Software and hardware obsolescence in the kernel

Posted Aug 30, 2020 14:58 UTC (Sun) by willy (subscriber, #9762)
In reply to: Software and hardware obsolescence in the kernel by epa
Parent article: Software and hardware obsolescence in the kernel

Imagine you have an array of pointers.

struct page *pages[15];

Do you really want to have each of these pointers be 10 bytes long? It doesn't even make sense for the pointers to be 12 bytes long. 16 bytes does make sense. And then you can use the top few bits for exciting things like ARM's pointer tagging.

Software and hardware obsolescence in the kernel

Posted Aug 30, 2020 18:10 UTC (Sun) by epa (subscriber, #39769) [Link] (2 responses)

Yes that’s exactly what I am saying, make the pointers 10 bytes long. Using 16 bytes would increase your hardware and electricity costs by 60% for little benefit. (Assuming a large system with lots of RAM)

Software and hardware obsolescence in the kernel

Posted Aug 30, 2020 18:12 UTC (Sun) by willy (subscriber, #9762) [Link] (1 responses)

There's a reason we didn't go to 48 bit pointers from 32 bit. I don't think you've thought through the consequences of 10 byte pointers.

Software and hardware obsolescence in the kernel

Posted Aug 30, 2020 23:26 UTC (Sun) by marcH (subscriber, #57642) [Link]

> I don't think you've thought through the consequences of 10 byte pointers.

Please share your thoughts (or references) if you have.

(as already mentioned by jem above:
https://stackoverflow.com/questions/6716946/why-do-x86-64... )