RFR : (bf) Backout 7190219 fix for jdk7u (JCK issue)

[Seán Coffey]

I haven't heard any more comments around this issue. I'll go ahead and 
back out this change from jdk7u on Friday unless I hear to the contrary.

regards,
Sean.

On 12/10/2012 08:48, Seán Coffey wrote:
> Jonathan,
>
> any thoughts here, are you ok for me to backout the current fix until 
> we resolve/relax the JCK testcase conditions  ?
>
> regards,
> Sean.
>
> On 08/10/12 09:19, Alan Bateman wrote:
>> On 08/10/2012 03:15, Jonathan Lu wrote:
>>> Hi Sean,
>>>
>>> Alan has got a fix for the behavior change introduced by 7190219, if 
>>> it solves the JCK failure, is it possible to back port 7199551 as 
>>> well instead of dropping this defect-fixing patch 7190219?
>>>
>>> Change set for 7199551,
>>> http://hg.openjdk.java.net/jdk8/tl/jdk/rev/3cfb621d5e7e
>> I don't think we should backport 7199551 to 7u. The reason is that it 
>> changes put($Type$Buffer) to throw ReadOnlyBufferException rather 
>> than BufferOverflowException for cases where someone attempts to put 
>> into a read-only buffer with insufficient space. Technically either 
>> exception should be okay but changing the exception in an update 
>> release often causes surprises. People expect change in major 
>> releases, not update releases.
>>
>> For the JCK issue then this is clearly a bug that needs to be fixed 
>> there but it does highlight the behavior change that I should have 
>> spotted during the review of your change. So for 7u then I think we 
>> either anti-delta 7190219 as Sean proposes, or alternative we 
>> changing put(String,int,int) to check if the buffer is read-only -- 
>> ie: partial backport of 7199551. Either is okay with me.
>>
>> -Alan.
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.openjdk.java.net/pipermail/nio-dev/attachments/20121018/149674ac/attachment.html