.h>
+#include <linux/circ_buf.h>
 
 #include "vfio_pci_private.h"
 
@@ -333,6 +334,41 @@ static const struct vfio_pci_regops vfio_pci_dma_fault_regops = {
 	.add_capability = vfio_pci_dma_fault_add_capability,
 };
 
+static int
+vfio_pci_iommu_dev_fault_handler(struct iommu_fault *fault, void *data)
+{
+	struct vfio_pci_device *vdev = (struct vfio_pci_device *)data;
+	struct vfio_region_dma_fault *reg =
+		(struct vfio_region_dma_fault *)vdev->fault_pages;
+	struct iommu_fault *new;
+	u32 head, tail, size;
+	int ret = -EINVAL;
+
+	if (WARN_ON(!reg))
+		return ret;
+
+	mutex_lock(&vdev->fault_queue_lock);
+
+	head = reg->head;
+	tail = reg->tail;
+	size = reg->nb_entries;
+
+	new = (struct iommu_fault *)(vdev->fault_pages + reg->offset +
+				     head * reg->entry_size);
+
+	if (CIRC_SPACE(head, tail, size) < 1) {
+		ret = -ENOSPC;
+		goto unlock;
+	}
+
+	*new = *fault;
+	reg->head = (head + 1) % size;
+	ret = 0;
+unlock:
+	mutex_unlock(&vdev->fault_queue_lock);
+	return ret;
+}
+
 #define DMA_FAULT_RING_LENGTH 512
 
 static int vfio_pci_dma_fault_init(struct vfio_pci_device *vdev)
@@ -377,6 +413,13 @@ static int vfio_pci_dma_fault_init(struct vfio_pci_device *vdev)
 	header->entry_size = sizeof(struct iommu_fault);
 	header->nb_entries = DMA_FAULT_RING_LENGTH;
 	header->offset = sizeof(struct vfio_region_dma_fault);
+
+	ret = iommu_register_device_fault_handler(&vdev->pdev->dev,
+					vfio_pci_iommu_dev_fault_handler,
+					vdev);
+	if (ret) /* the dma fault region is freed in vfio_pci_disable() */
+		goto out;
+
 	return 0;
 out:
 	kfree(vdev->fault_pages);
@@ -500,7 +543,7 @@ static void vfio_pci_disable(struct vfio_pci_device *vdev)
 	struct pci_dev *pdev = vdev->pdev;
 	struct vfio_pci_dummy_resource *dummy_res, *tmp;
 	struct vfio_pci_ioeventfd *ioeventfd, *ioeventfd_tmp;
-	int i, bar;
+	int i, bar, ret;
 
 	/* Stop the device from further DMA */
 	pci_clear_master(pdev);
@@ -509,6 +552,9 @@ static void vfio_pci_disable(struct vfio_pci_device *vdev)
 				VFIO_IRQ_SET_ACTION_TRIGGER,
 				vdev->irq_type, 0, 0, NULL);
 
+	ret = iommu_unregister_device_fault_handler(&vdev->pdev->dev);
+	WARN_ON(ret == -EBUSY);
+
 	/* Device closed, don't need mutex here */
 	list_for_each_entry_safe(ioeventfd, ioeventfd_tmp,
 				 &vdev->ioeventfds_list, next) {
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 06/13] vfio/pci: Allow to mmap the fault queue

[Eric Auger]

The DMA FAULT region contains the fault ring buffer.
There is benefit to let the userspace mmap this area.
Expose this mmappable area through a sparse mmap entry
and implement the mmap operation.

Signed-off-by: Eric Auger <eric.auger@redhat.com>

---

v8 -> v9:
- remove unused index local variable in vfio_pci_fault_mmap
---
 drivers/vfio/pci/vfio_pci.c | 61 +++++++++++++++++++++++++++++++++++--
 1 file changed, 58 insertions(+), 3 deletions(-)

diff --git a/drivers/vfio/pci/vfio_pci.c b/drivers/vfio/pci/vfio_pci.c
index 92840e0f46bf..eef76560be55 100644
--- a/drivers/vfio/pci/vfio_pci.c
+++ b/drivers/vfio/pci/vfio_pci.c
@@ -316,21 +316,75 @@ static void vfio_pci_dma_fault_release(struct vfio_pci_device *vdev,
 	kfree(vdev->fault_pages);
 }
 
+static int vfio_pci_dma_fault_mmap(struct vfio_pci_device *vdev,
+				   struct vfio_pci_region *region,
+				   struct vm_area_struct *vma)
+{
+	u64 phys_len, req_len, pgoff, req_start;
+	unsigned long long addr;
+	unsigned int ret;
+
+	phys_len = region->size;
+
+	req_len = vma->vm_end - vma->vm_start;
+	pgoff = vma->vm_pgoff &
+		((1U << (VFIO_PCI_OFFSET_SHIFT - PAGE_SHIFT)) - 1);
+	req_start = pgoff << PAGE_SHIFT;
+
+	/* only the second page of the producer fault region is mmappable */
+	if (req_start < PAGE_SIZE)
+		return -EINVAL;
+
+	if (req_start + req_len > phys_len)
+		return -EINVAL;
+
+	addr = virt_to_phys(vdev->fault_pages);
+	vma->vm_private_data = vdev;
+	vma->vm_pgoff = (addr >> PAGE_SHIFT) + pgoff;
+
+	ret = remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff,
+			      req_len, vma->vm_page_prot);
+	return ret;
+}
+
 static int vfio_pci_dma_fault_add_capability(struct vfio_pci_device *vdev,
 					     struct vfio_pci_region *region,
 					     struct vfio_info_cap *caps)
 {
+	struct vfio_region_info_cap_sparse_mmap *sparse = NULL;
 	struct vfio_region_info_cap_fault cap = {
 		.header.id = VFIO_REGION_INFO_CAP_DMA_FAULT,
 		.header.version = 1,
 		.version = 1,
 	};
-	return vfio_info_add_capability(caps, &cap.header, sizeof(cap));
+	size_t size = sizeof(*sparse) + sizeof(*sparse->areas);
+	int ret;
+
+	ret = vfio_info_add_capability(caps, &cap.header, sizeof(cap));
+	if (ret)
+		return ret;
+
+	sparse = kzalloc(size, GFP_KERNEL);
+	if (!sparse)
+		return -ENOMEM;
+
+	sparse->header.id = VFIO_REGION_INFO_CAP_SPARSE_MMAP;
+	sparse->header.version = 1;
+	sparse->nr_areas = 1;
+	sparse->areas[0].offset = PAGE_SIZE;
+	sparse->areas[0].size = region->size - PAGE_SIZE;
+
+	ret = vfio_info_add_capability(caps, &sparse->header, size);
+	if (ret)
+		kfree(sparse);
+
+	return ret;
 }
 
 static const struct vfio_pci_regops vfio_pci_dma_fault_regops = {
 	.rw		= vfio_pci_dma_fault_rw,
 	.release	= vfio_pci_dma_fault_release,
+	.mmap		= vfio_pci_dma_fault_mmap,
 	.add_capability = vfio_pci_dma_fault_add_capability,
 };
 
