mirror of
https://github.com/torvalds/linux
synced 2024-10-04 10:26:40 +00:00
b6900aeb19
The current cpu addr fixup mask for ARTPEC-6, GENMASK(27, 0), is wrong.
The correct cpu addr fixup mask for ARTPEC-6 is GENMASK(28, 0).
However, having a hardcoded cpu addr fixup mask in each driver is
arguably wrong.
A device tree property called something like "cpu-addr-fixup-mask"
would have been a better solution.
Introducing such a property is not needed though, since we already have
pp->cfg0_base and ep->phys_base, which is derived from already existing
device tree properties.
It is also worth noting that for ARTPEC-7, hardcoding the cpu addr fixup
mask is not possible, since it uses a High Address Bits Look Up Table,
which means that it can, at runtime, map the PCIe window to an arbitrary
address in the 32-bit address space.
By using pp->cfg0_base and ep->phys_base, we avoid hardcoding a mask
in each driver. This should work for ARTPEC-6, DRA7xx, and ARTPEC-7.
I have not changed the code in DRA7xx though, since their existing
code works, but if they want, they could use the same logic as
artpec6_pcie_cpu_addr_fixup, and thus remove their hardcoded mask.
The reason why the fixup mask is needed is explained in commit
|
||
|---|---|---|
| .. | ||
| Kconfig | ||
| Makefile | ||
| pci-dra7xx.c | ||
| pci-exynos.c | ||
| pci-imx6.c | ||
| pci-keystone-dw.c | ||
| pci-keystone.c | ||
| pci-keystone.h | ||
| pci-layerscape.c | ||
| pcie-armada8k.c | ||
| pcie-artpec6.c | ||
| pcie-designware-ep.c | ||
| pcie-designware-host.c | ||
| pcie-designware-plat.c | ||
| pcie-designware.c | ||
| pcie-designware.h | ||
| pcie-hisi.c | ||
| pcie-histb.c | ||
| pcie-kirin.c | ||
| pcie-qcom.c | ||
| pcie-spear13xx.c | ||