[foreign-jextract] RFR Add varargs support for jextract generated code

[sundararajan.athijegannathan at oracle.com]

Thanks. I'll fix that and push.

-Sundar

On 22/01/20 11:32 pm, Maurizio Cimadamore wrote:
>
> Looks good - minor nit - the constant holder can be computed in a 
> static initializer and stored in a static final constant.
>
> Maurizio
>
> On 22/01/2020 17:40, sundararajan.athijegannathan at oracle.com wrote:
>>
>> Fixed. Please review the updated webrev: 
>> http://cr.openjdk.java.net/~sundar/minimal_jextract_varargs/webrev.01/
>>
>> Thanks,
>>
>> -Sundar
>>
>> On 22/01/20 7:47 pm, Maurizio Cimadamore wrote:
>>>
>>> Looks good - but there's an issue in the template changes - the 
>>> template does explicit, platform-dependent layout imports:
>>>
>>> +import static jdk.incubator.foreign.MemoryLayouts.SysV.C_DOUBLE;
>>> +import static jdk.incubator.foreign.MemoryLayouts.SysV.C_LONG;
>>> +import static jdk.incubator.foreign.MemoryLayouts.SysV.C_POINTER;
>>> Now, when [1] will be fixed, you will be able to have a single set 
>>> of static imports which will auto-resolve to the right constants. I 
>>> think for now (and maybe in the future) we should use some kind of 
>>> templating scheme here e.g.
>>>
>>> import static ${C_LONG}
>>>
>>> etc.
>>>
>>> And then jextract will fill in with the layout it thinks are the 
>>> correct one (at extraction time).
>>>
>>> Maurizio
>>>
>>> [1] https://bugs.openjdk.java.net/browse/JDK-8237545
>>>
>>> On 22/01/2020 14:07, sundararajan.athijegannathan at oracle.com wrote:
>>>> Please review.
>>>>
>>>> Bug: https://bugs.openjdk.java.net/browse/JDK-8237646
>>>>
>>>> Webrev: 
>>>> http://cr.openjdk.java.net/~sundar/minimal_jextract_varargs/webrev.00/
>>>>
>>>> PS. Piggybacking to add ToolProvider for jextract tool and basic 
>>>> test framework with simple generation tests.
>>>>
>>>> Thanks,
>>>>
>>>> -Sundar
>>>>