On 06/03/2014 02:09 PM, yonatan cale wrote: > Ok, let me just make sure I'm with you on the same page, > We want to check where we are getting stuck in "scsi_scan_target" (in > scsi_scan.c) and in the functions that it calls (and they call). Right?
Yes. > (we are not debugging any other functions right now) > If you want to make it a little more complicated we could also try to figure out why when the initiator fails a inqiury with DID_TRANSPORT_FAILED the scsi layer does not seem to fail the command. This was part of the mystery where I thought it might not be a retry and was really the sequential scan code going crazy, but you ruled that out. I think the fact we see inquirys over and over in the logs is still related to the problem we are debugging. For this you would need to add debugging to the queueing and completion code which would be more complex, so I am not sure if you want to just start with the scanning code. > > On Tue, Jun 3, 2014 at 5:46 PM, Mike Christie <[email protected] > <mailto:[email protected]>> wrote: > > On 6/3/14, 9:29 AM, Mike Christie wrote: > > On 6/3/14, 6:46 AM, Cale, Yonatan wrote: > > Hi, > > Regarding debug prints in "__scsi_scan_target", here is what > I added: > > > I meant add more. We are trying to debug why the scan does not > complete. > For the printks you added below it does not tell us where we are > stuck > in the scan process. It could have got stuck after the printks > you added > there. > > Also, did you even hang in that run you provided the log output for > below? It looks like the probe and add for LUN0 returned > SCSI_SCAN_NO_RESPONSE, so it looks like in that run you did not drop > down to the report luns and sequential scan code like you did in > some > other runs. > > You need to just add some printks or do a kprobe module or > figure out > how to turn on scsi scan logging or a combo of both. > > > Ignore the above. I thought you were telling me the printks you > added in response to the mail I said to add more printks. As I > reread the mail, it sounds like you are just trying to tell me why > you think scsi_probe_and_add_lun always completes. > > Going forward you just need to add more printks or enable more > logging or do a kprobe module or some combo of both, so you can > narrow down where in the scan process we are getting stuck. > > > -- > You received this message because you are subscribed to the Google > Groups "open-iscsi" group. > To unsubscribe from this group and stop receiving emails from it, send > an email to [email protected] > <mailto:[email protected]>. > To post to this group, send email to [email protected] > <mailto:[email protected]>. > Visit this group at http://groups.google.com/group/open-iscsi. > For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "open-iscsi" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/open-iscsi. For more options, visit https://groups.google.com/d/optout.
