On Mon, Feb 03, 2025 at 06:46:49PM +1030, Qu Wenruo wrote: > The root cause is the content change during direct IO, and XFS/EXT4 doesn't > wait for folio writeback before dirtying the folio (if no AS_STABLE_WRITES > set). > That's a valid optimization, but that will cause contents change. > > (I know there is the AS_STABLE_WRITES, but I'm not sure if qemu will pass > that flag to virtio block devices inside the VM) It doesn't, and even if it did you can force guests to add it. But it would be an interesting experiment to support passing it through at least for virtio-blk. > And with btrfs' checksum calculation happening before submitting the real > bio, it means if the contents changed after the csum calculation and before > bio finished, we will got csum mismatch. btrfs checksums before submitting the bio. But that doesn't change the thing as you still have the same problem.
