Date
1 - 5 of 5
[PATCH v2 1/1] server extension: PCIe requirements
Mayuresh Chitale
This patch adds requirements for PCIe support for the server extension
Signed-off-by: Mayuresh Chitale <mchitale@...> --- riscv-platform-spec.adoc | 174 ++++++++++++++++++++++++++++++++++++++- 1 file changed, 172 insertions(+), 2 deletions(-) diff --git a/riscv-platform-spec.adoc b/riscv-platform-spec.adoc index 4418788..e738585 100644 --- a/riscv-platform-spec.adoc +++ b/riscv-platform-spec.adoc @@ -47,7 +47,21 @@ include::profiles.adoc[] |RVA22 | RISC-V Application 2022 |EE | Execution Environment |RV32GC | RISC-V 32-bit general purpose ISA described as RV32IMAFDC. -|RV64GC | RISC-V 64-bit general purpose ISA described as RV64IMAFDC. +|RV64GC | RISC-V 64-bit general purpose ISA described as RV64IMAFDC. +|PCIe | PCI Express +|ECAM | Enhanced Configuration Access Mechanism +|BAR | Base Address Register +|AER | Advanced Error Reporting +|CRS | Configuration Request Retry Status +|TLP | Transaction Layer Packet +|RCiEP | Root Complex Integrated Endpoint +|RCEC | Root Complex Event Collector +|PME | Power Management Event +|MSI | Message Signaled Interrupts +|MSI-X | Enhanced Message Signaled Interrupts +|INTx | PCIe Legacy Interrupts +|PMA | Physical Memory Attributes +|PBMT | Page Based Memory Types |=== === Specifications @@ -363,7 +377,163 @@ https://lists.riscv.org/g/tech-privileged/message/404[Sstc] extension. ** Platforms are required to delegate the supervisor timer interrupt to 'S' mode. If the 'H' extension is implemented then the platforms are required to delegate the virtual supervisor timer interrupt to 'VS' mode. -* PCI-E + +===== PCIe +Platforms are required to support at least PCIe Base Specification Revision 1.1 +footnote:[https://pcisig.com/specifications]. + +====== PCIe Config Space +* Platforms shall support access to the PCIe config space via ECAM as described +in the PCIe Base specification. +* The entire config space for a single PCIe domain should be accessible via a +single ECAM I/O region. +* Platform firmware should implement the MCFG table to allow the operating +systems to discover the supported PCIe domains and map the ECAM I/O region for +each domain. +* Platform software shall configure ECAM I/O regions such that the effective +memory type (PMA + PBMT) is UC. + +====== PCIe Memory Space +Platforms are required to map PCIe address space directly in the system address +space and not have any address translation for outbound accesses from harts or +for inbound accesses to any component in the system address space + +* PCIe Outbound Memory + +PCIe devices and bridges/switches frequently implement BARs which only support +32-bit addressing or support 64 bit addressing but do not support prefetchable +memory. To support mapping of such BARs, platforms are required to reserve +some space below 4G for each root port present in the system. + +[sidebar] +-- +[underline]*_Implementation Note_* + +Platform software would likely configure these per root port regions such that +their effective memory type (PMA + PBMT) is UC. Platforms would likely also +reserve some space above 4G to map BARs that support 64 bit addressing and +prefetchable memory which could be configured by the platform software as either +I/O or memory. +-- + +* PCIe Inbound Memory + +For security reasons, platforms must provide a mechanism controlled by M-mode +software to restrict inbound PCIe accesses from accessing regions of address +space intended to be accessible only to M-mode software. + +[sidebar] +-- +[underline]*_Implementation Note_* + +Such an access control mechanism could be analogous to the per-hart PMP +as described in the RISC-V Privileged Architectures specification. +-- + +====== PCIe Interrupts +* Platforms shall support both INTx and MSI/MSI-x interrupts. +* Integration with AIA + +TBD + +====== PCIe cache coherency +PCIe transactions that are not marked as No_snoop and access memory that is +cacheable by harts, as well as accesses to memory that is noncacheable by +harts, are I/O Coherent and no software coherency management is needed. +In contrast, PCIe transactions that are marked as No_snoop and access memory +that is cacheable by harts, must have coherency managed by software. + +====== PCIe Topology +Platforms are required to implement at least one of the following topologies +and the components required in that topology. + +[ditaa] +.... + + +----------+ +----------+ + | CPU | | CPU | + | | | | + +-----|----+ +-----|----+ + | | + | | + +-------------|------------+ +-------------|------------+ + | ROOT | COMPLEX | | ROOT | COMPLEX | + | | | | + | +------|-------+ | | +------|-------+ | + | | Host Bridge | | | | Host Bridge | | + | +------|-------+ | | +------|-------+ | + | | | | | | + | | BUS 0 | | | BUS 0 | + | |-------|------| | | +-----|-------+ | + | | | | | | ROOT PORT | | + | | | | | +-----|-------+ | + | +---|---+ +---|---+ | | | | + | | RCiEP | | RCEC | | | | PCIe Link | + | +-------+ +-------+ | | | | + | | +-------------|------------+ + +--------------------------+ | + | BUS 1 + RCiEP - Root complex integrated endpoint + RCEC - Root complex event collector +.... + +* Host Bridge + +Following are the requirements for host bridges: + +** Any read or write access by a hart to an ECAM I/O region shall be converted +by the host bridge into the corresponding PCIe config read or config write +request. +** Any read or write access by a hart to a PCIe outbound region shall be +forwarded by the host bridge to a BAR or prefetch/non-prefetch memory window, +if the address falls within the region claimed by the BAR or prefetch/ +non-prefetch memory window. Otherwise the host bridge shall return