Re: [PATCH v5 4/7] v4l2-ctl/compliance: Add simple routing test

Tomi Valkeinen <tomi.valkeinen@xxxxxxxxxxxxxxxx> · Wed, 19 Jul 2023 11:22:34 +0300

On 07/06/2023 14:57, Hans Verkuil wrote:
On 07/06/2023 13:48, Hans Verkuil wrote:
On 07/06/2023 13:35, Hans Verkuil wrote:
On 29/05/2023 10:50, Tomi Valkeinen wrote:
Add a very simple test for
VIDIOC_SUBDEV_G_ROUTING/VIDIOC_SUBDEV_S_ROUTING.

We can't (at least at the moment) really know here what kind of routings
the driver would accept, but we can test a VIDIOC_SUBDEV_G_ROUTING call,
followed by a VIDIOC_SUBDEV_S_ROUTING call with the routing we
received.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@xxxxxxxxxxxxxxxx>
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@xxxxxxxxxxxxxxxx>
---
  utils/v4l2-compliance/v4l2-compliance.cpp   | 12 ++++++++++++
  utils/v4l2-compliance/v4l2-compliance.h     |  1 +
  utils/v4l2-compliance/v4l2-test-subdevs.cpp | 16 ++++++++++++++++
  3 files changed, 29 insertions(+)

diff --git a/utils/v4l2-compliance/v4l2-compliance.cpp b/utils/v4l2-compliance/v4l2-compliance.cpp
index d7c10482..f082f569 100644
--- a/utils/v4l2-compliance/v4l2-compliance.cpp
+++ b/utils/v4l2-compliance/v4l2-compliance.cpp
@@ -1249,6 +1249,18 @@ void testNode(struct node &node, struct node &node_m2m_cap, struct node &expbuf_
  		node.is_passthrough_subdev = has_source && has_sink;
  
  		if (has_routes) {
+			printf("Sub-Device routing ioctls:\n");
+
+			for (unsigned which = V4L2_SUBDEV_FORMAT_TRY;
+				which <= V4L2_SUBDEV_FORMAT_ACTIVE; which++) {
+
+				printf("\ttest %s VIDIOC_SUBDEV_G_ROUTING/VIDIOC_SUBDEV_S_ROUTING: %s\n",
+				       which ? "Active" : "Try",
+				       ok(testSubDevRouting(&node, which)));
+			}
+
+			printf("\n");
+
  			for (unsigned which = V4L2_SUBDEV_FORMAT_TRY;
  				which <= V4L2_SUBDEV_FORMAT_ACTIVE; which++) {
  
diff --git a/utils/v4l2-compliance/v4l2-compliance.h b/utils/v4l2-compliance/v4l2-compliance.h
index 0cd43980..35b2274b 100644
--- a/utils/v4l2-compliance/v4l2-compliance.h
+++ b/utils/v4l2-compliance/v4l2-compliance.h
@@ -375,6 +375,7 @@ int testSubDevEnum(struct node *node, unsigned which, unsigned pad, unsigned str
  int testSubDevFormat(struct node *node, unsigned which, unsigned pad, unsigned stream);
  int testSubDevSelection(struct node *node, unsigned which, unsigned pad, unsigned stream);
  int testSubDevFrameInterval(struct node *node, unsigned pad, unsigned stream);
+int testSubDevRouting(struct node *node, unsigned which);
  
  // Buffer ioctl tests
  int testReqBufs(struct node *node);
diff --git a/utils/v4l2-compliance/v4l2-test-subdevs.cpp b/utils/v4l2-compliance/v4l2-test-subdevs.cpp
index 07192bda..962d9244 100644
--- a/utils/v4l2-compliance/v4l2-test-subdevs.cpp
+++ b/utils/v4l2-compliance/v4l2-test-subdevs.cpp
@@ -551,3 +551,19 @@ int testSubDevSelection(struct node *node, unsigned which, unsigned pad, unsigne
  
  	return have_sel ? 0 : ENOTTY;
  }
+
+int testSubDevRouting(struct node *node, unsigned which)
+{
+	struct v4l2_subdev_routing routing = {};
+	struct v4l2_subdev_route routes[256] = {};

NUM_ROUTES_MAX

+
+	routing.which = which;
+	routing.routes = (__u64)&routes;
+	routing.num_routes = 256;

NUM_ROUTES_MAX

Yes, a common NUM_ROUTES_MAX in utils/common/v4l2-info.h makes sense.

Actually, you should also test the corner cases of NUM_ROUTES_MAX + 1

Yes, that fails, but I don't think it's a good test as the maximum 
number of routes is not part of the uAPI, it's just an arbitrary sanity 
limit.

(that should fail, right?) and setting num_routes to 0 and check that
ENOSPC is returned and num_routes is updated.

Yes, I'll add that. However, if the subdev has 0 routes, the ioctl can 
succeed and num_routes is 0. I'll check that too.

Also verify that 'reserved' is zeroed (i.e. set it to 0xff here, then
check for 0 after the ioctl).

Ok.

I assume also that if num_routes is set to 256, then G_ROUTING is called,
num_routes is updated to the actual number of routes? The spec does not
actually state that.

Hmm, indeed, the doc doesn't say that num_routes is updated on success. 
I need to add that.

And what should happen when num_routes is set to 0 and S_ROUTING is called?
Would that clear all routes? Or is that an error? And can G_ROUTING actually
return num_routes == 0 as well if there are no routes defined?

Yes, it would clear the routes, and G_ROUTING would return num_routes == 
0. However, this is really up to the driver, and it can return an error 
for S_ROUTING with num_routes == 0.

Additional checks you can do is to verify that all sink/source pads are valid
and that 'flags' is valid.

Ok. But if I understand the code right, I can only do that when the user 
is testing a media device, not if the user is testing a single 
subdevice. I can do this part of the test when node->pads is non-NULL.

Another test is setting a pad or stream to an invalid value and verify that EINVAL
is returned.

There are no invalid stream values, but I can test for invalid pads. 
However, here I again need a media device as above.

Note that the spec says that E2BIG is returned for S_ROUTING, but it is returned
for G_ROUTING as well, that should be updated.

For the 256+ case? As I mentioned above, it's an arbitrary "hidden" 
limit, which can be increased in the kernel when needed. So, in theory, 
G_ROUTING should accept any num_routes value, but to avoid large 
allocations in the kernel side, it's at the moment limited to 256.

Maybe a better kernel side implementation would be to silently limit the 
kernel side arrays to 256, even if the userspace provides a larger array.

 Tomi







