Re: [PATCH v14] i2c: virtio: add a virtio i2c frontend driver

On Fri, Jul 09, 2021 at 09:14:07AM +0530, Viresh Kumar wrote:

On 09-07-21, 10:25, Jie Deng wrote:

quoted

Add an I2C bus driver for virtio para-virtualization.

The controller can be emulated by the backend driver in
any device model software by following the virtio protocol.

The device specification can be found on
https://lists.oasis-open.org/archives/virtio-comment/202101/msg00008.html.

By following the specification, people may implement different
backend drivers to emulate different controllers according to
their needs.

Co-developed-by: Conghui Chen <redacted>
Signed-off-by: Conghui Chen <redacted>
Signed-off-by: Jie Deng <redacted>
---
Changes v13 -> v14
	- Put the headers in virtio_i2c.h in alphabetical order.
	- Dropped I2C_FUNC_SMBUS_QUICK support.
	- Dropped few unnecessary variables and checks.
	- Use "num" everywhere instead of num or nr, to be consistent.
	- Added few comments which make the design more clear. 

Thanks a lot for following this up so far :)

quoted

+static int virtio_i2c_prepare_reqs(struct virtqueue *vq,
+				   struct virtio_i2c_req *reqs,
+				   struct i2c_msg *msgs, int num)
+{
+	struct scatterlist *sgs[3], out_hdr, msg_buf, in_hdr;
+	int i;
+
+	for (i = 0; i < num; i++) {
+		int outcnt = 0, incnt = 0;
+
+		/*
+		 * We don't support 0 length messages and so masked out
+		 * I2C_FUNC_SMBUS_QUICK in virtio_i2c_func().
+		 */
+		if (!msgs[i].len)
+			break;
+
+		/*
+		 * Only 7-bit mode supported for this moment. For the address
+		 * format, Please check the Virtio I2C Specification.
+		 */
+		reqs[i].out_hdr.addr = cpu_to_le16(msgs[i].addr << 1);
+
+		if (i != num - 1)
+			reqs[i].out_hdr.flags = cpu_to_le32(VIRTIO_I2C_FLAGS_FAIL_NEXT);
+
+		sg_init_one(&out_hdr, &reqs[i].out_hdr, sizeof(reqs[i].out_hdr));
+		sgs[outcnt++] = &out_hdr;
+
+		reqs[i].buf = i2c_get_dma_safe_msg_buf(&msgs[i], 1);
+		if (!reqs[i].buf)
+			break;
+
+		sg_init_one(&msg_buf, reqs[i].buf, msgs[i].len);
+
+		if (msgs[i].flags & I2C_M_RD)
+			sgs[outcnt + incnt++] = &msg_buf;
+		else
+			sgs[outcnt++] = &msg_buf;
+
+		sg_init_one(&in_hdr, &reqs[i].in_hdr, sizeof(reqs[i].in_hdr));
+		sgs[outcnt + incnt++] = &in_hdr;
+
+		if (virtqueue_add_sgs(vq, sgs, outcnt, incnt, &reqs[i], GFP_KERNEL)) {
+			i2c_put_dma_safe_msg_buf(reqs[i].buf, &msgs[i], false);
+			break;
+		}
+	}
+
+	return i;
+}

Wolfram, in case you wonder why we don't error out early as discussed earlier,
then ...

quoted

+static int virtio_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs,
+			   int num)
+{

...

quoted

+	/*
+	 * For the case where count < num, i.e. we weren't able to queue all the
+	 * msgs, ideally we should abort right away and return early, but some
+	 * of the messages are already sent to the remote I2C controller and the
+	 * virtqueue will be left in undefined state in that case. We kick the
+	 * remote here to clear the virtqueue, so we can try another set of
+	 * messages later on.
+	 */

... here is the reasoning for that.

Please see if you can still get it merged into 5.14-rc1/2. Thanks.

Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Tested-by: Viresh Kumar <viresh.kumar@linaro.org>

Well a new driver so maybe rc2 is still ok ...

Acked-by: Michael S. Tsirkin <mst@redhat.com>

-- 
viresh

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization