Linux 6.2 Likely To Enable Btrfs Async Discard By Default

Btrfs' async discard functionality will likely be turned on by default with the upcoming Linux 6.2 kernel cycle when running on an SSD.

Two years ago Btrfs introduced async discard support for moving the SSD TRIM/discard to the transaction commit. With the work done asynchronously, Facebook/Meta engineers found this to be more efficient and help reduce read latencies.

This feature can be turned on via the "discard=async" mount option but now that it's matured and helping in the real-world, Btrfs developers are planning to enable it by default. Async discard functionality would be automatically turned on whenever being mounted with a capable solid-state drive.

This patch by SUSE's David Sterba has been picked up in his Btrfs for-next branch ahead of the Linux 6.2 merge window in December.

Async discard will be turned on automatically for SSDs unless using the new "async=nodiscard" mount option for forcing off the support.

The default change ultimately stems from a two year old Fedora request for enabling async discard by default. That proposal also notes how Facebook/Meta has been using this option by default across their consumer-level hardware for some time now to great success.