Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

On 01/07/16 16:33, Mandy Chung wrote: I think the additional check is unecessary but it’s harmless if you prefre to keep that. Oh - I see - the permission check in ConfigurationData::merge can be removed: ConfigurationData is a private class and config is a private field. I will do that.

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

> On Jul 1, 2016, at 8:23 AM, Daniel Fuchs wrote: > > On 01/07/16 16:19, Mandy Chung wrote: >>> I'd prefer to keep the doPrivileged in LogManager so that >>> > Logger.mergeWithSystemLogger can call checkpermission(). >>> > >>> > From a conceptual point of view it's only

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

On 01/07/16 16:19, Mandy Chung wrote: I'd prefer to keep the doPrivileged in LogManager so that > Logger.mergeWithSystemLogger can call checkpermission(). > > From a conceptual point of view it's only when calling > this method from LogManager that we want to be privileged, > even though the

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

> On Jul 1, 2016, at 8:16 AM, Daniel Fuchs wrote: > > On 01/07/16 16:09, Mandy Chung wrote: >> I was thinking to move the entire doPrivileged block to >> mergeWithSystemLogger that will look like: >> >>if (cfg == system.config) { >>return; >>} >> >>

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

> On Jul 1, 2016, at 7:09 AM, Daniel Fuchs wrote: > > > OK. I was concerned about the cost of creating a doPrivileged call > for nothing - but then reflected that system classes usually call > getLogger only once for a specific logger - at which point the >

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

> On Jun 30, 2016, at 9:33 AM, Daniel Fuchs wrote: > > Indeed, good catch! I should have seen that :-( > > Here is a patch that should take care of the issue: > > http://cr.openjdk.java.net/~dfuchs/webrev_8159245/webrev.05 > > Thanks for your offline suggestions on

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

Hi Mandy, On 21/06/16 17:01, Mandy Chung wrote: On Jun 21, 2016, at 8:39 AM, Daniel Fuchs wrote: I don't understand this scenario. ConfigurationData should remain as simple as possible. Logger.getLogger() / LogManager.demandSystemLogger() will never return a logger

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

Hi Daniel, Got you point. Thank you -Hamlin On 2016/6/22 13:24, Daniel Fuchs wrote: Hi Hamlin, On 22/06/16 02:40, Hamlin Li wrote: Just another minor comment, I'm not sure if following line is necessary in Logger.java, as it's already been checked in LogManager.java line 577: 439 if (cfg

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

Hi Hamlin, On 22/06/16 02:40, Hamlin Li wrote: Just another minor comment, I'm not sure if following line is necessary in Logger.java, as it's already been checked in LogManager.java line 577: 439 if (cfg == other) return cfg; Yes, you're right - but it makes sense for the method to have it

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

Hi Daniel, Thank you for detailed explanation, new version makes sense to me. Just another minor comment, I'm not sure if following line is necessary in Logger.java, as it's already been checked in LogManager.java line 577: 439 if (cfg == other) return cfg; Thank you -Hamlin On 2016/6/21

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

> On Jun 21, 2016, at 8:39 AM, Daniel Fuchs wrote: > > I don't understand this scenario. > ConfigurationData should remain as simple as possible. > Logger.getLogger() / LogManager.demandSystemLogger() will never return > a logger before it has been configured. > When

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

On 21/06/16 16:12, Mandy Chung wrote: On Jun 21, 2016, at 4:54 AM, Daniel Fuchs wrote: Hi Hamlin, I was mistaken in my first assessment. The case where the system handler's list is not empty should only happen if by misfortune two different threads happen to

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

> On Jun 21, 2016, at 4:54 AM, Daniel Fuchs wrote: > > Hi Hamlin, > > I was mistaken in my first assessment. > > The case where the system handler's list is not empty > should only happen if by misfortune two different threads > happen to attempt to merge the same two

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

Hi Hamlin, I was mistaken in my first assessment. The case where the system handler's list is not empty should only happen if by misfortune two different threads happen to attempt to merge the same two configurations concurrently. Though of no consequence for level, filter, etc... (single

Re: RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

On 21/06/16 04:29, Hamlin Li wrote: Hi Daniel, I just has one simple question. In new version, in Logger::importConfig, userParentHandlers, filter, levelObject are copied from other(ConfigurationData) unconditionally, handlers in other are only *copied *when it's empty in system. In previous

RFR 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code.

Hi, Please find below a patch for: 8159245: Loggers created by system classes are not initialized correctly when configured programmatically from application code. Issue: https://bugs.openjdk.java.net/browse/JDK-8150173 Patch: