Skip to contentSkip to conversation

Fix BUG_ON in vfio_iommu_group_notifier()

PATCH Superseded Lu Baolu 5 participants 56 messages Jan 4, 2022 – Feb 23, 2022

Subsystems (10)
KVM PCI Driver core IOMMU +6 more subsystems

v5 of 8 Superseded Latest is v8 →

lore
Share & tools
lore

Lu Baolu

author
#0 lore raw

v5 cover Fix BUG_ON in vfio_iommu_group_notifier()

Hi folks,

The iommu group is the minimal isolation boundary for DMA. Devices in a group can access each other's MMIO registers via peer to peer DMA and also need share the same I/O address space.

Once the I/O address space is assigned to user control it is no longer available to the dma_map* API, which effectively makes the DMA API non-working.

Second, userspace can use DMA initiated by a device that it controls to access the MMIO spaces of other devices in the group. This allows userspace to indirectly attack any kernel owned device and it's driver.

Therefore groups must either be entirely under kernel control or userspace control, never a mixture. Unfortunately some systems have problems with the granularity of groups and there are a couple of important exceptions:

- pci_stub allows the admin to block driver binding on a device and make it permanently shared with userspace. Since PCI stub does not do DMA it is safe, however the admin must understand that using pci_stub allows userspace to attack whatever device it was bound it.

- PCI bridges are sometimes included in groups. Typically PCI bridges do not use DMA, and generally do not have MMIO regions.

Generally any device that does not have any MMIO registers is a possible candidate for an exception.

Currently vfio adopts a workaround to detect violations of the above restrictions by monitoring the driver core BOUND event, and hardwiring the above exceptions. Since there is no way for vfio to reject driver binding at this point, BUG_ON() is triggered if a violation is captured (kernel driver BOUND event on a group which already has some devices assigned to userspace). Aside from the bad user experience this opens a way for root userspace to crash the kernel, even in high integrity configurations, by manipulating the module binding and triggering the BUG_ON.

This series solves this problem by making the user/kernel ownership a core concept at the IOMMU layer. The driver core enforces kernel ownership while drivers are bound and violations now result in a error codes during probe, not BUG_ON failures.

Patch partitions:
[PATCH 1-7]: Detect DMA ownership conflicts during driver binding; [PATCH 8-10]: Add security context management for assigned devices; [PATCH 11-14]: Various cleanups.

This is also part one of three initial series for IOMMUFD:
* Move IOMMU Group security into the iommu layer
- Generic IOMMUFD implementation
- VFIO ability to consume IOMMUFD

Change log:
v1: initial post
- https://lore.kernel.org/linux-iommu/[email protected]/

v2:
- https://lore.kernel.org/linux-iommu/[email protected]/

- Move kernel dma ownership auto-claiming from driver core to bus

    callback. [Greg/Christoph/Robin/Jason]
    https://lore.kernel.org/linux-iommu/[email protected]/T/#m153706912b770682cb12e3c28f57e171aa1f9d0c

- Code and interface refactoring for iommu_set/release_dma_owner()

    interfaces. [Jason]
    https://lore.kernel.org/linux-iommu/[email protected]/T/#mea70ed8e4e3665aedf32a5a0a7db095bf680325e

- [NEW]Add new iommu_attach/detach_device_shared() interfaces for multiple devices group. [Robin/Jason]
https://lore.kernel.org/linux-iommu/[email protected]/T/#mea70ed8e4e3665aedf32a5a0a7db095bf680325e

- [NEW]Use iommu_attach/detach_device_shared() in drm/tegra drivers.

- Refactoring and description refinement.

v3:
- https://lore.kernel.org/linux-iommu/[email protected]/

- Rename bus_type::dma_unconfigure to bus_type::dma_cleanup. [Greg]

    https://lore.kernel.org/linux-iommu/[email protected]/T/#m6711e041e47cb0cbe3964fad0a3466f5ae4b3b9b

- Avoid _platform_dma_configure for platform_bus_type::dma_configure.

    [Greg]
    https://lore.kernel.org/linux-iommu/[email protected]/T/#m43fc46286611aa56a5c0eeaad99d539e5519f3f6

- Patch "0012-iommu-Add-iommu_at-de-tach_device_shared-for-mult.patch" and "0018-drm-tegra-Use-the-iommu-dma_owner-mechanism.patch" have been tested by Dmitry Osipenko <[email protected]>.

v4:
- https://lore.kernel.org/linux-iommu/[email protected]/
- Remove unnecessary tegra->domain chech in the tegra patch. (Jason)
- Remove DMA_OWNER_NONE. (Joerg)
- Change refcount to unsigned int. (Christoph)
- Move mutex lock into group set_dma_owner functions. (Christoph)
- Add kernel doc for iommu_attach/detach_domain_shared(). (Christoph)
- Move dma auto-claim into driver core. (Jason/Christoph)

v5:
- Move kernel dma ownership auto-claiming from driver core to bus callback. (Greg)
- Refactor the iommu interfaces to make them more specific. (Jason/Robin)
- Simplify the dma ownership implementation by removing the owner type. (Jason)
- Commit message refactoring for PCI drivers. (Bjorn)
- Move iommu_attach/detach_device() improvement patches into another series as there are a lot of code refactoring and cleanup staffs in various device drivers.

This is based on next branch of linux-iommu tree:
https://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu.git
and also available on github:
https://github.com/LuBaolu/intel-iommu/commits/iommu-dma-ownership-v5

Best regards,
baolu

Jason Gunthorpe (1):
vfio: Delete the unbound_list

Lu Baolu (13):
iommu: Add dma ownership management interfaces driver core: Add dma_cleanup callback in bus_type
amba: Stop sharing platform_dma_configure() driver core: platform: Add driver dma ownership management
amba: Add driver dma ownership management
bus: fsl-mc: Add driver dma ownership management
PCI: Add driver dma ownership management
PCI: pci_stub: Suppress kernel DMA ownership auto-claiming
PCI: portdrv: Suppress kernel DMA ownership auto-claiming
vfio: Set DMA ownership for VFIO devices
vfio: Remove use of vfio_group_viable()
vfio: Remove iommu group notifier
iommu: Remove iommu group changes notifier

 include/linux/amba/bus.h              |   1 +
 include/linux/device/bus.h            |   3 +
 include/linux/fsl/mc.h                |   5 +
 include/linux/iommu.h                 |  54 +++---
 include/linux/pci.h                   |   5 +
 include/linux/platform_device.h       |   3 +-
 drivers/amba/bus.c                    |  39 +++-
 drivers/base/dd.c                     |   5 +
 drivers/base/platform.c               |  23 ++-
 drivers/bus/fsl-mc/fsl-mc-bus.c       |  26 ++-
 drivers/iommu/iommu.c                 | 236 +++++++++++++++++--------
 drivers/pci/pci-driver.c              |  21 +++
 drivers/pci/pci-stub.c                |   1 +
 drivers/pci/pcie/portdrv_pci.c        |   2 +
 drivers/vfio/fsl-mc/vfio_fsl_mc.c     |   1 +
 drivers/vfio/pci/vfio_pci.c           |   1 +
 drivers/vfio/platform/vfio_amba.c     |   1 +
 drivers/vfio/platform/vfio_platform.c |   1 +
 drivers/vfio/vfio.c                   | 245 ++------------------------
 19 files changed, 335 insertions(+), 338 deletions(-)
Signature · 1 line
2.25.1
author
#1 lore raw

PATCH v5 1/14 iommu: Add dma ownership management interfaces

Multiple devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture.

This adds dma ownership management in iommu core and exposes several interfaces for the device drivers and the device userspace assignment framework (i.e. vfio), so that any conflict between user and kernel controlled DMA could be detected at the beginning.

The device driver oriented interfaces are,

	int iommu_device_use_dma_api(struct device *dev);
	void iommu_device_unuse_dma_api(struct device *dev);

Devices under kernel drivers control must call iommu_device_use_dma_api() before driver probes. The driver binding process must be aborted if it returns failure.

The vfio oriented interfaces are,

	int iommu_group_set_dma_owner(struct iommu_group *group,
				      void *owner);
	void iommu_group_release_dma_owner(struct iommu_group *group);
	bool iommu_group_dma_owner_claimed(struct iommu_group *group);

The device userspace assignment must be disallowed if the set dma owner interface returns failure.

---
 include/linux/iommu.h |  31 ++++++++
 drivers/iommu/iommu.c | 161 +++++++++++++++++++++++++++++++++++++++++-
 2 files changed, 189 insertions(+), 3 deletions(-)
include/linux/iommu.h +31 -0
index de0c57a567c8..568f285468cf 100644
@@ -682,6 +682,13 @@ struct iommu_sva *iommu_sva_bind_device(struct device *dev,
 void iommu_sva_unbind_device(struct iommu_sva *handle);
 u32 iommu_sva_get_pasid(struct iommu_sva *handle);
 
+int iommu_device_use_dma_api(struct device *dev);
+void iommu_device_unuse_dma_api(struct device *dev);
+
+int iommu_group_set_dma_owner(struct iommu_group *group, void *owner);
+void iommu_group_release_dma_owner(struct iommu_group *group);
+bool iommu_group_dma_owner_claimed(struct iommu_group *group);
+
 #else /* CONFIG_IOMMU_API */
 
 struct iommu_ops {};
@@ -1082,6 +1089,30 @@ static inline struct iommu_fwspec *dev_iommu_fwspec_get(struct device *dev)
 {
 	return NULL;
 }
+
+static inline int iommu_device_use_dma_api(struct device *dev)
+{
+	return 0;
+}
+
+static inline void iommu_device_unuse_dma_api(struct device *dev)
+{
+}
+
+static inline int
+iommu_group_set_dma_owner(struct iommu_group *group, void *owner)
+{
+	return -ENODEV;
+}
+
+static inline void iommu_group_release_dma_owner(struct iommu_group *group)
+{
+}
+
+static inline bool iommu_group_dma_owner_claimed(struct iommu_group *group)
+{
+	return false;
+}
 #endif /* CONFIG_IOMMU_API */
 
 /**
drivers/iommu/iommu.c +158 -3

188 lines not on this page · raw message

Signature · 1 line
2.25.1
#2 lore raw

Re: [PATCH v5 01/14] iommu: Add dma ownership management interfaces

Quoted 122 lines · Lu Baolu

Multiple devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture.

This adds dma ownership management in iommu core and exposes several interfaces for the device drivers and the device userspace assignment framework (i.e. vfio), so that any conflict between user and kernel controlled DMA could be detected at the beginning.

The device driver oriented interfaces are,

	int iommu_device_use_dma_api(struct device *dev);
	void iommu_device_unuse_dma_api(struct device *dev);

Devices under kernel drivers control must call iommu_device_use_dma_api() before driver probes. The driver binding process must be aborted if it returns failure.

The vfio oriented interfaces are,

	int iommu_group_set_dma_owner(struct iommu_group *group,
				      void *owner);
	void iommu_group_release_dma_owner(struct iommu_group *group);
	bool iommu_group_dma_owner_claimed(struct iommu_group *group);

The device userspace assignment must be disallowed if the set dma owner interface returns failure.

---
 include/linux/iommu.h |  31 ++++++++
 drivers/iommu/iommu.c | 161 +++++++++++++++++++++++++++++++++++++++++-
 2 files changed, 189 insertions(+), 3 deletions(-)
include/linux/iommu.h +31 -0
index de0c57a567c8..568f285468cf 100644
@@ -682,6 +682,13 @@ struct iommu_sva *iommu_sva_bind_device(struct device *dev,
 void iommu_sva_unbind_device(struct iommu_sva *handle);
 u32 iommu_sva_get_pasid(struct iommu_sva *handle);
 
+int iommu_device_use_dma_api(struct device *dev);
+void iommu_device_unuse_dma_api(struct device *dev);
+
+int iommu_group_set_dma_owner(struct iommu_group *group, void *owner);
+void iommu_group_release_dma_owner(struct iommu_group *group);
+bool iommu_group_dma_owner_claimed(struct iommu_group *group);
+
 #else /* CONFIG_IOMMU_API */
 
 struct iommu_ops {};
@@ -1082,6 +1089,30 @@ static inline struct iommu_fwspec *dev_iommu_fwspec_get(struct device *dev)
 {
 	return NULL;
 }
+
+static inline int iommu_device_use_dma_api(struct device *dev)
+{
+	return 0;
+}
+
+static inline void iommu_device_unuse_dma_api(struct device *dev)
+{
+}
+
+static inline int
+iommu_group_set_dma_owner(struct iommu_group *group, void *owner)
+{
+	return -ENODEV;
+}
+
+static inline void iommu_group_release_dma_owner(struct iommu_group *group)
+{
+}
+
+static inline bool iommu_group_dma_owner_claimed(struct iommu_group *group)
+{
+	return false;
+}
 #endif /* CONFIG_IOMMU_API */
 
 /**
drivers/iommu/iommu.c +9 -2
index 8b86406b7162..ff0c8c1ad5af 100644
@@ -48,6 +48,8 @@ struct iommu_group {
 	struct iommu_domain *default_domain;
 	struct iommu_domain *domain;
 	struct list_head entry;
+	unsigned int owner_cnt;
+	void *owner;
 };
 
 struct group_device {
@@ -289,7 +291,12 @@ int iommu_probe_device(struct device *dev)
 	mutex_lock(&group->mutex);
 	iommu_alloc_default_domain(group, dev);
 
-	if (group->default_domain) {
+	/*
+	 * If device joined an existing group which has been claimed
+	 * for none kernel DMA purpose, avoid attaching the default
+	 * domain.
+	 */
+	if (group->default_domain && !group->owner) {
 		ret = __iommu_attach_device(group->default_domain, dev);
 		if (ret) {
 			mutex_unlock(&group->mutex);
@@ -2320,7 +2327,7 @@ static int __iommu_attach_group(struct iommu_domain *domain,
 {
 	int ret;
 
-	if (group->default_domain && group->domain != group->default_domain)
+	if (group->domain && group->domain != group->default_domain)
 		return -EBUSY;
 
 	ret = __iommu_group_for_each_dev(group, domain,

drivers/iommu/iommu.c static void __iommu_detach_group(struct iommu_domain *domain, · line 2370

Quoted 8 lines
drivers/iommu/iommu.c +4 -1
@@ -2357,7 +2364,11 @@ static void __iommu_detach_group(struct iommu_domain *domain,
 {
 	int ret;
 
-	if (!group->default_domain) {
+	/*
+	 * If group has been claimed for none kernel DMA purpose, avoid
+	 * re-attaching the default domain.
+	 */

none kernel reads odd. But maybe drop that and just say 'claimed already' ala:

	/*
	 * If the group has been claimed already, do not re-attach the default
	 * domain.
	 */
#3 lore raw

Re: [PATCH v5 01/14] iommu: Add dma ownership management interfaces

Quoted 4 lines · Christoph Hellwig
Quoted 3 lines · Lu Baolu

Multiple devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture.

I guess the reason is that if a group contained a mixture, userspace could attack the kernel by programming a device to DMA to a device owned by the kernel?

Quoted 4 lines
Quoted 4 lines

This adds dma ownership management in iommu core and exposes several interfaces for the device drivers and the device userspace assignment framework (i.e. vfio), so that any conflict between user and kernel controlled DMA could be detected at the beginning.

Maybe I'm missing the point because I don't know what "conflict between user and kernel controlled DMA" is. Are you talking about both userspace and the kernel programming the same device to do DMA?

