From 3e0803345e764b921805f3987556abf44267faae Mon Sep 17 00:00:00 2001
From: Chris Foster <chris42f@gmail.com>
Date: Mon, 9 Aug 2021 12:46:49 +1000
Subject: [PATCH] Document threadsafety of isready(::Channel) and remove data
 race

Make it clear that isready() is not threadsafe, as it has a data race.
We could make it threadsafe with lock() but this would also make it
blocking and wouldn't make uses of it non-racy without a lock around the
surronding code. So for now just document the current state of affairs.

Also remove an optimistic use of `isready()` in `wait()` as it's a data
race.
---
 base/channels.jl  | 6 +++++-
 base/condition.jl | 1 -
 2 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/base/channels.jl b/base/channels.jl
index 1557504bbe21e..5fb09f38f7a01 100644
--- a/base/channels.jl
+++ b/base/channels.jl
@@ -416,6 +416,11 @@ immediately, does not block.
 
 For unbuffered channels returns `true` if there are tasks waiting
 on a [`put!`](@ref).
+
+!!! warning
+    This function isn't threadsafe. For multithreaded code, use it inside a
+    block of code which is synchronized with `lock(c)` to avoid race
+    conditions.
 """
 isready(c::Channel) = n_avail(c) > 0
 n_avail(c::Channel) = isbuffered(c) ? length(c.data) : length(c.cond_put.waitq)
@@ -427,7 +432,6 @@ unlock(c::Channel) = unlock(c.cond_take)
 trylock(c::Channel) = trylock(c.cond_take)
 
 function wait(c::Channel)
-    isready(c) && return
     lock(c)
     try
         while !isready(c)
diff --git a/base/condition.jl b/base/condition.jl
index be0f618865a48..ce725823dc3b5 100644
--- a/base/condition.jl
+++ b/base/condition.jl
@@ -79,7 +79,6 @@ lock(f, c::GenericCondition) = lock(f, c.lock)
 
 # have waiter wait for c
 function _wait2(c::GenericCondition, waiter::Task)
-    ct = current_task()
     assert_havelock(c)
     push!(c.waitq, waiter)
     return