WebSocket spec says "Frames denoted by bytes that have the high bit set (0x80 to 0xFF) have a leading length indicator, which is encoded as a series of 7-bit bytes stored in octets with the 8th bit being set for all but the last byte", but current implementation doesn't count the last byte with 8th bit off.
Created attachment 44370 [details] Fix bad legnth parsing in WebSocket.
style-queue ran check-webkit-style on attachment 44370 [details] without any errors.
+ int c = *p; Can "c" have a descriptive name? - msg += "\x80\x81\x81" + msg += "\x80\x81\x01" # skip 1*128+1 bytes. ChangeLog should explain why changing an existing test is OK. It is not obvious why/if the bug it was originally created for is still tested.
Created attachment 44388 [details] Fix wrong length parsing in WebSocket.
style-queue ran check-webkit-style on attachment 44388 [details] without any errors.
Committed r51829: <http://trac.webkit.org/changeset/51829>