Merge tag 'vfio-v6.18-rc4' of https://github.com/awilliam/linux-vfio

Pull VFIO fixes from Alex Williamson:

 - Fix overflows in vfio type1 backend for mappings at the end of the
   64-bit address space, resulting in leaked pinned memory.

   New selftest support included to avoid such issues in the future
   (Alex Mastro)

* tag 'vfio-v6.18-rc4' of https://github.com/awilliam/linux-vfio:
  vfio: selftests: add end of address space DMA map/unmap tests
  vfio: selftests: update DMA map/unmap helpers to support more test kinds
  vfio/type1: handle DMA map/unmap up to the addressable limit
  vfio/type1: move iova increment to unmap_unpin_*() caller
  vfio/type1: sanitize for overflow using check_*_overflow()

3 files changed, 198 insertions, 28 deletions

diff --git a/tools/testing/selftests/vfio/lib/include/vfio_util.h b/tools/testing/selftests/vfio/lib/include/vfio_util.h
index ed31606e01b7..240409bf5f8a 100644
--- a/tools/testing/selftests/vfio/lib/include/vfio_util.h
+++ b/tools/testing/selftests/vfio/lib/include/vfio_util.h
@@ -206,10 +206,29 @@ struct vfio_pci_device *vfio_pci_device_init(const char *bdf, const char *iommu_
 void vfio_pci_device_cleanup(struct vfio_pci_device *device);
 void vfio_pci_device_reset(struct vfio_pci_device *device);
 
-void vfio_pci_dma_map(struct vfio_pci_device *device,
-		      struct vfio_dma_region *region);
-void vfio_pci_dma_unmap(struct vfio_pci_device *device,
-			struct vfio_dma_region *region);
+int __vfio_pci_dma_map(struct vfio_pci_device *device,
+		       struct vfio_dma_region *region);
+int __vfio_pci_dma_unmap(struct vfio_pci_device *device,
+			 struct vfio_dma_region *region,
+			 u64 *unmapped);
+int __vfio_pci_dma_unmap_all(struct vfio_pci_device *device, u64 *unmapped);
+
+static inline void vfio_pci_dma_map(struct vfio_pci_device *device,
+				    struct vfio_dma_region *region)
+{
+	VFIO_ASSERT_EQ(__vfio_pci_dma_map(device, region), 0);
+}
+
+static inline void vfio_pci_dma_unmap(struct vfio_pci_device *device,
+				      struct vfio_dma_region *region)
+{
+	VFIO_ASSERT_EQ(__vfio_pci_dma_unmap(device, region, NULL), 0);
+}
+
+static inline void vfio_pci_dma_unmap_all(struct vfio_pci_device *device)
+{
+	VFIO_ASSERT_EQ(__vfio_pci_dma_unmap_all(device, NULL), 0);
+}
 
 void vfio_pci_config_access(struct vfio_pci_device *device, bool write,
 			    size_t config, size_t size, void *data);
diff --git a/tools/testing/selftests/vfio/lib/vfio_pci_device.c b/tools/testing/selftests/vfio/lib/vfio_pci_device.c
index 0921b2451ba5..a381fd253aa7 100644
--- a/tools/testing/selftests/vfio/lib/vfio_pci_device.c
+++ b/tools/testing/selftests/vfio/lib/vfio_pci_device.c
@@ -2,6 +2,7 @@
 #include <dirent.h>
 #include <fcntl.h>
 #include <libgen.h>
+#include <stdint.h>
 #include <stdlib.h>
 #include <string.h>
 #include <unistd.h>
@@ -141,7 +142,7 @@ static void vfio_pci_irq_get(struct vfio_pci_device *device, u32 index,
 	ioctl_assert(device->fd, VFIO_DEVICE_GET_IRQ_INFO, irq_info);
 }
 
-static void vfio_iommu_dma_map(struct vfio_pci_device *device,
+static int vfio_iommu_dma_map(struct vfio_pci_device *device,
 			       struct vfio_dma_region *region)
 {
 	struct vfio_iommu_type1_dma_map args = {
@@ -152,10 +153,13 @@ static void vfio_iommu_dma_map(struct vfio_pci_device *device,
 		.size = region->size,
 	};
 
-	ioctl_assert(device->container_fd, VFIO_IOMMU_MAP_DMA, &args);
+	if (ioctl(device->container_fd, VFIO_IOMMU_MAP_DMA, &args))
+		return -errno;
+
+	return 0;
 }
 
-static void iommufd_dma_map(struct vfio_pci_device *device,
+static int iommufd_dma_map(struct vfio_pci_device *device,
 			    struct vfio_dma_region *region)
 {
 	struct iommu_ioas_map args = {
@@ -169,54 +173,108 @@ static void iommufd_dma_map(struct vfio_pci_device *device,
 		.ioas_id = device->ioas_id,
 	};
 
