>>>>> "Christoph" == Christoph Hellwig <hch@xxxxxx> writes: Christoph> Currently blkdev_issue_zeroout cascades down from discards Christoph> (if the driver gurantees that discards zero data), to WRITE guarantees Christoph> SAME and then to a loop writing zeroes. Unfortunately we Christoph> ignore tun-time EOPNOTSUPP errors in the block layer run-time Christoph> blkdev_issue_discard helper to work around DM volumes that Christoph> may have mixed discard support underneath. Christoph> This path intoroduces a new BLKDEV_DISCARD_ZERO flag to patch introduces Christoph> blkdev_issue_discard that indicates we are called for zeroing Christoph> operation. This allows both to ignore the EOPNOTSUPP hack Christoph> and actually consolidating the discard_zeroes_data check into Christoph> the function. I like this better than the I/O error tracking approach. The blkdev_issue_write_same() -EOPNOTSUPP regression is still present, though. -- Martin K. Petersen Oracle Linux Engineering -- To unsubscribe from this list: send the line "unsubscribe linux-block" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
