Stream: git-wasmtime

Topic: wasmtime / PR #1894 reference types: Implement the `table...


view this post on Zulip Wasmtime GitHub notifications bot (Jun 17 2020 at 21:06):

fitzgen opened PR #1894 from table-size-and-table-grow to master:

(Still doesn't handle table.grow for funcrefs yet, as there are still some open questions about how to represent on-stack, in-Wasm funcrefs.)

Also de-duplicates / cleans up declaring builtin function indices a little bit, so that they aren't so tedious and repetitive.

Part of #929

view this post on Zulip Wasmtime GitHub notifications bot (Jun 17 2020 at 21:06):

fitzgen requested alexcrichton for a review on PR #1894.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 17 2020 at 21:50):

alexcrichton submitted PR Review.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 17 2020 at 21:50):

alexcrichton submitted PR Review.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 17 2020 at 21:50):

alexcrichton created PR Review Comment:

Presumably we're going to want to have the table size available inline for table.get, and in that case we probably won't have an intrinsic, right?

If that's the case could this go ahead and implement the JIT support necessary to do the load inline?

view this post on Zulip Wasmtime GitHub notifications bot (Jun 17 2020 at 21:50):

alexcrichton created PR Review Comment:

I'm a bit curious about the ownership here. Presumably it's not safe to gc at the beginning of this function, right? The init_value pointer is in the activations table for sure, but there's no guarantee it's reachable from any stack map?

view this post on Zulip Wasmtime GitHub notifications bot (Jun 17 2020 at 21:57):

fitzgen submitted PR Review.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 17 2020 at 21:57):

fitzgen created PR Review Comment:

Sure thing.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 17 2020 at 21:57):

fitzgen created PR Review Comment:

Correct. I should probably add a warning to the docs at the top of this file, like how we warn about traps and potential leaks from skipped drops.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 17 2020 at 21:57):

fitzgen submitted PR Review.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 17 2020 at 22:39):

fitzgen updated PR #1894 from table-size-and-table-grow to master:

(Still doesn't handle table.grow for funcrefs yet, as there are still some open questions about how to represent on-stack, in-Wasm funcrefs.)

Also de-duplicates / cleans up declaring builtin function indices a little bit, so that they aren't so tedious and repetitive.

Part of #929

view this post on Zulip Wasmtime GitHub notifications bot (Jun 17 2020 at 23:03):

fitzgen updated PR #1894 from table-size-and-table-grow to master:

(Still doesn't handle table.grow for funcrefs yet, as there are still some open questions about how to represent on-stack, in-Wasm funcrefs.)

Also de-duplicates / cleans up declaring builtin function indices a little bit, so that they aren't so tedious and repetitive.

Part of #929

view this post on Zulip Wasmtime GitHub notifications bot (Jun 18 2020 at 13:57):

alexcrichton submitted PR Review.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 18 2020 at 13:57):

alexcrichton merged PR #1894.


Last updated: Nov 22 2024 at 16:03 UTC