Stream: git-wasmtime

Topic: wasmtime / issue #4571 Cranellift: remove Baldrdash suppo...

view this post on Zulip Wasmtime GitHub notifications bot (Aug 01 2022 at 22:04):

cfallin commented on issue #4571:

There is something strange going on with the lldb test (gdb passes) where it can't find a line number for a breakpoint anymore. It seems like this should be unrelated and I'm wondering if it has to do with a silent version change somewhere. I'll poke at a bit, but if I can't work it out I may just disable the lldb test, since debug support in general is tier-3.

view this post on Zulip Wasmtime GitHub notifications bot (Aug 02 2022 at 04:43):

cfallin commented on issue #4571:

I was having difficulty investigating the lldb failure locally because it seemed flaky (inconsistently failed) until I realized this depended on me moving the branch around and rebuilding -- hence on the state of the compilation cache. It turns out our debugger tests invoke a nested wasmtime process under a debugger bug don't disable the compilation cache, and this appears to cause issues with missing debug info.

I'm not sure if this implies an issue with the caching key or not (cc @alexcrichton perhaps?) but let's see if the latest commit at least fixes the tests here...

view this post on Zulip Wasmtime GitHub notifications bot (Aug 02 2022 at 04:44):

cfallin edited a comment on issue #4571:

I was having difficulty investigating the lldb failure locally because it seemed flaky (inconsistently failed) until I realized this depended on me moving the branch around and rebuilding -- hence on the state of the compilation cache. It turns out our debugger tests invoke a nested wasmtime process under a debugger but don't disable the compilation cache, and this appears to cause issues with missing debug info.

I'm not sure if this implies an issue with the caching key or not (cc @alexcrichton perhaps?) but let's see if the latest commit at least fixes the tests here...

view this post on Zulip Wasmtime GitHub notifications bot (Aug 02 2022 at 04:45):

cfallin commented on issue #4571:

If this does fix the issue then I'm happy to either split out the last commit into another PR or include it here with your updated r+, @fitzgen...

Last updated: Nov 22 2024 at 16:03 UTC