>>>>> "Christoph" == Christoph Hellwig <hch@xxxxxx> writes: Christoph, Christoph> Unlike discard outside the zeroout path, write same is a data Christoph> integrity operation. Just like in the zero out case turning Christoph> an EOPNOTSUPP into 0 will get you data corruption, as the Christoph> caller will see a successful return for an operation that did Christoph> not actually write data to disk. Exactly. So why add the dreaded -EOPNOTSUPP special casing to blkdev_issue_write_same()? Christoph> Re stacking drivers and discard / write same: why does Christoph> blk_set_stacking_limits set discard_zeroes_data to 1 and Christoph> max_write_same_sectors to UINT_MAX? These seem like Christoph> inherently dangerous defaults. It's just shorthand for "stacking driver does not impose any limits, use whatever the low-level device sets". If the stacking driver has an actual constraint it is free to set a different limit prior to calling the stacking function. -- 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
