On Thu, 2015-01-08 at 14:49 -0800, Nicholas A. Bellinger wrote: > On Thu, 2015-01-08 at 09:37 -0500, Martin K. Petersen wrote: > > >>>>> "nab" == Nicholas A Bellinger <nab@xxxxxxxxxxxxxxx> writes: > > > > nab> IIRC, most modern hardware is reporting a large enough value for > > nab> queue_max_hw_sectors() to support 8 MB I/Os, but I'm thinking that > > nab> this could end up being problematic for older hardware that is > > nab> reporting much smaller values. > > > > Reporting queue_max_hw_sectors sounds sane to me. > > > > What's your concern wrt. older hardware? > > > > The target is still enforcing it's own hw_max_sectors in sbc_parse_cdb() > based upon what queue_max_hw_sectors() reports for IBLOCK, and will > throw an exception for I/Os who's sector count exceeds this maximum. > > The concern is when older hardware drivers are reporting say > queue_max_hw_sectors=128 with initiators are not actively honoring block > limits EVPD MAXIMUM TRANSFER LENGTH, that would result in I/Os over 64K > generating exception status. > > So the question is what is a sane minimum for IBLOCK's hw_max_sectors so > that large I/Os (say up to 8 MB) aren't rejected by sbc_parse_sbc(), and > don't trigger the subsequent checks in generic_make_request() -> > generic_make_request_checks(). > Christoph, any input on this..? --nab -- 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
