[foreign] RFR 8218679: binder throws NPE when binding sqlite3 interface
Sundararajan Athijegannathan
sundararajan.athijegannathan at oracle.com
Mon Feb 11 16:05:11 UTC 2019
Oops! Fixed it. Updated:
https://cr.openjdk.java.net/~sundar/8218679/webrev.02/
-Sundar
On 11/02/19, 8:53 PM, Maurizio Cimadamore wrote:
> Looks good - but when looking at the code some more, I found another
> place where things probably go wrong: Sequence.
>
> We have two subclasses in Layout: Address (subclass from Value) and
> Sequence (subclass from Group).
>
> Address correctly delegates hash/equals to super, Sequence does not,
> so I believe it can be prone to the same issue you are trying to fix
> (e.g. not taking annos into account).
>
> Maurizio
>
> On 11/02/2019 15:11, Sundararajan Athijegannathan wrote:
>> Fixed Unresolved.equals and hashCode as mentioned.
>>
>> Updated: https://cr.openjdk.java.net/~sundar/8218679/webrev.01/
>>
>> -Sundar
>>
>> On 11/02/19, 7:06 PM, Maurizio Cimadamore wrote:
>>> Looks good (and kind of surprising we made it thus far w/o seeing
>>> this!).
>>>
>>> Minor quibble - I'd tweak the Unresolved equals/hash to use
>>> layoutExpression + annos.
>>>
>>> Moving forward, we want jextract to generate a single artifact with
>>> single name resolution context - so, same layoutExpr -> same layout.
>>>
>>> Maurizio
>>>
>>> On 11/02/2019 13:18, Sundararajan Athijegannathan wrote:
>>>> Please review.
>>>>
>>>> Bug: https://bugs.openjdk.java.net/browse/JDK-8218679
>>>> Webrev: https://cr.openjdk.java.net/~sundar/8218679/webrev.00/
>>>>
>>>> Thanks
>>>> -Sundar
More information about the panama-dev
mailing list