Am 03.11.2014 um 09:18 schrieb Christoph Hellwig: > On Sun, Nov 02, 2014 at 02:00:55PM +0100, Richard Weinberger wrote: >> +#define UBIBLOCK_SG_COUNT 64 > > > Can you document why you choose this number? The default nr_request > for the old code would be 128.
Is 64 a problem? Beside of the fact that I forgot to set blk_queue_max_segments(). I used this number because 128 seemed a bit high and my goal was to keep the memory footprint small. This is also why I've set tag_set.queue_depth to 64. > Also I'll send a patch ASAP that allows drivers to ensure their > ->queue_rq is always called from a driver-specific workqueue, which > should simplify the bui block driver a lot. Sounds great! Thanks, //richard -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
