Skip to contentSkip to conversation

Check for endpoints in fwnode->secondary more sensibly

PATCH Superseded Daniel Scally 2 participants 8 messages Jul 22, 2021 – Jul 31, 2021

Subsystems (2)

v1 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Daniel Scally

author reviewer
#0 lore raw

cover Check for endpoints in fwnode->secondary more sensibly

Hello all

A while ago I patched fwnode_graph_get_endpoint_by_id() to check for endpoints against fwnode->secondary if none was found against the primary. It's actually better to do this in fwnode_graph_get_next_endpoint() instead, since that function is called by fwnode_graph_get_endpoint_by_id() and also directly called in a bunch of other places (primarily sensor drivers checking that they have endpoints connected during probe). This small series just adds the equivalent functionality to fwnode_graph_get_next_endpoint() and reverts the earlier commit.

Thanks
Dan

Daniel Scally (2):
device property: Check fwnode->secondary in
fwnode_graph_get_next_endpoint()
Revert "media: device property: Call fwnode_graph_get_endpoint_by_id()

    for fwnode->secondary"
 drivers/base/property.c | 30 +++++++++++++++++++++---------
 1 file changed, 21 insertions(+), 9 deletions(-)
Signature · 1 line
2.25.1
author reviewer
#1 lore raw

PATCH 1/2 device property: Check fwnode->secondary in fwnode_graph_get_next_endpoint()

Sensor drivers often check for an endpoint to make sure that they're connected to a consuming device like a CIO2 during .probe(). Some of those endpoints might be in the form of software_nodes assigned as a secondary to the device's fwnode_handle. Account for this possibility in fwnode_graph_get_next_endpoint() to avoid having to do it in the sensor drivers themselves.

---
 drivers/base/property.c | 21 ++++++++++++++++++++-
 1 file changed, 20 insertions(+), 1 deletion(-)
drivers/base/property.c +20 -1
index 1421e9548857..e3aceb3a9a0d 100644
@@ -1036,7 +1036,26 @@ struct fwnode_handle *
 fwnode_graph_get_next_endpoint(const struct fwnode_handle *fwnode,
 			       struct fwnode_handle *prev)
 {
-	return fwnode_call_ptr_op(fwnode, graph_get_next_endpoint, prev);
+	const struct fwnode_handle *parent;
+	struct fwnode_handle *ep;
+
+	/*
+	 * If this function is in a loop and the previous iteration returned
+	 * an endpoint from fwnode->secondary, then we need to use the secondary
+	 * as parent rather than @fwnode.
+	 */
+	if (prev)
+		parent = fwnode_graph_get_port_parent(prev);
+	else
+		parent = fwnode;
+
+	ep = fwnode_call_ptr_op(parent, graph_get_next_endpoint, prev);
+
+	if (IS_ERR_OR_NULL(ep) && !IS_ERR_OR_NULL(parent) &&
+	    !IS_ERR_OR_NULL(parent->secondary))
+		ep = fwnode_graph_get_next_endpoint(parent->secondary, NULL);
+
+	return ep;
 }
 EXPORT_SYMBOL_GPL(fwnode_graph_get_next_endpoint);
 
Signature · 1 line
2.25.1
#2 lore raw

Re: [PATCH 1/2] device property: Check fwnode->secondary in fwnode_graph_get_next_endpoint()

Quoted 6 lines · Daniel Scally

Sensor drivers often check for an endpoint to make sure that they're connected to a consuming device like a CIO2 during .probe(). Some of those endpoints might be in the form of software_nodes assigned as a secondary to the device's fwnode_handle. Account for this possibility in fwnode_graph_get_next_endpoint() to avoid having to do it in the sensor drivers themselves.

...

drivers/base/property.c struct fwnode_handle * · line 1055

Quoted 4 lines
drivers/base/property.c +4 -0
+	ep = fwnode_call_ptr_op(parent, graph_get_next_endpoint, prev);
+
+	if (IS_ERR_OR_NULL(ep) && !IS_ERR_OR_NULL(parent) &&
+	    !IS_ERR_OR_NULL(parent->secondary))

Nit-pick, I would put it like:

	if (!IS_ERR_OR_NULL(parent->secondary) && !IS_ERR_OR_NULL(parent) &&
	    IS_ERR_OR_NULL(ep))

or

	if (IS_ERR_OR_NULL(ep) &&
	    !IS_ERR_OR_NULL(parent->secondary) && !IS_ERR_OR_NULL(parent))

for the sake of logical split.