@@ -404,7 +458,8 @@ static int vfio_pci_dma_fault_init(struct vfio_pci_device *vdev)
 		VFIO_REGION_TYPE_NESTED,
 		VFIO_REGION_SUBTYPE_NESTED_DMA_FAULT,
 		&vfio_pci_dma_fault_regops, size,
-		VFIO_REGION_INFO_FLAG_READ | VFIO_REGION_INFO_FLAG_WRITE,
+		VFIO_REGION_INFO_FLAG_READ | VFIO_REGION_INFO_FLAG_WRITE |
+		VFIO_REGION_INFO_FLAG_MMAP,
 		vdev->fault_pages);
 	if (ret)
 		goto out;
@@ -412,7 +467,7 @@ static int vfio_pci_dma_fault_init(struct vfio_pci_device *vdev)
 	header = (struct vfio_region_dma_fault *)vdev->fault_pages;
 	header->entry_size = sizeof(struct iommu_fault);
 	header->nb_entries = DMA_FAULT_RING_LENGTH;
-	header->offset = sizeof(struct vfio_region_dma_fault);
+	header->offset = PAGE_SIZE;
 
 	ret = iommu_register_device_fault_handler(&vdev->pdev->dev,
 					vfio_pci_iommu_dev_fault_handler,
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 07/13] vfio: Use capability chains to handle device specific irq

[Eric Auger]

From: Tina Zhang <tina.zhang@intel.com>

Caps the number of irqs with fixed indexes and uses capability chains
to chain device specific irqs.

Signed-off-by: Tina Zhang <tina.zhang@intel.com>
Signed-off-by: Eric Auger <eric.auger@redhat.com>
[Eric: Put cap_offset at the end of the vfio_irq_info struct,
remove GFX IRQ at the moment and remove any reference to this latter
in the commit message]

---
---
 include/uapi/linux/vfio.h | 19 ++++++++++++++++++-
 1 file changed, 18 insertions(+), 1 deletion(-)

diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h
index ca8cc796e254..0728b6f3f348 100644
--- a/include/uapi/linux/vfio.h
+++ b/include/uapi/linux/vfio.h
@@ -708,11 +708,27 @@ struct vfio_irq_info {
 #define VFIO_IRQ_INFO_MASKABLE		(1 << 1)
 #define VFIO_IRQ_INFO_AUTOMASKED	(1 << 2)
 #define VFIO_IRQ_INFO_NORESIZE		(1 << 3)
+#define VFIO_IRQ_INFO_FLAG_CAPS		(1 << 4) /* Info supports caps */
 	__u32	index;		/* IRQ index */
 	__u32	count;		/* Number of IRQs within this index */
+	__u32	cap_offset;	/* Offset within info struct of first cap */
 };
 #define VFIO_DEVICE_GET_IRQ_INFO	_IO(VFIO_TYPE, VFIO_BASE + 9)
 
+/*
+ * The irq type capability allows IRQs unique to a specific device or
+ * class of devices to be exposed.
+ *
+ * The structures below define version 1 of this capability.
+ */
+#define VFIO_IRQ_INFO_CAP_TYPE      3
+
+struct vfio_irq_info_cap_type {
+	struct vfio_info_cap_header header;
+	__u32 type;     /* global per bus driver */
+	__u32 subtype;  /* type specific */
+};
+
 /**
  * VFIO_DEVICE_SET_IRQS - _IOW(VFIO_TYPE, VFIO_BASE + 10, struct vfio_irq_set)
  *
@@ -814,7 +830,8 @@ enum {
 	VFIO_PCI_MSIX_IRQ_INDEX,
 	VFIO_PCI_ERR_IRQ_INDEX,
 	VFIO_PCI_REQ_IRQ_INDEX,
-	VFIO_PCI_NUM_IRQS
+	VFIO_PCI_NUM_IRQS = 5	/* Fixed user ABI, IRQ indexes >=5 use   */
+				/* device specific cap to define content */
 };
 
 /*
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 08/13] vfio/pci: Add framework for custom interrupt indices

[Eric Auger]

Implement IRQ capability chain infrastructure. All interrupt
indexes beyond VFIO_PCI_NUM_IRQS are handled as extended
interrupts. They are registered with a specific type/subtype
and supported flags.

Signed-off-by: Eric Auger <eric.auger@redhat.com>

---

v11 -> v12:
- check !vdev->num_ext_irqs in vfio_pci_set_ext_irq_trigger()
  [Shameer, Qubingbing]
---
 drivers/vfio/pci/vfio_pci.c         | 99 +++++++++++++++++++++++------
 drivers/vfio/pci/vfio_pci_intrs.c   | 62 ++++++++++++++++++
 drivers/vfio/pci/vfio_pci_private.h | 14 ++++
 3 files changed, 157 insertions(+), 18 deletions(-)

diff --git a/drivers/vfio/pci/vfio_pci.c b/drivers/vfio/pci/vfio_pci.c
index eef76560be55..b3fc6ed4ed7a 100644
--- a/drivers/vfio/pci/vfio_pci.c
+++ b/drivers/vfio/pci/vfio_pci.c
@@ -610,6 +610,14 @@ static void vfio_pci_disable(struct vfio_pci_device *vdev)
 	ret = iommu_unregister_device_fault_handler(&vdev->pdev->dev);
 	WARN_ON(ret == -EBUSY);
 