Quoted 4 lines
Quoted 4 lines

The device driver oriented interfaces are,

	int iommu_device_use_dma_api(struct device *dev);
	void iommu_device_unuse_dma_api(struct device *dev);

Nit, do we care whether it uses the actual DMA API? Or is it just that iommu_device_use_dma_api() tells us the driver may program the device to do DMA?

Quoted 3 lines
Quoted 3 lines

Devices under kernel drivers control must call iommu_device_use_dma_api() before driver probes. The driver binding process must be aborted if it returns failure.

"Devices" don't call functions. Drivers do, or in this case, it looks like the bus DMA code (platform, amba, fsl, pci, etc).

These functions are EXPORT_SYMBOL_GPL(), but it looks like all the callers are built-in, so maybe the export is unnecessary?

You use "iommu"/"IOMMU" and "dma"/"DMA" interchangeably above. Would be easier to read if you picked one.

Quoted 9 lines
Quoted 9 lines

The vfio oriented interfaces are,

	int iommu_group_set_dma_owner(struct iommu_group *group,
				      void *owner);
	void iommu_group_release_dma_owner(struct iommu_group *group);
	bool iommu_group_dma_owner_claimed(struct iommu_group *group);

The device userspace assignment must be disallowed if the set dma owner interface returns failure.

Can you connect this back to the "never a mixture" from the beginning? If all you cared about was prevent an IOMMU group from containing devices with a mixture of kernel drivers and userspace drivers, I assume you could do that without iommu_device_use_dma_api(). So is this a way to *allow* a mixture under certain restricted conditions?

Another nit below.

Quoted 74 lines
Quoted 74 lines
---
 include/linux/iommu.h |  31 ++++++++
 drivers/iommu/iommu.c | 161 +++++++++++++++++++++++++++++++++++++++++-
 2 files changed, 189 insertions(+), 3 deletions(-)
include/linux/iommu.h +31 -0
index de0c57a567c8..568f285468cf 100644
@@ -682,6 +682,13 @@ struct iommu_sva *iommu_sva_bind_device(struct device *dev,
 void iommu_sva_unbind_device(struct iommu_sva *handle);
 u32 iommu_sva_get_pasid(struct iommu_sva *handle);
 
+int iommu_device_use_dma_api(struct device *dev);
+void iommu_device_unuse_dma_api(struct device *dev);
+
+int iommu_group_set_dma_owner(struct iommu_group *group, void *owner);
+void iommu_group_release_dma_owner(struct iommu_group *group);
+bool iommu_group_dma_owner_claimed(struct iommu_group *group);
+
 #else /* CONFIG_IOMMU_API */
 
 struct iommu_ops {};
@@ -1082,6 +1089,30 @@ static inline struct iommu_fwspec *dev_iommu_fwspec_get(struct device *dev)
 {
 	return NULL;
 }
+
+static inline int iommu_device_use_dma_api(struct device *dev)
+{
+	return 0;
+}
+
+static inline void iommu_device_unuse_dma_api(struct device *dev)
+{
+}
+
+static inline int
+iommu_group_set_dma_owner(struct iommu_group *group, void *owner)
+{
+	return -ENODEV;
+}
+
+static inline void iommu_group_release_dma_owner(struct iommu_group *group)
+{
+}
+
+static inline bool iommu_group_dma_owner_claimed(struct iommu_group *group)
+{
+	return false;
+}
 #endif /* CONFIG_IOMMU_API */
 
 /**
drivers/iommu/iommu.c +2 -0
index 8b86406b7162..ff0c8c1ad5af 100644
@@ -48,6 +48,8 @@ struct iommu_group {
 	struct iommu_domain *default_domain;
 	struct iommu_domain *domain;
 	struct list_head entry;
+	unsigned int owner_cnt;
+	void *owner;
 };
 
 struct group_device {
@@ -289,7 +291,12 @@ int iommu_probe_device(struct device *dev)
 	mutex_lock(&group->mutex);
 	iommu_alloc_default_domain(group, dev);
 
Quoted 5 lines

drivers/iommu/iommu.c int iommu_probe_device(struct device *dev) · line 297

Quoted 5 lines
drivers/iommu/iommu.c +4 -1
-	if (group->default_domain) {
+	/*
+	 * If device joined an existing group which has been claimed
+	 * for none kernel DMA purpose, avoid attaching the default
+	 * domain.

AOL: another "none kernel DMA purpose" that doesn't read well. Is this supposed to be "non-kernel"? What does "claimed for non-kernel DMA purpose" mean? What interface does that?

Quoted 31 lines
Quoted 15 lines
drivers/iommu/iommu.c +3 -1
+	 */
+	if (group->default_domain && !group->owner) {
 		ret = __iommu_attach_device(group->default_domain, dev);
 		if (ret) {
 			mutex_unlock(&group->mutex);
@@ -2320,7 +2327,7 @@ static int __iommu_attach_group(struct iommu_domain *domain,
 {
 	int ret;
 
-	if (group->default_domain && group->domain != group->default_domain)
+	if (group->domain && group->domain != group->default_domain)
 		return -EBUSY;
 
 	ret = __iommu_group_for_each_dev(group, domain,

drivers/iommu/iommu.c static void __iommu_detach_group(struct iommu_domain *domain, · line 2370

Quoted 8 lines
drivers/iommu/iommu.c +4 -1
@@ -2357,7 +2364,11 @@ static void __iommu_detach_group(struct iommu_domain *domain,
 {
 	int ret;
 
-	if (!group->default_domain) {
+	/*
+	 * If group has been claimed for none kernel DMA purpose, avoid
+	 * re-attaching the default domain.
+	 */

none kernel reads odd. But maybe drop that and just say 'claimed already' ala:

	/*
	 * If the group has been claimed already, do not re-attach the default
	 * domain.
	 */
#4 lore raw

Re: [PATCH v5 01/14] iommu: Add dma ownership management interfaces

Quoted 9 lines · Bjorn Helgaas
Quoted 4 lines · Christoph Hellwig
Quoted 3 lines · Lu Baolu

Multiple devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture.

I guess the reason is that if a group contained a mixture, userspace could attack the kernel by programming a device to DMA to a device owned by the kernel?

There are several of reasons, including what you guess, but for the design of the series now we can just focus on the group's domain.

If the kernel is using a device then the kernel driver uses the DMA API and the group's domain must always point to the default domain so long as a DMA API user exists. Hopefully it is clear to understand

Quoted 8 lines
Quoted 4 lines
Quoted 4 lines

The device driver oriented interfaces are,

	int iommu_device_use_dma_api(struct device *dev);
	void iommu_device_unuse_dma_api(struct device *dev);

Nit, do we care whether it uses the actual DMA API? Or is it just that iommu_device_use_dma_api() tells us the driver may program the device to do DMA?

As the main purpose, yes this is all about the DMA API because it asserts the group domain is the DMA API's domain.

There is a secondary purpose that has to do with the user/kernel attack you mentioned above. Maintaining the DMA API domain also prevents VFIO from allowing userspace to operate any device in the group which blocks P2P attacks to MMIO of other devices.

This is why, even if the driver doesn't use DMA, it should still do a iommu_device_use_dma_api(), except in the special cases where we don't care about P2P attacks (eg pci-stub, bridges, etc).

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

The vfio oriented interfaces are,

	int iommu_group_set_dma_owner(struct iommu_group *group,
				      void *owner);
	void iommu_group_release_dma_owner(struct iommu_group *group);
	bool iommu_group_dma_owner_claimed(struct iommu_group *group);
Quoted 8 lines
Quoted 2 lines
Quoted 2 lines

The device userspace assignment must be disallowed if the set dma owner interface returns failure.

Can you connect this back to the "never a mixture" from the beginning? If all you cared about was prevent an IOMMU group from containing devices with a mixture of kernel drivers and userspace drivers, I assume you could do that without iommu_device_use_dma_api(). So is this a way to *allow* a mixture under certain restricted conditions?

It is not about user/kernel, it is about arbitrating the shared group->domain against multiple different requests to set it to something else.

Lu, Given that the word 'user' was deleted from the API entirely it makes sense to reword these commit messages to focus less on user vs kernel and more on ownership of the domain pointer.

Jason
_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

author
#5 lore raw

Re: [PATCH v5 01/14] iommu: Add dma ownership management interfaces

Quoted 14 lines · Jason Gunthorpe
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

The vfio oriented interfaces are,

	int iommu_group_set_dma_owner(struct iommu_group *group,
				      void *owner);
	void iommu_group_release_dma_owner(struct iommu_group *group);
	bool iommu_group_dma_owner_claimed(struct iommu_group *group);
Quoted 7 lines
Quoted 2 lines
Quoted 2 lines

The device userspace assignment must be disallowed if the set dma owner interface returns failure.

Can you connect this back to the "never a mixture" from the beginning? If all you cared about was prevent an IOMMU group from containing devices with a mixture of kernel drivers and userspace drivers, I assume you could do that without iommu_device_use_dma_api(). So is this a way to*allow* a mixture under certain restricted conditions?

Quoted 7 lines

It is not about user/kernel, it is about arbitrating the shared group->domain against multiple different requests to set it to something else.

Lu, Given that the word 'user' was deleted from the API entirely it makes sense to reword these commit messages to focus less on user vs kernel and more on ownership of the domain pointer.

Sure. Will do it.

Best regards,
baolu

author
#6 lore raw

Re: [PATCH v5 01/14] iommu: Add dma ownership management interfaces

In reply to Jason Gunthorpe
Quoted 10 lines · Jason Gunthorpe
Quoted 7 lines
Quoted 4 lines
Quoted 4 lines

The device driver oriented interfaces are,

	int iommu_device_use_dma_api(struct device *dev);
	void iommu_device_unuse_dma_api(struct device *dev);

Nit, do we care whether it uses the actual DMA API? Or is it just that iommu_device_use_dma_api() tells us the driver may program the device to do DMA?

As the main purpose, yes this is all about the DMA API because it asserts the group domain is the DMA API's domain.

Quoted 9 lines

There is a secondary purpose that has to do with the user/kernel attack you mentioned above. Maintaining the DMA API domain also prevents VFIO from allowing userspace to operate any device in the group which blocks P2P attacks to MMIO of other devices.

This is why, even if the driver doesn't use DMA, it should still do a iommu_device_use_dma_api(), except in the special cases where we don't care about P2P attacks (eg pci-stub, bridges, etc).

By the way, use_dma_api seems hard to read. How about

iommu_device_use_default_dma()?

Best regards,
baolu

#7 lore raw

Re: [PATCH v5 01/14] iommu: Add dma ownership management interfaces

Quoted 16 lines · Lu Baolu
Quoted 15 lines · Jason Gunthorpe
Quoted 7 lines
Quoted 4 lines
Quoted 4 lines

The device driver oriented interfaces are,

	int iommu_device_use_dma_api(struct device *dev);
	void iommu_device_unuse_dma_api(struct device *dev);

Nit, do we care whether it uses the actual DMA API? Or is it just that iommu_device_use_dma_api() tells us the driver may program the device to do DMA?

As the main purpose, yes this is all about the DMA API because it asserts the group domain is the DMA API's domain.

There is a secondary purpose that has to do with the user/kernel attack you mentioned above. Maintaining the DMA API domain also prevents VFIO from allowing userspace to operate any device in the group which blocks P2P attacks to MMIO of other devices.

Quoted 8 lines
Quoted 4 lines

This is why, even if the driver doesn't use DMA, it should still do a iommu_device_use_dma_api(), except in the special cases where we don't care about P2P attacks (eg pci-stub, bridges, etc).

By the way, use_dma_api seems hard to read. How about

iommu_device_use_default_dma()?

You could just say "use default domain"

IMHO the way the iommu subsystem has its own wonky language is a little troublesome. In the rest of the kernel we call this the DMA API, while the iommu subsystem calls the domain that the DMA API uses the 'default domain' not the 'DMA API' domain.

Still, it is probably better to align to the iommu language - just be sure to put in the function comment that this API 'allows the driver to use the DMA API eg dma_map_sg()'

Jason

author
#8 lore raw

Re: [PATCH v5 01/14] iommu: Add dma ownership management interfaces

Quoted 28 lines · Jason Gunthorpe
Quoted 16 lines · Lu Baolu
Quoted 15 lines · Jason Gunthorpe
Quoted 7 lines
Quoted 4 lines
Quoted 4 lines

The device driver oriented interfaces are,

	int iommu_device_use_dma_api(struct device *dev);
	void iommu_device_unuse_dma_api(struct device *dev);

Nit, do we care whether it uses the actual DMA API? Or is it just that iommu_device_use_dma_api() tells us the driver may program the device to do DMA?

As the main purpose, yes this is all about the DMA API because it asserts the group domain is the DMA API's domain.

There is a secondary purpose that has to do with the user/kernel attack you mentioned above. Maintaining the DMA API domain also prevents VFIO from allowing userspace to operate any device in the group which blocks P2P attacks to MMIO of other devices.

Quoted 8 lines
Quoted 4 lines

This is why, even if the driver doesn't use DMA, it should still do a iommu_device_use_dma_api(), except in the special cases where we don't care about P2P attacks (eg pci-stub, bridges, etc).

By the way, use_dma_api seems hard to read. How about

iommu_device_use_default_dma()?

You could just say "use default domain"

Quoted 8 lines

IMHO the way the iommu subsystem has its own wonky language is a little troublesome. In the rest of the kernel we call this the DMA API, while the iommu subsystem calls the domain that the DMA API uses the 'default domain' not the 'DMA API' domain.

Still, it is probably better to align to the iommu language - just be sure to put in the function comment that this API 'allows the driver to use the DMA API eg dma_map_sg()'

iommu_device_use_default_domain() reads better. And add some comments to link "default domain" with "DMA API". Thanks!

Quoted 3 lines

Jason

Best regards,
baolu

author
#9 lore raw

Re: [PATCH v5 01/14] iommu: Add dma ownership management interfaces

In reply to Bjorn Helgaas

Hi Bjorn,

Quoted 7 lines · Bjorn Helgaas
Quoted 4 lines
Quoted 4 lines

This adds dma ownership management in iommu core and exposes several interfaces for the device drivers and the device userspace assignment framework (i.e. vfio), so that any conflict between user and kernel controlled DMA could be detected at the beginning.

Maybe I'm missing the point because I don't know what "conflict between user and kernel controlled DMA" is. Are you talking about both userspace and the kernel programming the same device to do DMA?

It's about both userspace and kernel programming different devices belonging to a same iommu group to do DMA. For example, PCI device A and B sit in a some iommu group. Userspace programs device A for DMA and a kernel driver programs device B. Userspace may intentionally or unintentionally program device A to access and change device B's MMIO with P2P DMA transactions which cause the kernel driver for device B result in an inconsistent state.

This may not be all, just an example.

Best regards,
baolu

author
#10 lore raw

Re: [PATCH v5 01/14] iommu: Add dma ownership management interfaces

In reply to Bjorn Helgaas
Quoted 8 lines · Bjorn Helgaas
Quoted 3 lines
Quoted 3 lines

Devices under kernel drivers control must call iommu_device_use_dma_api() before driver probes. The driver binding process must be aborted if it returns failure.

"Devices" don't call functions. Drivers do, or in this case, it looks like the bus DMA code (platform, amba, fsl, pci, etc).