-	ioctl_assert(device->iommufd, IOMMU_IOAS_MAP, &args);
+	if (ioctl(device->iommufd, IOMMU_IOAS_MAP, &args))
+		return -errno;
+
+	return 0;
 }
 
-void vfio_pci_dma_map(struct vfio_pci_device *device,
+int __vfio_pci_dma_map(struct vfio_pci_device *device,
 		      struct vfio_dma_region *region)
 {
+	int ret;
+
 	if (device->iommufd)
-		iommufd_dma_map(device, region);
+		ret = iommufd_dma_map(device, region);
 	else
-		vfio_iommu_dma_map(device, region);
+		ret = vfio_iommu_dma_map(device, region);
+
+	if (ret)
+		return ret;
 
 	list_add(&region->link, &device->dma_regions);
+
+	return 0;
 }
 
-static void vfio_iommu_dma_unmap(struct vfio_pci_device *device,
-				 struct vfio_dma_region *region)
+static int vfio_iommu_dma_unmap(int fd, u64 iova, u64 size, u32 flags,
+				u64 *unmapped)
 {
 	struct vfio_iommu_type1_dma_unmap args = {
 		.argsz = sizeof(args),
-		.iova = region->iova,
-		.size = region->size,
+		.iova = iova,
+		.size = size,
+		.flags = flags,
 	};
 
-	ioctl_assert(device->container_fd, VFIO_IOMMU_UNMAP_DMA, &args);
+	if (ioctl(fd, VFIO_IOMMU_UNMAP_DMA, &args))
+		return -errno;
+
+	if (unmapped)
+		*unmapped = args.size;
+
+	return 0;
 }
 
-static void iommufd_dma_unmap(struct vfio_pci_device *device,
-			      struct vfio_dma_region *region)
+static int iommufd_dma_unmap(int fd, u64 iova, u64 length, u32 ioas_id,
+			     u64 *unmapped)
 {
 	struct iommu_ioas_unmap args = {
 		.size = sizeof(args),
-		.iova = region->iova,
-		.length = region->size,
-		.ioas_id = device->ioas_id,
+		.iova = iova,
+		.length = length,
+		.ioas_id = ioas_id,
 	};
 
-	ioctl_assert(device->iommufd, IOMMU_IOAS_UNMAP, &args);
+	if (ioctl(fd, IOMMU_IOAS_UNMAP, &args))
+		return -errno;
+
+	if (unmapped)
+		*unmapped = args.length;
+
+	return 0;
 }
 
-void vfio_pci_dma_unmap(struct vfio_pci_device *device,
-			struct vfio_dma_region *region)
+int __vfio_pci_dma_unmap(struct vfio_pci_device *device,
+			 struct vfio_dma_region *region, u64 *unmapped)
 {
+	int ret;
+
+	if (device->iommufd)
+		ret = iommufd_dma_unmap(device->iommufd, region->iova,
+					region->size, device->ioas_id,
+					unmapped);
+	else
+		ret = vfio_iommu_dma_unmap(device->container_fd, region->iova,
+					   region->size, 0, unmapped);
+
+	if (ret)
+		return ret;
+
+	list_del_init(&region->link);
+
+	return 0;
+}
+
+int __vfio_pci_dma_unmap_all(struct vfio_pci_device *device, u64 *unmapped)
+{
+	int ret;
+	struct vfio_dma_region *curr, *next;
+
 	if (device->iommufd)
-		iommufd_dma_unmap(device, region);
+		ret = iommufd_dma_unmap(device->iommufd, 0, UINT64_MAX,
+					device->ioas_id, unmapped);
 	else
-		vfio_iommu_dma_unmap(device, region);
+		ret = vfio_iommu_dma_unmap(device->container_fd, 0, 0,
+					   VFIO_DMA_UNMAP_FLAG_ALL, unmapped);
+
+	if (ret)
+		return ret;
+
+	list_for_each_entry_safe(curr, next, &device->dma_regions, link)
+		list_del_init(&curr->link);
 
-	list_del(&region->link);
+	return 0;
 }
 
 static void vfio_pci_region_get(struct vfio_pci_device *device, int index,
diff --git a/tools/testing/selftests/vfio/vfio_dma_mapping_test.c b/tools/testing/selftests/vfio/vfio_dma_mapping_test.c
index ab19c54a774d..4f1ea79a200c 100644
--- a/tools/testing/selftests/vfio/vfio_dma_mapping_test.c
+++ b/tools/testing/selftests/vfio/vfio_dma_mapping_test.c
@@ -112,6 +112,8 @@ FIXTURE_VARIANT_ADD_ALL_IOMMU_MODES(anonymous, 0, 0);
 FIXTURE_VARIANT_ADD_ALL_IOMMU_MODES(anonymous_hugetlb_2mb, SZ_2M, MAP_HUGETLB | MAP_HUGE_2MB);
 FIXTURE_VARIANT_ADD_ALL_IOMMU_MODES(anonymous_hugetlb_1gb, SZ_1G, MAP_HUGETLB | MAP_HUGE_1GB);
 
