How high are he memory costs of polymorphic inline caches?

[Raffaello Giulietti]

On 2014-08-20 11:08, MacGregor, Duncan (GE Energy Management) wrote:
> The memory costs are highly dependent on what combinators are being used
> and and how many entries are in the cache, and whether you¹re caching and
> reusing any adaption needed to the base method handles. It¹s also going to
> change quite radically for the better in 8y40 when various patches have
> landed.
>
> A naïve implementation with some combinators wrapping each target method
> will normally result in about 100 bytes of retained heap per PIC entry,
> but there may be more attached to anonymous class constant pools and so
> forth. This memory usage can be reduced through careful caching and reuse,
> and as I said there are patches incoming for 8u40 which make a great
> improvement.
>
> Regards, Duncan.
>

Hi Duncan, sounds encouraging!

Thanks for the good news
Raffaello


-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3764 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://mail.openjdk.java.net/pipermail/mlvm-dev/attachments/20140820/9e657519/smime.p7s>