These functions are EXPORT_SYMBOL_GPL(), but it looks like all the callers are built-in, so maybe the export is unnecessary?

Yes. If all callers are built-in, we can remove this export.

Quoted 4 lines

You use "iommu"/"IOMMU" and "dma"/"DMA" interchangeably above. Would be easier to read if you picked one.

I should cleanup all these. I also realized that I forgot to cleanup some typos you pointed out in v4:

https://lore.kernel.org/linux-iommu/20211229211626.GA1701512@bhelgaas/

Sorry about it. I will take care of all these in the next version.

Best regards,
baolu

author
#11 lore raw

Re: [PATCH v5 01/14] iommu: Add dma ownership management interfaces

In reply to Bjorn Helgaas
Quoted 5 lines · Bjorn Helgaas
Quoted 5 lines
Quoted 5 lines
  struct group_device {
drivers/iommu/iommu.c +0 -0
@@ -289,7 +291,12 @@ int iommu_probe_device(struct device *dev)
  	mutex_lock(&group->mutex);
  	iommu_alloc_default_domain(group, dev);
 
Quoted 9 lines
Quoted 5 lines

drivers/iommu/iommu.c int iommu_probe_device(struct device *dev) · line 297

Quoted 5 lines
drivers/iommu/iommu.c +4 -1
-	if (group->default_domain) {
+	/*
+	 * If device joined an existing group which has been claimed
+	 * for none kernel DMA purpose, avoid attaching the default
+	 * domain.

AOL: another "none kernel DMA purpose" that doesn't read well. Is this supposed to be "non-kernel"? What does "claimed for non-kernel DMA purpose" mean? What interface does that?

It's hard to read. I will rephrase it like this:

/*
* If device joined an existing group which has been claimed, don't
* attach the default domain.
*/

Best regards,
baolu

author
#12 lore raw

Re: [PATCH v5 01/14] iommu: Add dma ownership management interfaces

In reply to Christoph Hellwig

Hi Christoph,

Quoted 135 lines · Christoph Hellwig
Quoted 122 lines · Lu Baolu

Multiple devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture.

This adds dma ownership management in iommu core and exposes several interfaces for the device drivers and the device userspace assignment framework (i.e. vfio), so that any conflict between user and kernel controlled DMA could be detected at the beginning.

The device driver oriented interfaces are,

	int iommu_device_use_dma_api(struct device *dev);
	void iommu_device_unuse_dma_api(struct device *dev);

Devices under kernel drivers control must call iommu_device_use_dma_api() before driver probes. The driver binding process must be aborted if it returns failure.

The vfio oriented interfaces are,

	int iommu_group_set_dma_owner(struct iommu_group *group,
				      void *owner);
	void iommu_group_release_dma_owner(struct iommu_group *group);
	bool iommu_group_dma_owner_claimed(struct iommu_group *group);

The device userspace assignment must be disallowed if the set dma owner interface returns failure.

---
  include/linux/iommu.h |  31 ++++++++
  drivers/iommu/iommu.c | 161 +++++++++++++++++++++++++++++++++++++++++-
  2 files changed, 189 insertions(+), 3 deletions(-)
include/linux/iommu.h +31 -0
index de0c57a567c8..568f285468cf 100644
@@ -682,6 +682,13 @@ struct iommu_sva *iommu_sva_bind_device(struct device *dev,
  void iommu_sva_unbind_device(struct iommu_sva *handle);
  u32 iommu_sva_get_pasid(struct iommu_sva *handle);
 
+int iommu_device_use_dma_api(struct device *dev);
+void iommu_device_unuse_dma_api(struct device *dev);
+
+int iommu_group_set_dma_owner(struct iommu_group *group, void *owner);
+void iommu_group_release_dma_owner(struct iommu_group *group);
+bool iommu_group_dma_owner_claimed(struct iommu_group *group);
+
  #else /* CONFIG_IOMMU_API */
 
  struct iommu_ops {};
@@ -1082,6 +1089,30 @@ static inline struct iommu_fwspec *dev_iommu_fwspec_get(struct device *dev)
  {
  	return NULL;
  }
+
+static inline int iommu_device_use_dma_api(struct device *dev)
+{
+	return 0;
+}
+
+static inline void iommu_device_unuse_dma_api(struct device *dev)
+{
+}
+
+static inline int
+iommu_group_set_dma_owner(struct iommu_group *group, void *owner)
+{
+	return -ENODEV;
+}
+
+static inline void iommu_group_release_dma_owner(struct iommu_group *group)
+{
+}
+
+static inline bool iommu_group_dma_owner_claimed(struct iommu_group *group)
+{
+	return false;
+}
  #endif /* CONFIG_IOMMU_API */
 
  /**
drivers/iommu/iommu.c +9 -2
index 8b86406b7162..ff0c8c1ad5af 100644
@@ -48,6 +48,8 @@ struct iommu_group {
  	struct iommu_domain *default_domain;
  	struct iommu_domain *domain;
  	struct list_head entry;
+	unsigned int owner_cnt;
+	void *owner;
  };
 
  struct group_device {
@@ -289,7 +291,12 @@ int iommu_probe_device(struct device *dev)
  	mutex_lock(&group->mutex);
  	iommu_alloc_default_domain(group, dev);
 
-	if (group->default_domain) {
+	/*
+	 * If device joined an existing group which has been claimed
+	 * for none kernel DMA purpose, avoid attaching the default
+	 * domain.
+	 */
+	if (group->default_domain && !group->owner) {
  		ret = __iommu_attach_device(group->default_domain, dev);
  		if (ret) {
  			mutex_unlock(&group->mutex);
@@ -2320,7 +2327,7 @@ static int __iommu_attach_group(struct iommu_domain *domain,
  {
  	int ret;
 
-	if (group->default_domain && group->domain != group->default_domain)
+	if (group->domain && group->domain != group->default_domain)
  		return -EBUSY;
 
  	ret = __iommu_group_for_each_dev(group, domain,

drivers/iommu/iommu.c static void __iommu_detach_group(struct iommu_domain *domain, · line 2370

Quoted 8 lines
drivers/iommu/iommu.c +4 -1
@@ -2357,7 +2364,11 @@ static void __iommu_detach_group(struct iommu_domain *domain,
  {
  	int ret;
 
-	if (!group->default_domain) {
+	/*
+	 * If group has been claimed for none kernel DMA purpose, avoid
+	 * re-attaching the default domain.
+	 */

none kernel reads odd. But maybe drop that and just say 'claimed already' ala:

Quoted 5 lines
	/*
	 * If the group has been claimed already, do not re-attach the default
	 * domain.
	 */

Sure!

Best regards,
baolu

author
#13 lore raw

PATCH v5 2/14 driver core: Add dma_cleanup callback in bus_type

In reply to Lu Baolu

The bus_type structure defines dma_configure() callback for bus drivers to configure DMA on the devices. This adds the paired dma_cleanup() callback and calls it during driver unbinding so that bus drivers can do some cleanup work.

One use case for this paired DMA callbacks is for the bus driver to check for DMA ownership conflicts during driver binding, where multiple devices belonging to a same IOMMU group (the minimum granularity of isolation and protection) may be assigned to kernel drivers or user space respectively.

Without this change, for example, the vfio driver has to listen to a bus BOUND_DRIVER event and then BUG_ON() in case of dma ownership conflict. This leads to bad user experience since careless driver binding operation may crash the system if the admin overlooks the group restriction. Aside from bad design, this leads to a security problem as a root user, even with lockdown=integrity, can force the kernel to BUG.

With this change, the bus driver could check and set the DMA ownership in driver binding process and fail on ownership conflicts. The DMA ownership should be released during driver unbinding.

---
 include/linux/device/bus.h | 3 +++
 drivers/base/dd.c          | 5 +++++
 2 files changed, 8 insertions(+)
include/linux/device/bus.h +3 -0
index a039ab809753..d8b29ccd07e5 100644
@@ -59,6 +59,8 @@ struct fwnode_handle;
  *		bus supports.
  * @dma_configure:	Called to setup DMA configuration on a device on
  *			this bus.
+ * @dma_cleanup:	Called to cleanup DMA configuration on a device on
+ *			this bus.
  * @pm:		Power management operations of this bus, callback the specific
  *		device driver's pm-ops.
  * @iommu_ops:  IOMMU specific operations for this bus, used to attach IOMMU
@@ -103,6 +105,7 @@ struct bus_type {
 	int (*num_vf)(struct device *dev);
 
 	int (*dma_configure)(struct device *dev);
+	void (*dma_cleanup)(struct device *dev);
 
 	const struct dev_pm_ops *pm;
 
drivers/base/dd.c +5 -0
index 9eaaff2f556c..2bcbd358eda3 100644
@@ -662,6 +662,8 @@ static int really_probe(struct device *dev, struct device_driver *drv)
 	if (dev->bus)
 		blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
 					     BUS_NOTIFY_DRIVER_NOT_BOUND, dev);
+	if (dev->bus->dma_cleanup)
+		dev->bus->dma_cleanup(dev);
 pinctrl_bind_failed:
 	device_links_no_driver(dev);
 	devres_release_all(dev);
@@ -1205,6 +1207,9 @@ static void __device_release_driver(struct device *dev, struct device *parent)
 		else if (drv->remove)
 			drv->remove(dev);
 
+		if (dev->bus->dma_cleanup)
+			dev->bus->dma_cleanup(dev);
+
 		device_links_driver_cleanup(dev);
 
 		devres_release_all(dev);
Signature · 1 line
2.25.1
#14 lore raw

Re: [PATCH v5 02/14] driver core: Add dma_cleanup callback in bus_type

All these bus callouts still looks horrible and just create tons of boilerplate code.

#15 lore raw

Re: [PATCH v5 02/14] driver core: Add dma_cleanup callback in bus_type

Quoted 2 lines · Christoph Hellwig

All these bus callouts still looks horrible and just create tons of boilerplate code.

Yes, Lu - Greg asked questions then didn't respond to their answers meaning he accepts them, you should stick with the v4 version.

Jason

#16 lore raw

Re: [PATCH v5 02/14] driver core: Add dma_cleanup callback in bus_type

Quoted 6 lines · Jason Gunthorpe
Quoted 2 lines · Christoph Hellwig

All these bus callouts still looks horrible and just create tons of boilerplate code.

Yes, Lu - Greg asked questions then didn't respond to their answers meaning he accepts them, you should stick with the v4 version.

Trying to catch up on emails from the break, that was way down my list of things to get back to as it's messy and non-obvious. I'll revisit it again after 5.17-rc1 is out, this is too late for that merge window anyway.

thanks,

greg k-h

author
#17 lore raw

Re: [PATCH v5 02/14] driver core: Add dma_cleanup callback in bus_type

Hi Greg,

Quoted 5 lines · Greg Kroah-Hartman
Quoted 4 lines · Jason Gunthorpe
Quoted 2 lines · Christoph Hellwig

All these bus callouts still looks horrible and just create tons of boilerplate code.

Quoted 7 lines
Quoted 2 lines

Yes, Lu - Greg asked questions then didn't respond to their answers meaning he accepts them, you should stick with the v4 version.

Trying to catch up on emails from the break, that was way down my list of things to get back to as it's messy and non-obvious. I'll revisit it again after 5.17-rc1 is out, this is too late for that merge window anyway.

In this series we want to add calls into the iommu subsystem during device driver binding/unbinding, so that the device DMA ownership conflict (kernel driver vs. user-space) could be detected and avoided before calling into device driver's .probe().

In this v5 series, we implemented this in the affected buses (amba/ platform/fsl-mc/pci) which are known to support assigning devices to user space through the vfio framework currently. And more buses are possible to be affected in the future if they also want to support device assignment. Christoph commented that this will create boilerplate code in various bus drivers.

Back to v4 of this series (please refer to below link [1]), we added this call in the driver core if buses have provided the dma_configure() callback (please refer to below link [2]).

Which would you prefer, or any other suggestions? We need your guide to move this series ahead. Please help to suggest.

[1] https://lore.kernel.org/linux-iommu/[email protected]/
[2] https://lore.kernel.org/linux-iommu/[email protected]/

Best regards,
baolu

#18 lore raw

Re: [PATCH v5 02/14] driver core: Add dma_cleanup callback in bus_type

Quoted 32 lines · Lu Baolu

Hi Greg,

Quoted 5 lines · Greg Kroah-Hartman
Quoted 4 lines · Jason Gunthorpe
Quoted 2 lines · Christoph Hellwig

All these bus callouts still looks horrible and just create tons of boilerplate code.

Quoted 7 lines
Quoted 2 lines

Yes, Lu - Greg asked questions then didn't respond to their answers meaning he accepts them, you should stick with the v4 version.

Trying to catch up on emails from the break, that was way down my list of things to get back to as it's messy and non-obvious. I'll revisit it again after 5.17-rc1 is out, this is too late for that merge window anyway.

In this series we want to add calls into the iommu subsystem during device driver binding/unbinding, so that the device DMA ownership conflict (kernel driver vs. user-space) could be detected and avoided before calling into device driver's .probe().

In this v5 series, we implemented this in the affected buses (amba/ platform/fsl-mc/pci) which are known to support assigning devices to user space through the vfio framework currently. And more buses are possible to be affected in the future if they also want to support device assignment. Christoph commented that this will create boilerplate code in various bus drivers.

Back to v4 of this series (please refer to below link [1]), we added this call in the driver core if buses have provided the dma_configure() callback (please refer to below link [2]).

Quoted 5 lines

Which would you prefer, or any other suggestions? We need your guide to move this series ahead. Please help to suggest.

[1] https://lore.kernel.org/linux-iommu/[email protected]/
[2] https://lore.kernel.org/linux-iommu/[email protected]/

Let me look over the series again this afternooon.

thanks,

greg k-h
_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

#19 lore raw

Re: [PATCH v5 02/14] driver core: Add dma_cleanup callback in bus_type

In reply to Christoph Hellwig
Quoted 2 lines · Christoph Hellwig

All these bus callouts still looks horrible and just create tons of boilerplate code.

I can't remember anymore what one vs. the other looks like. Having an explicit "opt-in" for a bus is good, in that no code breaks and only if you want to use this do you have to set the flag. Why do you want it to be "always"?

thanks,

greg k-h
_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

maintainer Reviewed-by on v5 2/14
#20 lore raw

Re: [PATCH v5 02/14] driver core: Add dma_cleanup callback in bus_type

In reply to Lu Baolu
Quoted 17 lines · Lu Baolu

The bus_type structure defines dma_configure() callback for bus drivers to configure DMA on the devices. This adds the paired dma_cleanup() callback and calls it during driver unbinding so that bus drivers can do some cleanup work.

One use case for this paired DMA callbacks is for the bus driver to check for DMA ownership conflicts during driver binding, where multiple devices belonging to a same IOMMU group (the minimum granularity of isolation and protection) may be assigned to kernel drivers or user space respectively.

Without this change, for example, the vfio driver has to listen to a bus BOUND_DRIVER event and then BUG_ON() in case of dma ownership conflict. This leads to bad user experience since careless driver binding operation may crash the system if the admin overlooks the group restriction. Aside from bad design, this leads to a security problem as a root user, even with lockdown=integrity, can force the kernel to BUG.

Quoted 5 lines

With this change, the bus driver could check and set the DMA ownership in driver binding process and fail on ownership conflicts. The DMA ownership should be released during driver unbinding.

