Re:Java's new AOT cache feature is in a rough state
by jayd16 on Thursday October 29, 2026 @11:59AM
(#50291163)
Should probably just kick in the JIT at that point, no?
Re:Java's new AOT cache feature is in a rough state
by odo1242 on Thursday October 29, 2026 @01:03PM
(#50292355)
Yea, it should probably warn and fallback to normal JIT
Re:Java's new AOT cache feature is in a rough state
by ahartmetz on Friday October 30, 2026 @01:50AM
(#50300295)
Standard operating procedure for caches: The cache key must contain all relevant criteria to decide whether the entry contains what you want. Code for a different architecture isn't it. In a VM that might get moved around, an extended cache key would solve a lot of problems, even preserving entries for other CPUs as long as there is space.
IMHO, that would be much better than running into crashes and having to figure out the cause yourself.