On Tue, 24 Mar 2026 11:22:16 GMT, Alan Bateman <[email protected]> wrote:
>> src/java.base/share/classes/java/lang/Thread.java line 1799:
>>
>>> 1797: String oldName = this.name;
>>> 1798: this.name = name;
>>> 1799: if (!isVirtual() && Thread.currentThread() == this &&
>>> !name.equals(oldName)) {
>>
>> Surely this slows down normal usage!
>
> I can't imagine this being common. I think part of the issue in this PR is
> that we are a bit in the dark as to the scenario that throws light on
> Thread.setName. Hopefully Roman can find out more.
I made an experiment, and found that adding this check (and nothing else) costs
5-8ns of performance for the normal case (. String comparisons are insanely
fast nowadays. This cost is drowned in the wins elsewhere in this PR - and
turns same-name scenario into a factual no-op.
| Benchmark | (length) | Baseline (ns/op) | +equals check (ns/op) |
Change |
|---------------|----------|------------------:|----------------------:|-------:|
| setName | 1 | 597.4 ± 1.7 | 610.7 ± 7.1 |
+2.2% |
| setName | 16 | 709.4 ± 3.4 | 714.3 ± 7.6 |
+0.7% |
| setName | 200 | 983.2 ± 2.0 | 988.3 ± 2.0 |
+0.5% |
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/30374#discussion_r2980859636
