Re: [classlib][luni] calling mmap

2009-09-29 Thread Mark Hindess
In message <25aac9fc0909290909v2e012fdfnc93e37eb32c79...@mail.gmail.com>, sebb writes: > > On 29/09/2009, Oliver Deakin wrote: > > Tim Ellison wrote: > > > > > On 29/Sep/2009 15:40, Oliver Deakin wrote: > > > [SNIP] > > > > > > > That's true, I hadn't spotted that. I also noticed a bug in the > >

Re: [jira] Commented: (HARMONY-6312) Concurrency problems in NIO

2009-09-29 Thread Jesse Wilson
On Tue, Sep 29, 2009 at 1:20 AM, Regis wrote: > I think "changes to the selected keys set" doesn't help to show more > things. > Because there is no much difference in processSelectResult. And the reason > for maintaining readableFDs/writableFDs and mapping is avoiding O(num(keys)) > in Selector.

Re: [classlib][luni] calling mmap

2009-09-29 Thread sebb
On 29/09/2009, Oliver Deakin wrote: > Tim Ellison wrote: > > > On 29/Sep/2009 15:40, Oliver Deakin wrote: > > > > > > > Tim Ellison wrote: > > > > > > > > > > Making size == 0 a special case that always succeeds (even if it is > > > > returning a bogus address) bypasses all the other checks that t

Re: [classlib][luni] calling mmap

2009-09-29 Thread Oliver Deakin
Tim Ellison wrote: On 29/Sep/2009 15:40, Oliver Deakin wrote: Tim Ellison wrote: Making size == 0 a special case that always succeeds (even if it is returning a bogus address) bypasses all the other checks that the mmap function call would have made. For example, what if the file descr

Re: [classlib][luni] calling mmap

2009-09-29 Thread Tim Ellison
On 29/Sep/2009 15:40, Oliver Deakin wrote: > Tim Ellison wrote: >> Making size == 0 a special case that always succeeds (even if it is >> returning a bogus address) bypasses all the other checks that the mmap >> function call would have made. >> >> For example, what if the file descriptor was inval

Re: [classlib][luni] calling mmap

2009-09-29 Thread Oliver Deakin
Tim Ellison wrote: Sorry for talking so long to raise this... On 16/Sep/2009 23:44, odea...@apache.org wrote: Author: odeakin Date: Wed Sep 16 22:44:49 2009 New Revision: 815994 URL: http://svn.apache.org/viewvc?rev=815994&view=rev Log: Apply patch for HARMONY-6315 ([classlib][nio] FileChan

[classlib][luni] calling mmap (was: Re: svn commit: r815994 ...)

2009-09-29 Thread Tim Ellison
Sorry for talking so long to raise this... On 16/Sep/2009 23:44, odea...@apache.org wrote: > Author: odeakin > Date: Wed Sep 16 22:44:49 2009 > New Revision: 815994 > > URL: http://svn.apache.org/viewvc?rev=815994&view=rev > Log: > Apply patch for HARMONY-6315 ([classlib][nio] FileChannel.map thr

Re: [jira] Created: (HARMONY-6344) Provide a lock-free version of ReferenceQueue

2009-09-29 Thread Tim Ellison
On 29/Sep/2009 10:13, Regis wrote: > James Gan (JIRA) wrote: >> Provide a lock-free version of ReferenceQueue >> - >> >> Key: HARMONY-6344 >> URL: https://issues.apache.org/jira/browse/HARMONY-6344 >> Project

Re: [jira] Created: (HARMONY-6344) Provide a lock-free version of ReferenceQueue

2009-09-29 Thread Regis
James Gan (JIRA) wrote: Provide a lock-free version of ReferenceQueue - Key: HARMONY-6344 URL: https://issues.apache.org/jira/browse/HARMONY-6344 Project: Harmony Issue Type: Improvement

Re: [jira] Commented: (HARMONY-6312) Concurrency problems in NIO

2009-09-29 Thread Regis
Jesse Wilson wrote: On Mon, Sep 28, 2009 at 1:58 AM, Regis Xu (JIRA) wrote: In my understanding, SelectorBenchmark.java try to simulate a "real" scenario of using selector, so I picked benchmark from HARMONY-4879 which *only* test Selector.selectNow(), the result: svn + no mapping clients/act