Stream: git-wasmtime

Topic: wasmtime / Issue #2266 machinst x64: New backend unwind


view this post on Zulip Wasmtime GitHub notifications bot (Oct 06 2020 at 19:57):

yurydelendik commented on Issue #2266:

I don't see how stackmaps in the CodeSink are populated. Keeping tests in "tests/all/gc.rs" disabled.

view this post on Zulip Wasmtime GitHub notifications bot (Oct 06 2020 at 20:04):

yurydelendik edited a comment on Issue #2266:

I don't see how stackmaps in the CodeSink are populated (for machinst). Keeping tests in "tests/all/gc.rs" disabled.

view this post on Zulip Wasmtime GitHub notifications bot (Oct 20 2020 at 21:51):

yurydelendik commented on Issue #2266:

not to use UnwindInfoKind?
make unwind info available without feature?

Cannot find solutions to address these two -- open for suggestions. Opening for review.

view this post on Zulip Wasmtime GitHub notifications bot (Oct 22 2020 at 03:40):

peterhuene commented on Issue #2266:

I also have no strong opinion on removal of the unwind feature. I would probably favor keeping it at this point.

Regarding UnwindInfoKind, I would probably remove it if possible as I don't see much of a use case for generating unwind information in a different format than the function's (ABI-specific) calling convention.

view this post on Zulip Wasmtime GitHub notifications bot (Oct 22 2020 at 09:01):

bjorn3 commented on Issue #2266:

macOS has both DWARF unwinding and compact unwind info. Both of which can be used with the same SystemV calling convention.

view this post on Zulip Wasmtime GitHub notifications bot (Oct 22 2020 at 19:11):

peterhuene commented on Issue #2266:

I'm :+1: with keeping it.

view this post on Zulip Wasmtime GitHub notifications bot (Oct 23 2020 at 19:45):

yurydelendik commented on Issue #2266:

merged with #2307 changes: that allowed to replace "mov reg, imm(rsp)" todo -- carrying over the review approval


Last updated: Nov 22 2024 at 16:03 UTC