an error. + +** Host bridge shall return all 1s in the following cases: +*** Config read to non existent functions and devices on root bus. +*** Config reads that receive Unsupported Request response from functions and +devices on the root bus. +* Root ports + +Following are the requirements for root ports. +** Root ports shall appear as PCI-PCI bridge to software. +** Root ports shall implement all registers of Type 1 header. +** Root ports shall implement all capabilities specified in the PCIe Base +specification for a root port. +** Root ports shall forward type 1 configuration access when the bus number in +the TLP is greater than the root port's secondary bus number and less than or +equal to the root port's subordinate bus number. +** Root ports shall convert type 1 configuration access to a type 0 +configuration access when bus number in the TLP is equal to the root port's +secondary bus number. +** Root ports shall respond to any type 0 configuration accesses it receives. +** Root ports shall forward memory accesses targeting its prefetch/non-prefetch +memory windows to downstream components. If address of the transaction does not +fall within the regions claimed by prefetch/non-prefetch memory windows then +the root port shall generate a Unsupported Request. +** Root port requester id or completer id shall be formed using the bdf of the +root port. +** The root ports shall support the CRS software visibility. +** The root port shall implement the AER capability. +** Root ports shall return all 1s in the following cases: +*** Config read to non existent functions and devices on secondary bus. +*** Config reads that receive Unsupported Request from downstream components. +*** Config read when root port's link is down. + +* RCiEP + +All the requirements for RCiEP in the PCIe Base specification shall be +implemented. +In addition the following requirements shall be met: +** If RCiEP is implemented then RCEC shall be implemented as well. All +requirements for RCEC specified in the PCIe Base specification shall be +implemented. RCEC is required to terminate the AER and PME messages from RCiEP. +** If both the topologies mentioned above are supported then RCiEP and RCEC +shall be implemented in a separate PCIe domain and shall be addressable via a +separate ECAM I/O region. + +===== PCIe Device Firmware Requirement +PCI expansion ROM code type 3 (UEFI) image must be provided by PCIe device for +OS/A server extension platform according to +https://pcisig.com/specifications/conventional/pci_firmware[PCI Firmware Specification Revision 3.3] +if that PCIe device is utilized during UEFI firmware boot process. The image +stored in PCI expansion ROM is an UEFI driver that must be compliant with +https://uefi.org/specifications[UEFI specification 2.9] 14.4.2 PCI Option ROMs. + +====== PCIe peer to peer transactions + +TBD ==== Secure Boot * TEE -- 2.17.1 |
|
Looks good. Reviewed-by: Abner Chang <abner.chang@...>Mayuresh Chitale <mchitale@...> 於 2021年7月2日 週五 上午12:50寫道: This patch adds requirements for PCIe support for the server extension |
|
atishp@...
On Thu, 2021-07-01 at 22:20 +0530, Mayuresh Chitale wrote:
This patch adds requirements for PCIe support for the serverMCFG table was not referenced earlier. +systems to discover the supported PCIe domains and map the ECAM I/OUC is not present in abbreviation section. Moreover, PMA + PBMT seems bit ambiguous to me. It wouldn't hurt to be more verbose here. +====== PCIe Memory Spacesame comment as above. Platforms would likely also/s/No_snoop/No Snoop ? +cacheable by harts, as well as accesses to memory that isApart from the above nit comments, looks good to me. Reviewed-by: Atish Patra <atish.patra@...> -- Regards, Atish |
|
Greg Favor
On Sat, Jul 3, 2021 at 8:03 AM Atish Patra <atish.patra@...> wrote: > +* Platform software shall configure ECAM I/O regions such that the The proposed RISC-V name for this memory type is "IO", but it is up in the air for the moment as to whether the the memory types supported by Svpbmt will have acronym names (i.e. IO and NC), or just use thier longer descriptive names, e.g. Non-cacheable, non-idempotent, strongly-ordered I/O memory for "IO". Overall it is probably better to make the above statement in a somewhat more generic manner terminology-wise. For example:
This implicitly allows for systems that do and don't support Svpbmt. > +Platform software would likely configure these per root port regions Same suggestion as above. > +====== PCIe cache coherency ARM uses "No_snoop", but others will use No-snoop and No Snoop. Avoiding the last form avoids any ambiguity when not using it followed by "bit", i.e. "No Snoop bit" is pretty clear, whereas referring to just "No Snoop" as a transaction attribute to some might be ambiguous as to whether "Snoop" or "No Snoop" is the attribute. Myself, I don't have a significant leaning, but one alternative to the above sentence could be:
or ...
I maybe lean a little bit towards that last option. That also avoids use of the linguistic double negative. Greg |
|
Mayuresh Chitale
On Sat, Jul 3, 2021 at 8:33 PM Atish Patra <Atish.Patra@...> wrote: On Thu, 2021-07-01 at 22:20 +0530, Mayuresh Chitale wrote: Actually it is present in the 'Required ACPI System Description Tables'. I can add a pointer to that table in the text above.
I will rephrase all instances of this as Greg mentioned in his latest email. > +====== PCIe Memory Space PCIe spec does refer to it as No Snoop but ARM BSA refers to it as No_snoop. I thought the latter was slightly better as it might be easier to understand it as a single attribute or flag :) > +cacheable by harts, as well as accesses to memory that is |
|