Stream: git-wasmtime

Topic: wasmtime / issue #5898 Fix codegen riscv64 atomic_rmw_loo...


view this post on Zulip Wasmtime GitHub notifications bot (Mar 02 2023 at 02:07):

jameysharp commented on issue #5898:

This is failing the test in tests/spec_testsuite/proposals/threads/atomic.wast on line 142, because atomic_rmw is expected to return the _old_ value at the address, not the newly-written value.

To fix #5884, I think you should write the result of AtomicOP::extract to dst instead of writing it to t0, and update later instructions accordingly.

view this post on Zulip Wasmtime GitHub notifications bot (Mar 02 2023 at 02:33):

yuyang-ok commented on issue #5898:

@jameysharp , Thanks, The atomic_rmw_loop kinda of little complex. I think I want to recheck it when I have time.

view this post on Zulip Wasmtime GitHub notifications bot (Mar 02 2023 at 02:55):

jameysharp commented on issue #5898:

Yeah, I had a difficult time understanding what the loop was doing. I suspect we can improve it, and I'd be happy to discuss it further when you have time.

view this post on Zulip Wasmtime GitHub notifications bot (Mar 02 2023 at 03:23):

yuyang-ok commented on issue #5898:

@jameysharp sure.

view this post on Zulip Wasmtime GitHub notifications bot (Mar 03 2023 at 01:15):

yuyang-ok commented on issue #5898:

There is a CI failure Without errors.

view this post on Zulip Wasmtime GitHub notifications bot (Mar 03 2023 at 02:00):

jameysharp commented on issue #5898:

I tried re-running just the failed CI jobs, but then extra jobs ran, which failed without useful logs, and I don't understand why. I'll have to look into this again next week if nobody else gets to it first.

view this post on Zulip Wasmtime GitHub notifications bot (Mar 03 2023 at 02:39):

yuyang-ok commented on issue #5898:

@jameysharp Ok,thanks.


Last updated: Oct 23 2024 at 20:03 UTC