Stream: git-wasmtime

Topic: wasmtime / PR #10478 asm: fix trace logging again

Wasmtime GitHub notifications bot (Mar 27 2025 at 01:38):

abrown requested wasmtime-compiler-reviewers for a review on PR #10478.

Wasmtime GitHub notifications bot (Mar 27 2025 at 01:38):

abrown requested cfallin for a review on PR #10478.

Wasmtime GitHub notifications bot (Mar 27 2025 at 01:38):

abrown opened PR #10478 from abrown:asm-more-trace-printing to bytecodealliance:main:

As pointed out in #10476, there are a few more locations in which the raw encoding of a register is used when pretty-printing memory operands. Using the raw encoding before register allocation results in a panic. This change ensures that the additional locations delegate to the to_string() method which is overriden in cranelift-codegen to know how to pretty-print virtual registers. Closes #10476.

<!--
Please make sure you include the following information:

• If this work has been discussed elsewhere, please include a link to that
  conversation. If it was discussed in an issue, just mention "issue #...".

• Explain why this change is needed. If the details are in an issue already,
  this can be brief.

Our development process is documented in the Wasmtime book:
https://docs.wasmtime.dev/contributing-development-process.html

Please ensure all communication follows the code of conduct:
https://github.com/bytecodealliance/wasmtime/blob/main/CODE_OF_CONDUCT.md
-->

Wasmtime GitHub notifications bot (Mar 27 2025 at 05:05):

cfallin submitted PR review:

Thanks!

Would it be possible to add some sort of test to make sure that trace-logging works for a whole compilation? Perhaps programmatically setting the log-level to trace with the log crate or related and then compiling some simple function body in a unit test, or similar.

Wasmtime GitHub notifications bot (Mar 27 2025 at 16:14):

abrown commented on PR #10478:

Would it be possible to add some sort of test to make sure that trace-logging works for a whole compilation? Perhaps programmatically setting the log-level to trace with the log crate or related and then compiling some simple function body in a unit test, or similar.

Should it just be a CLI test?

Wasmtime GitHub notifications bot (Mar 27 2025 at 16:26):

alexcrichton commented on PR #10478:

One thing I'll note is that this'd be easiest during tests/all/cli_tests.rs as that's where we can scope setting an env var to just one process, but that's disconnected from wast testing. What you can do is pick any of the test-programs modules/components and compile that with the env var and ensure that it doesn't panic.

Wasmtime GitHub notifications bot (Mar 27 2025 at 23:34):

abrown requested alexcrichton for a review on PR #10478.

Wasmtime GitHub notifications bot (Mar 27 2025 at 23:34):

abrown updated PR #10478.

Wasmtime GitHub notifications bot (Mar 27 2025 at 23:34):

abrown requested wasmtime-core-reviewers for a review on PR #10478.

Wasmtime GitHub notifications bot (Mar 28 2025 at 03:24):

alexcrichton submitted PR review.

Wasmtime GitHub notifications bot (Mar 28 2025 at 03:46):

alexcrichton merged PR #10478.

Last updated: Apr 17 2025 at 13:10 UTC