On the Lenovo X13s the named components don't have the single mapping
flag set, which we so far had expected to be there. This is allowed by
the IORT spec. In that case simply make use of the first

On the Lenovo X13s the named components don't have the single mapping
flag set, which we so far had expected to be there. This is allowed by
the IORT spec. In that case simply make use of the first mapping if
there is a single one.

With feedback from and ok kettenis@

show more ...

Compare equality of looked up node using pointers instead of
strings, as the component name provided in the IORT might be
specified in a different way than our ACPI stack would do.

ok kettenis@

Support SMMUv3 IORT nodes as well in the midlayers. This allows IOMMU
mappings to reach a future SMMUv3 implementation.

ok kettenis@, mlarkin@

constify struct cfattach

The IORT table usually only contains single mappings. While on the LX2K there
is a non-single mapping for the networking subsystem, we don't yet have support
for it and apparently Linux has trouble

The IORT table usually only contains single mappings. While on the LX2K there
is a non-single mapping for the networking subsystem, we don't yet have support
for it and apparently Linux has trouble with it as well. So for now we can
remove the code, because it uses an uninitialized variable and there's no easy
way to fix it without re-thinking the concept. The code got there in the first
place because it's a copy of the same pattern in our other IOMMU code.

ok jsg@

show more ...

While it seems like we can choose any I/O virtual address for peripheral
devices, this isn't really the case. It depends on the bus topology of
how devices are connected. In the case of PCIe, devic

While it seems like we can choose any I/O virtual address for peripheral
devices, this isn't really the case. It depends on the bus topology of
how devices are connected. In the case of PCIe, devices are assigned
addresses (in PCI BARs) from the PCI address spaces. Now if we take an
address from one of these address spaces for our IOVA, transfers from
from a PCI device to that address will terminate inside of the PCI bus.
This is because from the PCI buses' point-of-view, the address we chose
is part of its address space. To make sure we don't allocate addresses
from there, reserve the PCI addresses in the IOVA.

Note that smmu(4) currently gives each device its own IOVA. So the PCI
addresses will be reserved only in IOVA from PCI devices, and only the
addresses concerning the PCI bus it is connected to will be reserved.
All other devices behind an smmu(4) will not have any changes to their
IOVA.

ok kettenis@

show more ...

Add code to acpiiort(4) to look up named components in the IORT and
map them. This makes ACPI's call to acpi_iommu_device_map() do work
through acpiiort(4).

ok kettenis@

Change API of acpiiort(4). It was written as a hook before, taking the
PCI attach args and replacing the DMA tag inside. Our other IOMMU API
though takes a DMA tag and returns the old one or a new

Change API of acpiiort(4). It was written as a hook before, taking the
PCI attach args and replacing the DMA tag inside. Our other IOMMU API
though takes a DMA tag and returns the old one or a new one. To have
acpiiort(4) integrate better with non-PCI ACPI devices, change the API
so that it is more similar to the other API. This also makes the code
easier to understand.

ok kettenis@

show more ...

Add acpiiort(4), a driver for the ACPI I/O Remapping Table. This table
contains information which IOMMUs we have and how the devices are routed.

ok kettenis@