Re: [PATCH v5] iommu/amd: Enable PerfOpt automatically for integrated GPUs

From: Mario Limonciello

Date: Wed Nov 18 2026 - 13:54:37 EST


On 11/18/26 12:50, Jason Gunthorpe wrote:
PerfOpt is usable by integrated GPUs in identity mode. Enable it from
the IOMMU driver's domain attachment paths instead of requiring an
amdgpu policy knob and exported enable/disable interfaces.

Mark supported integrated GPUs as PerfOpt-eligible when selecting their
default identity domain. Skip GCR3 setup and ATS/PRI/PASID enablement for
these identity attachments, then enable PerfOpt after installing the DTE.
Clear PerfOpt before attaching an eligible GPU to a translated or blocked
domain, including unmanaged domains used by VFIO. Propagate failures to
clear bypass, but keep a completed identity attachment successful if the
optional PerfOpt enable fails.

Clear inherited PerfOpt state during initialization, including when a crash
kernel reuses device tables. Read feature support directly from hardware
because cached features may not be available yet, and stop initialization
if the initial clear fails.

Each IOMMU has at most one PerfOpt-eligible GPU, so track activation with a
boolean under the IOMMU lock. Preserve this state while clearing hardware
at suspend and shutdown, and restore the bit on resume only where it was
active. Do not enable PerfOpt unconditionally during early IOMMU
initialization.

Remove the amdgpu parameter, driver callbacks, and exported interfaces.
Translated attachments retain their normal device capability setup.

Suggested-by: Jason Gunthorpe <jgg@xxxxxxxx>
Signed-off-by: Mario Limonciello (AMD) <superm1@xxxxxxxxxx>
---
Changes in v5:
* Enable PerfOpt after identity attachment instead of during early IOMMU init
* Clear PerfOpt before translated or blocked attachments, allowing VFIO use
* Drop require_direct and retain normal capability setup for translated domains
* Replace refcounting with a per-IOMMU activation boolean protected by its lock
* Clear inherited PerfOpt during init, including crash-kernel device-table reuse
* Read hardware feature support before cached features are available and fail
initialization if the initial clear fails
* Preserve activation across suspend and restore it on resume only if active
* Keep identity attachment successful if optional PerfOpt enablement fails
* Link to v4: https://lore.kernel.org/linux-iommu/20261001182410.1525285-1-superm1@xxxxxxxxxx/
---
drivers/gpu/drm/amd/amdgpu/amdgpu.h | 1 -
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 50 ------
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 12 --
drivers/iommu/amd/amd_iommu.h | 3 +-
drivers/iommu/amd/amd_iommu_types.h | 6 +-
drivers/iommu/amd/init.c | 91 ++++++-----
drivers/iommu/amd/iommu.c | 252 +++++------------------------
include/linux/amd-iommu.h | 11 --
8 files changed, 95 insertions(+), 331 deletions(-)

Vasant are you going to pick this up? It seems good to me

Reviewed-by: Jason Gunthorpe <jgg@xxxxxxxxxx>


FWIW - it was picked up in b32ecac779bba6e7ac7caeae8c40db4280941634.