Re: RFR: JDK-8068803:Performance of LocalDate.plusDays could be better

2016-01-09 Thread nadeesh tv
Hi Stephen, It's already covered in public void test_plusDays_invalidTooLarge() and public void test_minusDays_maximum(). Thanks and Regards, Nadeesh On 1/9/2016 4:58 PM, Stephen Colebourne wrote: Thanks for the update. You should have a test case for the exception thrown by the

Re: RFR: JDK-8068803:Performance of LocalDate.plusDays could be better

2016-01-09 Thread Stephen Colebourne
Thanks for the update. You should have a test case for the exception thrown by the checkValidValue() Stephen On 9 January 2016 at 09:33, nadeesh tv wrote: > Hi Stephen/Roger, > > Please see the updated the webrev > http://cr.openjdk.java.net/~ntv/8068803/webrev.03/ > >

Re: RFR: JDK-8068803:Performance of LocalDate.plusDays could be better

2016-01-09 Thread Stephen Colebourne
I'm happy with this now, thanks Stephen On 9 January 2016 at 12:44, nadeesh tv wrote: > Hi Stephen, > > It's already covered in > > public void test_plusDays_invalidTooLarge() and public void > test_minusDays_maximum(). > > Thanks and Regards, > Nadeesh > > > On

Re: RFR: JDK-8068803:Performance of LocalDate.plusDays could be better

2016-01-09 Thread nadeesh tv
Hi Stephen/Roger, Please see the updated the webrev http://cr.openjdk.java.net/~ntv/8068803/webrev.03/ Explicit "YEAR.checkValidValue(year + 1);' added in 3rd case to handle the *invalidTooLarge year case (*LocalDate.of(Year.MAX_VALUE, 12, 31).plusDays(1)) Thanks and Regards, Nadeesh

Re: RFR: JDK-8068803:Performance of LocalDate.plusDays could be better

2016-01-08 Thread nadeesh tv
Hi all, Thanks Stephen for the comments Please see the updated webrev http://cr.openjdk.java.net/~ntv/8068803/webrev.02/ Regards, Nadeesh On 1/7/2016 6:15 PM, Stephen Colebourne wrote: I updated the benchmark with this change and another one:

Re: RFR: JDK-8068803:Performance of LocalDate.plusDays could be better

2016-01-08 Thread Roger Riggs
+1 (With Stephen's update below). Roger On 1/8/2016 6:56 AM, Stephen Colebourne wrote: As I mentioned in my email: Rather than doing: return withDayOfMonth((int) dom); or return LocalDate.of(year, month, (int) dom); you can do return new LocalDate(year, month, (int) dom); (there

Re: RFR: JDK-8068803:Performance of LocalDate.plusDays could be better

2016-01-08 Thread Stephen Colebourne
As I mentioned in my email: Rather than doing: return withDayOfMonth((int) dom); or return LocalDate.of(year, month, (int) dom); you can do return new LocalDate(year, month, (int) dom); (there are two occurrences) Stephen On 8 January 2016 at 10:56, nadeesh tv

RFR: JDK-8068803:Performance of LocalDate.plusDays could be better

2016-01-05 Thread nadeesh tv
Hi all, Please review a fix for https://bugs.openjdk.java.net/browse/JDK-8068803 web rev : http://cr.openjdk.java.net/~ntv/8068803/webrev.00/ Special thanks for Stephen for providing the source code patch -- Thanks and Regards, Nadeesh TV

Re: RFR: JDK-8068803:Performance of LocalDate.plusDays could be better

2016-01-05 Thread Roger Riggs
Hi Nadeesh, LocalDate.java: +1374: For the most common case of dom > 0 && <= 28, I would have explicitly and immediately returned the new LocalDate. if (dom > 0 && dom <= 28) { return LocalDate.of(year, month, (int) dom); } ... TCKLocalDate.java: - Since the test_plusDays_normal is