On Fri, Oct 25, 2024 at 05:05:03PM -0400, Frank Li wrote: > On Fri, Oct 25, 2024 at 03:48:18PM -0500, Bjorn Helgaas wrote: > > On Thu, Oct 24, 2024 at 04:41:42PM -0400, Frank Li wrote: > > > Endpoint Root complex > > > ┌───────┐ ┌─────────┐ > > > ┌─────┐ │ EP │ │ │ ┌─────┐ > > > │ │ │ Ctrl │ │ │ │ CPU │ > > > │ DDR │ │ │ │ ┌────┐ │ └──┬──┘ > > > │ │◄──────┼─ATU ◄─┼────────┼─┤BarN│◄─┼─────────┘ > > > │ │ │ │ │ └────┘ │ Outbound Transfer > > > └─────┘ │ │ │ │ > > > │ │ │ │ > > > │ │ │ │ > > > │ │ │ │ Inbound Transfer > > > │ │ │ │ ┌──▼──┐ > > > ┌───────┐ │ │ │ ┌───────┼─────►│DDR │ > > > │ │ outbound Transfer* │ │ │ └─────┘ > > > ┌─────┐ │ Bus ┼─────►│ ATU ─┬────────┼─┘ │ > > > │ │ │ Fabric│Bus │ │ PCI Addr │ > > > │ CPU ├───►│ │Addr │ │ 0xA000_0000 │ > > > │ │CPU │ │0x8000_0000 │ │ │ > > > └─────┘Addr└───────┘ │ │ │ │ > > > 0x7000_0000 └───────┘ └─────────┘ > > > > > > Add `bus_addr_base` to configure the outbound window address for CPU write. > > > The BUS fabric generally passes the same address to the PCIe EP controller, > > > but some BUS fabrics convert the address before sending it to the PCIe EP > > > controller. > > > > > > Above diagram, CPU write data to outbound windows address 0x7000_0000, > > > Bus fabric convert it to 0x8000_0000. ATU should use BUS address > > > 0x8000_0000 as input address and convert to PCI address 0xA000_0000. > > > > The above doesn't match what's in patch 1/4, and I think the version > > in 1/4 is better, so I'll comment there. > > > > To avoid confusion, it might be better not to duplicate it in 0/4 and > > 1/4. > > Yes, cover letter don't come into git tree. This part is common and > important, It is not good just said ref to patch1 commit message. > > Add do you have addition comment about this and > https://lore.kernel.org/imx/20241015-pci_fixup_addr-v5-0-ced556c85270@xxxxxxx/T/#t > > The both are the pave the road to clean up pci_fixup_addr(). I think it would be helpful to combine the "PCI: dwc: optimize RC host pci_fixup_addr()" series and the "bus_addr_base" parts of this series together into a single series because they are doing very similar things, and it's easier to review them together. And split the dt-bindings, PHY sub mode, and new endpoint support parts to their own series because they're not related to the address translation changes. Bjorn
