RFR: 8373595: A new ObjectMonitorTable implementation [v2]

[Fredrik Bredberg]

On Sat, 7 Feb 2026 04:04:33 GMT, Fei Yang <fyang at openjdk.org> wrote:

>> Fredrik Bredberg has updated the pull request incrementally with one additional commit since the last revision:
>> 
>>   Updated all platforms after review comments
>
> src/hotspot/cpu/riscv/gc/shenandoah/shenandoahBarrierSetAssembler_riscv.cpp line 480:
> 
>> 478:   __ beqz(obj, done);
>> 479: 
>> 480:   assert(obj != tmp, "need tmp");
> 
> Seems better if we move this `assert` to entry of this assembler routine and replace it with: `assert_different_registers(obj, tmp);`. Also `ZBarrierSetAssembler::try_resolve_weak_handle_in_c2` could be improved. So I am suggesting following add-on changes:
> 
> 
> diff --git a/src/hotspot/cpu/riscv/gc/shenandoah/shenandoahBarrierSetAssembler_riscv.cpp b/src/hotspot/cpu/riscv/gc/shenandoah/shenandoahBarrierSetAssembler_riscv.cpp
> index fcb5fa512ec..8d530d15ee5 100644
> --- a/src/hotspot/cpu/riscv/gc/shenandoah/shenandoahBarrierSetAssembler_riscv.cpp
> +++ b/src/hotspot/cpu/riscv/gc/shenandoah/shenandoahBarrierSetAssembler_riscv.cpp
> @@ -465,6 +465,8 @@ void ShenandoahBarrierSetAssembler::try_resolve_jobject_in_native(MacroAssembler
>  #ifdef COMPILER2
>  void ShenandoahBarrierSetAssembler::try_resolve_weak_handle_in_c2(MacroAssembler *masm, Register obj,
>                                                                    Register tmp, Label& slow_path) {
> +  assert_different_registers(obj, tmp);
> +
>    Label done;
> 
>    // Resolve weak handle using the standard implementation.
> @@ -473,7 +475,6 @@ void ShenandoahBarrierSetAssembler::try_resolve_weak_handle_in_c2(MacroAssembler
>    // Check if the reference is null, and if it is, take the fast path.
>    __ beqz(obj, done);
> 
> -  assert(obj != tmp, "need tmp");
>    Address gc_state(xthread, ShenandoahThreadLocalData::gc_state_offset());
>    __ lbu(tmp, gc_state);
> 
> diff --git a/src/hotspot/cpu/riscv/gc/z/zBarrierSetAssembler_riscv.cpp b/src/hotspot/cpu/riscv/gc/z/zBarrierSetAssembler_riscv.cpp
> index f2453af19d7..163271a2f11 100644
> --- a/src/hotspot/cpu/riscv/gc/z/zBarrierSetAssembler_riscv.cpp
> +++ b/src/hotspot/cpu/riscv/gc/z/zBarrierSetAssembler_riscv.cpp
> @@ -610,7 +610,9 @@ void ZBarrierSetAssembler::try_resolve_weak_handle_in_c2(MacroAssembler* masm, R
>    BarrierSetAssembler::try_resolve_weak_handle_in_c2(masm, obj, tmp, slow_path);
> 
>    // Check if the oop is bad, in which case we need to take the slow path.
> -  __ ld(tmp, mark_bad_mask_from_thread(xthread));
> +  __ relocate(barrier_Relocation::spec(), [&] {
> +    __ li16u(tmp, barrier_Relocation::unpatched);
> +  }, ZBarrierRelocationFormatMarkBadMask);
>    __ andr(tmp, obj, tmp);
>    __ bnez(tmp, slow_path);

Fixed

-------------

PR Review Comment: https://git.openjdk.org/jdk/pull/29383#discussion_r2813206165