Review request for JDK-8014230: Compilation incorrectly succeeds with inner class constructor with 254 parameters

[Eric McCorkle]

A new webrev has been posted that incorporates your and Alex's suggestions.
http://cr.openjdk.java.net/~emc/8014230/


On 06/20/13 18:30, Jonathan Gibbons wrote:
> Changes to Main are unnecessary and should be undone.  If you want to
> provide your own Log, stash it in the Context -- that is why that
> parameter is provided.
> 
> Also, why do you need to interpose on Log -- why not use Diagnostic API?
> 
> -- Jon
> 
> 
> On 06/19/2013 12:08 PM, Eric McCorkle wrote:
>> Added a new constructor to Main that allows tests to pass in their own
>> Log, facilitating much more precise tests of what errors the compilers
>> generates.
>>
>> I also rolled all the arg limit tests into a general framework.
>>
>> Webrev is here:
>> http://cr.openjdk.java.net/~emc/8014230/
>>
>> On 06/14/13 16:30, Jonathan Gibbons wrote:
>>> There are two solutions.
>>>
>>> For a one-off test, the standard technique is to use a golden file
>>> in conjunctions with rawDiagnostics.   The combination is
>>> "somewhat frail" but nowhere near "incredibly frail".  The technique
>>> has served us well for many years now.
>>>
>>> If you are driving javac through the Compiler API, you can register
>>> a DiagnosticListener, and verify the characteristics of the Diagnostic
>>> objects passed to report.
>>>
>>> -- Jon
>>>
>>>
>>> On 06/14/2013 01:14 PM, Eric McCorkle wrote:
>>>> Is there a more convenient API for checking error messages?  Golden
>>>> will
>>>> make for an incredibly frail test, I think.
>>>>
>>>> On 06/14/13 15:46, Jonathan Gibbons wrote:
>>>>> The change to Gen looks OK, but the tests look weak.  At a minimum, I
>>>>> would expect to see the test check the validity of the error message
>>>>> that is generated; even better would be to generate the test cases on
>>>>> the fly.
>>>>>
>>>>> -- Jon
>>>>>
>>>>> On 06/14/2013 11:29 AM, Eric McCorkle wrote:
>>>>>> Hello,
>>>>>>
>>>>>> Please review this patch, which addresses a problem with javac not
>>>>>> taking inner this parameters into account when determining if
>>>>>> there are
>>>>>> too many parameters to a function.
>>>>>>
>>>>>> The webrev is here:
>>>>>> http://cr.openjdk.java.net/~emc/8014230/
>>>>>>
>>>>>> The bug report is here:
>>>>>> http://bugs.sun.com/view_bug.do?bug_id=8014230
>>>>>>
>>>>>> Thanks,
>>>>>> Eric
> 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: eric_mccorkle.vcf
Type: text/x-vcard
Size: 314 bytes
Desc: not available
Url : http://mail.openjdk.java.net/pipermail/compiler-dev/attachments/20130621/35ed7bfa/eric_mccorkle.vcf