[LSF/MM/BPF TOPIC] File system checksum offload

hch at infradead.org hch at infradead.org
Wed Mar 19 22:48:11 PDT 2025


On Wed, Mar 19, 2025 at 11:36:28PM +0530, Kanchan Joshi wrote:
> I tried to describe that in the cover letter of the PoC:
> https://lore.kernel.org/linux-btrfs/20250129140207.22718-1-joshi.k@samsung.com/

I don't think it does as people have pointed that out multiple times.
Once you do checksums on the device they are not end to end, either in
the T10 DIF definition or that of file system checksums.  You leave a
huge glaring gap in the protection envelope.  So calling this "offload"
is extremely dishonest and misleading.




More information about the Linux-nvme mailing list