For those making increased use of Linux's Zswap as a compressed RAM cache for swap pages amid these times of sky high RAM prices, a new patch series was posted providing some incremental performance gains for Zswap by batching the writeback I/O.

Separate from the Zswap patches posted earlier this week providing [1]rather significant performance improvements , a separate patch series for Zswap posted to the Linux kernel mailing list are providing more incremental gains. In the presence of Zswap, the patches provided 3~4% time-savings when compiling the Linux kernel as a sample workload.

[2]

Alexandre Ghiti posted the patches to batch the writeback I/O for Zswap, explaining:

"Since commit 8f29aa226f82 ("mm/swap: introduce struct swap_io_ctx"), reclaim merges the swap writes of folios with consecutive slots into a single bio, but zswap writeback still issues one IO per entry. Patch 1 makes each zswap LRU walk batch its writeback IO. Patch 2 plugs the walk, so that the block layer can merge what the batch cannot."

[3]The patches are under review and so far the review comments have been favorable. With 3~4% build time savings for the Linux kernel in testing. every bit of improvement helps amid the increased interest around ZRAM and Zswap in today's environment.



[1] https://www.phoronix.com/news/Linux-Better-Zswap-Performance

[2] https://www.phoronix.com/image-viewer.php?id=2026&image=more_zswwap_lrg

[3] https://lore.kernel.org/all/20261007153507.50715-1-alex@ghiti.fr/