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

author
#21 lore raw

PATCH v5 3/14 amba: Stop sharing platform_dma_configure()

In reply to Lu Baolu

Stop sharing platform_dma_configure() helper as they are about to have their specific bus dma_configure callbacks.

---
 include/linux/platform_device.h |  2 --
 drivers/amba/bus.c              | 19 ++++++++++++++++++-
 drivers/base/platform.c         |  3 +--
 3 files changed, 19 insertions(+), 5 deletions(-)
include/linux/platform_device.h +0 -2
index 7c96f169d274..17fde717df68 100644
@@ -328,8 +328,6 @@ extern int platform_pm_restore(struct device *dev);
 #define platform_pm_restore		NULL
 #endif
 
-extern int platform_dma_configure(struct device *dev);
-
 #ifdef CONFIG_PM_SLEEP
 #define USE_PLATFORM_PM_SLEEP_OPS \
 	.suspend = platform_pm_suspend, \
drivers/amba/bus.c +18 -1
index 720aa6cdd402..850a0a8973c6 100644
@@ -20,6 +20,8 @@
 #include <linux/platform_device.h>
 #include <linux/reset.h>
 #include <linux/of_irq.h>
+#include <linux/of_device.h>
+#include <linux/acpi.h>
 
 #include <asm/irq.h>
 
@@ -251,6 +253,21 @@ static void amba_shutdown(struct device *dev)
 		drv->shutdown(to_amba_device(dev));
 }
 
+static int amba_dma_configure(struct device *dev)
+{
+	enum dev_dma_attr attr;
+	int ret = 0;
+
+	if (dev->of_node) {
+		ret = of_dma_configure(dev, dev->of_node, true);
+	} else if (has_acpi_companion(dev)) {
+		attr = acpi_get_dma_attr(to_acpi_device_node(dev->fwnode));
+		ret = acpi_dma_configure(dev, attr);
+	}
+
+	return ret;
+}
+
 #ifdef CONFIG_PM
 /*
  * Hooks to provide runtime PM of the pclk (bus clock).  It is safe to
@@ -319,7 +336,7 @@ struct bus_type amba_bustype = {
 	.probe		= amba_probe,
 	.remove		= amba_remove,
 	.shutdown	= amba_shutdown,
-	.dma_configure	= platform_dma_configure,
+	.dma_configure	= amba_dma_configure,
 	.pm		= &amba_pm,
 };
 EXPORT_SYMBOL_GPL(amba_bustype);
drivers/base/platform.c +1 -2
index 598acf93a360..b4d36b46ab2e 100644
@@ -1449,8 +1449,7 @@ static void platform_shutdown(struct device *_dev)
 		drv->shutdown(dev);
 }
 
-
-int platform_dma_configure(struct device *dev)
+static int platform_dma_configure(struct device *dev)
 {
 	enum dev_dma_attr attr;
 	int ret = 0;
Signature · 1 line
2.25.1
author
#22 lore raw

PATCH v5 4/14 driver core: platform: Add driver dma ownership management

In reply to Lu Baolu

Multiple platform devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture. This checks and sets DMA ownership during driver binding, and release the ownership during driver unbinding.

The device driver may set a new flag (no_kernel_api_dma) to skip calling iommu_device_use_dma_api() during the binding process. For instance, the userspace framework drivers (vfio etc.) which need to manually claim their own dma ownership when assigning the device to userspace.

---
 include/linux/platform_device.h |  1 +
 drivers/base/platform.c         | 20 ++++++++++++++++++++
 2 files changed, 21 insertions(+)
include/linux/platform_device.h +1 -0
index 17fde717df68..8f0eaafcef47 100644
@@ -210,6 +210,7 @@ struct platform_driver {
 	struct device_driver driver;
 	const struct platform_device_id *id_table;
 	bool prevent_deferred_probe;
+	bool no_kernel_api_dma;
 };
 
 #define to_platform_driver(drv)	(container_of((drv), struct platform_driver, \
drivers/base/platform.c +20 -0
index b4d36b46ab2e..d5171e44d903 100644
@@ -30,6 +30,7 @@
 #include <linux/property.h>
 #include <linux/kmemleak.h>
 #include <linux/types.h>
+#include <linux/iommu.h>
 
 #include "base.h"
 #include "power/power.h"
@@ -1451,9 +1452,16 @@ static void platform_shutdown(struct device *_dev)
 
 static int platform_dma_configure(struct device *dev)
 {
+	struct platform_driver *drv = to_platform_driver(dev->driver);
 	enum dev_dma_attr attr;
 	int ret = 0;
 
+	if (!drv->no_kernel_api_dma) {
+		ret = iommu_device_use_dma_api(dev);
+		if (ret)
+			return ret;
+	}
+
 	if (dev->of_node) {
 		ret = of_dma_configure(dev, dev->of_node, true);
 	} else if (has_acpi_companion(dev)) {
@@ -1461,9 +1469,20 @@ static int platform_dma_configure(struct device *dev)
 		ret = acpi_dma_configure(dev, attr);
 	}
 
+	if (ret && !drv->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);
+
 	return ret;
 }
 
+static void platform_dma_cleanup(struct device *dev)
+{
+	struct platform_driver *drv = to_platform_driver(dev->driver);
+
+	if (!drv->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);
+}
+
 static const struct dev_pm_ops platform_dev_pm_ops = {
 	SET_RUNTIME_PM_OPS(pm_generic_runtime_suspend, pm_generic_runtime_resume, NULL)
 	USE_PLATFORM_PM_SLEEP_OPS
@@ -1478,6 +1497,7 @@ struct bus_type platform_bus_type = {
 	.remove		= platform_remove,
 	.shutdown	= platform_shutdown,
 	.dma_configure	= platform_dma_configure,
+	.dma_cleanup	= platform_dma_cleanup,
 	.pm		= &platform_dev_pm_ops,
 };
 EXPORT_SYMBOL_GPL(platform_bus_type);
Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
#23 lore raw

Re: [PATCH v5 04/14] driver core: platform: Add driver dma ownership management

Quoted 59 lines · Lu Baolu

Multiple platform devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture. This checks and sets DMA ownership during driver binding, and release the ownership during driver unbinding.

The device driver may set a new flag (no_kernel_api_dma) to skip calling iommu_device_use_dma_api() during the binding process. For instance, the userspace framework drivers (vfio etc.) which need to manually claim their own dma ownership when assigning the device to userspace.

---
 include/linux/platform_device.h |  1 +
 drivers/base/platform.c         | 20 ++++++++++++++++++++
 2 files changed, 21 insertions(+)
include/linux/platform_device.h +1 -0
index 17fde717df68..8f0eaafcef47 100644
@@ -210,6 +210,7 @@ struct platform_driver {
 	struct device_driver driver;
 	const struct platform_device_id *id_table;
 	bool prevent_deferred_probe;
+	bool no_kernel_api_dma;
 };
 
 #define to_platform_driver(drv)	(container_of((drv), struct platform_driver, \
drivers/base/platform.c +8 -0
index b4d36b46ab2e..d5171e44d903 100644
@@ -30,6 +30,7 @@
 #include <linux/property.h>
 #include <linux/kmemleak.h>
 #include <linux/types.h>
+#include <linux/iommu.h>
 
 #include "base.h"
 #include "power/power.h"
@@ -1451,9 +1452,16 @@ static void platform_shutdown(struct device *_dev)
 
 static int platform_dma_configure(struct device *dev)
 {
+	struct platform_driver *drv = to_platform_driver(dev->driver);
 	enum dev_dma_attr attr;
 	int ret = 0;
 
+	if (!drv->no_kernel_api_dma) {
+		ret = iommu_device_use_dma_api(dev);
+		if (ret)
+			return ret;
+	}
+
 	if (dev->of_node) {
 		ret = of_dma_configure(dev, dev->of_node, true);
 	} else if (has_acpi_companion(dev)) {

drivers/base/platform.c static int platform_dma_configure(struct device *dev) · line 1473

Quoted 5 lines
drivers/base/platform.c +2 -0
@@ -1461,9 +1469,20 @@ static int platform_dma_configure(struct device *dev)
 		ret = acpi_dma_configure(dev, attr);
 	}
 
+	if (ret && !drv->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);

So you are now going to call this for every platform driver _unless_ they set this flag?

And having "negative" flags is rough to parse at times. Yes, we have "prevent_deferred_probe" already here, so maybe keep this in the same nameing scheme and use "prevent_dma_api"?

And it would be good to document this somewhere as to what this means.

thanks,

greg k-h
_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

#24 lore raw

Re: [PATCH v5 04/14] driver core: platform: Add driver dma ownership management

Quoted 5 lines · Greg Kroah-Hartman

Review

Quoted 2 lines
(quoted diff) +2 -0
+	if (ret && !drv->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);

So you are now going to call this for every platform driver _unless_ they set this flag?

Yes, it is necessary because VFIO supports platform devices as well and needs to ensure security. Conflicting kernel driver attachements must be blocked, just like for PCI.

Jason

#25 lore raw

Re: [PATCH v5 04/14] driver core: platform: Add driver dma ownership management

Quoted 11 lines · Jason Gunthorpe
Quoted 5 lines · Greg Kroah-Hartman

Review

Quoted 2 lines
(quoted diff) +2 -0
+	if (ret && !drv->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);

So you are now going to call this for every platform driver _unless_ they set this flag?

Yes, it is necessary because VFIO supports platform devices as well and needs to ensure security. Conflicting kernel driver attachements must be blocked, just like for PCI.

A platform device shouldn't be using VFIO, but ugh, oh well, that ship has sailed :(

And stop it with the "security" mess, do not give people a false sense of it here please.

thanks,

greg k-h

#26 lore raw

Re: [PATCH v5 04/14] driver core: platform: Add driver dma ownership management

Quoted 9 lines · Greg Kroah-Hartman
Quoted 8 lines · Jason Gunthorpe
Quoted 5 lines · Greg Kroah-Hartman

Review

Quoted 2 lines
(quoted diff) +2 -0
+	if (ret && !drv->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);

So you are now going to call this for every platform driver _unless_ they set this flag?

Quoted 6 lines
Quoted 3 lines

Yes, it is necessary because VFIO supports platform devices as well and needs to ensure security. Conflicting kernel driver attachements must be blocked, just like for PCI.

A platform device shouldn't be using VFIO, but ugh, oh well, that ship has sailed :(

I don't know why you say that, but yes, this is was set long ago.

Quoted 2 lines

And stop it with the "security" mess, do not give people a false sense of it here please.

I'm confused by what you mean. This is all about what we tend to refer to as DMA security - meaning a device's DMA can be controled by a hostile environment and not impact the integrity of the kernel.

Jason

author
#27 lore raw

PATCH v5 5/14 amba: Add driver dma ownership management

In reply to Lu Baolu

Multiple amba devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture. This checks and sets DMA ownership during driver binding, and release the ownership during driver unbinding.

The device driver may set a new flag (no_kernel_api_dma) to skip calling iommu_device_use_dma_api() during the binding process. For instance, the userspace framework drivers (vfio etc.) which need to manually claim their own dma ownership when assigning the device to userspace.

---
 include/linux/amba/bus.h |  1 +
 drivers/amba/bus.c       | 20 ++++++++++++++++++++
 2 files changed, 21 insertions(+)
include/linux/amba/bus.h +1 -0
index edfcf7a14dcd..a4f374e2fd35 100644
@@ -79,6 +79,7 @@ struct amba_driver {
 	void			(*remove)(struct amba_device *);
 	void			(*shutdown)(struct amba_device *);
 	const struct amba_id	*id_table;
+	bool no_kernel_api_dma;
 };
 
 /*
drivers/amba/bus.c +20 -0
index 850a0a8973c6..42b20ecfa50c 100644
@@ -22,6 +22,7 @@
 #include <linux/of_irq.h>
 #include <linux/of_device.h>
 #include <linux/acpi.h>
+#include <linux/iommu.h>
 
 #include <asm/irq.h>
 
@@ -255,9 +256,16 @@ static void amba_shutdown(struct device *dev)
 
 static int amba_dma_configure(struct device *dev)
 {
+	struct amba_driver *drv = to_amba_driver(dev->driver);
 	enum dev_dma_attr attr;
 	int ret = 0;
 
+	if (!drv->no_kernel_api_dma) {
+		ret = iommu_device_use_dma_api(dev);
+		if (ret)
+			return ret;
+	}
+
 	if (dev->of_node) {
 		ret = of_dma_configure(dev, dev->of_node, true);
 	} else if (has_acpi_companion(dev)) {
@@ -265,9 +273,20 @@ static int amba_dma_configure(struct device *dev)
 		ret = acpi_dma_configure(dev, attr);
 	}
 
+	if (ret && !drv->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);
+
 	return ret;
 }
 
+static void amba_dma_cleanup(struct device *dev)
+{
+	struct amba_driver *drv = to_amba_driver(dev->driver);
+
+	if (!drv->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);
+}
+
 #ifdef CONFIG_PM
 /*
  * Hooks to provide runtime PM of the pclk (bus clock).  It is safe to
@@ -337,6 +356,7 @@ struct bus_type amba_bustype = {
 	.remove		= amba_remove,
 	.shutdown	= amba_shutdown,
 	.dma_configure	= amba_dma_configure,
+	.dma_cleanup	= amba_dma_cleanup,
 	.pm		= &amba_pm,
 };
 EXPORT_SYMBOL_GPL(amba_bustype);
Signature · 1 line
2.25.1
author
#28 lore raw

PATCH v5 6/14 bus: fsl-mc: Add driver dma ownership management

In reply to Lu Baolu

Multiple fsl-mc devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture. This checks and sets DMA ownership during driver binding, and release the ownership during driver unbinding.

The device driver may set a new flag (no_kernel_api_dma) to skip calling iommu_device_use_dma_api() during the binding process. For instance, the userspace framework drivers (vfio etc.) which need to manually claim their own dma ownership when assigning the device to userspace.

