[PATCH v2 01/14] block: Add blk_validate_atomic_write_op_size()

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



FSes will not check if the BIO submitted for an atomic write adheres to the
request_queue limits, so check in the BIO submission path.

Signed-off-by: John Garry <john.g.garry@xxxxxxxxxx>
---
We really should return -EINVAL to userspce, so can be have BLK_STS_INVAL
or similar?
 block/blk-core.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/block/blk-core.c b/block/blk-core.c
index de771093b526..a1b095a68498 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -718,6 +718,18 @@ void submit_bio_noacct_nocheck(struct bio *bio)
 		__submit_bio_noacct(bio);
 }
 
+static blk_status_t blk_validate_atomic_write_op_size(struct request_queue *q,
+						 struct bio *bio)
+{
+	if (bio->bi_iter.bi_size > queue_atomic_write_unit_max_bytes(q))
+		return BLK_STS_IOERR;
+
+	if (bio->bi_iter.bi_size % queue_atomic_write_unit_min_bytes(q))
+		return BLK_STS_IOERR;
+
+	return BLK_STS_OK;
+}
+
 /**
  * submit_bio_noacct - re-submit a bio to the block device layer for I/O
  * @bio:  The bio describing the location in memory and on the device.
@@ -775,6 +787,11 @@ void submit_bio_noacct(struct bio *bio)
 	switch (bio_op(bio)) {
 	case REQ_OP_READ:
 	case REQ_OP_WRITE:
+		if (bio->bi_opf & REQ_ATOMIC) {
+			status = blk_validate_atomic_write_op_size(q, bio);
+			if (status != BLK_STS_OK)
+				goto end_io;
+		}
 		break;
 	case REQ_OP_FLUSH:
 		/*
-- 
2.31.1





[Index of Archives]     [XFS Filesystem Development (older mail)]     [Linux Filesystem Development]     [Linux Audio Users]     [Yosemite Trails]     [Linux Kernel]     [Linux RAID]     [Linux SCSI]


  Powered by Linux