[lworld] RFR: allow inline classes to have abstract supers

[Frederic Parain]

On Mon, 20 Apr 2020 20:27:12 GMT, Harold Seigel <hseigel at openjdk.org> wrote:

>> Looks good to me.
>> Have you already created the follow-up CR to implement the support for inline friendly abstract classes?
>> Do you have an idea if changes lines 179-181 of verificationTypes.cpp will trigger more class loading then today? If
>> yes, could this become an issue?
>> Fred
>
> Hi Fred,
> 
> Thanks for reviewing this change.
> 
> I have not created the new CR, but will shortly.
> 
> The changes to lines 179-181 should not trigger more class loading than
> today.  Whenever the verifier needs to check if type A is assignable to
> type B, it has to load both types, unless they are the same, or type B
> is java.lang.Object or an interface. However, this load would not be
> needed if the only valid super class for an inline type was
> java.lang.Object.
> 
> Thanks, Harold
> 
> On 4/20/2020 4:09 PM, Frederic Parain wrote:
>>
>> *@fparain* approved this pull request.
>>
>> Looks good to me.
>> Have you already created the follow-up CR to implement the support for
>> inline friendly abstract classes?
>> Do you have an idea if changes lines 179-181 of verificationTypes.cpp
>> will trigger more class loading then today? If yes, could this become
>> an issue?
>>
>> Fred
>>
>> —
>> You are receiving this because you were mentioned.
>> Reply to this email directly, view it on GitHub
>> <https://github.com/openjdk/valhalla/pull/24#pullrequestreview-396775095>,
>> or unsubscribe
>> <https://github.com/notifications/unsubscribe-auth/AOXZ7OTVTZ3PJYMI6LTD2ILRNSTWVANCNFSM4MMWWNRA>.
>>

Harold,

Thank you for the clarifications.

Fred

-------------

PR: https://git.openjdk.java.net/valhalla/pull/24