---
 include/linux/fsl/mc.h          |  5 +++++
 drivers/bus/fsl-mc/fsl-mc-bus.c | 26 ++++++++++++++++++++++++--
 2 files changed, 29 insertions(+), 2 deletions(-)
include/linux/fsl/mc.h +5 -0
index e026f6c48b49..e85a5f18104d 100644
@@ -32,6 +32,10 @@ struct fsl_mc_io;
  * @shutdown: Function called at shutdown time to quiesce the device
  * @suspend: Function called when a device is stopped
  * @resume: Function called when a device is resumed
+ * @no_kernel_api_dma: Device driver doesn't use kernel DMA API for DMA.
+ *		Drivers which don't require DMA or want to manually claim the
+ *		owner type (e.g. userspace driver frameworks) could set this
+ *		flag.
  *
  * Generic DPAA device driver object for device drivers that are registered
  * with a DPRC bus. This structure is to be embedded in each device-specific
@@ -45,6 +49,7 @@ struct fsl_mc_driver {
 	void (*shutdown)(struct fsl_mc_device *dev);
 	int (*suspend)(struct fsl_mc_device *dev, pm_message_t state);
 	int (*resume)(struct fsl_mc_device *dev);
+	bool no_kernel_api_dma;
 };
 
 #define to_fsl_mc_driver(_drv) \
drivers/bus/fsl-mc/fsl-mc-bus.c +24 -2
index 8fd4a356a86e..9ee85f43822a 100644
@@ -140,15 +140,36 @@ static int fsl_mc_dma_configure(struct device *dev)
 {
 	struct device *dma_dev = dev;
 	struct fsl_mc_device *mc_dev = to_fsl_mc_device(dev);
+	struct fsl_mc_driver *mc_drv = to_fsl_mc_driver(dev->driver);
 	u32 input_id = mc_dev->icid;
+	int ret;
+
+	if (!mc_drv->no_kernel_api_dma) {
+		ret = iommu_device_use_dma_api(dev);
+		if (ret)
+			return ret;
+	}
 
 	while (dev_is_fsl_mc(dma_dev))
 		dma_dev = dma_dev->parent;
 
 	if (dev_of_node(dma_dev))
-		return of_dma_configure_id(dev, dma_dev->of_node, 0, &input_id);
+		ret = of_dma_configure_id(dev, dma_dev->of_node, 0, &input_id);
+	else
+		ret = acpi_dma_configure_id(dev, DEV_DMA_COHERENT, &input_id);
+
+	if (ret && !mc_drv->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);
+
+	return ret;
+}
+
+static void fsl_mc_dma_cleanup(struct device *dev)
+{
+	struct fsl_mc_driver *mc_drv = to_fsl_mc_driver(dev->driver);
 
-	return acpi_dma_configure_id(dev, DEV_DMA_COHERENT, &input_id);
+	if (!mc_drv->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);
 }
 
 static ssize_t modalias_show(struct device *dev, struct device_attribute *attr,
@@ -312,6 +333,7 @@ struct bus_type fsl_mc_bus_type = {
 	.match = fsl_mc_bus_match,
 	.uevent = fsl_mc_bus_uevent,
 	.dma_configure  = fsl_mc_dma_configure,
+	.dma_cleanup = fsl_mc_dma_cleanup,
 	.dev_groups = fsl_mc_dev_groups,
 	.bus_groups = fsl_mc_bus_groups,
 };
Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author
#29 lore raw

PATCH v5 7/14 PCI: Add driver dma ownership management

In reply to Lu Baolu

Multiple PCI devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture. This checks and sets DMA ownership during driver binding, and release the ownership during driver unbinding.

The device driver may set a new flag (no_kernel_api_dma) to skip calling iommu_device_use_dma_api() during the binding process. For instance, the userspace framework drivers (vfio etc.) which need to manually claim their own dma ownership when assigning the device to userspace.

---
 include/linux/pci.h      |  5 +++++
 drivers/pci/pci-driver.c | 21 +++++++++++++++++++++
 2 files changed, 26 insertions(+)
include/linux/pci.h +5 -0
index 18a75c8e615c..d29a990e3f02 100644
@@ -882,6 +882,10 @@ struct module;
  *              created once it is bound to the driver.
  * @driver:	Driver model structure.
  * @dynids:	List of dynamically added device IDs.
+ * @no_kernel_api_dma: Device driver doesn't use kernel DMA API for DMA.
+ *		Drivers which don't require DMA or want to manually claim the
+ *		owner type (e.g. userspace driver frameworks) could set this
+ *		flag.
  */
 struct pci_driver {
 	struct list_head	node;
@@ -900,6 +904,7 @@ struct pci_driver {
 	const struct attribute_group **dev_groups;
 	struct device_driver	driver;
 	struct pci_dynids	dynids;
+	bool no_kernel_api_dma;
 };
 
 static inline struct pci_driver *to_pci_driver(struct device_driver *drv)
drivers/pci/pci-driver.c +21 -0
index 588588cfda48..4e003ea12718 100644
@@ -20,6 +20,7 @@
 #include <linux/of_device.h>
 #include <linux/acpi.h>
 #include <linux/dma-map-ops.h>
+#include <linux/iommu.h>
 #include "pci.h"
 #include "pcie/portdrv.h"
 
@@ -1590,9 +1591,16 @@ static int pci_bus_num_vf(struct device *dev)
  */
 static int pci_dma_configure(struct device *dev)
 {
+	struct pci_driver *driver = to_pci_driver(dev->driver);
 	struct device *bridge;
 	int ret = 0;
 
+	if (!driver->no_kernel_api_dma) {
+		ret = iommu_device_use_dma_api(dev);
+		if (ret)
+			return ret;
+	}
+
 	bridge = pci_get_host_bridge_device(to_pci_dev(dev));
 
 	if (IS_ENABLED(CONFIG_OF) && bridge->parent &&
@@ -1605,9 +1613,21 @@ static int pci_dma_configure(struct device *dev)
 	}
 
 	pci_put_host_bridge_device(bridge);
+
+	if (ret && !driver->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);
+
 	return ret;
 }
 
+static void pci_dma_cleanup(struct device *dev)
+{
+	struct pci_driver *driver = to_pci_driver(dev->driver);
+
+	if (!driver->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);
+}
+
 struct bus_type pci_bus_type = {
 	.name		= "pci",
 	.match		= pci_bus_match,
@@ -1621,6 +1641,7 @@ struct bus_type pci_bus_type = {
 	.pm		= PCI_PM_OPS_PTR,
 	.num_vf		= pci_bus_num_vf,
 	.dma_configure	= pci_dma_configure,
+	.dma_cleanup	= pci_dma_cleanup,
 };
 EXPORT_SYMBOL(pci_bus_type);
 
Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
#30 lore raw

Re: [PATCH v5 07/14] PCI: Add driver dma ownership management

Quoted 22 lines · Lu Baolu

Multiple PCI devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture. This checks and sets DMA ownership during driver binding, and release the ownership during driver unbinding.

The device driver may set a new flag (no_kernel_api_dma) to skip calling iommu_device_use_dma_api() during the binding process. For instance, the userspace framework drivers (vfio etc.) which need to manually claim their own dma ownership when assigning the device to userspace.

---
 include/linux/pci.h      |  5 +++++
 drivers/pci/pci-driver.c | 21 +++++++++++++++++++++
 2 files changed, 26 insertions(+)
include/linux/pci.h +0 -0
index 18a75c8e615c..d29a990e3f02 100644

include/linux/pci.h struct module; · line 888

Quoted 7 lines
include/linux/pci.h +4 -0
@@ -882,6 +882,10 @@ struct module;
  *              created once it is bound to the driver.
  * @driver:	Driver model structure.
  * @dynids:	List of dynamically added device IDs.
+ * @no_kernel_api_dma: Device driver doesn't use kernel DMA API for DMA.
+ *		Drivers which don't require DMA or want to manually claim the
+ *		owner type (e.g. userspace driver frameworks) could set this
+ *		flag.

Again with the bikeshedding, but this name is a bit odd. Of course it's in the kernel, this is all kernel code, so you can drop that. And again, "negative" flags are rough. So maybe just "prevent_dma"?

thanks,

greg k-h
_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

#31 lore raw

Re: [PATCH v5 07/14] PCI: Add driver dma ownership management

Quoted 17 lines · Greg Kroah-Hartman
Quoted 16 lines · Lu Baolu

Multiple PCI devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture. This checks and sets DMA ownership during driver binding, and release the ownership during driver unbinding.

The device driver may set a new flag (no_kernel_api_dma) to skip calling iommu_device_use_dma_api() during the binding process. For instance, the userspace framework drivers (vfio etc.) which need to manually claim their own dma ownership when assigning the device to userspace.

 include/linux/pci.h      |  5 +++++
 drivers/pci/pci-driver.c | 21 +++++++++++++++++++++
 2 files changed, 26 insertions(+)
Quoted 15 lines

include/linux/pci.h struct module; · line 888

Quoted 11 lines
include/linux/pci.h +5 -0
index 18a75c8e615c..d29a990e3f02 100644
+++ b/include/linux/pci.h
@@ -882,6 +882,10 @@ struct module;
  *              created once it is bound to the driver.
  * @driver:	Driver model structure.
  * @dynids:	List of dynamically added device IDs.
+ * @no_kernel_api_dma: Device driver doesn't use kernel DMA API for DMA.
+ *		Drivers which don't require DMA or want to manually claim the
+ *		owner type (e.g. userspace driver frameworks) could set this
+ *		flag.

Again with the bikeshedding, but this name is a bit odd. Of course it's in the kernel, this is all kernel code, so you can drop that. And again, "negative" flags are rough. So maybe just "prevent_dma"?

That is misleading too, it is not that DMA is prevented, but that the kernel's dma_api has not been setup.

Though I agree the name as-is isn't great, I think the comment is good.

Jason

#32 lore raw

Re: [PATCH v5 07/14] PCI: Add driver dma ownership management

Quoted 34 lines · Jason Gunthorpe
Quoted 17 lines · Greg Kroah-Hartman
Quoted 16 lines · Lu Baolu

Multiple PCI devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture. This checks and sets DMA ownership during driver binding, and release the ownership during driver unbinding.

The device driver may set a new flag (no_kernel_api_dma) to skip calling iommu_device_use_dma_api() during the binding process. For instance, the userspace framework drivers (vfio etc.) which need to manually claim their own dma ownership when assigning the device to userspace.

 include/linux/pci.h      |  5 +++++
 drivers/pci/pci-driver.c | 21 +++++++++++++++++++++
 2 files changed, 26 insertions(+)
Quoted 15 lines

include/linux/pci.h struct module; · line 888

Quoted 11 lines
include/linux/pci.h +5 -0
index 18a75c8e615c..d29a990e3f02 100644
+++ b/include/linux/pci.h
@@ -882,6 +882,10 @@ struct module;
  *              created once it is bound to the driver.
  * @driver:	Driver model structure.
  * @dynids:	List of dynamically added device IDs.
+ * @no_kernel_api_dma: Device driver doesn't use kernel DMA API for DMA.
+ *		Drivers which don't require DMA or want to manually claim the
+ *		owner type (e.g. userspace driver frameworks) could set this
+ *		flag.

Again with the bikeshedding, but this name is a bit odd. Of course it's in the kernel, this is all kernel code, so you can drop that. And again, "negative" flags are rough. So maybe just "prevent_dma"?

Quoted 2 lines

That is misleading too, it is not that DMA is prevented, but that the kernel's dma_api has not been setup.

"has not been" or "will not be"?

What you want to prevent is the iommu core claiming the device automatically, right? So how about "prevent_iommu_dma"?

naming is hard,

greg k-h

#33 lore raw

Re: [PATCH v5 07/14] PCI: Add driver dma ownership management

Quoted 35 lines · Greg Kroah-Hartman
Quoted 34 lines · Jason Gunthorpe
Quoted 32 lines · Greg Kroah-Hartman
Quoted 20 lines · Lu Baolu

Multiple PCI devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture. This checks and sets DMA ownership during driver binding, and release the ownership during driver unbinding.

The device driver may set a new flag (no_kernel_api_dma) to skip calling iommu_device_use_dma_api() during the binding process. For instance, the userspace framework drivers (vfio etc.) which need to manually claim their own dma ownership when assigning the device to userspace.

 include/linux/pci.h      |  5 +++++
 drivers/pci/pci-driver.c | 21 +++++++++++++++++++++
 2 files changed, 26 insertions(+)
include/linux/pci.h +1 -0
index 18a75c8e615c..d29a990e3f02 100644
+++ b/include/linux/pci.h

include/linux/pci.h struct module; · line 888

Quoted 7 lines
include/linux/pci.h +4 -0
@@ -882,6 +882,10 @@ struct module;
  *              created once it is bound to the driver.
  * @driver:	Driver model structure.
  * @dynids:	List of dynamically added device IDs.
+ * @no_kernel_api_dma: Device driver doesn't use kernel DMA API for DMA.
+ *		Drivers which don't require DMA or want to manually claim the
+ *		owner type (e.g. userspace driver frameworks) could set this
+ *		flag.

Again with the bikeshedding, but this name is a bit odd. Of course it's in the kernel, this is all kernel code, so you can drop that. And again, "negative" flags are rough. So maybe just "prevent_dma"?

Quoted 4 lines
Quoted 2 lines

That is misleading too, it is not that DMA is prevented, but that the kernel's dma_api has not been setup.

"has not been" or "will not be"?

"has not been" as that action was supposed to happen before probe(), but the flag skips it.

A driver that sets this flag can still decide to enable the dma API on its own. eg tegra drivers do this.

Quoted 2 lines

What you want to prevent is the iommu core claiming the device automatically, right? So how about "prevent_iommu_dma"?

"claim" is not a good description. iommu always "claims" the device - eg sets a domain, sets the dev and bus parameters, etc.

This really is only about setting up the in-kernel dma api, eg allowing dma_map_sg()/etc to work.

dma api is just one way to operate the iommu, there are others too.

Think of this flag as
false = the driver is going to use the dma api (most common) true = the driver will decide how to use the iommu by itself

Does it help think of a clearer name?

Jason

#34 lore raw

Re: [PATCH v5 07/14] PCI: Add driver dma ownership management

Quoted 39 lines · Jason Gunthorpe
Quoted 38 lines · Greg Kroah-Hartman
Quoted 36 lines · Jason Gunthorpe
Quoted 17 lines · Greg Kroah-Hartman
Quoted 16 lines · Lu Baolu

Multiple PCI devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture. This checks and sets DMA ownership during driver binding, and release the ownership during driver unbinding.

The device driver may set a new flag (no_kernel_api_dma) to skip calling iommu_device_use_dma_api() during the binding process. For instance, the userspace framework drivers (vfio etc.) which need to manually claim their own dma ownership when assigning the device to userspace.

 include/linux/pci.h      |  5 +++++
 drivers/pci/pci-driver.c | 21 +++++++++++++++++++++
 2 files changed, 26 insertions(+)
Quoted 15 lines

include/linux/pci.h struct module; · line 888

Quoted 11 lines
include/linux/pci.h +5 -0
index 18a75c8e615c..d29a990e3f02 100644
+++ b/include/linux/pci.h
@@ -882,6 +882,10 @@ struct module;
  *              created once it is bound to the driver.
  * @driver:	Driver model structure.
  * @dynids:	List of dynamically added device IDs.
+ * @no_kernel_api_dma: Device driver doesn't use kernel DMA API for DMA.
+ *		Drivers which don't require DMA or want to manually claim the
+ *		owner type (e.g. userspace driver frameworks) could set this
+ *		flag.

Again with the bikeshedding, but this name is a bit odd. Of course it's in the kernel, this is all kernel code, so you can drop that. And again, "negative" flags are rough. So maybe just "prevent_dma"?

That is misleading too, it is not that DMA is prevented, but that the kernel's dma_api has not been setup.

Quoted 7 lines
Quoted 1 line

"has not been" or "will not be"?

"has not been" as that action was supposed to happen before probe(), but the flag skips it.

A driver that sets this flag can still decide to enable the dma API on its own. eg tegra drivers do this.

So you are just forcing the driver to manage this all on their own, so how about, "driver_managed_dma", or even shorter "managed_dma"?

Quoted 9 lines
Quoted 2 lines

What you want to prevent is the iommu core claiming the device automatically, right? So how about "prevent_iommu_dma"?

"claim" is not a good description. iommu always "claims" the device - eg sets a domain, sets the dev and bus parameters, etc.

This really is only about setting up the in-kernel dma api, eg allowing dma_map_sg()/etc to work.

Quoted 7 lines

dma api is just one way to operate the iommu, there are others too.

Think of this flag as
false = the driver is going to use the dma api (most common) true = the driver will decide how to use the iommu by itself

Does it help think of a clearer name?

