On Tue, 28 Apr 2026 21:55:24 GMT, Vladimir Ivanov <[email protected]> wrote:
>> Quan Anh Mai has updated the pull request incrementally with one additional
>> commit since the last revision:
>>
>> Add diagnostic flag, change initial value to NodeCountInliningCutoff
>
> src/hotspot/share/opto/bytecodeInfo.cpp line 34:
>
>> 32: #include "jfr/jfrEvents.hpp"
>> 33: #include "oops/objArrayKlass.hpp"
>> 34: #include "opto/c2_globals.hpp"
>
> Is it redundant?
Removed.
> src/hotspot/share/opto/bytecodeInfo.cpp line 401:
>
>> 399: // parsing in the caller, and node liveness is more easily
>> determined.
>> 400: if (C->over_inlining_cutoff()) {
>> 401: bool give_up = true;
>
> I suggest to extract the logic into a helper method (e.g.,
> `Compile::should_delay_over_inlining_cutoff()`).
Done!
> src/hotspot/share/opto/compile.cpp line 50:
>
>> 48: #include "opto/addnode.hpp"
>> 49: #include "opto/block.hpp"
>> 50: #include "opto/c2_globals.hpp"
>
> Redundant?
Removed.
> test/micro/org/openjdk/bench/java/lang/foreign/FFMStructAccessTest.java line
> 41:
>
>> 39: @BenchmarkMode(Mode.AverageTime)
>> 40: @OutputTimeUnit(TimeUnit.NANOSECONDS)
>> 41: public class FFMStructAccessTest {
>
> Can you turn one of the microbenchmarks into a regression test?
I have come up with a regression test that I think is more robust since it does
not need to rely on the FFM API.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/30874#discussion_r3158595989
PR Review Comment: https://git.openjdk.org/jdk/pull/30874#discussion_r3158594708
PR Review Comment: https://git.openjdk.org/jdk/pull/30874#discussion_r3158595699
PR Review Comment: https://git.openjdk.org/jdk/pull/30874#discussion_r3158600306
