Skip to contentSkip to conversation

device property: fix child iteration issues with secondary fwnodes

PATCH In review Andy Shevchenko 4 participants 11 messages Jun 11 – Jul 20

Subsystems (4)
lore
Share & tools
lore

Andy Shevchenko

author reviewer
#0 lore raw

v4 cover device property: fix child iteration issues with secondary fwnodes

This series fixes two issues in the fwnode child iteration logic when a secondary fwnode is present.

The first issue is a refcount imbalance in software_node_get_next_child(). When a software node is used as a secondary fwnode, the iteration code may incorrectly decrement the refcount of child nodes that do not belong to the software node hierarchy. This results in refcount underflow and possible use-after-free.

The second issue is an infinite loop in fwnode_for_each_child_node(), caused by improper handling of iteration state across primary and secondary fwnodes. When iterating over children from both primary and secondary fwnodes, the code may incorrectly resume iteration from the primary fwnode even when the current child belongs to the secondary, leading to repeated traversal and a loop.

Both issues are triggered when mixing different fwnode types through the secondary mechanism, and stem from incorrect assumptions about ownership and traversal context of child nodes.

Changes in v4:
- amended the fix and test case (Andy)
- added patch 2 to align other implementations with RAII approach (Andy)
- tested on Intel Galileo board for which the initial code was developed (Andy)
- Link to v3: https://patch.msgid.link/[email protected]

Changes in v3:
- remove software node patch
- add a kunit test case suggested by Andy Shevchenko
- Link to v2: https://patch.msgid.link/[email protected]

Changes in v2:
- use __free() to cleanup parent fwnode
- Link to v1: https://lore.kernel.org/r/[email protected]

Andy Shevchenko (1):
device property: Refactor to use RAII approach

Xu Yang (2):
device property: fix infinite loop in fwnode_for_each_child_node() device property: add test cases for fwnode_for_each_child_node()

 drivers/base/property.c                 |  41 ++++---
 drivers/base/test/Kconfig               |   1 +
 drivers/base/test/property-entry-test.c | 136 ++++++++++++++++++++++++
 3 files changed, 161 insertions(+), 17 deletions(-)
Signature · 1 line
2.50.1
author reviewer
#1 lore raw

PATCH v4 1/3 device property: fix infinite loop in fwnode_for_each_child_node()

From: Xu Yang <[email protected]>

When iterate over children of a fwnode that has a secondary fwnode, fwnode_get_next_child_node() can enter an infinite loop if the secondary fwnode has more than one child.

                       Parent        Child
      (Primary fwnode)   FWa:   {FWa1, FWa2, FWa3}
    (Secondary fwnode)   FWb:   {FWb1, FWb2}

In this case:

 ┌─> fwnode_get_next_child_node(FWa, FWa1)
 │    - fwnode_call_ptr_op(FWa, get_next_child_node, FWa1) returns FWa2
 │
 │   ...
 │
 │   fwnode_get_next_child_node(FWa, FWa3)
 │    - fwnode_call_ptr_op(FWa, get_next_child_node, FWa3) returns NULL
 │    - fwnode_call_ptr_op(FWb, get_next_child_node, FWa3) returns FWb1
 │
 │   fwnode_get_next_child_node(FWa, FWb1)
 │    - fwnode_call_ptr_op(FWa, get_next_child_node, FWb1) returns FWa1
 └────┘

This cause fwnode_for_each_child_node() to loop indefinitely, reapeatedly output {FWa1, FWa2, FWa3, FWb1, FWa1, ...}.

The root cause is that when the current child (FWb1) belongs to the secondary fwnode, calling get_next_child_node() on the parimary fwnode incorrectly returns the first child (FWa1) again instead of NULL.

Fix this by dynamically checking the parent fwnode of the current child before calling get_next_child_node(). This approach follows the pattern established in commit b5b41ab6b0c1 ("device property: Check fwnode->secondary in fwnode_graph_get_next_endpoint()").

Fixes 2692c614f8f0 ("device property: Allow secondary lookup in fwnode_get_next_child_node()")
---
 drivers/base/property.c | 19 ++++++++++++++++---
 1 file changed, 16 insertions(+), 3 deletions(-)
drivers/base/property.c +16 -3
index 8e0148a37fff..f7b30d9c8716 100644
@@ -807,18 +807,31 @@ struct fwnode_handle *
 fwnode_get_next_child_node(const struct fwnode_handle *fwnode,
 			   struct fwnode_handle *child)
 {
+	const struct fwnode_handle *parent;
+	struct fwnode_handle *child_parent __free(fwnode_handle) = NULL;
 	struct fwnode_handle *next;
 
-	if (IS_ERR_OR_NULL(fwnode))
+	/*
+	 * If this function is in a loop and the previous iteration returned
+	 * an child from fwnode->secondary, then we need to use the secondary
+	 * as parent rather than @fwnode.
+	 */
+	if (child) {
+		child_parent = fwnode_get_parent(child);
+		parent = child_parent;
+	} else {
+		parent = fwnode;
+	}
+	if (IS_ERR_OR_NULL(parent))
 		return NULL;
 
