Hello, sorry for a rude top-post and please CC me in a main discussion afterward. Since yesterday since I found the same issue with C602 and S3710 in mine production and stood in horror for a couple of moments, I made a small amount of runs with fio to check if the issue is here or not for certain configuration, here are preliminary results: - isci in all LTS are affected, if I could use this wording in appearance of an obviously broken hardware counterparts, - setting device`s qd=1 is *not* fixing the issue for me, - setting noncq for a specific link closes the deal. Therefore, I`d like to propose to either add hook from scsi code to set horkage_on on setting queue_depth to 1 and removing it back at runtime, or to reword libata faq [1], as for me there is a clear difference between actual NCQ disablement for a link/controller and setting queue depth to 1, the bus resets are still appearing in a second case. 1. https://ata.wiki.kernel.org/index.php/Libata_FAQ -- To unsubscribe from this list: send the line "unsubscribe linux-ide" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