See above, you want a driver author to know instantly what this is and not have to look anything up. "I_will_manage_the_dma_myself_as_I_really_know_what_I_am_doing" might be good, but a bit too long :)

thanks,

greg k-h

#35 lore raw

Re: [PATCH v5 07/14] PCI: Add driver dma ownership management

Quoted 5 lines · Greg Kroah-Hartman
Quoted 2 lines

A driver that sets this flag can still decide to enable the dma API on its own. eg tegra drivers do this.

So you are just forcing the driver to manage this all on their own, so how about, "driver_managed_dma", or even shorter "managed_dma"?

Yeah, I like "driver_managed_dma" alot, it captures the entire idea

Thanks,
Jason

author
#36 lore raw

Re: [PATCH v5 07/14] PCI: Add driver dma ownership management

Quoted 9 lines · Jason Gunthorpe
Quoted 5 lines · Greg Kroah-Hartman
Quoted 2 lines

A driver that sets this flag can still decide to enable the dma API on its own. eg tegra drivers do this.

So you are just forcing the driver to manage this all on their own, so how about, "driver_managed_dma", or even shorter "managed_dma"?

Yeah, I like "driver_managed_dma" alot, it captures the entire idea

This makes a lot of sense. For most drivers, they don't need to care about this flag as all DMAs are handled through the kernel DMA API. For VFIO or similar drivers, they know how to manage the DMA themselves and set this flag so that the IOMMU layer will allow them to setup and manage their own I/O address space.

If there is no better naming, I'd like to use this and add some comments for device drivers developers.

Thanks a lot to Greg and Jason.

Best regards,
baolu

#37 lore raw

Re: [PATCH v5 07/14] PCI: Add driver dma ownership management

In reply to Lu Baolu

In subject,

s/dma/DMA/ to match the other patches

Quoted 10 lines · Lu Baolu

Multiple PCI devices may be placed in the same IOMMU group because they cannot be isolated from each other. These devices must either be entirely under kernel control or userspace control, never a mixture. This checks and sets DMA ownership during driver binding, and release the ownership during driver unbinding.

The device driver may set a new flag (no_kernel_api_dma) to skip calling iommu_device_use_dma_api() during the binding process. For instance, the userspace framework drivers (vfio etc.) which need to manually claim their own dma ownership when assigning the device to userspace.

s/vfio/VFIO/ when used as an acronym (occurs in several patches)

Quoted 4 lines

+ * @no_kernel_api_dma: Device driver doesn't use kernel DMA API for DMA.
+ * Drivers which don't require DMA or want to manually claim the
+ * owner type (e.g. userspace driver frameworks) could set this
+ * flag.

s/Drivers which/Drivers that/

drivers/pci/pci-driver.c static int pci_bus_num_vf(struct device *dev) · line 1598

Quoted 7 lines

static int pci_dma_configure(struct device *dev)
{

drivers/pci/pci-driver.c +2 -0
+	struct pci_driver *driver = to_pci_driver(dev->driver);
 	struct device *bridge;
 	int ret = 0;
 
+	if (!driver->no_kernel_api_dma) {

Ugh. Double negative, totally agree this needs a better name that reverses the sense. Every place you use it, you negate it again.

drivers/pci/pci-driver.c static int pci_dma_configure(struct device *dev) · line 1628

Quoted 9 lines
drivers/pci/pci-driver.c +8 -0
+	if (ret && !driver->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);
 
+static void pci_dma_cleanup(struct device *dev)
+{
+	struct pci_driver *driver = to_pci_driver(dev->driver);
+
+	if (!driver->no_kernel_api_dma)
+		iommu_device_unuse_dma_api(dev);

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

#38 lore raw

Re: [PATCH v5 07/14] PCI: Add driver dma ownership management

Quoted 10 lines · Bjorn Helgaas

drivers/pci/pci-driver.c static int pci_bus_num_vf(struct device *dev) · line 1598

Quoted 7 lines

static int pci_dma_configure(struct device *dev)
{

drivers/pci/pci-driver.c +2 -0
+	struct pci_driver *driver = to_pci_driver(dev->driver);
 	struct device *bridge;
 	int ret = 0;
 
+	if (!driver->no_kernel_api_dma) {

Ugh. Double negative, totally agree this needs a better name that reverses the sense. Every place you use it, you negate it again.

Greg came up with driver_managed_dma which is in the v6 version:

https://lore.kernel.org/all/[email protected]/

Thanks,
Jason

author
#39 lore raw

PATCH v5 8/14 PCI: pci_stub: Suppress kernel DMA ownership auto-claiming

In reply to Lu Baolu

The pci_dma_configure() marks the iommu_group as containing only devices with kernel drivers that manage DMA. Avoid this default behavior for the pci_stub because it does not program any DMA itself. This allows the pci_stub still able to be used by the admin to block driver binding after applying the DMA ownership to vfio.

---
 drivers/pci/pci-stub.c | 1 +
 1 file changed, 1 insertion(+)
drivers/pci/pci-stub.c +1 -0
index e408099fea52..1f4c0a58b819 100644
@@ -36,6 +36,7 @@ static struct pci_driver stub_driver = {
 	.name		= "pci-stub",
 	.id_table	= NULL,	/* only dynamic id's */
 	.probe		= pci_stub_probe,
+	.no_kernel_api_dma = true,
 };
 
 static int __init pci_stub_init(void)
Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author
#40 lore raw

PATCH v5 9/14 PCI: portdrv: Suppress kernel DMA ownership auto-claiming

In reply to Lu Baolu

If a switch lacks ACS P2P Request Redirect, a device below the switch can bypass the IOMMU and DMA directly to other devices below the switch, so all the downstream devices must be in the same IOMMU group as the switch itself. The existing vfio framework allows the portdrv driver to be bound to the bridge while its downstream devices are assigned to user space. The pci_dma_configure() marks the iommu_group as containing only devices with kernel drivers that manage DMA. Avoid this default behavior for the portdrv driver in order for compatibility with the current vfio policy.

---
 drivers/pci/pcie/portdrv_pci.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/pci/pcie/portdrv_pci.c +2 -0
index 35eca6277a96..2116f821c005 100644
@@ -202,6 +202,8 @@ static struct pci_driver pcie_portdriver = {
 
 	.err_handler	= &pcie_portdrv_err_handler,
 
+	.no_kernel_api_dma = true,
+
 	.driver.pm	= PCIE_PORTDRV_PM_OPS,
 };
 
Signature · 1 line
2.25.1
#41 lore raw

Re: [PATCH v5 09/14] PCI: portdrv: Suppress kernel DMA ownership auto-claiming

Quoted 4 lines · Lu Baolu

If a switch lacks ACS P2P Request Redirect, a device below the switch can bypass the IOMMU and DMA directly to other devices below the switch, so all the downstream devices must be in the same IOMMU group as the switch itself.

Help me think through what's going on here. IIUC, we put devices in the same IOMMU group when they can interfere with each other in any way (DMA, config access, etc).

(We said "DMA" above, but I guess this would also apply to config requests, right?)

*This* patch doesn't check for any ACS features. Can you connect the dots for me? I guess the presence or absence of P2P Request Redirect determines the size of the IOMMU group. And the following says something about what is allowed in the group? And .no_kernel_api_dma allows an exception to the general rule?

Quoted 2 lines

The existing vfio framework allows the portdrv driver to be bound to the bridge while its downstream devices are assigned to user space.

I.e., the existing VFIO framework allows a switch to be in the same IOMMU group as the devices below it, even though the switch has a kernel driver and the other devices may have userspace drivers?

Is this a function of VFIO design or of the IOMMU driver?

Quoted 3 lines

The pci_dma_configure() marks the iommu_group as containing only devices with kernel drivers that manage DMA. Avoid this default behavior for the portdrv driver in order for compatibility with the current vfio policy.

I assume "IOMMU group" means the same as "iommu_group"; maybe we can use one of them consistently?

Quoted 23 lines
---
 drivers/pci/pcie/portdrv_pci.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/pci/pcie/portdrv_pci.c +2 -0
index 35eca6277a96..2116f821c005 100644
@@ -202,6 +202,8 @@ static struct pci_driver pcie_portdriver = {
 
 	.err_handler	= &pcie_portdrv_err_handler,
 
+	.no_kernel_api_dma = true,
+
 	.driver.pm	= PCIE_PORTDRV_PM_OPS,
 };
 
Signature · 1 line
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

#42 lore raw

Re: [PATCH v5 09/14] PCI: portdrv: Suppress kernel DMA ownership auto-claiming

Quoted 6 lines · Bjorn Helgaas
Quoted 2 lines

The existing vfio framework allows the portdrv driver to be bound to the bridge while its downstream devices are assigned to user space.

I.e., the existing VFIO framework allows a switch to be in the same IOMMU group as the devices below it, even though the switch has a kernel driver and the other devices may have userspace drivers?

Yes, this patch exists to maintain current VFIO behavior which has this same check.

I belive the basis for VFIO doing this is that the these devices cannot do DMA, so don't care about the DMA API or the group->domain, and do not expose MMIO memory so do not care about the P2P attack.

A comment in the code to this effect would be good, IMHO.

Jason

#43 lore raw

Re: [PATCH v5 09/14] PCI: portdrv: Suppress kernel DMA ownership auto-claiming

Quoted 9 lines · Jason Gunthorpe
Quoted 6 lines · Bjorn Helgaas
Quoted 2 lines

The existing vfio framework allows the portdrv driver to be bound to the bridge while its downstream devices are assigned to user space.

I.e., the existing VFIO framework allows a switch to be in the same IOMMU group as the devices below it, even though the switch has a kernel driver and the other devices may have userspace drivers?

Quoted 6 lines

Yes, this patch exists to maintain current VFIO behavior which has this same check.

I belive the basis for VFIO doing this is that the these devices cannot do DMA, so don't care about the DMA API or the group->domain, and do not expose MMIO memory so do not care about the P2P attack.

"These devices" means bridges, right? Not sure why we wouldn't care about the P2P attack.

PCIe switches use MSI or MSI-X for hotplug, PME, etc, so they do DMA for that. Is that not relevant here?

Is there something that *prohibits* a bridge from having device-specific functionality including DMA?

I know some bridges have device-specific BARs for performance counters and the like.

#44 lore raw

Re: [PATCH v5 09/14] PCI: portdrv: Suppress kernel DMA ownership auto-claiming

Quoted 13 lines · Bjorn Helgaas
Quoted 12 lines · Jason Gunthorpe
Quoted 6 lines · Bjorn Helgaas
Quoted 2 lines

The existing vfio framework allows the portdrv driver to be bound to the bridge while its downstream devices are assigned to user space.

I.e., the existing VFIO framework allows a switch to be in the same IOMMU group as the devices below it, even though the switch has a kernel driver and the other devices may have userspace drivers?

Yes, this patch exists to maintain current VFIO behavior which has this same check.

Quoted 6 lines
Quoted 3 lines

I belive the basis for VFIO doing this is that the these devices cannot do DMA, so don't care about the DMA API or the group->domain, and do not expose MMIO memory so do not care about the P2P attack.

"These devices" means bridges, right? Not sure why we wouldn't care about the P2P attack.

Yes bridges. I said "I belive" because VFIO was changed to ignore bridges a long time ago and the security rational was a bit unclear in the commit.

See 5f096b14d421 ("vfio: Whitelist PCI bridges")

Quoted 2 lines

PCIe switches use MSI or MSI-X for hotplug, PME, etc, so they do DMA for that. Is that not relevant here?

Alex's comment in the above commit notes about interrupts, but I think the answer today is that it does not matter.

For platforms like x86 that keep interrupts and DMA seperate it works fine.

For platforms that comingle the iommu_domain and interrupts (do some exist?) we expect the platform to do whatever is necessary at iommu_domain attach time to ensure interrupts continue to work. (AFAICT at least)

In other words we don't have an API restriction to use iommu_device_use_dma_api() to use request_irq().

So the main concern should be P2P attacks on bridge MMIO registers.

Quoted 2 lines

Is there something that *prohibits* a bridge from having device-specific functionality including DMA?

I'm not sure, I think not, but the 'Bus Master Enable' language does have a different definition for Type 1..

However, it doesn't matter - the question here is not about what the device HW is capable of, but what does the kernel driver do. The portdrv does not use the DMA API, so that alone is half the requirement to skip the iommu_device_use_dma_api() call.

Some future device-specific bridge driver that is able to issue the device-specific MMIO's and operate the DMA API should be coded to use iommu_device_use_dma_api(), probably by using a different device_driver for the bridge.

Quoted 2 lines

I know some bridges have device-specific BARs for performance counters and the like.

Yep.

IMHO it is probably not so great as-is..

However, this patch is just porting the status-quo of commit 5f09 into this new framework.

What this new framework does allow is for portdrv to police itself. eg it could check if there is a MMIO BAR and call iommu_device_use_dma_api() out of caution. It could also have an allowance list of devices where we know hostile writes to the MMIO are known harmless.

Without knowing more about what motivated 5f09 it is hard to say, other than it has been 7 years like this and nobody has complained yet :)

Jason

author
#45 lore raw

Re: [PATCH v5 09/14] PCI: portdrv: Suppress kernel DMA ownership auto-claiming

In reply to Bjorn Helgaas

Hi Bjorn,

Quoted 11 lines · Bjorn Helgaas
Quoted 4 lines · Lu Baolu

If a switch lacks ACS P2P Request Redirect, a device below the switch can bypass the IOMMU and DMA directly to other devices below the switch, so all the downstream devices must be in the same IOMMU group as the switch itself.

Help me think through what's going on here. IIUC, we put devices in the same IOMMU group when they can interfere with each other in any way (DMA, config access, etc).

(We said "DMA" above, but I guess this would also apply to config requests, right?)

I am not sure whether devices could interfere each other through config space access. The IOMMU hardware only protects and isolates DMA accesses, so that userspace could control DMA directly. The config accesses will always be intercepted by VFIO. Hence, I don't see a problem.

Quoted 4 lines

*This* patch doesn't check for any ACS features. Can you connect the dots for me? I guess the presence or absence of P2P Request Redirect determines the size of the IOMMU group. And the following says

It's done in iommu core (drivers/iommu/iommu.c):

/*
* Use standard PCI bus topology, isolation features, and DMA alias quirks
* to find or create an IOMMU group for a device.
*/
struct iommu_group *pci_device_group(struct device *dev)

Quoted 3 lines

something about what is allowed in the group? And .no_kernel_api_dma allows an exception to the general rule?

Yes.

Quoted 5 lines
Quoted 3 lines

The pci_dma_configure() marks the iommu_group as containing only devices with kernel drivers that manage DMA. Avoid this default behavior for the portdrv driver in order for compatibility with the current vfio policy.

I assume "IOMMU group" means the same as "iommu_group"; maybe we can use one of them consistently?

Sure.

Best regards,
baolu

#46 lore raw

Re: [PATCH v5 09/14] PCI: portdrv: Suppress kernel DMA ownership auto-claiming

Quoted 10 lines · Lu Baolu
Quoted 9 lines · Bjorn Helgaas
Quoted 4 lines · Lu Baolu

If a switch lacks ACS P2P Request Redirect, a device below the switch can bypass the IOMMU and DMA directly to other devices below the switch, so all the downstream devices must be in the same IOMMU group as the switch itself.

Help me think through what's going on here. IIUC, we put devices in the same IOMMU group when they can interfere with each other in any way (DMA, config access, etc).

Quoted 8 lines
Quoted 2 lines

(We said "DMA" above, but I guess this would also apply to config requests, right?)

