Hello, On Tue, Aug 18, 2020 at 09:12:13PM -0700, Akshat Jain wrote: > To answer your question: > 1. Yes we will start working on a patch and send it for review. > 2. We found this issue during our code review. > > Another information I need to bring to your attention is that many > user libraries today decode the fixed format sense block based on the > format today's kernel (ata_gen_passthru_sense) provides. Rather than > the format specified in the SCSI Primary commands - 4 specification. > If we were to correct the field offsets for the fixed format sense block. > It may break such libraries. How do you assess the impact > of such a change? So, I don't know. Given that nobody tripped over it till now, I doubt fixing it would cause a lot of trouble but at the same time it isn't a real problem, again, given that nobody has tripped over it yet. Maybe just fix it and see how it goes? Thanks. -- tejun
