A caller is supposed to pass in REQ_NOWAIT if we can't block for any given operation, but O_DIRECT for block devices just ignore this. Hence we'll block for various resource shortages on the block layer side, like having to wait for requests.
Use the new REQ_NOWAIT_INLINE to ask for this error to be returned inline, so we can handle it appropriately and return -EAGAIN to the caller. Signed-off-by: Jens Axboe <[email protected]> --- fs/block_dev.c | 41 ++++++++++++++++++++++++++++++++++++++--- 1 file changed, 38 insertions(+), 3 deletions(-) diff --git a/fs/block_dev.c b/fs/block_dev.c index f00b569a9f89..7c4da2e2e5be 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -344,15 +344,24 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages) struct bio *bio; bool is_poll = (iocb->ki_flags & IOCB_HIPRI) != 0; bool is_read = (iov_iter_rw(iter) == READ), is_sync; + bool nowait = (iocb->ki_flags & IOCB_NOWAIT) != 0; loff_t pos = iocb->ki_pos; blk_qc_t qc = BLK_QC_T_NONE; + gfp_t gfp; int ret = 0; if ((pos | iov_iter_alignment(iter)) & (bdev_logical_block_size(bdev) - 1)) return -EINVAL; - bio = bio_alloc_bioset(GFP_KERNEL, nr_pages, &blkdev_dio_pool); + if (nowait) + gfp = GFP_NOWAIT; + else + gfp = GFP_KERNEL; + + bio = bio_alloc_bioset(gfp, nr_pages, &blkdev_dio_pool); + if (!bio) + return -EAGAIN; dio = container_of(bio, struct blkdev_dio, bio); dio->is_sync = is_sync = is_sync_kiocb(iocb); @@ -398,6 +407,14 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages) task_io_account_write(bio->bi_iter.bi_size); } + /* + * Tell underlying layer to not block for resource shortage. + * And if we would have blocked, return error inline instead + * of through the bio->bi_end_io() callback. + */ + if (nowait) + bio->bi_opf |= (REQ_NOWAIT | REQ_NOWAIT_INLINE); + dio->size += bio->bi_iter.bi_size; pos += bio->bi_iter.bi_size; @@ -411,6 +428,10 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages) } qc = submit_bio(bio); + if (qc == BLK_QC_T_EAGAIN) { + ret = -EAGAIN; + goto err; + } if (polled) WRITE_ONCE(iocb->ki_cookie, qc); @@ -431,8 +452,17 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages) atomic_inc(&dio->ref); } - submit_bio(bio); - bio = bio_alloc(GFP_KERNEL, nr_pages); + qc = submit_bio(bio); + if (qc == BLK_QC_T_EAGAIN) { + ret = -EAGAIN; + goto err; + } + + bio = bio_alloc(gfp, nr_pages); + if (!bio) { + ret = -EAGAIN; + goto err; + } } if (!is_poll) @@ -452,6 +482,7 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages) } __set_current_state(TASK_RUNNING); +out: if (!ret) ret = blk_status_to_errno(dio->bio.bi_status); if (likely(!ret)) @@ -459,6 +490,10 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages) bio_put(&dio->bio); return ret; +err: + if (!is_poll) + blk_finish_plug(&plug); + goto out; } static ssize_t -- 2.17.1