+	for (i = 0; i < vdev->num_ext_irqs; i++)
+		vfio_pci_set_irqs_ioctl(vdev, VFIO_IRQ_SET_DATA_NONE |
+					VFIO_IRQ_SET_ACTION_TRIGGER,
+					VFIO_PCI_NUM_IRQS + i, 0, 0, NULL);
+	vdev->num_ext_irqs = 0;
+	kfree(vdev->ext_irqs);
+	vdev->ext_irqs = NULL;
+
 	/* Device closed, don't need mutex here */
 	list_for_each_entry_safe(ioeventfd, ioeventfd_tmp,
 				 &vdev->ioeventfds_list, next) {
@@ -825,6 +833,9 @@ static int vfio_pci_get_irq_count(struct vfio_pci_device *vdev, int irq_type)
 			return 1;
 	} else if (irq_type == VFIO_PCI_REQ_IRQ_INDEX) {
 		return 1;
+	} else if (irq_type >= VFIO_PCI_NUM_IRQS &&
+		   irq_type < VFIO_PCI_NUM_IRQS + vdev->num_ext_irqs) {
+		return 1;
 	}
 
 	return 0;
@@ -1011,7 +1022,7 @@ static long vfio_pci_ioctl(void *device_data,
 			info.flags |= VFIO_DEVICE_FLAGS_RESET;
 
 		info.num_regions = VFIO_PCI_NUM_REGIONS + vdev->num_regions;
-		info.num_irqs = VFIO_PCI_NUM_IRQS;
+		info.num_irqs = VFIO_PCI_NUM_IRQS + vdev->num_ext_irqs;
 
 		ret = vfio_pci_info_zdev_add_caps(vdev, &caps);
 		if (ret && ret != -ENODEV) {
@@ -1187,36 +1198,87 @@ static long vfio_pci_ioctl(void *device_data,
 
 	} else if (cmd == VFIO_DEVICE_GET_IRQ_INFO) {
 		struct vfio_irq_info info;
+		struct vfio_info_cap caps = { .buf = NULL, .size = 0 };
+		unsigned long capsz;
 
 		minsz = offsetofend(struct vfio_irq_info, count);
 
+		/* For backward compatibility, cannot require this */
+		capsz = offsetofend(struct vfio_irq_info, cap_offset);
+
 		if (copy_from_user(&info, (void __user *)arg, minsz))
 			return -EFAULT;
 
-		if (info.argsz < minsz || info.index >= VFIO_PCI_NUM_IRQS)
+		if (info.argsz < minsz ||
+			info.index >= VFIO_PCI_NUM_IRQS + vdev->num_ext_irqs)
 			return -EINVAL;
 
-		switch (info.index) {
-		case VFIO_PCI_INTX_IRQ_INDEX ... VFIO_PCI_MSIX_IRQ_INDEX:
-		case VFIO_PCI_REQ_IRQ_INDEX:
-			break;
-		case VFIO_PCI_ERR_IRQ_INDEX:
-			if (pci_is_pcie(vdev->pdev))
-				break;
-			fallthrough;
-		default:
-			return -EINVAL;
-		}
+		if (info.argsz >= capsz)
+			minsz = capsz;
 
 		info.flags = VFIO_IRQ_INFO_EVENTFD;
 
-		info.count = vfio_pci_get_irq_count(vdev, info.index);
-
-		if (info.index == VFIO_PCI_INTX_IRQ_INDEX)
+		switch (info.index) {
+		case VFIO_PCI_INTX_IRQ_INDEX:
 			info.flags |= (VFIO_IRQ_INFO_MASKABLE |
 				       VFIO_IRQ_INFO_AUTOMASKED);
-		else
+			break;
+		case VFIO_PCI_MSI_IRQ_INDEX ... VFIO_PCI_MSIX_IRQ_INDEX:
+		case VFIO_PCI_REQ_IRQ_INDEX:
 			info.flags |= VFIO_IRQ_INFO_NORESIZE;
+			break;
+		case VFIO_PCI_ERR_IRQ_INDEX:
+			info.flags |= VFIO_IRQ_INFO_NORESIZE;
+			if (!pci_is_pcie(vdev->pdev))
+				return -EINVAL;
+			break;
+		default:
+		{
+			struct vfio_irq_info_cap_type cap_type = {
+				.header.id = VFIO_IRQ_INFO_CAP_TYPE,
+				.header.version = 1 };
+			int ret, i;
+
+			if (info.index >= VFIO_PCI_NUM_IRQS +
+						vdev->num_ext_irqs)
+				return -EINVAL;
+			info.index = array_index_nospec(info.index,
+							VFIO_PCI_NUM_IRQS +
+							vdev->num_ext_irqs);
+			i = info.index - VFIO_PCI_NUM_IRQS;
+
+			info.flags = vdev->ext_irqs[i].flags;
+			cap_type.type = vdev->ext_irqs[i].type;
+			cap_type.subtype = vdev->ext_irqs[i].subtype;
+
+			ret = vfio_info_add_capability(&caps,
+					&cap_type.header,
+					sizeof(cap_type));
+			if (ret)
+				return ret;
+		}
+		}
+
+		info.count = vfio_pci_get_irq_count(vdev, info.index);
+
+		if (caps.size) {
+			info.flags |= VFIO_IRQ_INFO_FLAG_CAPS;
+			if (info.argsz < sizeof(info) + caps.size) {
+				info.argsz = sizeof(info) + caps.size;
+				info.cap_offset = 0;
+			} else {
+				vfio_info_cap_shift(&caps, sizeof(info));
+				if (copy_to_user((void __user *)arg +
+						  sizeof(info), caps.buf,
+						  caps.size)) {
+					kfree(caps.buf);
+					return -EFAULT;
+				}
+				info.cap_offset = sizeof(info);
+			}
+
+			kfree(caps.buf);
+		}
 
 		return copy_to_user((void __user *)arg, &info, minsz) ?
 			-EFAULT : 0;
@@ -1235,7 +1297,8 @@ static long vfio_pci_ioctl(void *device_data,
 		max = vfio_pci_get_irq_count(vdev, hdr.index);
 
 		ret = vfio_set_irqs_validate_and_prepare(&hdr, max,
-						 VFIO_PCI_NUM_IRQS, &data_size);
+				VFIO_PCI_NUM_IRQS + vdev->num_ext_irqs,
+				&data_size);
 		if (ret)
 			return ret;
 
diff --git a/drivers/vfio/pci/vfio_pci_intrs.c b/drivers/vfio/pci/vfio_pci_intrs.c
index 869dce5f134d..d67995fe872f 100644
--- a/drivers/vfio/pci/vfio_pci_intrs.c
+++ b/drivers/vfio/pci/vfio_pci_intrs.c
@@ -19,6 +19,7 @@
 #include <linux/vfio.h>
 #include <linux/wait.h>
 #include <linux/slab.h>
+#include <linux/nospec.h>
 
 #include "vfio_pci_private.h"
 
@@ -635,6 +636,24 @@ static int vfio_pci_set_req_trigger(struct vfio_pci_device *vdev,
 					       count, flags, data);
 }
 
