Tracking issue for phase 4 advancement

[sbc100]

• Update spec text ([spec] Update syntax spec for memory64 #50)
• Implement table64 extension  #51
• Rebase interpreter changes
• JS API Updates
• Spec tests
• Implemented in at least two Web VMs
  • v8: https://crbug.com/41480462 / https://crbug.com/364917766
• Implemented in at least one toolchains
• BigInt parameters for the JS API? #68
  • Update JS API in v8: https://g-issues.chromium.org/issues/358381633
• Text format: new abbreviation for inline data? #29
• Restricting offsets to match the index type #76
• Rename Index Type to something else, perhaps Offset Type #67
• Tests needed for recent spec updates #80

[rossberg]

A while ago I heard about the intent to include 64-bit tables as well, in order to be able to compile function pointers consistently. What became of that?

[sbc100]

That request came from @matthias-blume who is working on an experimental wasm to native compiler without a sandbox.

We have a workaround in llvm that truncate all function pointers before call indirect. It add an instruction for every call indirect in the program so it could save a bit on code size and complexity if we could remove it.

Do you think it makes sense to roll it into this proposal?

[rossberg]

Yeah, from my perspective, Wasm64 is incomplete without it, and it leaves the language in an odd space. Do you think it would still be realistic to extend the proposal?

[tlively]

What's the status of the spec document? It would be good to get the changes in well ahead of the phase 4 vote at the f2f meeting.

[sbc100]

I need to pick up this open PR: #50

[wooden-worm]

Now with only the table64 extension remained, can we get an estimate please?

[sbc100]

We are just waiting on the second implementation of table64 now (in spidermonkey). My understanding is that is underway, so we should be able to vote on phase in next one or two meetings.

[bvisness]

@sbc100 Should the following two issues be added to this checklist?

• Text format: new abbreviation for inline data? #29 (explicitly marked as release blocker)
• Rename Index Type to something else, perhaps Offset Type #67

[rossberg]

@bvisness, I'd think so.

[bvisness]

@sbc100 More issues / PRs that I think should be on this list:

• Restricting offsets to match the index type #76
• Intent use number / bigint union type whatwg/webidl#1426

There are also the other scattered spec PRs I've been working on, but those are basically spec bugfixes and I don't think we need to track them here.

[bvisness]

#80 should be on this list as well, presumably.

[sbc100]

Indeed. Added #80, hopefully we can get those updates done by next meeting and vote on this.

[sbc100]

We are now phase 4!