>>>>> "Doug" == Douglas Gilbert <dgilbert@xxxxxxxxxxxx> writes: Doug> No PI, why not? This was tested on lk 3.9.4 with your patches Doug> applied. I've been communicating with Akinobu offline about this. The reason is that I really only implemented DIX support in scsi_debug. The dif module parameter is only there to select the reported protection type. The main problem with scsi_debug is that it's both initiator and target in one. And there isn't any notion of a data transfer between the initiator and the target. We just fill out the scatterlists with stuff from the in-memory buffer. To support DIX/DIF correctly we'd have to have the notion of a handoff between the front and back of scsi_debug. And we'd have to support all combinations of DIX and DIF with the relevant slicing and dicing of data and PI buffers. I have some patches pending as part of my next DIF/DIX update that makes some of these things more palatable at the block/SCSI level. Akinobu voiced interest in finishing the scsi_debug work on top of my code. -- Martin K. Petersen Oracle Linux Engineering -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