+static int vfio_pci_set_ext_irq_trigger(struct vfio_pci_device *vdev,
+					unsigned int index, unsigned int start,
+					unsigned int count, uint32_t flags,
+					void *data)
+{
+	int i;
+
+	if (start != 0 || count > 1 || !vdev->num_ext_irqs)
+		return -EINVAL;
+
+	index = array_index_nospec(index,
+				   VFIO_PCI_NUM_IRQS + vdev->num_ext_irqs);
+	i = index - VFIO_PCI_NUM_IRQS;
+
+	return vfio_pci_set_ctx_trigger_single(&vdev->ext_irqs[i].trigger,
+					       count, flags, data);
+}
+
 int vfio_pci_set_irqs_ioctl(struct vfio_pci_device *vdev, uint32_t flags,
 			    unsigned index, unsigned start, unsigned count,
 			    void *data)
@@ -684,6 +703,13 @@ int vfio_pci_set_irqs_ioctl(struct vfio_pci_device *vdev, uint32_t flags,
 			break;
 		}
 		break;
+	default:
+		switch (flags & VFIO_IRQ_SET_ACTION_TYPE_MASK) {
+		case VFIO_IRQ_SET_ACTION_TRIGGER:
+			func = vfio_pci_set_ext_irq_trigger;
+			break;
+		}
+		break;
 	}
 
 	if (!func)
@@ -691,3 +717,39 @@ int vfio_pci_set_irqs_ioctl(struct vfio_pci_device *vdev, uint32_t flags,
 
 	return func(vdev, index, start, count, flags, data);
 }
+
+int vfio_pci_get_ext_irq_index(struct vfio_pci_device *vdev,
+			       unsigned int type, unsigned int subtype)
+{
+	int i;
+
+	for (i = 0; i <  vdev->num_ext_irqs; i++) {
+		if (vdev->ext_irqs[i].type == type &&
+		    vdev->ext_irqs[i].subtype == subtype) {
+			return i;
+		}
+	}
+	return -EINVAL;
+}
+
+int vfio_pci_register_irq(struct vfio_pci_device *vdev,
+			  unsigned int type, unsigned int subtype,
+			  u32 flags)
+{
+	struct vfio_ext_irq *ext_irqs;
+
+	ext_irqs = krealloc(vdev->ext_irqs,
+			    (vdev->num_ext_irqs + 1) * sizeof(*ext_irqs),
+			    GFP_KERNEL);
+	if (!ext_irqs)
+		return -ENOMEM;
+
+	vdev->ext_irqs = ext_irqs;
+
+	vdev->ext_irqs[vdev->num_ext_irqs].type = type;
+	vdev->ext_irqs[vdev->num_ext_irqs].subtype = subtype;
+	vdev->ext_irqs[vdev->num_ext_irqs].flags = flags;
+	vdev->ext_irqs[vdev->num_ext_irqs].trigger = NULL;
+	vdev->num_ext_irqs++;
+	return 0;
+}
diff --git a/drivers/vfio/pci/vfio_pci_private.h b/drivers/vfio/pci/vfio_pci_private.h
index 76cb3bd11375..e2a999107bcb 100644
--- a/drivers/vfio/pci/vfio_pci_private.h
+++ b/drivers/vfio/pci/vfio_pci_private.h
@@ -77,6 +77,13 @@ struct vfio_pci_region {
 	u32				flags;
 };
 
+struct vfio_ext_irq {
+	u32				type;
+	u32				subtype;
+	u32				flags;
+	struct eventfd_ctx		*trigger;
+};
+
 struct vfio_pci_dummy_resource {
 	struct resource		resource;
 	int			index;
@@ -111,6 +118,8 @@ struct vfio_pci_device {
 	struct vfio_pci_irq_ctx	*ctx;
 	int			num_ctx;
 	int			irq_type;
+	struct vfio_ext_irq	*ext_irqs;
+	int			num_ext_irqs;
 	int			num_regions;
 	struct vfio_pci_region	*region;
 	u8			msi_qmax;
@@ -154,6 +163,11 @@ struct vfio_pci_device {
 
 extern void vfio_pci_intx_mask(struct vfio_pci_device *vdev);
 extern void vfio_pci_intx_unmask(struct vfio_pci_device *vdev);
+extern int vfio_pci_register_irq(struct vfio_pci_device *vdev,
+				 unsigned int type, unsigned int subtype,
+				 u32 flags);
+extern int vfio_pci_get_ext_irq_index(struct vfio_pci_device *vdev,
+				      unsigned int type, unsigned int subtype);
 
 extern int vfio_pci_set_irqs_ioctl(struct vfio_pci_device *vdev,
 				   uint32_t flags, unsigned index,
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 09/13] vfio: Add new IRQ for DMA fault reporting

[Eric Auger]

Add a new IRQ type/subtype to get notification on nested
stage DMA faults.

Signed-off-by: Eric Auger <eric.auger@redhat.com>
---
 include/uapi/linux/vfio.h | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h
index 0728b6f3f348..ad7c275b4074 100644
--- a/include/uapi/linux/vfio.h
+++ b/include/uapi/linux/vfio.h
@@ -729,6 +729,9 @@ struct vfio_irq_info_cap_type {
 	__u32 subtype;  /* type specific */
 };
 
+#define VFIO_IRQ_TYPE_NESTED				(1)
+#define VFIO_IRQ_SUBTYPE_DMA_FAULT			(1)
+
 /**
  * VFIO_DEVICE_SET_IRQS - _IOW(VFIO_TYPE, VFIO_BASE + 10, struct vfio_irq_set)
  *
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 10/13] vfio/pci: Register and allow DMA FAULT IRQ signaling

[Eric Auger]

Register the VFIO_IRQ_TYPE_NESTED/VFIO_IRQ_SUBTYPE_DMA_FAULT
IRQ that allows to signal a nested mode DMA fault.

Signed-off-by: Eric Auger <eric.auger@redhat.com>

---

v10 -> v11:
- the irq now is registered in vfio_pci_dma_fault_init()
  in case the domain is nested
---
 drivers/vfio/pci/vfio_pci.c | 21 ++++++++++++++++++++-
 1 file changed, 20 insertions(+), 1 deletion(-)

diff --git a/drivers/vfio/pci/vfio_pci.c b/drivers/vfio/pci/vfio_pci.c
index b3fc6ed4ed7a..72d7c667b64c 100644
--- a/drivers/vfio/pci/vfio_pci.c
+++ b/drivers/vfio/pci/vfio_pci.c
@@ -396,6 +396,7 @@ vfio_pci_iommu_dev_fault_handler(struct iommu_fault *fault, void *data)
 		(struct vfio_region_dma_fault *)vdev->fault_pages;
 	struct iommu_fault *new;
 	u32 head, tail, size;
+	int ext_irq_index;
 	int ret = -EINVAL;
 
 	if (WARN_ON(!reg))
@@ -420,7 +421,19 @@ vfio_pci_iommu_dev_fault_handler(struct iommu_fault *fault, void *data)
 	ret = 0;
 unlock:
 	mutex_unlock(&vdev->fault_queue_lock);
-	return ret;
+	if (ret)
+		return ret;
+
+	ext_irq_index = vfio_pci_get_ext_irq_index(vdev, VFIO_IRQ_TYPE_NESTED,
+						   VFIO_IRQ_SUBTYPE_DMA_FAULT);
+	if (ext_irq_index < 0)
+		return -EINVAL;
+
+	mutex_lock(&vdev->igate);
+	if (vdev->ext_irqs[ext_irq_index].trigger)
+		eventfd_signal(vdev->ext_irqs[ext_irq_index].trigger, 1);
+	mutex_unlock(&vdev->igate);
+	return 0;
 }
 
