[External] : Re: AOT code usage restrictions

[Vladimir Kozlov]

Thank you, Charles

Our thinking is to generate "generic" AOT code, when asked for it, 
regardless on what machine your are training/creating AOT cache.
VM itself will set corresponding generic CPU features for that, no need 
to run on specific hardware. But you do need to run on machine which 
have these "generic" features.

Thanks,
Vladimir

On 1/22/26 7:13 PM, Charles Oliver Nutter wrote:
> Commenting from a user perspective, since we very much want to ship AOT 
> caches for JRuby users...
> 
> The restriction seems fine largely because I don't know how else you'd 
> work around it. But I think with this restriction there needs to be a 
> clear way to AOT for a lowest common denominator without having that 
> exact hardware on hand. I don't want to have to find some janky old x86 
> machine to train a portable cache on.
> 
> On Thu, Jan 22, 2026, 20:05 Vladimir Kozlov <vladimir.kozlov at oracle.com 
> <mailto:vladimir.kozlov at oracle.com>> wrote:
> 
>     Hi,
> 
>     I would like to get opinion from community regarding AOT code usage
>     restrictions for JEP https://openjdk.org/jeps/8335368 <https://
>     openjdk.org/jeps/8335368>.
> 
>     For current AOT features in JDK 25 and 26 it is not the issue - CPU
>     features do not affect them.
> 
>     CPU features affect generated AOT code - the instructions set is
>     defined
>     by machine on which AOT training is done. But CPU instructions on
>     machine where AOT cache is used could be different.
> 
>     Our current proposal is next:
> 
>     1. The best startup and peak performance with AOT code could be
>     achieved
>     when training and production runs use the same type of machines (same
>     CPU and OS).
> 
>     2. AOT code will not be used if CPU features on machine for production
>     run does not match one used for training. Other AOT features (cached
>     classes and profiling info) will still be used.
> 
>     3. In a future, when user requested during training run (with VM's
>     flag[s]), we will allow to use AOT code generated for old CPU
>     instructions (for example, AVX2) on machines with new instructions (for
>     example, AVX512) by restricting online (normal JIT) code generation to
>     old CPU instructions. This will allow to use the same cached AOT
>     code on
>     variety of machines. But you may not get the best peak performance on
>     newer machines.
> 
>     What do you think about this proposal?
> 
>     Thanks,
>     Vladimir K
>