Bump minor version.

lib/io/event/version.rb

@@ -7,6 +7,6 @@
 class IO
 	# @namespace
 	module Event
-		VERSION = "1.8.4"
+		VERSION = "1.9.0"
 	end
 end

readme.md

@@ -18,6 +18,10 @@ Please see the [project documentation](https://socketry.github.io/io-event/) for
 
 Please see the [project releases](https://socketry.github.io/io-event/releases/index) for all releases.
 
+### v1.9.0
+
+  - [Improved `IO::Event::Profiler` for detecting stalls.](https://socketry.github.io/io-event/releases/index#improved-io::event::profiler-for-detecting-stalls.)
+
 ### v1.8.0
 
   - [Detecting fibers that are stalling the event loop.](https://socketry.github.io/io-event/releases/index#detecting-fibers-that-are-stalling-the-event-loop.)

releases.md

@@ -1,12 +1,12 @@
 # Releases
 
-## Unreleased
+## v1.9.0
 
 ### Improved `IO::Event::Profiler` for detecting stalls.
 
 A new `IO::Event::Profiler` class has been added to help detect stalls in the event loop. The previous approach was insufficient to detect all possible stalls. This new approach uses the `RUBY_EVENT_FIBER_SWITCH` event to track context switching by the scheduler, and can detect stalls no matter how they occur.
 
-```ruby
+``` ruby
 profiler = IO::Event::Profiler.new
 
 profiler.start
@@ -20,9 +20,9 @@ profiler.stop
 
 A default profiler is exposed using `IO::Event::Profiler.default` which is controlled by the following environment variables:
 
-- `IO_EVENT_PROFILER=true` - Enable the profiler, otherwise `IO::Event::Profiler.default` will return `nil`.
-- `IO_EVENT_PROFILER_LOG_THRESHOLD` - Specify the threshold in seconds for logging a stall. Defaults to `0.01`.
-- `IO_EVENT_PROFILER_TRACK_CALLS` - Track the method call for each event, in order to log specifically which method is causing the stall. Defaults to `true`.
+  - `IO_EVENT_PROFILER=true` - Enable the profiler, otherwise `IO::Event::Profiler.default` will return `nil`.
+  - `IO_EVENT_PROFILER_LOG_THRESHOLD` - Specify the threshold in seconds for logging a stall. Defaults to `0.01`.
+  - `IO_EVENT_PROFILER_TRACK_CALLS` - Track the method call for each event, in order to log specifically which method is causing the stall. Defaults to `true`.
 
 The previous environment variables `IO_EVENT_SELECTOR_STALL_LOG_THRESHOLD` and `IO_EVENT_SELECTOR_STALL_LOG` no longer have any effect.