crazy idea: what if swap, in particular zram (RAM compression), was enhanced with eBPF bytecode invoked during page reclaim? for pages with a single (RW) mapping, you could e.g. ask userspace which subsets of the page must be preserved (if the heap metadata isn't swapped out)
-
-
heck, if you had a way to redirect attempts to swap heap metadata to non-metadata pages, you could ensure that the metadata almost always stays swapped in, and then you could store page swapping state in the userspace heap metadata from eBPF
Show this thread -
and then you could have a sorted linked list of preferred victim pages in userspace, which always only contains pages that are actually dirty+present and are therefore valid candidates for reclaim
Show this thread
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.