Bug 254207

Summary:

media/media-source/media-source-remove-readystate is a failure on GTK/WPE and media/media-source/media-source-monitor-playing-event.html timeouts on WPE

Product:

WebKit

Reporter:

Jean-Yves Avenard [:jya] <jean-yves.avenard>

Component:

Media

Assignee:

Nobody <webkit-unassigned>

Status:

NEW ---

Severity:

Normal

CC:

aboya, clopez

Priority:

Version:

WebKit Nightly Build

Hardware:

Unspecified

OS:

Linux

See Also:

https://bugs.webkit.org/show_bug.cgi?id=254198

Attachments:

Description	Flags
TestCase	none

Description Jean-Yves Avenard [:jya] 2023-03-21 04:32:58 PDT

Created attachment 465538 [details]
TestCase

Test added in bug 254198 is a constant failure and I can't make much sense of it seeing that readyState calculations should occur in platform agnostic code.

The test calls `SourceBuffer.appendBuffer()` until there's over 10s of data added to the buffered range. Yet readyState stays at 1 (HAVE_METADATA)

```
--- /home/ews/worker/GTK-WK2-Tests-EWS/build/layout-test-results/media/media-source/media-source-remove-readystate-expected.txt
+++ /home/ews/worker/GTK-WK2-Tests-EWS/build/layout-test-results/media/media-source/media-source-remove-readystate-actual.txt
@@ -6,7 +6,7 @@
 RUN(sourceBuffer.appendBuffer(loader.initSegment()))
 EVENT(update)
 Appending 10s Data
-EXPECTED (video.readyState > '1') OK
+EXPECTED (video.readyState > '1'), OBSERVED '1' FAIL
 RUN(video.currentTime = 2.5)
 EVENT(seeked)
 RUN(source.endOfStream())
@@ -15,6 +15,6 @@
 RUN(sourceBuffer.remove(video.currentTime, video.currentTime + 5))
 EVENT(waiting)
 EVENT(update)
-EXPECTED (video.readyState == '1') OK
+EXPECTED (video.readyState == '1'), OBSERVED '2' FAIL
 END OF TEST
 ```

Comment 1 Jean-Yves Avenard [:jya] 2023-03-22 01:32:24 PDT

The test found in `LayoutTests/media/media-source/media-source-monitor-playing-event.html` is incorrect too.

There's a sleep(1s) right after `sourceBuffer.remove(0, 10)` to check the `readyState` but it should have been updated even before the `updateend` event got fired.

I have marked it as failed for now.

Comment 2 Carlos Alberto Lopez Perez 2023-08-31 10:43:01 PDT

(In reply to Jean-Yves Avenard [:jya] from comment #1)
> The test found in
> `LayoutTests/media/media-source/media-source-monitor-playing-event.html` is
> incorrect too.
> 
> There's a sleep(1s) right after `sourceBuffer.remove(0, 10)` to check the
> `readyState` but it should have been updated even before the `updateend`
> event got fired.
> 
> I have marked it as failed for now.

The test media/media-source/media-source-monitor-playing-event.html passes fine on GTK now, but on WPE timeouts 90% of the times.