Hi Mike, On 10/29/2015 11:18 AM, Mike Snitzer wrote:
Sorry but I fail to see why your request to revert is viable.
No problem. Thanks for moving this for a discussion on a proper fix. I'm somewhat new to kernel and SCSI workings and its community process, so that's certainly appreciated.
Wouldn't the correct fix be to train scsi_verify_blk_ioctl() to work even without SYS_CAP_RAWIO?
I see it would fix the problem as well, but I don't happen to know all of its usages, so I'd have to defer to question to someone who knows more of it.
I'm not doubting that the commit caused problems for the case you care about (unprivledged users issueing ioctls) but that is an entirely different issue that needs to be discussed more directly (with the broader linux-scsi community, now cc'd).
Sure. Thanks for opening the discussion.
p.s. leaving full context for linux-scsi so they don't need to track down other mails (btw, thanks for the detailed patch header but it enabled me to be skeptical of your request to revert):
You're welcome. If it's been useful for rejecting this patch and getting a better one later, it's worth it. :) Kind regards, -- Mauricio Faria de Oliveira IBM Linux Technology Center -- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
