XFS and ext4 filesystems have received a bit of good news recently, by way of a 5% performance increase. This change was made by ByteDance engineer Fengnan Chang and was committed by Christian Brauner.

"Skip the memset of the iomap in iomap_iter() once the iteration is done," said Brauner in a recent git pull. Brauner continued, "In high-IOPS scenarios (4k randread NVMe polling via io_uring), the pointless memset wasted memory write bandwidth; this improves IOPS by about 5% on ext4 and XFS."

IOPS stands for Input/Output Operations Per Second and is the speed at which a drive is able to send and receive data.

You might think that Chang optimized those three lines of code to improve IOPS, but that is not the case. All Chang did was move those lines so that they ran after the process finished, instead of before. By moving those three lines, IOPS for ext4 and XFS filesystems received a 5% performance increase.

Although this might seem like a simple change, it would take considerable skill to spot the issue and how it would improve the performance of the read and write speeds of a drive.

According to Phoronix, "that same pull also adds the VFS infrastructure needed for implementing FS-VERITY support for XFS with a post-EOF Merkle tree."

Bravo to Chang for making this tiny code change that will benefit Linux systems worldwide.