[v13,00/10] iommu: I/O page faults for SMMUv3

Message

Jean-Philippe Brucker March 2, 2021, 9:26 a.m. UTC

Add stall support to the SMMUv3 driver, along with a common I/O Page
Fault handler.

Since v12 [1]:
* Fixed failure path of arm_smmu_insert_master(), in patch 07 (Keqian
  Zhu)
* In arm_smmu_handle_evt(), patch 10, don't report IPA field on stage-1
  faults, and report accurate fault reason (Eric Auger)
* Fix possible use-after-free in arm_smmu_handle_evt(), patch 10: if a
  master is removed while we handle its events, we could in theory
  dereference a freed master struct. Hold streams_mutex while using a
  master struct obtained with arm_smmu_find_master().


Future work regarding IOPF:
* Keep stall disabled by default, only enable it per CD when drivers
  request it [2][3].
* Add PRI support to SMMUv3.
* Route all recoverable faults through io-pgfault.c, so we can track
  partial faults better [4].
* Nested IOPF [5].

[1] https://lore.kernel.org/linux-iommu/[e-mail]/
[2] https://lore.kernel.org/linux-iommu/[e-mail]/
[3] https://lore.kernel.org/linux-iommu/YAhui7UOw7743shI@myrica/
[4] https://lore.kernel.org/kvm/YB0f5Yno9frihQq4@myrica/
[5] https://lore.kernel.org/linux-acpi/[e-mail]/

Jean-Philippe Brucker (10):
  iommu: Fix comment for struct iommu_fwspec
  iommu/arm-smmu-v3: Use device properties for pasid-num-bits
  iommu: Separate IOMMU_DEV_FEAT_IOPF from IOMMU_DEV_FEAT_SVA
  iommu/vt-d: Support IOMMU_DEV_FEAT_IOPF
  uacce: Enable IOMMU_DEV_FEAT_IOPF
  iommu: Add a page fault handler
  iommu/arm-smmu-v3: Maintain a SID->device structure
  dt-bindings: document stall property for IOMMU masters
  ACPI/IORT: Enable stall support for platform devices
  iommu/arm-smmu-v3: Add stall support for platform devices

 drivers/iommu/Makefile                        |   1 +
 .../devicetree/bindings/iommu/iommu.txt       |  18 +
 drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h   |  56 ++-
 drivers/iommu/iommu-sva-lib.h                 |  53 ++
 include/linux/iommu.h                         |  26 +-
 drivers/acpi/arm64/iort.c                     |  15 +-
 .../iommu/arm/arm-smmu-v3/arm-smmu-v3-sva.c   |  59 ++-
 drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c   | 354 ++++++++++++--
 drivers/iommu/intel/iommu.c                   |  11 +-
 drivers/iommu/io-pgfault.c                    | 461 ++++++++++++++++++
 drivers/iommu/of_iommu.c                      |   5 -
 drivers/misc/uacce/uacce.c                    |  39 +-
 12 files changed, 1024 insertions(+), 74 deletions(-)
 create mode 100644 drivers/iommu/io-pgfault.c

Comments

Krishna Reddy March 18, 2021, 12:25 a.m. UTC | #1

Tested-by: Krishna Reddy <[e-mail]>

Validated v13 patches in context of nested translations validation for NVMe PCIe device assigned to Guest VM.
V12 patches(v13 is yet to be tested) has been tested for SVA functionality on Native OS and is functional. 

-KR

Jean-Philippe Brucker March 30, 2021, 5:17 p.m. UTC | #2

Hi Will, Joerg,

On Tue, Mar 02, 2021 at 10:26:36AM +0100, Jean-Philippe Brucker wrote:
> Add stall support to the SMMUv3 driver, along with a common I/O Page
> Fault handler.

I only have review/ack tags and one assert_lockdep to add to this series.
Should I send a v14 now or wait a little longer?  Just making sure we can
get at least patches 1-6 in v5.13.

Thanks,
Jean

Will Deacon April 1, 2021, 8:57 a.m. UTC | #3

On Tue, Mar 30, 2021 at 07:17:55PM +0200, Jean-Philippe Brucker wrote:
> On Tue, Mar 02, 2021 at 10:26:36AM +0100, Jean-Philippe Brucker wrote:
> > Add stall support to the SMMUv3 driver, along with a common I/O Page
> > Fault handler.
> 
> I only have review/ack tags and one assert_lockdep to add to this series.
> Should I send a v14 now or wait a little longer?  Just making sure we can
> get at least patches 1-6 in v5.13.

Easiest thing is probably to send a v14 containing just the patches
targetting v5.13, so that Joerg can queue them up.

Will