Bug 517818. When nsMediaCache::Update wants to read from the end of the stream, don't resume if we're already suspended. And if the channel is not already positioned at the end, suspend instead of seeking to the end, since there is no point in seeking to the end and it may cause harmful side effects. r=doublec

--HG--
extra : rebase_source : e816c61621f5dd7c670687cf1a70d2eccb488dd6

content/media/nsMediaCache.cpp

@@ -1130,11 +1130,18 @@ nsMediaCache::Update()
     PRBool enableReading;
     if (stream->mStreamLength >= 0 &&
         desiredOffset >= stream->mStreamLength) {
-      // We're at the end of the stream. Nothing to read, but we don't
-      // need to suspend, we may as well just keep reading and hit EOF
-      // (or discover more data if the server lied to us).
+      // We want to read at the end of the stream, where there's nothing to
+      // read. We don't want to try to read if we're suspended, because that
+      // might create a new channel and seek unnecessarily (and incorrectly,
+      // since HTTP doesn't allow seeking to the actual EOF), and we don't want
+      // to suspend if we're not suspended and already reading at the end of
+      // the stream, since there just might be more data than the server
+      // advertised with Content-Length, and we may as well keep reading.
+      // But we don't want to seek to the end of the stream if we're not
+      // already there.
       LOG(PR_LOG_DEBUG, ("Stream %p at end of stream", stream));
-      enableReading = PR_TRUE;
+      enableReading = !stream->mCacheSuspended &&
+        desiredOffset == stream->mChannelOffset;
     } else if (desiredOffset < stream->mStreamOffset) {
       // We're reading to try to catch up to where the current stream
       // reader wants to be. Better not stop.