mbox series

[v2,00/10] IOMMU memory observability

Message ID 20231130201504.2322355-1-pasha.tatashin@soleen.com (mailing list archive)
Headers show
Series IOMMU memory observability | expand

Message

Pasha Tatashin Nov. 30, 2023, 8:14 p.m. UTC
From: Pasha Tatashin <tatashin@google.com>

----------------------------------------------------------------------
Changelog
----------------------------------------------------------------------
v2:
- Added Reviewed-by Janne Grunau
- Sync with 6.7.0-rc3, 3b47bc037bd44f142ac09848e8d3ecccc726be99
- Separated form the series patches:
vhost-vdpa: account iommu allocations
https://lore.kernel.org/all/20231130200447.2319543-1-pasha.tatashin@soleen.com
vfio: account iommu allocations
https://lore.kernel.org/all/20231130200900.2320829-1-pasha.tatashin@soleen.com
as suggested by Jason Gunthorpe
- Fixed SPARC build issue detected by kernel test robot
- Drop the following patches as they do account iommu page tables:
iommu/dma: use page allocation function provided by iommu-pages.h
iommu/fsl: use page allocation function provided by iommu-pages.h
iommu/iommufd: use page allocation function provided by iommu-pages.h
as suggested by Robin Murphy. These patches are not related to IOMMU
page tables. We might need to do a separate work to support DMA
observability.
- Remove support iommu/io-pgtable-arm-v7s as the 2nd level pages are
under a page size, thanks Robin Murphy for pointing this out.

----------------------------------------------------------------------
Description
----------------------------------------------------------------------
IOMMU subsystem may contain state that is in gigabytes. Majority of that
state is iommu page tables. Yet, there is currently, no way to observe
how much memory is actually used by the iommu subsystem.

This patch series solves this problem by adding both observability to
all pages that are allocated by IOMMU, and also accountability, so
admins can limit the amount if via cgroups.

The system-wide observability is using /proc/meminfo:
SecPageTables:    438176 kB

Contains IOMMU and KVM memory.

Per-node observability:
/sys/devices/system/node/nodeN/meminfo
Node N SecPageTables:    422204 kB

Contains IOMMU and KVM memory memory in the given NUMA node.

Per-node IOMMU only observability:
/sys/devices/system/node/nodeN/vmstat
nr_iommu_pages 105555

Contains number of pages IOMMU allocated in the given node.

Accountability: using sec_pagetables cgroup-v2 memory.stat entry.

With the change, iova_stress[1] stops as limit is reached:

# ./iova_stress
iova space:     0T      free memory:   497G
iova space:     1T      free memory:   495G
iova space:     2T      free memory:   493G
iova space:     3T      free memory:   491G

stops as limit is reached.

This series encorporates suggestions that came from the discussion
at LPC [2].
----------------------------------------------------------------------
[1] https://github.com/soleen/iova_stress
[2] https://lpc.events/event/17/contributions/1466
----------------------------------------------------------------------
Previous versions
v1: https://lore.kernel.org/all/20231128204938.1453583-1-pasha.tatashin@soleen.com
----------------------------------------------------------------------

Pasha Tatashin (10):
  iommu/vt-d: add wrapper functions for page allocations
  iommu/amd: use page allocation function provided by iommu-pages.h
  iommu/io-pgtable-arm: use page allocation function provided by
    iommu-pages.h
  iommu/io-pgtable-dart: use page allocation function provided by
    iommu-pages.h
  iommu/exynos: use page allocation function provided by iommu-pages.h
  iommu/rockchip: use page allocation function provided by iommu-pages.h
  iommu/sun50i: use page allocation function provided by iommu-pages.h
  iommu/tegra-smmu: use page allocation function provided by
    iommu-pages.h
  iommu: observability of the IOMMU allocations
  iommu: account IOMMU allocated memory

 Documentation/admin-guide/cgroup-v2.rst |   2 +-
 Documentation/filesystems/proc.rst      |   4 +-
 drivers/iommu/amd/amd_iommu.h           |   8 -
 drivers/iommu/amd/init.c                |  91 +++++-----
 drivers/iommu/amd/io_pgtable.c          |  13 +-
 drivers/iommu/amd/io_pgtable_v2.c       |  20 +-
 drivers/iommu/amd/iommu.c               |  13 +-
 drivers/iommu/exynos-iommu.c            |  14 +-
 drivers/iommu/intel/dmar.c              |  10 +-
 drivers/iommu/intel/iommu.c             |  47 ++---
 drivers/iommu/intel/iommu.h             |   2 -
 drivers/iommu/intel/irq_remapping.c     |  10 +-
 drivers/iommu/intel/pasid.c             |  12 +-
 drivers/iommu/intel/svm.c               |   7 +-
 drivers/iommu/io-pgtable-arm.c          |   7 +-
 drivers/iommu/io-pgtable-dart.c         |  37 ++--
 drivers/iommu/iommu-pages.h             | 231 ++++++++++++++++++++++++
 drivers/iommu/rockchip-iommu.c          |  14 +-
 drivers/iommu/sun50i-iommu.c            |   7 +-
 drivers/iommu/tegra-smmu.c              |  18 +-
 include/linux/mmzone.h                  |   5 +-
 mm/vmstat.c                             |   3 +
 22 files changed, 390 insertions(+), 185 deletions(-)
 create mode 100644 drivers/iommu/iommu-pages.h

