process: fix error when runtime is shut down #7672
Conversation
Darksonn
added
A-tokio
There was a problem hiding this comment.
Emm, it looks like there is still a issue.
Imagine the behavior change of std::io::Error::description has landed in Rust 1.91.0. And downstream compile their application using Rust 1.91.0 and older tokio (<= 1.47.1), will this causes bugs? If so, we have to bump the MSRV to 1.91.0.
|
We cannot bump the MSRV of old minor releases. But we can backport this to LTS releases. |
There was a problem hiding this comment.
From the downstream perspective, bumping the Rust toolchain to the next stable release will introduce a new bug into their application just because they didn't bump the tokio to the latest LTS release. Emm, this looks surprising and not downstream-friendly.
I think this fix is ok, there is nothing else we can do to improve the downstream experience on the tokio side.
|
I want to get a release out, so I'm merging this into master. We can backport it later. |
…okio-rs#7672) (cherry picked from commit e150d4e)
…okio-rs#7672) (cherry picked from commit 9e5527d)
| fmt_eq.remainder.is_empty() && !fmt_eq.unequal | ||
| } | ||
|
|
||
| #[allow(deprecated)] |
There was a problem hiding this comment.
Can that allow(deprecated) be removed, since the new code doesn't call the deprecated description method anymore?
|
What is the impact of this change on applications using broken versions of tokio? Just worse diagnostics, or does it impact actual functionality? |
|
The impact is that (Specifically the case where |
Fixes: #7671
See:
https://users.rust-lang.org/t/compare-error-description-without-allocating/134630