From 2fa159a87f14dd16cdce39b8e8dd78a674ad733c Mon Sep 17 00:00:00 2001 From: Serge Barral Date: Tue, 7 May 2024 12:44:42 +0200 Subject: [PATCH] Fix comments --- asynchronix/src/executor/task.rs | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/asynchronix/src/executor/task.rs b/asynchronix/src/executor/task.rs index 9236de9..3b3bd5f 100644 --- a/asynchronix/src/executor/task.rs +++ b/asynchronix/src/executor/task.rs @@ -287,12 +287,12 @@ where /// `Waker::will_wake` as the latter tests the pointers to the virtual tables /// for equality. /// -/// Forcing the function to be inlined appears to solve this problem, but we may -/// want to investigate more robust methods. Tokio has [switched][1] to a single -/// non-generic virtual table declared as `static`, which then delegates each -/// call with another virtual call. This does ensure that `Waker::will_wake` -/// will always work, but the double indirection is a bit unfortunate and its -/// cost would need to be evaluated. +/// Preventing the function from being inlined appears to solve this problem, +/// but we may want to investigate more robust methods. For unrelated reasons, +/// Tokio has switched [1] to a single non-generic virtual table declared as +/// `static` which then delegates each call to another virtual call. This does +/// ensure that `Waker::will_wake` will always work, but the double indirection +/// is a bit unfortunate and its cost would need to be evaluated. /// /// [1]: https://github.com/tokio-rs/tokio/pull/5213 #[inline(never)]