On Mon, 16 Mar 2026 08:10:55 GMT, Damon Fenacci <[email protected]> wrote:
>> src/hotspot/share/opto/library_call.cpp line 1143:
>>
>>> 1141: if (too_many_traps(Deoptimization::Reason_intrinsic)) {
>>> 1142: return false;
>>> 1143: }
>>
>> @dafedafe, can this be removed too?
>
> It looks OK.
Removed in 0f5f0a342f8.
>> src/hotspot/share/opto/library_call.cpp line 1288:
>>
>>> 1286: if (too_many_traps(Deoptimization::Reason_intrinsic)) {
>>> 1287: return false;
>>> 1288: }
>>
>> @dafedafe, do we still need this too?
>
> This one too.
Removed in 0f5f0a342f8.
>> src/hotspot/share/opto/library_call.cpp line 1391:
>>
>>> 1389: if (too_many_traps(Deoptimization::Reason_intrinsic)) {
>>> 1390: return false;
>>> 1391: }
>>
>> @dafedafe, this one?
>
> Not this one: there is an `uncommon_trap` below.
Right. Not touching it.
>> src/hotspot/share/opto/library_call.cpp line 1464:
>>
>>> 1462: if (too_many_traps(Deoptimization::Reason_intrinsic)) {
>>> 1463: return false;
>>> 1464: }
>>
>> @dafedafe, what about this one?
>
> This one too.
>
> Note: I made a rough check of the code after the three `too_many_traps`
> checks that can be removed. It should be OK to remove them but it would be
> good if you could have another pass as well.
0f5f0a342f8 implements the recent set of deoptimization-related changes. I've
skimmed through it a couple of times, haven't spotted any red flags. I'll
report the Tier 1-2 status soon.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/30139#discussion_r2939823623
PR Review Comment: https://git.openjdk.org/jdk/pull/30139#discussion_r2939824099
PR Review Comment: https://git.openjdk.org/jdk/pull/30139#discussion_r2939824846
PR Review Comment: https://git.openjdk.org/jdk/pull/30139#discussion_r2939841713
