Skip to content

Conversation

merlimat
Copy link

@merlimat merlimat commented Oct 4, 2018

In same case it might beneficial to be able to compress/decompress from direct memory buffer that don't necessarily belong to a ByteBuffer instance.

In that case it's possible, in unsafe Java world, to have a long that represent a pointer to a memory location. Passing that through JNI would be the most efficient way to compress that data.

@merlimat
Copy link
Author

@odaira Can you take a look at this? It's on the same line as what described in #106

@merlimat
Copy link
Author

@odaira Any chance on this one? :)

@odaira
Copy link
Member

odaira commented Apr 15, 2019

@merlimat Sorry for the slow process. I'll include this to 1.6.0 once the review is done.

@merlimat
Copy link
Author

👍 Thanks!

@odaira
Copy link
Member

odaira commented May 15, 2019

Sorry, I must release 1.6.0 as soon as possible, so let me push this off until 1.7.0.

@emkornfield
Copy link

Any plans on merging this?

@p4654545
Copy link

p4654545 commented Apr 6, 2021

+1

Our case is to compress/decompress blocks of memory where both source and destination are allocated with Unsafe.allocateMemory().

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants