RFR: 8013395 StringBuffer.toString performance regression impacting embedded benchmarks
david.r.chase at oracle.com
Fri May 10 13:21:26 UTC 2013
On 2013-05-10, at 2:47 AM, David Holmes <david.holmes at oracle.com> wrote:
> Right this was our first thought too, but the specification for toString states that a new String is created. So to go this path we'd also have to push through a spec change for StringBuilder/StringBuffer. Given this is an obscure corner case I wanted to go the most minimal route possible. The copy-constructor doesn't copy the array which is what 6239985 was complaining about.
Any chance that we can concurrently start a push for a spec change? Any place in the spec where we're picky about the reference identity of immutable objects is a candidate for revision; this is an issue for value types, if/when we ever get those into the language.
More information about the core-libs-dev