block: improve handling of the magic discard payload

Instead of allocating a single unused biovec for discard requests, send
them down without any payload.  Instead we allow the driver to add a
"special" payload using a biovec embedded into struct request (unioned
over other fields never used while in the driver), and overloading
the number of segments for this case.

This has a couple of advantages:

 - we don't have to allocate the bio_vec
 - the amount of special casing for discard requests in the block
   layer is significantly reduced
 - using this same scheme for other request types is trivial,
   which will be important for implementing the new WRITE_ZEROES
   op on devices where it actually requires a payload (e.g. SCSI)
 - we can get rid of playing games with the request length, as
   we'll never touch it and completions will work just fine
 - it will allow us to support ranged discard operations in the
   future by merging non-contiguous discard bios into a single
   request
 - last but not least it removes a lot of code

This patch is the common base for my WIP series for ranges discards and to
remove discard_zeroes_data in favor of always using REQ_OP_WRITE_ZEROES,
so it would be good to get it in quickly.

Signed-off-by: Christoph Hellwig <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>

block/bio.c

@@ -1840,15 +1840,7 @@ struct bio *bio_split(struct bio *bio, int sectors,
 	BUG_ON(sectors <= 0);
 	BUG_ON(sectors >= bio_sectors(bio));
 
-	/*
-	 * Discards need a mutable bio_vec to accommodate the payload
-	 * required by the DSM TRIM and UNMAP commands.
-	 */
-	if (bio_op(bio) == REQ_OP_DISCARD || bio_op(bio) == REQ_OP_SECURE_ERASE)
-		split = bio_clone_bioset(bio, gfp, bs);
-	else
-		split = bio_clone_fast(bio, gfp, bs);
-
+	split = bio_clone_fast(bio, gfp, bs);
 	if (!split)
 		return NULL;
 

block/blk-core.c

@@ -1475,38 +1475,6 @@ void blk_put_request(struct request *req)
 }
 EXPORT_SYMBOL(blk_put_request);
 
-/**
- * blk_add_request_payload - add a payload to a request
- * @rq: request to update
- * @page: page backing the payload
- * @offset: offset in page
- * @len: length of the payload.
- *
- * This allows to later add a payload to an already submitted request by
- * a block driver.  The driver needs to take care of freeing the payload
- * itself.
- *
- * Note that this is a quite horrible hack and nothing but handling of
- * discard requests should ever use it.
- */
-void blk_add_request_payload(struct request *rq, struct page *page,
-		int offset, unsigned int len)
-{
-	struct bio *bio = rq->bio;
-
-	bio->bi_io_vec->bv_page = page;
-	bio->bi_io_vec->bv_offset = offset;
-	bio->bi_io_vec->bv_len = len;
-
-	bio->bi_iter.bi_size = len;
-	bio->bi_vcnt = 1;
-	bio->bi_phys_segments = 1;
-
-	rq->__data_len = rq->resid_len = len;
-	rq->nr_phys_segments = 1;
-}
-EXPORT_SYMBOL_GPL(blk_add_request_payload);
-
 bool bio_attempt_back_merge(struct request_queue *q, struct request *req,
 			    struct bio *bio)
 {
@@ -2642,6 +2610,8 @@ bool blk_update_request(struct request *req, int error, unsigned int nr_bytes)
 		return false;
 	}
 
+	WARN_ON_ONCE(req->rq_flags & RQF_SPECIAL_PAYLOAD);
+
 	req->__data_len -= total_bytes;
 
 	/* update sector only for requests with clear definition of sector */

block/blk-lib.c

@@ -80,7 +80,7 @@ int __blkdev_issue_discard(struct block_device *bdev, sector_t sector,
 			req_sects = end_sect - sector;
 		}
 
