You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 37 Next »

Status at a glance:

To Reach Definition-of-Done Plan status: extension lifecycle and milestone definitions

  • Copy  Non-ISA Platform Ratification Template 
  • keep specification status spreadsheet up-to-date ( projected dates, DoD component effort sizing, DoD tasks your group needs resource help with,...)
  • Keep dashboard  Non-ISA Extensions On Deck for Freeze Milestone up-to-date
  • Develop final charter and plan with the group and send to Chairs (tech-chairs@lists.riscv.org) for approval within 4 weeks
  • Fill in DoD Plan tab in checklist and send to chairs for approval of any deliverables that don't match the defaults along with any requested waivers
  • Freeze-time preliminary POC definition approved by IC
  • Develop Rationale document and send to chairs for approval
  • Notify TSC and Chairs when you complete this milestone and include any waivers and an updated DoD spreadsheet.
  • Plan: RISCV has prioritized RVA22.  expect CLIC to possibly be ratified with RVM23 with other embedded extensions like Zfinx, new compressed, rv32e, p-extension.  So CLIC still on roadmap but probably next gen RISC-V grouping.  Continue working on issues and refining spec.
  • Hypervisor support?
    • See issue #92 discussion.  Lastest discussion: The AIA spec is being developed to be the standard support for interrupts for standard hypervisors. Possible hypervisor directions for fast interrupts for embedded include: 1) a minimal CLIC extension that adds HS mode as another privilege layer in the stack, so the hypervisor could optionally receive interrupts and optionally enable/hide interrupts from lower privilege modes. This would allow the 2-stage translation and other features of the hypervisor to be used in a CLIC-based system, but would not support sending interrupts directly to descheduled guest OS contexts (have to route via hypervisor interrupt). 2) support multiple descheduled CLIC contexts to which interrupts could be sent directly while they were sleeping, which would be a much bigger project. 3) add an extension to AIA (instead of working on CLIC) to somehow reduce interrupt latency using the standard hypervisor stack.
  • Impact of N-extension on CLIC?  Update CLIC spec to not rely on N-extension but add commentary on if N-extension is ratified how it would work with CLIC.

Charter

Specification

  • Latest Draft Fast Interrupt Specification (v0.9-draft-20230411)
  • What's next:
    • 28 outstanding issues.
      • 12 active spec issues.
        • big items: hypervisor, cliccfg reformat proposal
      • 6 issues to track Definition of Done tasks
      • 6 issues kept open for future enhancements (post rev1.0) 
    • Major items on Freeze Checklist: 

      • RISC-V SAIL 

      Saravana introduction.  will be working on interrupts and SAIL.  From networking background.  Coming up to speed on SAIL model.  Went through cookbook model.  Starting with PLIC.  PLIC is non-ISA.  interrupt registers  PLIC is almost completely MMIO addresses.  wires from PLIC into interrupt registers.  That I/f is sort of architectural.  hook up pseudo event generator to wiggle the wires.  if hook up PLIC have to connect to PLIC. 

      From a hart perspective there are a lot of interrupt wires coming in.  how do they get there.  incoming write addr turns into wiggling a wire.   

      Event generator relative to retired count of a hart. Assert the wire 5 instruction retires from now. As independent from implementation as possible.  instruction retires are relatively deterministic. 

      SAIL status – 1 more pull request for vector.  do have backlog.  7 or 8 prs that are pending.  some PRs may affect the same file.  instruction extension usually just a separate file.  CLIC will add CSRs and is more complicated (touch at least 3 files).  RNMI.  PRs priority things that are ratified. 

      • RISC-V Tests 

      If there is a PR for ACT should have been tested and compared to Spike.  riscv-non-isa/riscv-arch-test (github.com) 

      Ovpsim may be a good alternative to spike with initial development while spike is being updated.  have to do in environment that generates interrupts. 

      Regression testing – interaction with CLINT. 

      Would CLIC get its own folder under riscv-arch-test/riscv-test-suite/rv32i? 

      In one place but say execute if RV32I or RV64I or don’t run test. 

      Test plan – need to write something that stimulates clic sources, turns into wires.   

      ACT – read/understood/implemented the spec.  not DV coverage (lots of combinations).  but do want some cross-coverage.  

      • RISC-V Tests Input 

      riscv-software-src/riscv-config: RISC-V Configuration Validator (github.com) 

      Schema_isa.yaml - gives allowed configuration on risc-v and allowed values.  WARL fields give allowed ranges.  10k lines in file. 

