RFR: 8213795: Force explicit null check on patching placeholder offset

Vladimir Kozlov vladimir.kozlov at oracle.com
Wed Nov 14 17:30:32 UTC 2018

Looks good.

So it is not only C1 but Interpreter too. Comment is good.


On 11/14/18 3:06 AM, Roman Kennke wrote:
> Hi Andrew, Vladimir,
> See comment in updated webrev. Does it explain the situation well?
> http://cr.openjdk.java.net/~rkennke/JDK-8213795/webrev.01/
> Thanks,
> Roman
>> On 13/11/2018 23:21, Vladimir Kozlov wrote:
>>> Originally I just want more detailed comment, nothing more. Current
>>> comment is too general.
>>> I asked Graal's engineers and they said Graal does not use -1 offset.
>>> This is really C1 only "feature".
>>> The comment should reflect that.
>> Vladimir is right that a comment fingering C1 for this behaviour will be
>> enough. Scoping the provenance of those -1 values is the right step #1.
>> Fixing the code to show how they actually get lobbed out of that
>> particular ball park can be left for a later clean-up.
>> So, I'm still ok with the patch once the comment is updated.
>> regards,
>> Andrew Dinn
>> -----------
>> Senior Principal Software Engineer
>> Red Hat UK Ltd
>> Registered in England and Wales under Company Registration No. 03798903
>> Directors: Michael Cunningham, Michael ("Mike") O'Neill, Eric Shander

More information about the hotspot-compiler-dev mailing list