Quoted 1 line
drivers/base/property.c +1 -0
+		ep = fwnode_graph_get_next_endpoint(parent->secondary, NULL);
Signature · 2 lines
With Best Regards,
Andy Shevchenko
author reviewer
#3 lore raw

Re: [PATCH 1/2] device property: Check fwnode->secondary in fwnode_graph_get_next_endpoint()

Quoted 18 lines · Andy Shevchenko
Quoted 6 lines · Daniel Scally

Sensor drivers often check for an endpoint to make sure that they're connected to a consuming device like a CIO2 during .probe(). Some of those endpoints might be in the form of software_nodes assigned as a secondary to the device's fwnode_handle. Account for this possibility in fwnode_graph_get_next_endpoint() to avoid having to do it in the sensor drivers themselves.

...

drivers/base/property.c struct fwnode_handle * · line 1055

Quoted 4 lines
drivers/base/property.c +4 -0
+	ep = fwnode_call_ptr_op(parent, graph_get_next_endpoint, prev);
+
+	if (IS_ERR_OR_NULL(ep) && !IS_ERR_OR_NULL(parent) &&
+	    !IS_ERR_OR_NULL(parent->secondary))

Nit-pick, I would put it like:

	if (!IS_ERR_OR_NULL(parent->secondary) && !IS_ERR_OR_NULL(parent) &&
	    IS_ERR_OR_NULL(ep))
Quoted 6 lines

or

	if (IS_ERR_OR_NULL(ep) &&
	    !IS_ERR_OR_NULL(parent->secondary) && !IS_ERR_OR_NULL(parent))

for the sake of logical split.

OK; I'll do the second one, feel like it's better to have ep as the first check.

Quoted 2 lines
Quoted 2 lines
drivers/base/property.c +1 -0
+		ep = fwnode_graph_get_next_endpoint(parent->secondary, NULL);
#4 lore raw

Re: [PATCH 1/2] device property: Check fwnode->secondary in fwnode_graph_get_next_endpoint()

Quoted 22 lines · Daniel Scally
Quoted 20 lines · Andy Shevchenko
Quoted 6 lines · Daniel Scally

Sensor drivers often check for an endpoint to make sure that they're connected to a consuming device like a CIO2 during .probe(). Some of those endpoints might be in the form of software_nodes assigned as a secondary to the device's fwnode_handle. Account for this possibility in fwnode_graph_get_next_endpoint() to avoid having to do it in the sensor drivers themselves.

...

drivers/base/property.c struct fwnode_handle * · line 1055

Quoted 4 lines
drivers/base/property.c +4 -0
+	ep = fwnode_call_ptr_op(parent, graph_get_next_endpoint, prev);
+
+	if (IS_ERR_OR_NULL(ep) && !IS_ERR_OR_NULL(parent) &&
+	    !IS_ERR_OR_NULL(parent->secondary))

Nit-pick, I would put it like:

	if (!IS_ERR_OR_NULL(parent->secondary) && !IS_ERR_OR_NULL(parent) &&
	    IS_ERR_OR_NULL(ep))

or

Quoted 8 lines
Quoted 4 lines
	if (IS_ERR_OR_NULL(ep) &&
	    !IS_ERR_OR_NULL(parent->secondary) && !IS_ERR_OR_NULL(parent))

for the sake of logical split.

OK; I'll do the second one, feel like it's better to have ep as the first check.

Fine, but also I have just noticed that parent should be checked before parent->secondary.

Something like this

	if (IS_ERR_OR_NULL(ep) &&
	    !IS_ERR_OR_NULL(parent) && IS_ERR_OR_NULL(parent->secondary))
Quoted 1 line
Quoted 1 line
Quoted 1 line
drivers/base/property.c +1 -0
+		ep = fwnode_graph_get_next_endpoint(parent->secondary, NULL);
Signature · 2 lines
With Best Regards,
Andy Shevchenko
author reviewer
#5 lore raw

Re: [PATCH 1/2] device property: Check fwnode->secondary in fwnode_graph_get_next_endpoint()

Quoted 33 lines · Andy Shevchenko
Quoted 22 lines · Daniel Scally
Quoted 20 lines · Andy Shevchenko
Quoted 6 lines · Daniel Scally

Sensor drivers often check for an endpoint to make sure that they're connected to a consuming device like a CIO2 during .probe(). Some of those endpoints might be in the form of software_nodes assigned as a secondary to the device's fwnode_handle. Account for this possibility in fwnode_graph_get_next_endpoint() to avoid having to do it in the sensor drivers themselves.

