Re: RFR: 8237878: Improve ModuleLoaderMap datastructures

2020-02-13 Thread Claes Redestad
Hi, On 2020-02-13 09:12, Peter Levart wrote: This change is ok as it stands, but I'm afraid it is not in the spirit of Valhalla. I wouldn't really know what the spirit of Valhalla will end up being, yet. :-) Unboxing the values have a measurable cost in the interpreter, and can trigger some v

Re: RFR: 8237878: Improve ModuleLoaderMap datastructures

2020-02-13 Thread Peter Levart
Hi Claes, I hope I'm not to late to comment on this. This change is ok as it stands, but I'm afraid it is not in the spirit of Valhalla. As I understand, you rely on the fact that Integer instances in the low range of values are cached and you then use identity comparison in the following met

Re: RFR: 8237878: Improve ModuleLoaderMap datastructures

2020-02-10 Thread Claes Redestad
On 2020-02-10 12:34, Alan Bateman wrote: On 10/02/2020 09:04, Claes Redestad wrote: : So how about: http://cr.openjdk.java.net/~redestad/8237878/open.02/ Thanks for restoring the use of Function. Changing Module::defineClass to invoke a method on ModuleLoaderMap is okay but the method ne

Re: RFR: 8237878: Improve ModuleLoaderMap datastructures

2020-02-10 Thread Alan Bateman
On 10/02/2020 09:04, Claes Redestad wrote: : So how about: http://cr.openjdk.java.net/~redestad/8237878/open.02/ Thanks for restoring the use of Function. Changing Module::defineClass to invoke a method on ModuleLoaderMap is okay but the method needs to something like "isBuiltinMapper" becau

Re: RFR: 8237878: Improve ModuleLoaderMap datastructures

2020-02-10 Thread Claes Redestad
On 2020-02-09 17:49, Alan Bateman wrote: On 06/02/2020 13:48, Claes Redestad wrote: : New webrev: http://cr.openjdk.java.net/~redestad/8237878/open.01/ The archiving looks good but I'd prefer if this patch didn't change the usages of Function to ModuleLoaderMap.Mapper - that's an implemen

Re: RFR: 8237878: Improve ModuleLoaderMap datastructures

2020-02-09 Thread Alan Bateman
On 06/02/2020 13:48, Claes Redestad wrote: : New webrev: http://cr.openjdk.java.net/~redestad/8237878/open.01/ The archiving looks good but I'd prefer if this patch didn't change the usages of Function to ModuleLoaderMap.Mapper - that's an implementation class that should not be used outside

Re: RFR: 8237878: Improve ModuleLoaderMap datastructures

2020-02-06 Thread forax
- Mail original - > De: "Claes Redestad" > À: "Remi Forax" > Cc: "core-libs-dev" > Envoyé: Jeudi 6 Février 2020 14:48:38 > Objet: Re: RFR: 8237878: Improve ModuleLoaderMap datastructures > Hi Rémi, > > > On 2020-02-06 14:08,

Re: RFR: 8237878: Improve ModuleLoaderMap datastructures

2020-02-06 Thread Claes Redestad
e it. Rémi - Mail original - De: "Claes Redestad" À: "core-libs-dev" Envoyé: Jeudi 6 Février 2020 13:37:59 Objet: RFR: 8237878: Improve ModuleLoaderMap datastructures Hi, refactor ModuleLoaderMap to allow the datastructure to be archived, then archive it. Webre

Re: RFR: 8237878: Improve ModuleLoaderMap datastructures

2020-02-06 Thread Remi Forax
ink it's better than having a mix between codes that use var and codes that doesn't use it. Rémi - Mail original - > De: "Claes Redestad" > À: "core-libs-dev" > Envoyé: Jeudi 6 Février 2020 13:37:59 > Objet: RFR: 8237878: Improve ModuleLoaderMa

RFR: 8237878: Improve ModuleLoaderMap datastructures

2020-02-06 Thread Claes Redestad
Hi, refactor ModuleLoaderMap to allow the datastructure to be archived, then archive it. Webrev: http://cr.openjdk.java.net/~redestad/8237878/open.00/ Bug:https://bugs.openjdk.java.net/browse/JDK-8237878 Testing: tier1-3 /Claes