 #define DMA_FAULT_RING_LENGTH 512
@@ -475,6 +488,12 @@ static int vfio_pci_dma_fault_init(struct vfio_pci_device *vdev)
 	if (ret) /* the dma fault region is freed in vfio_pci_disable() */
 		goto out;
 
+	ret = vfio_pci_register_irq(vdev, VFIO_IRQ_TYPE_NESTED,
+				    VFIO_IRQ_SUBTYPE_DMA_FAULT,
+				    VFIO_IRQ_INFO_EVENTFD);
+	if (ret) /* the fault handler is also freed in vfio_pci_disable() */
+		goto out;
+
 	return 0;
 out:
 	kfree(vdev->fault_pages);
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 11/13] vfio: Document nested stage control

[Eric Auger]

The VFIO API was enhanced to support nested stage control: a bunch of
new iotcls, one DMA FAULT region and an associated specific IRQ.

Let's document the process to follow to set up nested mode.

Signed-off-by: Eric Auger <eric.auger@redhat.com>

---

v11 -> v12:
s/VFIO_REGION_INFO_CAP_PRODUCER_FAULT/VFIO_REGION_INFO_CAP_DMA_FAULT

v8 -> v9:
- new names for SET_MSI_BINDING and SET_PASID_TABLE
- new layout for the DMA FAULT memory region and specific IRQ

v2 -> v3:
- document the new fault API

v1 -> v2:
- use the new ioctl names
- add doc related to fault handling
---
 Documentation/driver-api/vfio.rst | 77 +++++++++++++++++++++++++++++++
 1 file changed, 77 insertions(+)

diff --git a/Documentation/driver-api/vfio.rst b/Documentation/driver-api/vfio.rst
index f1a4d3c3ba0b..14e41324237d 100644
--- a/Documentation/driver-api/vfio.rst
+++ b/Documentation/driver-api/vfio.rst
@@ -239,6 +239,83 @@ group and can access them as follows::
 	/* Gratuitous device reset and go... */
 	ioctl(device, VFIO_DEVICE_RESET);
 
+IOMMU Dual Stage Control
+------------------------
+
+Some IOMMUs support 2 stages/levels of translation. "Stage" corresponds to
+the ARM terminology while "level" corresponds to Intel's VTD terminology. In
+the following text we use either without distinction.
+
+This is useful when the guest is exposed with a virtual IOMMU and some
+devices are assigned to the guest through VFIO. Then the guest OS can use
+stage 1 (IOVA -> GPA), while the hypervisor uses stage 2 for VM isolation
+(GPA -> HPA).
+
+The guest gets ownership of the stage 1 page tables and also owns stage 1
+configuration structures. The hypervisor owns the root configuration structure
+(for security reason), including stage 2 configuration. This works as long
+configuration structures and page table format are compatible between the
+virtual IOMMU and the physical IOMMU.
+
+Assuming the HW supports it, this nested mode is selected by choosing the
+VFIO_TYPE1_NESTING_IOMMU type through:
+
+ioctl(container, VFIO_SET_IOMMU, VFIO_TYPE1_NESTING_IOMMU);
+
+This forces the hypervisor to use the stage 2, leaving stage 1 available for
+guest usage.
+
+Once groups are attached to the container, the guest stage 1 translation
+configuration data can be passed to VFIO by using
+
+ioctl(container, VFIO_IOMMU_SET_PASID_TABLE, &pasid_table_info);
+
+This allows to combine the guest stage 1 configuration structure along with
+the hypervisor stage 2 configuration structure. Stage 1 configuration
+structures are dependent on the IOMMU type.
+
+As the stage 1 translation is fully delegated to the HW, translation faults
+encountered during the translation process need to be propagated up to
+the virtualizer and re-injected into the guest.
+
+The userspace must be prepared to receive faults. The VFIO-PCI device
+exposes one dedicated DMA FAULT region: it contains a ring buffer and
+its header that allows to manage the head/tail indices. The region is
+identified by the following index/subindex:
+- VFIO_REGION_TYPE_NESTED/VFIO_REGION_SUBTYPE_NESTED_DMA_FAULT
+
+The DMA FAULT region exposes a VFIO_REGION_INFO_CAP_DMA_FAULT
+region capability that allows the userspace to retrieve the ABI version
+of the fault records filled by the host.
+
+On top of that region, the userspace can be notified whenever a fault
+occurs at the physical level. It can use the VFIO_IRQ_TYPE_NESTED/
+VFIO_IRQ_SUBTYPE_DMA_FAULT specific IRQ to attach the eventfd to be
+signalled.
+
+The ring buffer containing the fault records can be mmapped. When
+the userspace consumes a fault in the queue, it should increment
+the consumer index to allow new fault records to replace the used ones.
+
+The queue size and the entry size can be retrieved in the header.
+The tail index should never overshoot the producer index as in any
+other circular buffer scheme. Also it must be less than the queue size
+otherwise the change fails.
+
+When the guest invalidates stage 1 related caches, invalidations must be
+forwarded to the host through
+ioctl(container, VFIO_IOMMU_CACHE_INVALIDATE, &inv_data);
+Those invalidations can happen at various granularity levels, page, context, ...
+
+The ARM SMMU specification introduces another challenge: MSIs are translated by
+both the virtual SMMU and the physical SMMU. To build a nested mapping for the
+IOVA programmed into the assigned device, the guest needs to pass its IOVA/MSI
+doorbell GPA binding to the host. Then the hypervisor can build a nested stage 2
+binding eventually translating into the physical MSI doorbell.
+
+This is achieved by calling
+ioctl(container, VFIO_IOMMU_SET_MSI_BINDING, &guest_binding);
+
 VFIO User API
 -------------------------------------------------------------------------------
 
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 12/13] vfio/pci: Register a DMA fault response region

[Eric Auger]

In preparation for vSVA, let's register a DMA fault response region,
where the userspace will push the page responses and increment the
head of the buffer. The kernel will pop those responses and inject them
on iommu side.

