PLIC on SAIL model

[ptprasanna]

Technical Group

PLIC

ratification-pkg

PLIC

Technical Liaison

Prasanna

Task Category

SAIL

Task Sub Category

• gcc
• binutils
• gdb
• intrinsics
• Java
• KVM
• ld
• llvm
• Linux kernel
• QEMU
• Spike

Ratification Target

2022

Statement of Work (SOW)

Component names:

Requirements:
Developing PLIC extension of Riscv Specification, on Sail model using sail programming language. By implementing memory mapped registers on sail for handling the externally generated interrupts (platform generated interrupts).

Deliverables:

• Component_1 and changes expected.
• Component_2 and changes expected.
  ...

Acceptance Criteria:

• Component_1 and measured result.
• Component_2 and measured result.
  ...

Projected timeframe: (best guess date)

SOW Signoffs:

• Task group liaison sign-off
• Development partner sign-off
• Golden Model SIG sign-off

Waiver

• Freeze
• Ratification

Pull Request Details

No response

[jjscheel]

Hi, @ptprasanna, I've added the SOW template to your first comment and moved your text into the requirements section. Let's edit it there as-needed.

Thanks!

[jrtc27]

What's the point of having a PLIC without any devices attached to it that can generate interrupts?

[jjscheel]

Today's DevPartners meeting suggested that this work is on-going to create a PLIC infrastructure in SAIL and it will include a basic framework for testing interrupts.

The open question for the DevPartner's meeting is to determine how much to track here so as not to duplicate work. At this time, I view this item as FYI for us.

[ptprasanna]

PLIC : Algorithm has been completed and coding in progress. Hoping to get this done by this week and will send the code to Bill for review.

Testing : Sail(PLIC) will be listening on a certain port. And an external software will send the interrupt messages to PLIC on that port. We'll use system call for this.

[jjscheel]

@ptprasanna, do we have a target date by which all work is anticipated to be done, including PR acceptance?

[jrtc27]

NB: Bill doesn't review the code, the community does

[jjscheel]

@jrtc27, indeed. I suspect that because the person is new, they're actually "pre-reviewing" the code with Bill. Everyone understands that the community will review the code, comment, and ultimately accept the PR.