...

drivers/base/property.c struct fwnode_handle * · line 1055

Quoted 4 lines
drivers/base/property.c +4 -0
+  ep = fwnode_call_ptr_op(parent, graph_get_next_endpoint, prev);
+
+  if (IS_ERR_OR_NULL(ep) && !IS_ERR_OR_NULL(parent) &&
+      !IS_ERR_OR_NULL(parent->secondary))

Nit-pick, I would put it like:

    if (!IS_ERR_OR_NULL(parent->secondary) && !IS_ERR_OR_NULL(parent) &&
        IS_ERR_OR_NULL(ep))

or

Quoted 8 lines
Quoted 4 lines
    if (IS_ERR_OR_NULL(ep) &&
        !IS_ERR_OR_NULL(parent->secondary) && !IS_ERR_OR_NULL(parent))

for the sake of logical split.

OK; I'll do the second one, feel like it's better to have ep as the first check.

Quoted 9 lines

Fine, but also I have just noticed that parent should be checked before parent->secondary.

Something like this

        if (IS_ERR_OR_NULL(ep) &&
            !IS_ERR_OR_NULL(parent) && IS_ERR_OR_NULL(parent->secondary))
Quoted 1 line
Quoted 1 line
Quoted 1 line

+ ep = fwnode_graph_get_next_endpoint(parent->secondary, NULL);

Yes, no problem. I'll send a v2 when I can, It will likely be another week or so though, my computer's in a cardboard box.

Quoted 6 lines

--
With Best Regards,
Andy Shevchenko

author reviewer
#6 lore raw

PATCH 2/2 Revert "media: device property: Call fwnode_graph_get_endpoint_by_id() for fwnode->secondary"

In reply to Daniel Scally

This reverts commit acd418bfcfc415cf5e6414b6d1c6acfec850f290. Checking for endpoints against fwnode->secondary in fwnode_graph_get_next_endpoint() is a better way to do this since that function is also used in a bunch of other places, for instance sensor drivers checking that they do have an endpoint connected during probe.

---
 drivers/base/property.c | 9 +--------
 1 file changed, 1 insertion(+), 8 deletions(-)
drivers/base/property.c +1 -8
index e3aceb3a9a0d..689276fb0e45 100644
@@ -1234,14 +1234,7 @@ fwnode_graph_get_endpoint_by_id(const struct fwnode_handle *fwnode,
 		best_ep_id = fwnode_ep.id;
 	}
 
-	if (best_ep)
-		return best_ep;
-
-	if (fwnode && !IS_ERR_OR_NULL(fwnode->secondary))
-		return fwnode_graph_get_endpoint_by_id(fwnode->secondary, port,
-						       endpoint, flags);
-
-	return NULL;
+	return best_ep;
 }
 EXPORT_SYMBOL_GPL(fwnode_graph_get_endpoint_by_id);
 
Signature · 1 line
2.25.1
reviewer Reviewed-by on the v1 series
#7 lore raw
In reply to Daniel Scally
Quoted 10 lines · Daniel Scally

Hello all

A while ago I patched fwnode_graph_get_endpoint_by_id() to check for endpoints against fwnode->secondary if none was found against the primary. It's actually better to do this in fwnode_graph_get_next_endpoint() instead, since that function is called by fwnode_graph_get_endpoint_by_id() and also directly called in a bunch of other places (primarily sensor drivers checking that they have endpoints connected during probe). This small series just adds the equivalent functionality to fwnode_graph_get_next_endpoint() and reverts the earlier commit.

Makes sense to me (one nit-pick to patch 1, though).

Quoted 12 lines

Daniel Scally (2):
device property: Check fwnode->secondary in
fwnode_graph_get_next_endpoint()
Revert "media: device property: Call fwnode_graph_get_endpoint_by_id()

    for fwnode->secondary"
 drivers/base/property.c | 30 +++++++++++++++++++++---------
 1 file changed, 21 insertions(+), 9 deletions(-)
Signature · 1 line
2.25.1
Signature · 2 lines
With Best Regards,
Andy Shevchenko