      Examples/rivc32i_isa.yaml is compared against schema_isa.yaml to see if it is allowed. 

      When add new CSRs, add PR to add to schema_isa.yaml and then will be checked against example implementation. 

      Schemas/schema_platforml.yaml - memory mapped registers (like clicintctl, etc.) 

      Run python scripts that look at your implementation and see if it is valid.  e.g., if have d then have f- extenstion, etc. checks if WARL field is valid. 

      Write anything, read legal.  mapping from what is illegal to what is legal.  mapping is arbitrary.  so prefer (easiest) if implementers implement when write something illegal, don’t write.  that works easiest for describing in this file. 

Encoding/OpCode consistency review

  • Need to propose new CLIC CSR Registers and addresses
  • What's next: when outstanding issues are reduced, start planning for review
  • tech-chairs@lists.riscv.org - when spec is solid but not a final spec - primarily want to nail down opcode/CSR assignment and have a solid draft spec (but not a final spec ready for official Arch Review)
  • Also, to remind people of what gets reviewed (as is appropriate for a given extension), see the following list.  In addition to the extension spec, please submit information about the PoCs and about utility/efficiency (although we don't need all the gory detail - a paragraph or so for each can be fine).  For items considered to not be consequential, a sentence or so explaining why should suffice.
    • Consistency with the RISC-V architecture and philosophy
    • Documentation clarity and completeness
      • Including proper distinction between normative and non-normative text
    • Motivation and rationale for the features, instructions, and CSRs
    • Utility and efficiency (relative to existing architectural features and mechanisms)
      • Is there enough value or benefit to justify the cost of implementation
      • Is the cost in terms of area, timing, and complexity reasonable
    • Proof of Concept (PoC)
      • Software PoC to ensure feature completeness and appropriateness for intended use cases
      • Hardware PoC to demonstrate reasonable implementability
    • Inappropriate references to protected IP (i.e. covered by patents, copyright, etc.)

Architecture Tests

  • Deterministic Test plan for the fast-interrupt is available.  Discussion on-going on how to add async/undeterministic testing of interrupts.
  • YAML config needs to be created.  See info here.
  • Discussion in Arch tests group to add automation (docker?) to validate check-in so that arch-test-suite is run against sail, spike, gcc/toolchain, with versions used recorded.  So sail and spike will need to work for CLIC before CLIC tests can be added to riscv-config github.

Compilers / Toolchains

GCC and Binutils

  • No new instructions are added.  Needs to be aware CSR names?  Need to choose arch string. Given that the key CSRs are in M-mode, it should probably be named something like "Smclic"

LLVM

  • No new instructions are added.  Needs to be aware of CSR names? 

Simulators

Though all listed under "simulators", these are actually a collection of formal model / virtual machine / architectural simulators / DV simulators etc.

SAIL

  • TBD - Waiver until SAIL privilege model is completed?  e.g. add CLIC support to OS like Zephyr.
  • which github? https://github.com/riscv/sail-riscv is now the official location

Spike

  • TBD

riscvOVPSimPlus

QEMU

Proof-of-Concept implementations

Hardware

Project NameBase ArchitectureLevel of implementationNotes
area-optimized coreRV32/64

RTL simulation, FPGA Implementation, Synthesis

closed / commercial source   https://www.seagate.com/innovation/risc-v/
high-performance coreRV32RTL simulation, FPGA Implementation, Synthesisclosed / commercial source   https://www.seagate.com/innovation/risc-v/
microcontroller-class coreRV32IMAFCRTL, fully synthesizableApache License, Version 2.0 https://github.com/T-head-Semi/opene906/blob/main/doc/opene906_datasheet.pdf
E2/S2 seriesRV32/64RTL, fully synthesizablehttps://www.sifive.com/core-designer
N22RV32RTL, fully synthesizablehttp://www.andestech.com/en/products-solutions/andescore-processors/riscv-n22/
BM-310/BI-651RV32/64RTL, fully synthesizablehttps://cloudbear.ru/bm_310.html      https://cloudbear.ru/bi_651.html
n200/n900/nx900/ux900RV32/64RTL, fully synthesizable  (ECLIC)https://www.nucleisys.com/product.php?site=n200      https://www.nucleisys.com/product.php?site=n900

Software

Project/MaintainerDescription






ABI Extensions (no new ABI required)

  • No labels