Christoph Hellwig wrote:
> The ->rw_page method is a special purpose bypass of the usual bio
> handling path that is limited to single-page reads and writes and
> synchronous which causes a lot of extra code in the drivers, callers
> and the block layer.
>
> The only remaining user is the MM swap code. Switch that swap code to
> simply submit a single-vec on-stack bio an synchronously wait on it
> based on a newly added QUEUE_FLAG_SYNCHRONOUS flag set by the drivers
> that currently implement ->rw_page instead. While this touches one
> extra cache line and executes extra code, it simplifies the block
> layer and drivers and ensures that all feastures are properly supported
> by all drivers, e.g. right now ->rw_page bypassed cgroup writeback
> entirely.
>
> Signed-off-by: Christoph Hellwig <hch@xxxxxx>
> ---
> block/bdev.c | 78 -----------------------------------
> drivers/block/brd.c | 15 +------
> drivers/block/zram/zram_drv.c | 61 +--------------------------
> drivers/nvdimm/btt.c | 16 +------
> drivers/nvdimm/pmem.c | 24 +----------
> include/linux/blkdev.h | 12 +++---
> mm/page_io.c | 53 ++++++++++++++----------
> mm/swapfile.c | 2 +-
> 8 files changed, 44 insertions(+), 217 deletions(-)
>
[..]
> diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
> index 89f51d68c68ad6..1bffe8f44ae9a8 100644
> --- a/include/linux/blkdev.h
> +++ b/include/linux/blkdev.h
> @@ -555,6 +555,7 @@ struct request_queue {
> #define QUEUE_FLAG_IO_STAT 7 /* do disk/partitions IO accounting */
> #define QUEUE_FLAG_NOXMERGES 9 /* No extended merges */
> #define QUEUE_FLAG_ADD_RANDOM 10 /* Contributes to random pool */
> +#define QUEUE_FLAG_SYNCHRONOUS 11 /* always complets in submit context */
s/complets/completes/
> #define QUEUE_FLAG_SAME_FORCE 12 /* force complete on same CPU */
> #define QUEUE_FLAG_INIT_DONE 14 /* queue is initialized */
> #define QUEUE_FLAG_STABLE_WRITES 15 /* don't modify blks until WB is done */
> @@ -1252,6 +1253,12 @@ static inline bool bdev_nonrot(struct block_device *bdev)
> return blk_queue_nonrot(bdev_get_queue(bdev));
> }
>
Other than that, this looks good and passes regression:
Reviewed-by: Dan Williams <dan.j.williams@xxxxxxxxx>