Signed-off-by: Eric Auger <eric.auger@redhat.com>

---

v11 -> v12:
- use DMA_FAULT_RESPONSE cap [Shameer]
- struct vfio_pci_device dma_fault_response_wq field introduced in
  this patch
- return 0 if the domain is NULL
- pass an int pointer to iommu_domain_get_attr
---
 drivers/vfio/pci/vfio_pci.c         | 125 ++++++++++++++++++++++++++--
 drivers/vfio/pci/vfio_pci_private.h |   6 ++
 drivers/vfio/pci/vfio_pci_rdwr.c    |  39 +++++++++
 include/uapi/linux/vfio.h           |  32 +++++++
 4 files changed, 193 insertions(+), 9 deletions(-)

diff --git a/drivers/vfio/pci/vfio_pci.c b/drivers/vfio/pci/vfio_pci.c
index 72d7c667b64c..560b1a830726 100644
--- a/drivers/vfio/pci/vfio_pci.c
+++ b/drivers/vfio/pci/vfio_pci.c
@@ -316,9 +316,20 @@ static void vfio_pci_dma_fault_release(struct vfio_pci_device *vdev,
 	kfree(vdev->fault_pages);
 }
 
-static int vfio_pci_dma_fault_mmap(struct vfio_pci_device *vdev,
-				   struct vfio_pci_region *region,
-				   struct vm_area_struct *vma)
+static void
+vfio_pci_dma_fault_response_release(struct vfio_pci_device *vdev,
+				    struct vfio_pci_region *region)
+{
+	if (vdev->dma_fault_response_wq)
+		destroy_workqueue(vdev->dma_fault_response_wq);
+	kfree(vdev->fault_response_pages);
+	vdev->fault_response_pages = NULL;
+}
+
+static int __vfio_pci_dma_fault_mmap(struct vfio_pci_device *vdev,
+				     struct vfio_pci_region *region,
+				     struct vm_area_struct *vma,
+				     u8 *pages)
 {
 	u64 phys_len, req_len, pgoff, req_start;
 	unsigned long long addr;
@@ -331,14 +342,14 @@ static int vfio_pci_dma_fault_mmap(struct vfio_pci_device *vdev,
 		((1U << (VFIO_PCI_OFFSET_SHIFT - PAGE_SHIFT)) - 1);
 	req_start = pgoff << PAGE_SHIFT;
 
-	/* only the second page of the producer fault region is mmappable */
+	/* only the second page of the fault region is mmappable */
 	if (req_start < PAGE_SIZE)
 		return -EINVAL;
 
 	if (req_start + req_len > phys_len)
 		return -EINVAL;
 
-	addr = virt_to_phys(vdev->fault_pages);
+	addr = virt_to_phys(pages);
 	vma->vm_private_data = vdev;
 	vma->vm_pgoff = (addr >> PAGE_SHIFT) + pgoff;
 
@@ -347,13 +358,29 @@ static int vfio_pci_dma_fault_mmap(struct vfio_pci_device *vdev,
 	return ret;
 }
 
-static int vfio_pci_dma_fault_add_capability(struct vfio_pci_device *vdev,
-					     struct vfio_pci_region *region,
-					     struct vfio_info_cap *caps)
+static int vfio_pci_dma_fault_mmap(struct vfio_pci_device *vdev,
+				   struct vfio_pci_region *region,
+				   struct vm_area_struct *vma)
+{
+	return __vfio_pci_dma_fault_mmap(vdev, region, vma, vdev->fault_pages);
+}
+
+static int
+vfio_pci_dma_fault_response_mmap(struct vfio_pci_device *vdev,
+				struct vfio_pci_region *region,
+				struct vm_area_struct *vma)
+{
+	return __vfio_pci_dma_fault_mmap(vdev, region, vma, vdev->fault_response_pages);
+}
+
+static int __vfio_pci_dma_fault_add_capability(struct vfio_pci_device *vdev,
+					       struct vfio_pci_region *region,
+					       struct vfio_info_cap *caps,
+					       u32 cap_id)
 {
 	struct vfio_region_info_cap_sparse_mmap *sparse = NULL;
 	struct vfio_region_info_cap_fault cap = {
-		.header.id = VFIO_REGION_INFO_CAP_DMA_FAULT,
+		.header.id = cap_id,
 		.header.version = 1,
 		.version = 1,
 	};
@@ -381,6 +408,23 @@ static int vfio_pci_dma_fault_add_capability(struct vfio_pci_device *vdev,
 	return ret;
 }
 
+static int vfio_pci_dma_fault_add_capability(struct vfio_pci_device *vdev,
+					     struct vfio_pci_region *region,
+					     struct vfio_info_cap *caps)
+{
+	return __vfio_pci_dma_fault_add_capability(vdev, region, caps,
+						   VFIO_REGION_INFO_CAP_DMA_FAULT);
+}
+
+static int
+vfio_pci_dma_fault_response_add_capability(struct vfio_pci_device *vdev,
+					   struct vfio_pci_region *region,
+					   struct vfio_info_cap *caps)
+{
+	return __vfio_pci_dma_fault_add_capability(vdev, region, caps,
+						   VFIO_REGION_INFO_CAP_DMA_FAULT_RESPONSE);
+}
+
 static const struct vfio_pci_regops vfio_pci_dma_fault_regops = {
 	.rw		= vfio_pci_dma_fault_rw,
 	.release	= vfio_pci_dma_fault_release,
@@ -388,6 +432,13 @@ static const struct vfio_pci_regops vfio_pci_dma_fault_regops = {
 	.add_capability = vfio_pci_dma_fault_add_capability,
 };
 
+static const struct vfio_pci_regops vfio_pci_dma_fault_response_regops = {
+	.rw		= vfio_pci_dma_fault_response_rw,
+	.release	= vfio_pci_dma_fault_response_release,
+	.mmap		= vfio_pci_dma_fault_response_mmap,
+	.add_capability = vfio_pci_dma_fault_response_add_capability,
+};
+
 static int
 vfio_pci_iommu_dev_fault_handler(struct iommu_fault *fault, void *data)
 {
@@ -501,6 +552,57 @@ static int vfio_pci_dma_fault_init(struct vfio_pci_device *vdev)
 	return ret;
 }
 