+#undef FIXTURE_VARIANT_ADD_IOMMU_MODE
+
 FIXTURE_SETUP(vfio_dma_mapping_test)
 {
 	self->device = vfio_pci_device_init(device_bdf, variant->iommu_mode);
@@ -129,6 +131,7 @@ TEST_F(vfio_dma_mapping_test, dma_map_unmap)
 	struct vfio_dma_region region;
 	struct iommu_mapping mapping;
 	u64 mapping_size = size;
+	u64 unmapped;
 	int rc;
 
 	region.vaddr = mmap(NULL, size, PROT_READ | PROT_WRITE, flags, -1, 0);
@@ -184,7 +187,9 @@ TEST_F(vfio_dma_mapping_test, dma_map_unmap)
 	}
 
 unmap:
-	vfio_pci_dma_unmap(self->device, &region);
+	rc = __vfio_pci_dma_unmap(self->device, &region, &unmapped);
+	ASSERT_EQ(rc, 0);
+	ASSERT_EQ(unmapped, region.size);
 	printf("Unmapped IOVA 0x%lx\n", region.iova);
 	ASSERT_EQ(INVALID_IOVA, __to_iova(self->device, region.vaddr));
 	ASSERT_NE(0, iommu_mapping_get(device_bdf, region.iova, &mapping));
@@ -192,6 +197,94 @@ unmap:
 	ASSERT_TRUE(!munmap(region.vaddr, size));
 }
 
+FIXTURE(vfio_dma_map_limit_test) {
+	struct vfio_pci_device *device;
+	struct vfio_dma_region region;
+	size_t mmap_size;
+};
+
+FIXTURE_VARIANT(vfio_dma_map_limit_test) {
+	const char *iommu_mode;
+};
+
+#define FIXTURE_VARIANT_ADD_IOMMU_MODE(_iommu_mode)			       \
+FIXTURE_VARIANT_ADD(vfio_dma_map_limit_test, _iommu_mode) {		       \
+	.iommu_mode = #_iommu_mode,					       \
+}
+
+FIXTURE_VARIANT_ADD_ALL_IOMMU_MODES();
+
+#undef FIXTURE_VARIANT_ADD_IOMMU_MODE
+
+FIXTURE_SETUP(vfio_dma_map_limit_test)
+{
+	struct vfio_dma_region *region = &self->region;
+	u64 region_size = getpagesize();
+
+	/*
+	 * Over-allocate mmap by double the size to provide enough backing vaddr
+	 * for overflow tests
+	 */
+	self->mmap_size = 2 * region_size;
+
+	self->device = vfio_pci_device_init(device_bdf, variant->iommu_mode);
+	region->vaddr = mmap(NULL, self->mmap_size, PROT_READ | PROT_WRITE,
+			     MAP_ANONYMOUS | MAP_PRIVATE, -1, 0);
+	ASSERT_NE(region->vaddr, MAP_FAILED);
+
+	/* One page prior to the end of address space */
+	region->iova = ~(iova_t)0 & ~(region_size - 1);
+	region->size = region_size;
+}
+
+FIXTURE_TEARDOWN(vfio_dma_map_limit_test)
+{
+	vfio_pci_device_cleanup(self->device);
+	ASSERT_EQ(munmap(self->region.vaddr, self->mmap_size), 0);
+}
+
+TEST_F(vfio_dma_map_limit_test, unmap_range)
+{
+	struct vfio_dma_region *region = &self->region;
+	u64 unmapped;
+	int rc;
+
+	vfio_pci_dma_map(self->device, region);
+	ASSERT_EQ(region->iova, to_iova(self->device, region->vaddr));
+
+	rc = __vfio_pci_dma_unmap(self->device, region, &unmapped);
+	ASSERT_EQ(rc, 0);
+	ASSERT_EQ(unmapped, region->size);
+}
+
+TEST_F(vfio_dma_map_limit_test, unmap_all)
+{
+	struct vfio_dma_region *region = &self->region;
+	u64 unmapped;
+	int rc;
+
+	vfio_pci_dma_map(self->device, region);
+	ASSERT_EQ(region->iova, to_iova(self->device, region->vaddr));
+
+	rc = __vfio_pci_dma_unmap_all(self->device, &unmapped);
+	ASSERT_EQ(rc, 0);
+	ASSERT_EQ(unmapped, region->size);
+}
+
+TEST_F(vfio_dma_map_limit_test, overflow)
+{
+	struct vfio_dma_region *region = &self->region;
+	int rc;
+
+	region->size = self->mmap_size;
+
+	rc = __vfio_pci_dma_map(self->device, region);
+	ASSERT_EQ(rc, -EOVERFLOW);
+
+	rc = __vfio_pci_dma_unmap(self->device, region, NULL);
+	ASSERT_EQ(rc, -EOVERFLOW);
+}
+
 int main(int argc, char *argv[])
 {
 	device_bdf = vfio_selftests_get_bdf(&argc, argv);