[P4-design] Meeting minutes (Re: 4/10 draft agenda)

Calin Cascaval cascaval at barefootnetworks.com
Mon Apr 10 17:52:32 EDT 2017


Language compatibility

What do we say at the workshop about backward compatibility

 - multiple implementations put more pressure on keeping stable releases more stable

 - hard to make promises that we’re not breaking compatibility

 - make an effort to keep compatibility w/o actually promising 

- do we limit the scope to P4_14/16? 

- do we know what changes are coming that potentially break compatibility? 

    - composability and modularity 

    - control plane 

- is there a threshold for changes?

PR 158 - Add table of allowed member types in headers, structs, and tuples

structs and tuples should behave the same – compiler non-conformant. tuples behave as listed, structs do not
nesting? target restrictions on total size, etc.
varbit? only one allowed in a header. Computed as the difference between total size and the rest of the fields
error vs. match_kind in structs? match_kind is pretty much always used only as a type for table matches, and thus there is no need to store in variables. 
AI:
Andy to update with rationale
update the table with header_union when approved (part of header_union PR)
Issue 150 - Global instances

impact on resource allocation and sharing
instantiations whether inline in a package decl or at top level should be the same
instantiations should be done by either the program or the architecture. Should we restrict at one? Answers
these tables/controls are not dead code, since the control plane can access them
compiler should be careful to distinguish between control- and data-plane inlining. Control-plane inlining should behave as the control was not inlined.
architecture dependent
AI: add a note of clarification to the spec

Issue 138 - Require unambiguous names

multiple aliases and what order they apply?
annotations are applied in program order, the last one remains exposed and all others become hidden
need clarification
Chris: proposes replacing rename with alias (and globalrename with globalalias)
AI: Cole to introduce in the spec

Issue 151 - Action calls from parsers

we should add functions to parsers rather than hijacking actions which are part of the match-action abstraction (which parsers are not)
AI: add lambdas to the next version of the language

PR 127, Issue 124 - adds specification for table entries

AI: PR approval


--
Thanks, Calin

> On Apr 10, 2017, at 12:20, Nate Foster <jnfoster at cs.cornell.edu> wrote:
> 
> Here's a final agenda for today (with two additions from @jafingerhut and bumping header unions).
> 
> -N
> 
> * Status and Roadmap [5 minutes]
> 
> * Stability / Backwards Compatibility [10 minutes]
> 
> * P4 Types [10 minutes]
> https://github.com/p4lang/p4-spec/pull/158 <https://github.com/p4lang/p4-spec/pull/158>
> 
> * Calling Actions from Parsers [10 minutes]
> https://github.com/p4lang/p4-spec/issues/151 <https://github.com/p4lang/p4-spec/issues/151>
> 
> * Global Instances (Continued) [15 minutes]
> https://github.com/p4lang/p4-spec/issues/150 <https://github.com/p4lang/p4-spec/issues/150>
> 
> * Control-Plane Names (Continued) [15 minutes]
> https://github.com/p4lang/p4-spec/issues/138 <https://github.com/p4lang/p4-spec/issues/138>
> 
> * Table entries (Continued) [15 minutes]
> https://github.com/p4lang/p4-spec/pull/127 <https://github.com/p4lang/p4-spec/pull/127>
> 
> * Other business [10 minutes]
> 
> On Sun, Apr 9, 2017 at 4:38 PM, Nate Foster <jnfoster at cs.cornell.edu <mailto:jnfoster at cs.cornell.edu>> wrote:
> We meet again tomorrow from 1:30pm-3:00pm PT in Gates 260. We have a lot to discuss so I plan to start the meeting right at 1:30pm.
> 
> A draft agenda is below. Please let me know if you have anything you'd like to add to it. I'll circulate a final version tomorrow morning.
> 
> Draft Agenda:
> 
> * Status and Roadmap [5 minutes]
> 
> * Stability / Backwards Compatibility [15 minutes]
> 
> * Global Instances (Continued) [15 minutes]
> https://github.com/p4lang/p4-spec/issues/150 <https://github.com/p4lang/p4-spec/issues/150>
> 
> * Control-Plane Names (Continued) [15 minutes]
> https://github.com/p4lang/p4-spec/issues/138 <https://github.com/p4lang/p4-spec/issues/138>
> 
> * Header Unions (Continued) [15 minutes]
> 
> * Table entries (Continued) [15 minutes]
> https://github.com/p4lang/p4-spec/pull/127 <https://github.com/p4lang/p4-spec/pull/127>
> 
> * Other business [10 minutes]
> 
> Call-in details:
> * By computer: https://cornell.zoom.us/j/959432658 <https://cornell.zoom.us/j/959432658>
> * By phone: +1 646 558 8656 <tel:(646)%20558-8656> (US Toll) or +1 408 638 0968 <tel:(408)%20638-0968> (US Toll). The Meeting ID is 959 432 658. International numbers are available at https://cornell.zoom.us/zoomconference?m=p414IDyY2yeauZygZLycHH0Vnc0wYEFg <https://cornell.zoom.us/zoomconference?m=p414IDyY2yeauZygZLycHH0Vnc0wYEFg>
> _______________________________________________
> P4-design mailing list
> P4-design at lists.p4.org
> http://lists.p4.org/mailman/listinfo/p4-design_lists.p4.org
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.p4.org/pipermail/p4-design_lists.p4.org/attachments/20170410/2e4c572e/attachment-0002.html>


More information about the P4-design mailing list