Vitaly, > With this patch applied over 6.6.16 problem still persists. I figured. Couldn't understand why any of these changes would lead to the symptoms you are experiencing. I think my original patch just exposed a timing issue. Since your device has a SATL, we should never be sending a WRITE SAME command in the first place. I have another patch in the pipeline which, in combination with Christoph's atomic queue limit update series, should close this gap. I'll reach out when it's ready for testing. Thanks for your help! -- Martin K. Petersen Oracle Linux Engineering
