Igor Chorążewicz

@haowu14 the idea is that each FileSegment can represent a different memory type, and so each will have a different base path. For example, our most common use case is...

The implementation in first commit suffers from a few problems: - the item is added back to the MMContainer if eviction fails (becomes "hot") - the item can potentially be...

@therealgymmy So, the only needed fix for correctness is to change the memory order on reads (at least for non-chained items)? Why the relaxed memory order is enough right now?...

I've responded to comments, run the tests you suggested and reworked the code. I realized there was one other problem: incrementing ref count under the mmContainer lock was causing races...

@therealgymmy Thank you for the feedback! I managed to reproduce the issue locally and will work on fixing it. > Did you mean under the hash-table lock? Seems like the...

@therealgymmy I found out what the issue was: I was passing `candidate` to releaseBackToAllocator which after my changes represented the parent Item. We should pass pointer to the child we...

As for ecryptfs this can be a bug in the filesystem, because although msync returns EINTR there is no signal showed under strace. As for the manpage, I found following...

I don't think it;s fixed - we only have a workaround on pmemkv side.

This could be useful for batch inserts in libpmemobj-cpp containers. We could perform insert in 2 phases - in first phase we just iterate over the structure and snapshot the...

This is quite similar to approaches for checking data consistency in concurrent environments, for example: https://github.com/facebook/CacheLib/blob/v2022.04.11.00/cachelib/cachebench/consistency/ValueHistory.h