RFR: 8246451: Reduce overhead of normalizing file paths with trailing slash

[Lance Andersen]

Hi Claes,

This looks OK :-)

Best
Lance

> On Jun 3, 2020, at 7:54 AM, Claes Redestad <claes.redestad at oracle.com> wrote:
> 
> Hi,
> 
> I'd like to fix a bug in the microbenchmark I pushed for JDK-8246338
> where the FileOpen.trailingSlash variant doesn't do what it intends to.
> 
> There's also a profitable low-hanging optimization down that path to
> substring rather than take the detour via StringBuilder[1]
> 
> Bug:    https://bugs.openjdk.java.net/browse/JDK-8246451
> Webrev: http://cr.openjdk.java.net/~redestad/8246451/open.00/
> 
> Testing: tier1
> 
> Thanks!
> 
> /Claes
> 
> [1]
> Benchmark                                                Mode  Cnt Score     Error   Units
> FileOpen.trailingSlash                                   avgt    5 0.054 ±   0.010   us/op
> FileOpen.trailingSlash:·gc.alloc.rate.norm               avgt    5 160.013 ±   0.003    B/op
> 
> FileOpen.trailingSlash                                   avgt    5 0.036 ±   0.003   us/op
> FileOpen.trailingSlash:·gc.alloc.rate.norm               avgt    5 96.008 ±   0.002    B/op

 <http://oracle.com/us/design/oracle-email-sig-198324.gif>
 <http://oracle.com/us/design/oracle-email-sig-198324.gif> <http://oracle.com/us/design/oracle-email-sig-198324.gif>
 <http://oracle.com/us/design/oracle-email-sig-198324.gif>Lance Andersen| Principal Member of Technical Staff | +1.781.442.2037
Oracle Java Engineering 
1 Network Drive 
Burlington, MA 01803
Lance.Andersen at oracle.com <mailto:Lance.Andersen at oracle.com>