+#define DMA_FAULT_RESPONSE_RING_LENGTH 512
+
+static int vfio_pci_dma_fault_response_init(struct vfio_pci_device *vdev)
+{
+	struct vfio_region_dma_fault_response *header;
+	struct iommu_domain *domain;
+	int nested, ret;
+	size_t size;
+
+	domain = iommu_get_domain_for_dev(&vdev->pdev->dev);
+	if (!domain)
+		return 0;
+
+	ret = iommu_domain_get_attr(domain, DOMAIN_ATTR_NESTING, &nested);
+	if (ret || !nested)
+		return ret;
+
+	mutex_init(&vdev->fault_response_queue_lock);
+
+	/*
+	 * We provision 1 page for the header and space for
+	 * DMA_FAULT_RING_LENGTH fault records in the ring buffer.
+	 */
+	size = ALIGN(sizeof(struct iommu_page_response) *
+		     DMA_FAULT_RESPONSE_RING_LENGTH, PAGE_SIZE) + PAGE_SIZE;
+
+	vdev->fault_response_pages = kzalloc(size, GFP_KERNEL);
+	if (!vdev->fault_response_pages)
+		return -ENOMEM;
+
+	ret = vfio_pci_register_dev_region(vdev,
+		VFIO_REGION_TYPE_NESTED,
+		VFIO_REGION_SUBTYPE_NESTED_DMA_FAULT_RESPONSE,
+		&vfio_pci_dma_fault_response_regops, size,
+		VFIO_REGION_INFO_FLAG_READ | VFIO_REGION_INFO_FLAG_WRITE |
+		VFIO_REGION_INFO_FLAG_MMAP,
+		vdev->fault_response_pages);
+	if (ret)
+		goto out;
+
+	header = (struct vfio_region_dma_fault_response *)vdev->fault_response_pages;
+	header->entry_size = sizeof(struct iommu_page_response);
+	header->nb_entries = DMA_FAULT_RESPONSE_RING_LENGTH;
+	header->offset = PAGE_SIZE;
+
+	return 0;
+out:
+	vdev->fault_response_pages = NULL;
+	return ret;
+}
+
 static int vfio_pci_enable(struct vfio_pci_device *vdev)
 {
 	struct pci_dev *pdev = vdev->pdev;
@@ -603,6 +705,10 @@ static int vfio_pci_enable(struct vfio_pci_device *vdev)
 	if (ret)
 		goto disable_exit;
 
+	ret = vfio_pci_dma_fault_response_init(vdev);
+	if (ret)
+		goto disable_exit;
+
 	vfio_pci_probe_mmaps(vdev);
 
 	return 0;
@@ -2228,6 +2334,7 @@ static int vfio_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	INIT_LIST_HEAD(&vdev->ioeventfds_list);
 	mutex_init(&vdev->vma_lock);
 	INIT_LIST_HEAD(&vdev->vma_list);
+	INIT_LIST_HEAD(&vdev->dummy_resources_list);
 	init_rwsem(&vdev->memory_lock);
 
 	ret = vfio_add_group_dev(&pdev->dev, &vfio_pci_ops, vdev);
diff --git a/drivers/vfio/pci/vfio_pci_private.h b/drivers/vfio/pci/vfio_pci_private.h
index e2a999107bcb..f7b1e7fb86e5 100644
--- a/drivers/vfio/pci/vfio_pci_private.h
+++ b/drivers/vfio/pci/vfio_pci_private.h
@@ -144,7 +144,10 @@ struct vfio_pci_device {
 	struct eventfd_ctx	*err_trigger;
 	struct eventfd_ctx	*req_trigger;
 	u8			*fault_pages;
+	u8			*fault_response_pages;
+	struct workqueue_struct *dma_fault_response_wq;
 	struct mutex		fault_queue_lock;
+	struct mutex		fault_response_queue_lock;
 	struct list_head	dummy_resources_list;
 	struct mutex		ioeventfds_lock;
 	struct list_head	ioeventfds_list;
@@ -189,6 +192,9 @@ extern long vfio_pci_ioeventfd(struct vfio_pci_device *vdev, loff_t offset,
 extern size_t vfio_pci_dma_fault_rw(struct vfio_pci_device *vdev,
 				    char __user *buf, size_t count,
 				    loff_t *ppos, bool iswrite);
+extern size_t vfio_pci_dma_fault_response_rw(struct vfio_pci_device *vdev,
+					     char __user *buf, size_t count,
+					     loff_t *ppos, bool iswrite);
 
 extern int vfio_pci_init_perm_bits(void);
 extern void vfio_pci_uninit_perm_bits(void);
diff --git a/drivers/vfio/pci/vfio_pci_rdwr.c b/drivers/vfio/pci/vfio_pci_rdwr.c
index 164120607469..efde0793360b 100644
--- a/drivers/vfio/pci/vfio_pci_rdwr.c
+++ b/drivers/vfio/pci/vfio_pci_rdwr.c
@@ -400,6 +400,45 @@ size_t vfio_pci_dma_fault_rw(struct vfio_pci_device *vdev, char __user *buf,
 	return ret;
 }
 
+size_t vfio_pci_dma_fault_response_rw(struct vfio_pci_device *vdev, char __user *buf,
+				      size_t count, loff_t *ppos, bool iswrite)
+{
+	unsigned int i = VFIO_PCI_OFFSET_TO_INDEX(*ppos) - VFIO_PCI_NUM_REGIONS;
+	loff_t pos = *ppos & VFIO_PCI_OFFSET_MASK;
+	void *base = vdev->region[i].data;
+	int ret = -EFAULT;
+
+	if (pos >= vdev->region[i].size)
+		return -EINVAL;
+
+	count = min(count, (size_t)(vdev->region[i].size - pos));
+
+	if (iswrite) {
+		struct vfio_region_dma_fault_response *header =
+			(struct vfio_region_dma_fault_response *)base;
+		uint32_t  new_head;
+
+		if (pos != 0 || count != 4)
+			return -EINVAL;
+
+		if (copy_from_user((void *)&new_head, buf, count))
+			return -EFAULT;
+
+		if (new_head >= header->nb_entries)
+			return -EINVAL;
+
+		mutex_lock(&vdev->fault_response_queue_lock);
+		header->head = new_head;
+		mutex_unlock(&vdev->fault_response_queue_lock);
+	} else {
+		if (copy_to_user(buf, base + pos, count))
+			return -EFAULT;
+	}
+	*ppos += count;
+	ret = count;
+	return ret;
+}
+
 static void vfio_pci_ioeventfd_do_write(struct vfio_pci_ioeventfd *ioeventfd,
 					bool test_mem)
 {
diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h
index ad7c275b4074..bf7c62b47ce2 100644
--- a/include/uapi/linux/vfio.h
+++ b/include/uapi/linux/vfio.h
@@ -352,6 +352,7 @@ struct vfio_region_info_cap_type {
 
 /* sub-types for VFIO_REGION_TYPE_NESTED */
 #define VFIO_REGION_SUBTYPE_NESTED_DMA_FAULT	(1)
+#define VFIO_REGION_SUBTYPE_NESTED_DMA_FAULT_RESPONSE	(2)
 
