On Wed, 16 Sep 2026 01:29:53 GMT, Vladimir Kozlov <[email protected]> wrote:

>> src/hotspot/share/code/aotCodeCache.hpp line 222:
>> 
>>> 220: 
>>> 221:   address* _extrs_addr;
>>> 222:   address* _stubs_addr;
>> 
>> Speaking of UB, `init_extrs` allocates `_stubs_addr` but does not initialize 
>> it.  It’s probably a wonderful fact that nobody will ever touch the toxic 
>> bits of that array, unless they really know somebody else already planted a 
>> defined address in it.  Still, it feels like jazz dancing at the top of a 
>> cliff over a snakepit.  Can we please zero out `_stubs_addr` on allocation?
>
> Done. I also cleared _extrs_addr array on allocation. Not all slots are 
> filled.

Thinking more about a future failure mode – there may be future edits to stubs 
that don’t synch with config tests.  If a zero ever comes from a query to the 
stubs table, will an assertion catch it before it is used in code?

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

PR Review Comment: https://git.openjdk.org/jdk/pull/30778#discussion_r4028488438

Reply via email to