@@ -187,22 +187,38 @@ static void bounce_end_io_read_isa(struct bio *bio)
__bounce_end_io_read(bio, isa_page_pool);
}
+static inline bool need_split(struct request_queue *q, struct bio *bio)
+{
+ return bio_sectors(bio) > BIO_SP_MAX_SECTORS;
+}
+
+static inline bool need_bounce(struct request_queue *q, struct bio *bio)
+{
+ struct bvec_iter iter;
+ struct bio_vec bv;
+
+ bio_for_each_segment_mp(bv, bio, iter) {
+ unsigned nr = (bv.bv_offset + bv.bv_len - 1) >>
+ PAGE_SHIFT;
+
+ if (page_to_pfn(bv.bv_page) + nr > queue_bounce_pfn(q))
+ return true;
+ }
+ return false;
+}
+
static void __blk_queue_bounce(struct request_queue *q, struct bio **bio_orig,
mempool_t *pool)
{
struct bio *bio;
int rw = bio_data_dir(*bio_orig);
- struct bio_vec *to, from;
- struct bvec_iter iter;
+ struct bio_vec *to;
unsigned i;
- bio_for_each_segment(from, *bio_orig, iter)
- if (page_to_pfn(from.bv_page) > queue_bounce_pfn(q))
- goto bounce;
+ if (!need_bounce(q, *bio_orig))
+ return;
- return;
-bounce:
- bio = bio_clone_bioset(*bio_orig, GFP_NOIO, fs_bio_set);
+ bio = bio_clone_bioset_sp(*bio_orig, GFP_NOIO, fs_bio_set);
bio_for_each_segment_all(to, bio, i) {
struct page *page = to->bv_page;
@@ -267,9 +283,23 @@ void blk_queue_bounce(struct request_queue *q, struct bio **bio_orig)
pool = isa_page_pool;
}
+ if (!need_bounce(q, *bio_orig))
+ return;
+
/*
* slow path
+ *
+ * REQ_PC bio won't reach splitting because multipage bvecs
+ * isn't enabled for REQ_PC.
*/
+ if (need_split(q, *bio_orig)) {
+ struct bio *split = bio_split(*bio_orig,
+ BIO_SP_MAX_SECTORS,
+ GFP_NOIO, q->bio_split);
+ bio_chain(split, *bio_orig);
+ generic_make_request(*bio_orig);
+ *bio_orig = split;
+ }
__blk_queue_bounce(q, bio_orig, pool);
}
This patch trys to split the incoming multipage bvecs bio, so that the splitted bio can be held into one singlepage bvecs bio. Signed-off-by: Ming Lei <tom.leiming@gmail.com> --- block/bounce.c | 46 ++++++++++++++++++++++++++++++++++++++-------- 1 file changed, 38 insertions(+), 8 deletions(-)