Patch approved. I will include it in my next push.
> Hi Hema,
> as far as I can imagine, the problem is that the test uses character
> strings, which do not contain valid SJIS characters. Some change in the
> server seems to behave different on invalid characters than before.
> IMHO this is acceptable. Only valid input needs to produce predictable
> I propose the following patch, which uses valid SJIS characters (I
> hope): http://lists.mysql.com/commits/100981
> Please approve.