I am not sure whether devices could interfere each other through config space access. The IOMMU hardware only protects and isolates DMA accesses, so that userspace could control DMA directly. The config accesses will always be intercepted by VFIO. Hence, I don't see a problem.

I was wondering about config accesses generated by an endpoint, e.g., an endpoint doing config writes to a peer or the upstream bridge.

But I think that is prohibited by spec - PCIe r5.0, sec 7.3.3, says "Propagation of Configuration Requests from Downstream to Upstream as well as peer-to-peer are not supported" and "Configuration Requests are initiated only by the Host Bridge, including those passed through the SFI CAM mechanism."

Bjorn

author
#47 lore raw

Re: [PATCH v5 09/14] PCI: portdrv: Suppress kernel DMA ownership auto-claiming

Quoted 20 lines · Bjorn Helgaas
Quoted 10 lines · Lu Baolu
Quoted 9 lines · Bjorn Helgaas
Quoted 4 lines · Lu Baolu

If a switch lacks ACS P2P Request Redirect, a device below the switch can bypass the IOMMU and DMA directly to other devices below the switch, so all the downstream devices must be in the same IOMMU group as the switch itself.

Help me think through what's going on here. IIUC, we put devices in the same IOMMU group when they can interfere with each other in any way (DMA, config access, etc).

Quoted 8 lines
Quoted 2 lines

(We said "DMA" above, but I guess this would also apply to config requests, right?)

I am not sure whether devices could interfere each other through config space access. The IOMMU hardware only protects and isolates DMA accesses, so that userspace could control DMA directly. The config accesses will always be intercepted by VFIO. Hence, I don't see a problem.

Quoted 8 lines

I was wondering about config accesses generated by an endpoint, e.g., an endpoint doing config writes to a peer or the upstream bridge.

But I think that is prohibited by spec - PCIe r5.0, sec 7.3.3, says "Propagation of Configuration Requests from Downstream to Upstream as well as peer-to-peer are not supported" and "Configuration Requests are initiated only by the Host Bridge, including those passed through the SFI CAM mechanism."

That's clear. Thank you for the clarification.

Quoted 3 lines

Bjorn

Best regards,
baolu

author
#48 lore raw

PATCH v5 10/14 vfio: Set DMA ownership for VFIO devices

In reply to Lu Baolu

Claim group dma ownership when an iommu group is set to a container, and release DMA_OWNER_USER once the iommu group is unset from the container.

---
 drivers/vfio/fsl-mc/vfio_fsl_mc.c     |  1 +
 drivers/vfio/pci/vfio_pci.c           |  1 +
 drivers/vfio/platform/vfio_amba.c     |  1 +
 drivers/vfio/platform/vfio_platform.c |  1 +
 drivers/vfio/vfio.c                   | 10 +++++++++-
 5 files changed, 13 insertions(+), 1 deletion(-)
drivers/vfio/fsl-mc/vfio_fsl_mc.c +1 -0
index 6e2e62c6f47a..f133971d1c1f 100644
@@ -588,6 +588,7 @@ static struct fsl_mc_driver vfio_fsl_mc_driver = {
 		.name	= "vfio-fsl-mc",
 		.owner	= THIS_MODULE,
 	},
+	.no_kernel_api_dma = true,
 };
 
 static int __init vfio_fsl_mc_driver_init(void)
drivers/vfio/pci/vfio_pci.c +1 -0
index a5ce92beb655..8cd1e5a855da 100644
@@ -193,6 +193,7 @@ static struct pci_driver vfio_pci_driver = {
 	.remove			= vfio_pci_remove,
 	.sriov_configure	= vfio_pci_sriov_configure,
 	.err_handler		= &vfio_pci_core_err_handlers,
+	.no_kernel_api_dma = true,
 };
 
 static void __init vfio_pci_fill_ids(void)
drivers/vfio/platform/vfio_amba.c +1 -0
index badfffea14fb..39407ca143d5 100644
@@ -95,6 +95,7 @@ static struct amba_driver vfio_amba_driver = {
 		.name = "vfio-amba",
 		.owner = THIS_MODULE,
 	},
+	.no_kernel_api_dma = true,
 };
 
 module_amba_driver(vfio_amba_driver);
drivers/vfio/platform/vfio_platform.c +1 -0
index 68a1c87066d7..46848a58fc99 100644
@@ -76,6 +76,7 @@ static struct platform_driver vfio_platform_driver = {
 	.driver	= {
 		.name	= "vfio-platform",
 	},
+	.no_kernel_api_dma = true,
 };
 
 module_platform_driver(vfio_platform_driver);
drivers/vfio/vfio.c +9 -1
index 735d1d344af9..ef22fa312c2a 100644
@@ -1198,6 +1198,8 @@ static void __vfio_group_unset_container(struct vfio_group *group)
 		driver->ops->detach_group(container->iommu_data,
 					  group->iommu_group);
 
+	iommu_group_release_dma_owner(group->iommu_group);
+
 	group->container = NULL;
 	wake_up(&group->container_q);
 	list_del(&group->container_next);
@@ -1282,13 +1284,19 @@ static int vfio_group_set_container(struct vfio_group *group, int container_fd)
 		goto unlock_out;
 	}
 
+	ret = iommu_group_set_dma_owner(group->iommu_group, f.file);
+	if (ret)
+		goto unlock_out;
+
 	driver = container->iommu_driver;
 	if (driver) {
 		ret = driver->ops->attach_group(container->iommu_data,
 						group->iommu_group,
 						group->type);
-		if (ret)
+		if (ret) {
+			iommu_group_release_dma_owner(group->iommu_group);
 			goto unlock_out;
+		}
 	}
 
 	group->container = container;
Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author
#49 lore raw

PATCH v5 11/14 vfio: Remove use of vfio_group_viable()

In reply to Lu Baolu

As DMA USER ownership is claimed for the iommu group when a vfio group is added to a vfio container, the vfio group viability is guaranteed as long as group->container_users > 0. Remove those unnecessary group viability checks which are only hit when group->container_users is not zero.

The only remaining reference is in GROUP_GET_STATUS, which could be called at any time when group fd is valid. Here we just replace the vfio_group_viable() by directly calling iommu core to get viability status.

---
 drivers/vfio/vfio.c | 18 ++++++------------
 1 file changed, 6 insertions(+), 12 deletions(-)
drivers/vfio/vfio.c +6 -12
index ef22fa312c2a..789861853676 100644
@@ -1313,12 +1313,6 @@ static int vfio_group_set_container(struct vfio_group *group, int container_fd)
 	return ret;
 }
 
-static bool vfio_group_viable(struct vfio_group *group)
-{
-	return (iommu_group_for_each_dev(group->iommu_group,
-					 group, vfio_dev_viable) == 0);
-}
-
 static int vfio_group_add_container_user(struct vfio_group *group)
 {
 	if (!atomic_inc_not_zero(&group->container_users))
@@ -1328,7 +1322,7 @@ static int vfio_group_add_container_user(struct vfio_group *group)
 		atomic_dec(&group->container_users);
 		return -EPERM;
 	}
-	if (!group->container->iommu_driver || !vfio_group_viable(group)) {
+	if (!group->container->iommu_driver) {
 		atomic_dec(&group->container_users);
 		return -EINVAL;
 	}
@@ -1346,7 +1340,7 @@ static int vfio_group_get_device_fd(struct vfio_group *group, char *buf)
 	int ret = 0;
 
 	if (0 == atomic_read(&group->container_users) ||
-	    !group->container->iommu_driver || !vfio_group_viable(group))
+	    !group->container->iommu_driver)
 		return -EINVAL;
 
 	if (group->type == VFIO_NO_IOMMU && !capable(CAP_SYS_RAWIO))
@@ -1438,11 +1432,11 @@ static long vfio_group_fops_unl_ioctl(struct file *filep,
 
 		status.flags = 0;
 
-		if (vfio_group_viable(group))
-			status.flags |= VFIO_GROUP_FLAGS_VIABLE;
-
 		if (group->container)
-			status.flags |= VFIO_GROUP_FLAGS_CONTAINER_SET;
+			status.flags |= VFIO_GROUP_FLAGS_CONTAINER_SET |
+					VFIO_GROUP_FLAGS_VIABLE;
+		else if (!iommu_group_dma_owner_claimed(group->iommu_group))
+			status.flags |= VFIO_GROUP_FLAGS_VIABLE;
 
 		if (copy_to_user((void __user *)arg, &status, minsz))
 			return -EFAULT;
Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author
#50 lore raw

PATCH v5 12/14 vfio: Delete the unbound_list

In reply to Lu Baolu

From: Jason Gunthorpe <[email protected]>

commit 60720a0fc646 ("vfio: Add device tracking during unbind") added the unbound list to plug a problem with KVM where KVM_DEV_VFIO_GROUP_DEL relied on vfio_group_get_external_user() succeeding to return the vfio_group from a group file descriptor. The unbound list allowed vfio_group_get_external_user() to continue to succeed in edge cases.

However commit 5d6dee80a1e9 ("vfio: New external user group/file match") deleted the call to vfio_group_get_external_user() during KVM_DEV_VFIO_GROUP_DEL. Instead vfio_external_group_match_file() is used to directly match the file descriptor to the group pointer.

This in turn avoids the call down to vfio_dev_viable() during KVM_DEV_VFIO_GROUP_DEL and also avoids the trouble the first commit was trying to fix.

There are no other users of vfio_dev_viable() that care about the time after vfio_unregister_group_dev() returns, so simply delete the unbound_list entirely.

---
 drivers/vfio/vfio.c | 74 ++-------------------------------------------
 1 file changed, 2 insertions(+), 72 deletions(-)
drivers/vfio/vfio.c +2 -72
index 789861853676..2877ec47e32a 100644
@@ -62,11 +62,6 @@ struct vfio_container {
 	bool				noiommu;
 };
 
-struct vfio_unbound_dev {
-	struct device			*dev;
-	struct list_head		unbound_next;
-};
-
 struct vfio_group {
 	struct device 			dev;
 	struct cdev			cdev;
@@ -79,8 +74,6 @@ struct vfio_group {
 	struct notifier_block		nb;
 	struct list_head		vfio_next;
 	struct list_head		container_next;
-	struct list_head		unbound_list;
-	struct mutex			unbound_lock;
 	atomic_t			opened;
 	wait_queue_head_t		container_q;
 	enum vfio_group_type		type;
@@ -340,16 +333,8 @@ vfio_group_get_from_iommu(struct iommu_group *iommu_group)
 static void vfio_group_release(struct device *dev)
 {
 	struct vfio_group *group = container_of(dev, struct vfio_group, dev);
-	struct vfio_unbound_dev *unbound, *tmp;
-
-	list_for_each_entry_safe(unbound, tmp,
-				 &group->unbound_list, unbound_next) {
-		list_del(&unbound->unbound_next);
-		kfree(unbound);
-	}
 
 	mutex_destroy(&group->device_lock);
-	mutex_destroy(&group->unbound_lock);
 	iommu_group_put(group->iommu_group);
 	ida_free(&vfio.group_ida, MINOR(group->dev.devt));
 	kfree(group);
@@ -381,8 +366,6 @@ static struct vfio_group *vfio_group_alloc(struct iommu_group *iommu_group,
 	refcount_set(&group->users, 1);
 	INIT_LIST_HEAD(&group->device_list);
 	mutex_init(&group->device_lock);
-	INIT_LIST_HEAD(&group->unbound_list);
-	mutex_init(&group->unbound_lock);
 	init_waitqueue_head(&group->container_q);
 	group->iommu_group = iommu_group;
 	/* put in vfio_group_release() */
@@ -571,19 +554,8 @@ static int vfio_dev_viable(struct device *dev, void *data)
 	struct vfio_group *group = data;
 	struct vfio_device *device;
 	struct device_driver *drv = READ_ONCE(dev->driver);
-	struct vfio_unbound_dev *unbound;
-	int ret = -EINVAL;
 
-	mutex_lock(&group->unbound_lock);
-	list_for_each_entry(unbound, &group->unbound_list, unbound_next) {
-		if (dev == unbound->dev) {
-			ret = 0;
-			break;
-		}
-	}
-	mutex_unlock(&group->unbound_lock);
-
-	if (!ret || !drv || vfio_dev_driver_allowed(dev, drv))
+	if (!drv || vfio_dev_driver_allowed(dev, drv))
 		return 0;
 
 	device = vfio_group_get_device(group, dev);
@@ -592,7 +564,7 @@ static int vfio_dev_viable(struct device *dev, void *data)
 		return 0;
 	}
 
-	return ret;
+	return -EINVAL;
 }
 
 /*
@@ -634,7 +606,6 @@ static int vfio_iommu_group_notifier(struct notifier_block *nb,
 {
 	struct vfio_group *group = container_of(nb, struct vfio_group, nb);
 	struct device *dev = data;
-	struct vfio_unbound_dev *unbound;
 
 	switch (action) {
 	case IOMMU_GROUP_NOTIFY_ADD_DEVICE:
@@ -663,28 +634,6 @@ static int vfio_iommu_group_notifier(struct notifier_block *nb,
 			__func__, iommu_group_id(group->iommu_group),
 			dev->driver->name);
 		break;
-	case IOMMU_GROUP_NOTIFY_UNBOUND_DRIVER:
-		dev_dbg(dev, "%s: group %d unbound from driver\n", __func__,
-			iommu_group_id(group->iommu_group));
-		/*
-		 * XXX An unbound device in a live group is ok, but we'd
-		 * really like to avoid the above BUG_ON by preventing other
-		 * drivers from binding to it.  Once that occurs, we have to
-		 * stop the system to maintain isolation.  At a minimum, we'd
-		 * want a toggle to disable driver auto probe for this device.
-		 */
-
-		mutex_lock(&group->unbound_lock);
-		list_for_each_entry(unbound,
-				    &group->unbound_list, unbound_next) {
-			if (dev == unbound->dev) {
-				list_del(&unbound->unbound_next);
-				kfree(unbound);
-				break;
-			}
-		}
-		mutex_unlock(&group->unbound_lock);
-		break;
 	}
 	return NOTIFY_OK;
 }
@@ -889,29 +838,10 @@ static struct vfio_device *vfio_device_get_from_name(struct vfio_group *group,
 void vfio_unregister_group_dev(struct vfio_device *device)
 {
 	struct vfio_group *group = device->group;
-	struct vfio_unbound_dev *unbound;
 	unsigned int i = 0;
 	bool interrupted = false;
 	long rc;
 
-	/*
-	 * When the device is removed from the group, the group suddenly
-	 * becomes non-viable; the device has a driver (until the unbind
-	 * completes), but it's not present in the group.  This is bad news
-	 * for any external users that need to re-acquire a group reference
-	 * in order to match and release their existing reference.  To
-	 * solve this, we track such devices on the unbound_list to bridge
-	 * the gap until they're fully unbound.
-	 */
-	unbound = kzalloc(sizeof(*unbound), GFP_KERNEL);
-	if (unbound) {
-		unbound->dev = device->dev;
-		mutex_lock(&group->unbound_lock);
-		list_add(&unbound->unbound_next, &group->unbound_list);
-		mutex_unlock(&group->unbound_lock);
-	}
-	WARN_ON(!unbound);
-
 	vfio_device_put(device);
 	rc = try_wait_for_completion(&device->comp);
 	while (rc <= 0) {
Signature · 1 line
2.25.1
author
#51 lore raw

PATCH v5 13/14 vfio: Remove iommu group notifier

In reply to Lu Baolu

The iommu core and driver core have been enhanced to avoid unsafe driver binding to a live group after iommu_group_set_dma_owner(PRIVATE_USER) has been called. There's no need to register iommu group notifier. This removes the iommu group notifer which contains BUG_ON() and WARN().