-		bio = next_bio(bio, 1, gfp_mask);
+		bio = next_bio(bio, 0, gfp_mask);
 		bio->bi_iter.bi_sector = sector;
 		bio->bi_bdev = bdev;
 		bio_set_op_attrs(bio, op, 0);

block/blk-merge.c

@@ -241,18 +241,13 @@ static unsigned int __blk_recalc_rq_segments(struct request_queue *q,
 	if (!bio)
 		return 0;
 
-	/*
-	 * This should probably be returning 0, but blk_add_request_payload()
-	 * (Christoph!!!!)
-	 */
 	switch (bio_op(bio)) {
 	case REQ_OP_DISCARD:
 	case REQ_OP_SECURE_ERASE:
-	case REQ_OP_WRITE_SAME:
 	case REQ_OP_WRITE_ZEROES:
+		return 0;
+	case REQ_OP_WRITE_SAME:
 		return 1;
-	default:
-		break;
 	}
 
 	fbio = bio;
@@ -410,39 +405,21 @@ new_segment:
 	*bvprv = *bvec;
 }
 
+static inline int __blk_bvec_map_sg(struct request_queue *q, struct bio_vec bv,
+		struct scatterlist *sglist, struct scatterlist **sg)
+{
+	*sg = sglist;
+	sg_set_page(*sg, bv.bv_page, bv.bv_len, bv.bv_offset);
+	return 1;
+}
+
 static int __blk_bios_map_sg(struct request_queue *q, struct bio *bio,
 			     struct scatterlist *sglist,
 			     struct scatterlist **sg)
 {
 	struct bio_vec bvec, bvprv = { NULL };
 	struct bvec_iter iter;
-	int nsegs, cluster;
-
-	nsegs = 0;
-	cluster = blk_queue_cluster(q);
-
-	switch (bio_op(bio)) {
-	case REQ_OP_DISCARD:
-	case REQ_OP_SECURE_ERASE:
-	case REQ_OP_WRITE_ZEROES:
-		/*
-		 * This is a hack - drivers should be neither modifying the
-		 * biovec, nor relying on bi_vcnt - but because of
-		 * blk_add_request_payload(), a discard bio may or may not have
-		 * a payload we need to set up here (thank you Christoph) and
-		 * bi_vcnt is really the only way of telling if we need to.
-		 */
-		if (!bio->bi_vcnt)
-			return 0;
-		/* Fall through */
-	case REQ_OP_WRITE_SAME:
-		*sg = sglist;
-		bvec = bio_iovec(bio);
-		sg_set_page(*sg, bvec.bv_page, bvec.bv_len, bvec.bv_offset);
-		return 1;
-	default:
-		break;
-	}
+	int cluster = blk_queue_cluster(q), nsegs = 0;
 
 	for_each_bio(bio)
 		bio_for_each_segment(bvec, bio, iter)
@@ -462,7 +439,11 @@ int blk_rq_map_sg(struct request_queue *q, struct request *rq,
 	struct scatterlist *sg = NULL;
 	int nsegs = 0;
 
-	if (rq->bio)
+	if (rq->rq_flags & RQF_SPECIAL_PAYLOAD)
+		nsegs = __blk_bvec_map_sg(q, rq->special_vec, sglist, &sg);
+	else if (rq->bio && bio_op(rq->bio) == REQ_OP_WRITE_SAME)
+		nsegs = __blk_bvec_map_sg(q, bio_iovec(rq->bio), sglist, &sg);
+	else if (rq->bio)
 		nsegs = __blk_bios_map_sg(q, rq->bio, sglist, &sg);
 
 	if (unlikely(rq->rq_flags & RQF_COPY_USER) &&
@@ -495,7 +476,7 @@ int blk_rq_map_sg(struct request_queue *q, struct request *rq,
 	 * Something must have been wrong if the figured number of
 	 * segment is bigger than number of req's physical segments
 	 */
-	WARN_ON(nsegs > rq->nr_phys_segments);
+	WARN_ON(nsegs > blk_rq_nr_phys_segments(rq));
 
 	return nsegs;
 }