 /**
  * struct vfio_region_gfx_edid - EDID region layout.
@@ -1030,6 +1031,17 @@ struct vfio_region_info_cap_fault {
 	__u32 version;
 };
 
+/*
+ * Capability exposed by the DMA fault response region
+ * @version: ABI version
+ */
+#define VFIO_REGION_INFO_CAP_DMA_FAULT_RESPONSE	7
+
+struct vfio_region_info_cap_fault_response {
+	struct vfio_info_cap_header header;
+	__u32 version;
+};
+
 /*
  * DMA Fault Region Layout
  * @tail: index relative to the start of the ring buffer at which the
@@ -1050,6 +1062,26 @@ struct vfio_region_dma_fault {
 	__u32   head;
 };
 
+/*
+ * DMA Fault Response Region Layout
+ * @head: index relative to the start of the ring buffer at which the
+ *        producer (userspace) insert responses into the buffer
+ * @entry_size: fault ring buffer entry size in bytes
+ * @nb_entries: max capacity of the fault ring buffer
+ * @offset: ring buffer offset relative to the start of the region
+ * @tail: index relative to the start of the ring buffer at which the
+ *        consumer (kernel) finds the next item in the buffer
+ */
+struct vfio_region_dma_fault_response {
+	/* Write-Only */
+	__u32   head;
+	/* Read-Only */
+	__u32   entry_size;
+	__u32	nb_entries;
+	__u32	offset;
+	__u32   tail;
+};
+
 /* -------- API for Type1 VFIO IOMMU -------- */
 
 /**
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

[PATCH v13 13/13] vfio/pci: Inject page response upon response region fill

[Eric Auger]

When the userspace increments the head of the page response
buffer ring, let's push the response into the iommu layer.
This is done through a workqueue that pops the responses from
the ring buffer and increment the tail.

Signed-off-by: Eric Auger <eric.auger@redhat.com>
---
 drivers/vfio/pci/vfio_pci.c         | 40 +++++++++++++++++++++++++++++
 drivers/vfio/pci/vfio_pci_private.h |  7 +++++
 drivers/vfio/pci/vfio_pci_rdwr.c    |  1 +
 3 files changed, 48 insertions(+)

diff --git a/drivers/vfio/pci/vfio_pci.c b/drivers/vfio/pci/vfio_pci.c
index 560b1a830726..bb4a0e1e39bf 100644
--- a/drivers/vfio/pci/vfio_pci.c
+++ b/drivers/vfio/pci/vfio_pci.c
@@ -552,6 +552,32 @@ static int vfio_pci_dma_fault_init(struct vfio_pci_device *vdev)
 	return ret;
 }
 
+static void dma_response_inject(struct work_struct *work)
+{
+	struct vfio_pci_dma_fault_response_work *rwork =
+		container_of(work, struct vfio_pci_dma_fault_response_work, inject);
+	struct vfio_region_dma_fault_response *header = rwork->header;
+	struct vfio_pci_device *vdev = rwork->vdev;
+	struct iommu_page_response *resp;
+	u32 tail, head, size;
+
+	mutex_lock(&vdev->fault_response_queue_lock);
+
+	tail = header->tail;
+	head = header->head;
+	size = header->nb_entries;
+
+	while (CIRC_CNT(head, tail, size) >= 1) {
+		resp = (struct iommu_page_response *)(vdev->fault_response_pages + header->offset +
+						tail * header->entry_size);
+
+		/* TODO: properly handle the return value */
+		iommu_page_response(&vdev->pdev->dev, resp);
+		header->tail = tail = (tail + 1) % size;
+	}
+	mutex_unlock(&vdev->fault_response_queue_lock);
+}
+
 #define DMA_FAULT_RESPONSE_RING_LENGTH 512
 
 static int vfio_pci_dma_fault_response_init(struct vfio_pci_device *vdev)
@@ -597,8 +623,22 @@ static int vfio_pci_dma_fault_response_init(struct vfio_pci_device *vdev)
 	header->nb_entries = DMA_FAULT_RESPONSE_RING_LENGTH;
 	header->offset = PAGE_SIZE;
 
+	vdev->response_work = kzalloc(sizeof(*vdev->response_work), GFP_KERNEL);
+	if (!vdev->response_work)
+		goto out;
+	vdev->response_work->header = header;
+	vdev->response_work->vdev = vdev;
+
+	/* launch the thread that will extract the response */
+	INIT_WORK(&vdev->response_work->inject, dma_response_inject);
+	vdev->dma_fault_response_wq =
+		create_singlethread_workqueue("vfio-dma-fault-response");
+	if (!vdev->dma_fault_response_wq)
+		return -ENOMEM;
+
 	return 0;
 out:
+	kfree(vdev->fault_response_pages);
 	vdev->fault_response_pages = NULL;
 	return ret;
 }
diff --git a/drivers/vfio/pci/vfio_pci_private.h b/drivers/vfio/pci/vfio_pci_private.h
index f7b1e7fb86e5..835fbb221dea 100644
--- a/drivers/vfio/pci/vfio_pci_private.h
+++ b/drivers/vfio/pci/vfio_pci_private.h
@@ -52,6 +52,12 @@ struct vfio_pci_irq_ctx {
 	struct irq_bypass_producer	producer;
 };
 
+struct vfio_pci_dma_fault_response_work {
+	struct work_struct inject;
+	struct vfio_region_dma_fault_response *header;
+	struct vfio_pci_device *vdev;
+};
+
 struct vfio_pci_device;
 struct vfio_pci_region;
 
@@ -146,6 +152,7 @@ struct vfio_pci_device {
 	u8			*fault_pages;
 	u8			*fault_response_pages;
 	struct workqueue_struct *dma_fault_response_wq;
+	struct vfio_pci_dma_fault_response_work *response_work;
 	struct mutex		fault_queue_lock;
 	struct mutex		fault_response_queue_lock;
 	struct list_head	dummy_resources_list;
diff --git a/drivers/vfio/pci/vfio_pci_rdwr.c b/drivers/vfio/pci/vfio_pci_rdwr.c
index efde0793360b..78c494fe35cc 100644
--- a/drivers/vfio/pci/vfio_pci_rdwr.c
+++ b/drivers/vfio/pci/vfio_pci_rdwr.c
@@ -430,6 +430,7 @@ size_t vfio_pci_dma_fault_response_rw(struct vfio_pci_device *vdev, char __user
 		mutex_lock(&vdev->fault_response_queue_lock);
 		header->head = new_head;
 		mutex_unlock(&vdev->fault_response_queue_lock);
+		queue_work(vdev->dma_fault_response_wq, &vdev->response_work->inject);
 	} else {
 		if (copy_to_user(buf, base + pos, count))
 			return -EFAULT;
-- 
2.26.3

_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm