Ubuntu-nvidia-bos-7.0-7.0.0-2005.5~24.04.1

NVIDIA: VR: SAUCE: firmware: smccc: lfa: handle LFA_BUSY and improve …

…SMC retry pacing

Align LFA_PRIME and LFA_ACTIVATE polling with DEN0147 and reduce
long udelay-based busy-waiting where the calling context allows it.

LFA_BUSY:
- PRIME (§2.5): unexpected for this driver; log with pr_warn and return
  (concurrent LFA_PRIME on another CPU). No retry loop.
- ACTIVATE (§2.6): activation was postponed; retry within the time
  budget using an implementation-defined delay. Use udelay
  (LFA_SMC_RETRY_SLEEP_MIN_US) for every attempt so the same path is
  valid under stop_machine / IRQs off (usleep_range would not be). On
  budget expiry, return -ETIMEDOUT.

LFA_ACTIVATE call_again:
- drop udelay between SMC calls; progress is driven by re-invoking
  LFA_ACTIVATE. Use cpu_relax() between attempts (include
  linux/preempt.h).

LFA_PRIME call_again:
- replace tight udelay polling with usleep_range() over
  LFA_SMC_RETRY_SLEEP_{MIN,MAX}_US so multi-second priming does not
  burn CPU.

Consolidate timing:
- one 20s wall-clock budget (LFA_SMC_BUDGET_US) for both PRIME and
  ACTIVATE retry loops; shared min/max for PRIME usleep_range and
  ACTIVATE LFA_BUSY udelay length.

Fix prime_fw_image() success path returning an uninitialized value; return 0
on success.

Signed-off-by: Vedashree Vidwans <[email protected]>
Signed-off-by: Nirmoy Das <[email protected]>
Acked-by: Matthew R. Ochs <[email protected]>
Acked-by: Carol L. Soto <[email protected]>
Signed-off-by: Matthew R. Ochs <[email protected]>

iommu: Skip PASID validation for devices without PASID capability

Generally PASID support requires ACS settings that usually create
single device groups, but there are some niche cases where we can get
multi-device groups and still have working PASID support. The primary
issue is that PCI switches are not required to treat PASID tagged TLPs
specially so appropriate ACS settings are required to route all TLPs to
the host bridge if PASID is going to work properly.

pci_enable_pasid() does check that each device that will use PASID has
the proper ACS settings to achieve this routing.

However, no-PASID devices can be combined with PASID capable devices
within the same topology using non-uniform ACS settings. In this case
the no-PASID devices may not have strict route to host ACS flags and
end up being grouped with the PASID devices.

This configuration fails to allow use of the PASID within the iommu
core code which wrongly checks if the no-PASID device supports PASID.

Fix this by ignoring no-PASID devices during the PASID validation. They
will never issue a PASID TLP anyhow so they can be ignored.

Fixes: c404f55 ("iommu: Validate the PASID in iommu_attach_device_pasid()")
Cc: [email protected]
(backported from commit b3f6fcd)
Signed-off-by: Tushar Dave <[email protected]>
Reviewed-by: Lu Baolu <[email protected]>
Reviewed-by: Vasant Hegde <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Joerg Roedel <[email protected]>
Acked-by: Koba Ko <[email protected]>
Acked-by: Matthew R. Ochs <[email protected]>
Signed-off-by: Matthew R. Ochs <[email protected]>

Ubuntu-nvidia-7.0-7.0.0-1006.6~24.04.1

Ubuntu-nvidia-7.0-7.0.0-1005.5~24.04.6

Ubuntu-nvidia-bos-7.0-7.0.0-2004.4~24.04.3

Ubuntu-nvidia-7.0-7.0.0-1005.5~24.04.5

NVIDIA: VR: SAUCE: firmware: smccc: lfa: handle LFA_BUSY and improve …

…SMC retry pacing

Align LFA_PRIME and LFA_ACTIVATE polling with DEN0147 and reduce
long udelay-based busy-waiting where the calling context allows it.

LFA_BUSY:
- PRIME (§2.5): unexpected for this driver; log with pr_warn and return
  (concurrent LFA_PRIME on another CPU). No retry loop.
- ACTIVATE (§2.6): activation was postponed; retry within the time
  budget using an implementation-defined delay. Use udelay
  (LFA_SMC_RETRY_SLEEP_MIN_US) for every attempt so the same path is
  valid under stop_machine / IRQs off (usleep_range would not be). On
  budget expiry, return -ETIMEDOUT.

LFA_ACTIVATE call_again:
- drop udelay between SMC calls; progress is driven by re-invoking
  LFA_ACTIVATE. Use cpu_relax() between attempts (include
  linux/preempt.h).

LFA_PRIME call_again:
- replace tight udelay polling with usleep_range() over
  LFA_SMC_RETRY_SLEEP_{MIN,MAX}_US so multi-second priming does not
  burn CPU.

Consolidate timing:
- one 20s wall-clock budget (LFA_SMC_BUDGET_US) for both PRIME and
  ACTIVATE retry loops; shared min/max for PRIME usleep_range and
  ACTIVATE LFA_BUSY udelay length.

Fix prime_fw_image() success path returning an uninitialized value; return 0
on success.

Signed-off-by: Vedashree Vidwans <[email protected]>
Signed-off-by: Nirmoy Das <[email protected]>
Acked-by: Matthew R. Ochs <[email protected]>
Acked-by: Carol L. Soto <[email protected]>
Signed-off-by: Matthew R. Ochs <[email protected]>

iommu: Skip PASID validation for devices without PASID capability

Generally PASID support requires ACS settings that usually create
single device groups, but there are some niche cases where we can get
multi-device groups and still have working PASID support. The primary
issue is that PCI switches are not required to treat PASID tagged TLPs
specially so appropriate ACS settings are required to route all TLPs to
the host bridge if PASID is going to work properly.

pci_enable_pasid() does check that each device that will use PASID has
the proper ACS settings to achieve this routing.

However, no-PASID devices can be combined with PASID capable devices
within the same topology using non-uniform ACS settings. In this case
the no-PASID devices may not have strict route to host ACS flags and
end up being grouped with the PASID devices.

This configuration fails to allow use of the PASID within the iommu
core code which wrongly checks if the no-PASID device supports PASID.

Fix this by ignoring no-PASID devices during the PASID validation. They
will never issue a PASID TLP anyhow so they can be ignored.

Fixes: c404f55 ("iommu: Validate the PASID in iommu_attach_device_pasid()")
Cc: [email protected]
(backported from commit b3f6fcd)
Signed-off-by: Tushar Dave <[email protected]>
Reviewed-by: Lu Baolu <[email protected]>
Reviewed-by: Vasant Hegde <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Joerg Roedel <[email protected]>
Acked-by: Koba Ko <[email protected]>
Acked-by: Matthew R. Ochs <[email protected]>
Signed-off-by: Matthew R. Ochs <[email protected]>

NVIDIA: SAUCE: ci: add github action

Signed-off-by: Andrea Righi <[email protected]>