From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Martin K. Petersen" Subject: Re: An oops will occur while SCSI core is being used in 3.4-rc1 Date: Wed, 11 Apr 2012 20:48:59 -0400 Message-ID: References: <4F83EC71.90904@acm.org> <4F8461E3.3050808@cs.wisc.edu> <4F846398.5030609@cs.wisc.edu> <4F846506.4060801@cs.wisc.edu> <4F85B3E8.7040704@acm.org> <4F85E312.6070205@acm.org> Mime-Version: 1.0 Content-Type: text/plain Return-path: In-Reply-To: <4F85E312.6070205-HInyCGIudOg@public.gmane.org> (Bart Van Assche's message of "Wed, 11 Apr 2012 20:01:22 +0000") Sender: linux-usb-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Bart Van Assche Cc: "Martin K. Petersen" , Mike Christie , Elric Fu , James Bottomley , linux-scsi-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Sarah Sharp , Felipe Balbi , Alex He , Andiry Xu , Greg KH , Linux USB Mailing List , Alan Stern List-Id: linux-scsi@vger.kernel.org >>>>> "Bart" == Bart Van Assche writes: Bart> That comment didn't make sense - request.rq_disk is not modified Bart> after it has been initialized. Correct. Bart> Yet it would make me more happy if scsi_send_eh_cmnd() could be Bart> modified such that it doesn't have to access struct scsi_driver. Well, the whole point of the original patch was that we wanted to enable upper level drivers to interact with the current (SPC-centric) error handling. I understand your concern given the recent flurry of device lifetime issues. But I think the issues my patch tries to tackle are important enough to warrant the change. And we definitely don't want to start mucking with block device-specific commands in the generic error handling path. -- Martin K. Petersen Oracle Linux Engineering -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html