On Tue, 27 Oct 2026 18:24:11 GMT, Aleksey Shipilev <[email protected]> wrote:
> The uncommon trap count regression seems to be gone! > > I can see one config where first iterations are considerably faster when I > set a giant PCR, but the steady state is still slower to achieve. I think I'd expect that: the counters are updated so rarely that the compilers are starved of information. I don't think there has to be any more explanation for it. > BTW, this fails right away with ` -XX:+AggressiveProfileReduction`: Fixed. I removed one "optimization" that seems to confuse the C1 register allocator but doesn't help much. ------------- PR Comment: https://git.openjdk.org/jdk/pull/28541#issuecomment-6503363604