Comments

David Rientjes Dec. 24, 2023, 9:49 p.m. UTC | #1
On Thu, 30 Nov 2023, Pasha Tatashin wrote:

> IOMMU subsystem may contain state that is in gigabytes. Majority of that
> state is iommu page tables. Yet, there is currently, no way to observe
> how much memory is actually used by the iommu subsystem.
> 
> This patch series solves this problem by adding both observability to
> all pages that are allocated by IOMMU, and also accountability, so
> admins can limit the amount if via cgroups.
> 
> The system-wide observability is using /proc/meminfo:
> SecPageTables:    438176 kB
> 
> Contains IOMMU and KVM memory.
> 
> Per-node observability:
> /sys/devices/system/node/nodeN/meminfo
> Node N SecPageTables:    422204 kB
> 
> Contains IOMMU and KVM memory memory in the given NUMA node.
> 
> Per-node IOMMU only observability:
> /sys/devices/system/node/nodeN/vmstat
> nr_iommu_pages 105555
> 
> Contains number of pages IOMMU allocated in the given node.
> 
> Accountability: using sec_pagetables cgroup-v2 memory.stat entry.
> 
> With the change, iova_stress[1] stops as limit is reached:
> 
> # ./iova_stress
> iova space:     0T      free memory:   497G
> iova space:     1T      free memory:   495G
> iova space:     2T      free memory:   493G
> iova space:     3T      free memory:   491G
> 
> stops as limit is reached.
> 

I think this is *very* useful to provide visibility into a significant 
amount of memory that we currently cannot observe on a host.  It can help 
to uncover bugs and shed light onto a particularly large amount of memory 
that would otherwise be mysterious.

Joerg, Will, Robin, I think this series would go through the 
git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu.git tree since it 
depends on a common framework for all other IOMMU implementations to then 
use?

Any concerns about this patch series?  It would be very useful for us to 
create visibility into this memory.


Pasha: any chance of adding a selftest that can be run that will test the 
value of nr_iommu_pages?  I could imagine in the future that a bug could 
be introduced where either an allocation or free is done through 
alloc_pages() directly and its paired alloc/free function now results in a 
leak or underflow.

> This series encorporates suggestions that came from the discussion
> at LPC [2].
> ----------------------------------------------------------------------
> [1] https://github.com/soleen/iova_stress
> [2] https://lpc.events/event/17/contributions/1466
> ----------------------------------------------------------------------
> Previous versions
> v1: https://lore.kernel.org/all/20231128204938.1453583-1-pasha.tatashin@soleen.com
> ----------------------------------------------------------------------
> 
> Pasha Tatashin (10):
>   iommu/vt-d: add wrapper functions for page allocations
>   iommu/amd: use page allocation function provided by iommu-pages.h
>   iommu/io-pgtable-arm: use page allocation function provided by
>     iommu-pages.h
>   iommu/io-pgtable-dart: use page allocation function provided by
>     iommu-pages.h
>   iommu/exynos: use page allocation function provided by iommu-pages.h
>   iommu/rockchip: use page allocation function provided by iommu-pages.h
>   iommu/sun50i: use page allocation function provided by iommu-pages.h
>   iommu/tegra-smmu: use page allocation function provided by
>     iommu-pages.h
>   iommu: observability of the IOMMU allocations
>   iommu: account IOMMU allocated memory
> 
>  Documentation/admin-guide/cgroup-v2.rst |   2 +-
>  Documentation/filesystems/proc.rst      |   4 +-
>  drivers/iommu/amd/amd_iommu.h           |   8 -
>  drivers/iommu/amd/init.c                |  91 +++++-----
>  drivers/iommu/amd/io_pgtable.c          |  13 +-
>  drivers/iommu/amd/io_pgtable_v2.c       |  20 +-
>  drivers/iommu/amd/iommu.c               |  13 +-
>  drivers/iommu/exynos-iommu.c            |  14 +-
>  drivers/iommu/intel/dmar.c              |  10 +-
>  drivers/iommu/intel/iommu.c             |  47 ++---
>  drivers/iommu/intel/iommu.h             |   2 -
>  drivers/iommu/intel/irq_remapping.c     |  10 +-
>  drivers/iommu/intel/pasid.c             |  12 +-
>  drivers/iommu/intel/svm.c               |   7 +-
>  drivers/iommu/io-pgtable-arm.c          |   7 +-
>  drivers/iommu/io-pgtable-dart.c         |  37 ++--
>  drivers/iommu/iommu-pages.h             | 231 ++++++++++++++++++++++++
>  drivers/iommu/rockchip-iommu.c          |  14 +-
>  drivers/iommu/sun50i-iommu.c            |   7 +-
>  drivers/iommu/tegra-smmu.c              |  18 +-
>  include/linux/mmzone.h                  |   5 +-
>  mm/vmstat.c                             |   3 +
>  22 files changed, 390 insertions(+), 185 deletions(-)
>  create mode 100644 drivers/iommu/iommu-pages.h
> 
> -- 
> 2.43.0.rc2.451.g8631bc7472-goog
> 
> 
>