The commit 5f096b14d421b ("vfio: Whitelist PCI bridges") allowed all pcieport drivers to be bound with devices while the group is assigned to user space. This is not always safe. For example, The shpchp_core driver relies on the PCI MMIO access for the controller functionality. With its downstream devices assigned to the userspace, the MMIO might be changed through user initiated P2P accesses without any notification. This might break the kernel driver integrity and lead to some unpredictable consequences. As the result, currently we only allow the portdrv driver.

For any bridge driver, in order to avoiding default kernel DMA ownership claiming, we should consider:

1) Does the bridge driver use DMA? Calling pci_set_master() or a dma_map_* API is a sure indicate the driver is doing DMA

2) If the bridge driver uses MMIO, is it tolerant to hostile userspace also touching the same MMIO registers via P2P DMA attacks?

Conservatively if the driver maps an MMIO region at all, we can say that it fails the test.

---
 drivers/vfio/vfio.c | 147 --------------------------------------------
 1 file changed, 147 deletions(-)
drivers/vfio/vfio.c +0 -147
index 2877ec47e32a..dce42ede3190 100644
@@ -71,7 +71,6 @@ struct vfio_group {
 	struct vfio_container		*container;
 	struct list_head		device_list;
 	struct mutex			device_lock;
-	struct notifier_block		nb;
 	struct list_head		vfio_next;
 	struct list_head		container_next;
 	atomic_t			opened;
@@ -274,8 +273,6 @@ void vfio_unregister_iommu_driver(const struct vfio_iommu_driver_ops *ops)
 }
 EXPORT_SYMBOL_GPL(vfio_unregister_iommu_driver);
 
-static int vfio_iommu_group_notifier(struct notifier_block *nb,
-				     unsigned long action, void *data);
 static void vfio_group_get(struct vfio_group *group);
 
 /*
@@ -395,13 +392,6 @@ static struct vfio_group *vfio_create_group(struct iommu_group *iommu_group,
 		goto err_put;
 	}
 
-	group->nb.notifier_call = vfio_iommu_group_notifier;
-	err = iommu_group_register_notifier(iommu_group, &group->nb);
-	if (err) {
-		ret = ERR_PTR(err);
-		goto err_put;
-	}
-
 	mutex_lock(&vfio.group_lock);
 
 	/* Did we race creating this group? */
@@ -422,7 +412,6 @@ static struct vfio_group *vfio_create_group(struct iommu_group *iommu_group,
 
 err_unlock:
 	mutex_unlock(&vfio.group_lock);
-	iommu_group_unregister_notifier(group->iommu_group, &group->nb);
 err_put:
 	put_device(&group->dev);
 	return ret;
@@ -447,7 +436,6 @@ static void vfio_group_put(struct vfio_group *group)
 	cdev_device_del(&group->cdev, &group->dev);
 	mutex_unlock(&vfio.group_lock);
 
-	iommu_group_unregister_notifier(group->iommu_group, &group->nb);
 	put_device(&group->dev);
 }
 
@@ -503,141 +491,6 @@ static struct vfio_device *vfio_group_get_device(struct vfio_group *group,
 	return NULL;
 }
 
-/*
- * Some drivers, like pci-stub, are only used to prevent other drivers from
- * claiming a device and are therefore perfectly legitimate for a user owned
- * group.  The pci-stub driver has no dependencies on DMA or the IOVA mapping
- * of the device, but it does prevent the user from having direct access to
- * the device, which is useful in some circumstances.
- *
- * We also assume that we can include PCI interconnect devices, ie. bridges.
- * IOMMU grouping on PCI necessitates that if we lack isolation on a bridge
- * then all of the downstream devices will be part of the same IOMMU group as
- * the bridge.  Thus, if placing the bridge into the user owned IOVA space
- * breaks anything, it only does so for user owned devices downstream.  Note
- * that error notification via MSI can be affected for platforms that handle
- * MSI within the same IOVA space as DMA.
- */
-static const char * const vfio_driver_allowed[] = { "pci-stub" };
-
-static bool vfio_dev_driver_allowed(struct device *dev,
-				    struct device_driver *drv)
-{
-	if (dev_is_pci(dev)) {
-		struct pci_dev *pdev = to_pci_dev(dev);
-
-		if (pdev->hdr_type != PCI_HEADER_TYPE_NORMAL)
-			return true;
-	}
-
-	return match_string(vfio_driver_allowed,
-			    ARRAY_SIZE(vfio_driver_allowed),
-			    drv->name) >= 0;
-}
-
-/*
- * A vfio group is viable for use by userspace if all devices are in
- * one of the following states:
- *  - driver-less
- *  - bound to a vfio driver
- *  - bound to an otherwise allowed driver
- *  - a PCI interconnect device
- *
- * We use two methods to determine whether a device is bound to a vfio
- * driver.  The first is to test whether the device exists in the vfio
- * group.  The second is to test if the device exists on the group
- * unbound_list, indicating it's in the middle of transitioning from
- * a vfio driver to driver-less.
- */
-static int vfio_dev_viable(struct device *dev, void *data)
-{
-	struct vfio_group *group = data;
-	struct vfio_device *device;
-	struct device_driver *drv = READ_ONCE(dev->driver);
-
-	if (!drv || vfio_dev_driver_allowed(dev, drv))
-		return 0;
-
-	device = vfio_group_get_device(group, dev);
-	if (device) {
-		vfio_device_put(device);
-		return 0;
-	}
-
-	return -EINVAL;
-}
-
-/*
- * Async device support
- */
-static int vfio_group_nb_add_dev(struct vfio_group *group, struct device *dev)
-{
-	struct vfio_device *device;
-
-	/* Do we already know about it?  We shouldn't */
-	device = vfio_group_get_device(group, dev);
-	if (WARN_ON_ONCE(device)) {
-		vfio_device_put(device);
-		return 0;
-	}
-
-	/* Nothing to do for idle groups */
-	if (!atomic_read(&group->container_users))
-		return 0;
-
-	/* TODO Prevent device auto probing */
-	dev_WARN(dev, "Device added to live group %d!\n",
-		 iommu_group_id(group->iommu_group));
-
-	return 0;
-}
-
-static int vfio_group_nb_verify(struct vfio_group *group, struct device *dev)
-{
-	/* We don't care what happens when the group isn't in use */
-	if (!atomic_read(&group->container_users))
-		return 0;
-
-	return vfio_dev_viable(dev, group);
-}
-
-static int vfio_iommu_group_notifier(struct notifier_block *nb,
-				     unsigned long action, void *data)
-{
-	struct vfio_group *group = container_of(nb, struct vfio_group, nb);
-	struct device *dev = data;
-
-	switch (action) {
-	case IOMMU_GROUP_NOTIFY_ADD_DEVICE:
-		vfio_group_nb_add_dev(group, dev);
-		break;
-	case IOMMU_GROUP_NOTIFY_DEL_DEVICE:
-		/*
-		 * Nothing to do here.  If the device is in use, then the
-		 * vfio sub-driver should block the remove callback until
-		 * it is unused.  If the device is unused or attached to a
-		 * stub driver, then it should be released and we don't
-		 * care that it will be going away.
-		 */
-		break;
-	case IOMMU_GROUP_NOTIFY_BIND_DRIVER:
-		dev_dbg(dev, "%s: group %d binding to driver\n", __func__,
-			iommu_group_id(group->iommu_group));
-		break;
-	case IOMMU_GROUP_NOTIFY_BOUND_DRIVER:
-		dev_dbg(dev, "%s: group %d bound to driver %s\n", __func__,
-			iommu_group_id(group->iommu_group), dev->driver->name);
-		BUG_ON(vfio_group_nb_verify(group, dev));
-		break;
-	case IOMMU_GROUP_NOTIFY_UNBIND_DRIVER:
-		dev_dbg(dev, "%s: group %d unbinding from driver %s\n",
-			__func__, iommu_group_id(group->iommu_group),
-			dev->driver->name);
-		break;
-	}
-	return NOTIFY_OK;
-}
-
 /*
  * VFIO driver API
  */
Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author
#52 lore raw

PATCH v5 14/14 iommu: Remove iommu group changes notifier

In reply to Lu Baolu

The iommu group changes notifer is not referenced in the tree. Remove it to avoid dead code.

---
 include/linux/iommu.h | 23 -------------
 drivers/iommu/iommu.c | 75 -------------------------------------------
 2 files changed, 98 deletions(-)
include/linux/iommu.h +0 -23
index 568f285468cf..408a6d2b3034 100644
@@ -403,13 +403,6 @@ static inline void iommu_iotlb_gather_init(struct iommu_iotlb_gather *gather)
 	};
 }
 
-#define IOMMU_GROUP_NOTIFY_ADD_DEVICE		1 /* Device added */
-#define IOMMU_GROUP_NOTIFY_DEL_DEVICE		2 /* Pre Device removed */
-#define IOMMU_GROUP_NOTIFY_BIND_DRIVER		3 /* Pre Driver bind */
-#define IOMMU_GROUP_NOTIFY_BOUND_DRIVER		4 /* Post Driver bind */
-#define IOMMU_GROUP_NOTIFY_UNBIND_DRIVER	5 /* Pre Driver unbind */
-#define IOMMU_GROUP_NOTIFY_UNBOUND_DRIVER	6 /* Post Driver unbind */
-
 extern int bus_set_iommu(struct bus_type *bus, const struct iommu_ops *ops);
 extern int bus_iommu_probe(struct bus_type *bus);
 extern bool iommu_present(struct bus_type *bus);
@@ -482,10 +475,6 @@ extern int iommu_group_for_each_dev(struct iommu_group *group, void *data,
 extern struct iommu_group *iommu_group_get(struct device *dev);
 extern struct iommu_group *iommu_group_ref_get(struct iommu_group *group);
 extern void iommu_group_put(struct iommu_group *group);
-extern int iommu_group_register_notifier(struct iommu_group *group,
-					 struct notifier_block *nb);
-extern int iommu_group_unregister_notifier(struct iommu_group *group,
-					   struct notifier_block *nb);
 extern int iommu_register_device_fault_handler(struct device *dev,
 					iommu_dev_fault_handler_t handler,
 					void *data);
@@ -885,18 +874,6 @@ static inline void iommu_group_put(struct iommu_group *group)
 {
 }
 
-static inline int iommu_group_register_notifier(struct iommu_group *group,
-						struct notifier_block *nb)
-{
-	return -ENODEV;
-}
-
-static inline int iommu_group_unregister_notifier(struct iommu_group *group,
-						  struct notifier_block *nb)
-{
-	return 0;
-}
-
 static inline
 int iommu_register_device_fault_handler(struct device *dev,
 					iommu_dev_fault_handler_t handler,
drivers/iommu/iommu.c +0 -75

125 lines not on this page · raw message

Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
#53 lore raw
In reply to Lu Baolu
Quoted 11 lines · Lu Baolu

v5:
- Move kernel dma ownership auto-claiming from driver core to bus callback. (Greg)
- Refactor the iommu interfaces to make them more specific. (Jason/Robin)
- Simplify the dma ownership implementation by removing the owner type. (Jason)
- Commit message refactoring for PCI drivers. (Bjorn)
- Move iommu_attach/detach_device() improvement patches into another series as there are a lot of code refactoring and cleanup staffs in various device drivers.

Since you already have the code you should make this 'other series' right now. It should delete iommu_group_attach() and fix iommu_device_attach().

You also didn't really do my suggestion, this messes up the normal __iommu_attach_group()/__iommu_detach_group() instead of adding the clear to purpose iommu_replace_group() for VFIO to use. This just makes it more difficult to normalize the APIs.

Otherwise it does seem to have turned out to be more understandable.

Jason

author
#54 lore raw

Hi Jason,

Quoted 14 lines · Jason Gunthorpe
Quoted 11 lines · Lu Baolu

v5:
- Move kernel dma ownership auto-claiming from driver core to bus callback. (Greg)
- Refactor the iommu interfaces to make them more specific. (Jason/Robin)
- Simplify the dma ownership implementation by removing the owner type. (Jason)
- Commit message refactoring for PCI drivers. (Bjorn)
- Move iommu_attach/detach_device() improvement patches into another series as there are a lot of code refactoring and cleanup staffs in various device drivers.

Quoted 3 lines

Since you already have the code you should make this 'other series' right now. It should delete iommu_group_attach() and fix iommu_device_attach().

Yes. I am doing the functional and compile tests. I will post it once I complete the testing.

Quoted 5 lines

You also didn't really do my suggestion, this messes up the normal __iommu_attach_group()/__iommu_detach_group() instead of adding the clear to purpose iommu_replace_group() for VFIO to use. This just makes it more difficult to normalize the APIs.

I didn't forget that. :-) It's part of the new series.

Quoted 5 lines

Otherwise it does seem to have turned out to be more understandable.

Jason

Best regards,
baolu

author
#55 lore raw
In reply to Lu Baolu
Quoted 47 lines · Lu Baolu

Hi folks,

The iommu group is the minimal isolation boundary for DMA. Devices in a group can access each other's MMIO registers via peer to peer DMA and also need share the same I/O address space.

Once the I/O address space is assigned to user control it is no longer available to the dma_map* API, which effectively makes the DMA API non-working.

Second, userspace can use DMA initiated by a device that it controls to access the MMIO spaces of other devices in the group. This allows userspace to indirectly attack any kernel owned device and it's driver.

Therefore groups must either be entirely under kernel control or userspace control, never a mixture. Unfortunately some systems have problems with the granularity of groups and there are a couple of important exceptions:

- pci_stub allows the admin to block driver binding on a device and make it permanently shared with userspace. Since PCI stub does not do DMA it is safe, however the admin must understand that using pci_stub allows userspace to attack whatever device it was bound it.

- PCI bridges are sometimes included in groups. Typically PCI bridges do not use DMA, and generally do not have MMIO regions.

Generally any device that does not have any MMIO registers is a possible candidate for an exception.

Currently vfio adopts a workaround to detect violations of the above restrictions by monitoring the driver core BOUND event, and hardwiring the above exceptions. Since there is no way for vfio to reject driver binding at this point, BUG_ON() is triggered if a violation is captured (kernel driver BOUND event on a group which already has some devices assigned to userspace). Aside from the bad user experience this opens a way for root userspace to crash the kernel, even in high integrity configurations, by manipulating the module binding and triggering the BUG_ON.

This series solves this problem by making the user/kernel ownership a core concept at the IOMMU layer. The driver core enforces kernel ownership while drivers are bound and violations now result in a error codes during probe, not BUG_ON failures.

Patch partitions:

Quoted 8 lines
   [PATCH 1-7]: Detect DMA ownership conflicts during driver binding;
   [PATCH 8-10]: Add security context management for assigned devices;
   [PATCH 11-14]: Various cleanups.

This is also part one of three initial series for IOMMUFD:
* Move IOMMU Group security into the iommu layer
- Generic IOMMUFD implementation
- VFIO ability to consume IOMMUFD

Thank you very much for your comments. A new version of this series has been posted here:

https://lore.kernel.org/linux-iommu/[email protected]/

Best regards,
baolu