On Mon, 23 Mar 2026 07:17:39 GMT, Thomas Stuefe <[email protected]> wrote:
>> src/hotspot/cpu/aarch64/macroAssembler_aarch64.cpp line 765:
>>
>>> 763: #ifdef ASSERT
>>> 764: // TraceBytecodes does not use r12 but saves it over the call, so
>>> don't verify
>>> 765: // if (!TraceBytecodes) verify_heapbase("call_VM_base: heap base
>>> corrupted?");
>>
>> Only `if()` was commented before. Now you put `verify_heapbase()` under
>> comment too.
>> Even so `verify_heapbase()` is empty on this platform we should not change
>> this.
>
> Hi, @vnkozlov, I am not sure I understand. The full line - including
> `verify_heapbase` - was commented out previously and remains commented out.
> Is GitHub's diff viewer acting up again?
>
> All I did was to correct the statement in case someone removes the comment to
> trace things: since UseCompressedClassPointers is now always true,
> `((UseCompressedOops || UseCompressedClassPointers) && !TraceBytecodes)`
> becomes `(!TraceByteCodes)`
Sorry about that. GitHub split line into 2 and I thought they are separate
lines.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/28366#discussion_r2978387904
