Thread (16 messages) 16 messages, 2 authors, 2022-09-27

Re: [PATCH for-next v7 3/5] nvme: refactor nvme_alloc_user_request

From: Christoph Hellwig <hch@lst.de>
Date: 2022-09-20 12:02:36
Also in: io-uring, linux-nvme

On Fri, Sep 09, 2022 at 03:51:34PM +0530, Kanchan Joshi wrote:
quoted hunk ↗ jump to hunk
Separate this out to two functions with reduced number of arguments.
_
Signed-off-by: Kanchan Joshi <redacted>
---
 drivers/nvme/host/ioctl.c | 116 ++++++++++++++++++++++----------------
 1 file changed, 66 insertions(+), 50 deletions(-)
diff --git a/drivers/nvme/host/ioctl.c b/drivers/nvme/host/ioctl.c
index 548aca8b5b9f..cb2fa4db50dd 100644
--- a/drivers/nvme/host/ioctl.c
+++ b/drivers/nvme/host/ioctl.c
@@ -65,18 +65,10 @@ static int nvme_finish_user_metadata(struct request *req, void __user *ubuf,
 }
 
 static struct request *nvme_alloc_user_request(struct request_queue *q,
+		struct nvme_command *cmd, unsigned timeout,
 		blk_opf_t rq_flags, blk_mq_req_flags_t blk_flags)
I think we can also drop the timeout flag here, which seems like it
can be handled cleaner in the callers.
to set it can just do that.
+static int nvme_map_user_request(struct request *req, void __user *ubuffer,
+		unsigned bufflen, void __user *meta_buffer, unsigned meta_len,
+		u32 meta_seed, void **metap, bool vec)
+{
+	struct request_queue *q = req->q;
+	struct nvme_ns *ns = q->queuedata;
+	struct block_device *bdev = ns ? ns->disk->part0 : NULL;
+	struct bio *bio = NULL;
+	void *meta = NULL;
+	int ret;
+
+	if (!ubuffer || !bufflen)
+		return 0;
I'd leave these in the callers and not call the helper if there is
no data to transfer.
+
+	if (!vec)
+		ret = blk_rq_map_user(q, req, NULL, ubuffer, bufflen,
+			GFP_KERNEL);
+	else {
+		struct iovec fast_iov[UIO_FASTIOV];
+		struct iovec *iov = fast_iov;
+		struct iov_iter iter;
+
+		ret = import_iovec(rq_data_dir(req), ubuffer, bufflen,
+				UIO_FASTIOV, &iov, &iter);
+		if (ret < 0)
 			goto out;
+		ret = blk_rq_map_user_iov(q, req, NULL, &iter, GFP_KERNEL);
+		kfree(iov);
To me some of this almost screams like lifting the vectored vs
not to the block layer into a separate helper.
+	}
+	bio = req->bio;
+	if (ret)
+		goto out_unmap;
This seems incorrect, we don't need to unmap if blk_rq_map_user*
failed.
+	if (bdev)
+		bio_set_dev(bio, bdev);
I think we can actually drop this now - bi_bdev should only be used
by the non-passthrough path these days.
+	if (bdev && meta_buffer && meta_len) {
+		meta = nvme_add_user_metadata(bio, meta_buffer, meta_len,
+				meta_seed, req_op(req) == REQ_OP_DRV_OUT);
+		if (IS_ERR(meta)) {
+			ret = PTR_ERR(meta);
+			goto out_unmap;
 		}
+		req->cmd_flags |= REQ_INTEGRITY;
+		*metap = meta;
And if we pass the request to nvme_add_user_metadata, that can set
REQ_INTEGRITY.  And we don't need this second helper at all.
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help