On Fri, 17 Apr 2026 09:41:34 GMT, Stefan Karlsson <[email protected]> wrote:
>> src/hotspot/share/interpreter/oopMapCache.cpp line 157:
>>
>>> 155: if (has_valid_mask() && mask_size() > small_mask_limit) {
>>> 156: assert(_bit_mask[0] != 0, "should have pointer to C heap");
>>> 157: FREE_C_HEAP_ARRAY((uintptr_t*)_bit_mask[0]);
>>
>> ?
>
> I should have mentioned what this is. This part of the code puts a
> (uintptr_t*) in the first element of _bit_mask if the bit mask is too large
> to fit the static _bit_mask array. It is messing with the type system and we
> should probably rewrite this to use a union instead.
Sorry for not being clear about this, and I should have removed this comment. I
did look at this after writing the comment and before submitting my review, and
have the same impression that this should be a union or something. This is
really ugly.
Extra CR is totally fine.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/30786#discussion_r3100349161