 	/* Try to find a child in primary fwnode */
-	next = fwnode_call_ptr_op(fwnode, get_next_child_node, child);
+	next = fwnode_call_ptr_op(parent, get_next_child_node, child);
 	if (next)
 		return next;
 
 	/* When no more children in primary, continue with secondary */
-	return fwnode_call_ptr_op(fwnode->secondary, get_next_child_node, child);
+	return fwnode_get_next_child_node(parent->secondary, NULL);
 }
 EXPORT_SYMBOL_GPL(fwnode_get_next_child_node);
 
Signature · 1 line
2.50.1
Tested-by on v4 1/3
#2 lore raw

Re: [PATCH v4 1/3] device property: fix infinite loop in fwnode_for_each_child_node()

Quoted 37 lines · Andy Shevchenko

From: Xu Yang <[email protected]>

When iterate over children of a fwnode that has a secondary fwnode, fwnode_get_next_child_node() can enter an infinite loop if the secondary fwnode has more than one child.

                       Parent        Child
      (Primary fwnode)   FWa:   {FWa1, FWa2, FWa3}
    (Secondary fwnode)   FWb:   {FWb1, FWb2}

In this case:

 ┌─> fwnode_get_next_child_node(FWa, FWa1)
 │    - fwnode_call_ptr_op(FWa, get_next_child_node, FWa1) returns FWa2
 │
 │   ...
 │
 │   fwnode_get_next_child_node(FWa, FWa3)
 │    - fwnode_call_ptr_op(FWa, get_next_child_node, FWa3) returns NULL
 │    - fwnode_call_ptr_op(FWb, get_next_child_node, FWa3) returns FWb1
 │
 │   fwnode_get_next_child_node(FWa, FWb1)
 │    - fwnode_call_ptr_op(FWa, get_next_child_node, FWb1) returns FWa1
 └────┘

This cause fwnode_for_each_child_node() to loop indefinitely, reapeatedly output {FWa1, FWa2, FWa3, FWb1, FWa1, ...}.

The root cause is that when the current child (FWb1) belongs to the secondary fwnode, calling get_next_child_node() on the parimary fwnode incorrectly returns the first child (FWa1) again instead of NULL.

Fix this by dynamically checking the parent fwnode of the current child before calling get_next_child_node(). This approach follows the pattern established in commit b5b41ab6b0c1 ("device property: Check fwnode->secondary in fwnode_graph_get_next_endpoint()").

Quoted 5 lines
Fixes 2692c614f8f0 ("device property: Allow secondary lookup in fwnode_get_next_child_node()")

Thanks,
Xu Yang

Quoted 46 lines
---
 drivers/base/property.c | 19 ++++++++++++++++---
 1 file changed, 16 insertions(+), 3 deletions(-)
drivers/base/property.c +16 -3
index 8e0148a37fff..f7b30d9c8716 100644
@@ -807,18 +807,31 @@ struct fwnode_handle *
 fwnode_get_next_child_node(const struct fwnode_handle *fwnode,
 			   struct fwnode_handle *child)
 {
+	const struct fwnode_handle *parent;
+	struct fwnode_handle *child_parent __free(fwnode_handle) = NULL;
 	struct fwnode_handle *next;
 
-	if (IS_ERR_OR_NULL(fwnode))
+	/*
+	 * If this function is in a loop and the previous iteration returned
+	 * an child from fwnode->secondary, then we need to use the secondary
+	 * as parent rather than @fwnode.
+	 */
+	if (child) {
+		child_parent = fwnode_get_parent(child);
+		parent = child_parent;
+	} else {
+		parent = fwnode;
+	}
+	if (IS_ERR_OR_NULL(parent))
 		return NULL;
 
 	/* Try to find a child in primary fwnode */
-	next = fwnode_call_ptr_op(fwnode, get_next_child_node, child);
+	next = fwnode_call_ptr_op(parent, get_next_child_node, child);
 	if (next)
 		return next;
 
 	/* When no more children in primary, continue with secondary */
-	return fwnode_call_ptr_op(fwnode->secondary, get_next_child_node, child);
+	return fwnode_get_next_child_node(parent->secondary, NULL);
 }
 EXPORT_SYMBOL_GPL(fwnode_get_next_child_node);
 
Signature · 1 line
2.50.1
author reviewer
#3 lore raw

PATCH v4 2/3 device property: Refactor to use RAII approach

In reply to Andy Shevchenko

In a couple of functions code can be made cleaner with help of __free() macro. Refactor these to use RAII approach.

---
 drivers/base/property.c | 22 ++++++++--------------
 1 file changed, 8 insertions(+), 14 deletions(-)
drivers/base/property.c +8 -14
index f7b30d9c8716..808e8a90c125 100644
@@ -7,10 +7,10 @@
  *          Mika Westerberg <[email protected]>
  */
 
+#include <linux/cleanup.h>
 #include <linux/device.h>
 #include <linux/err.h>
 #include <linux/export.h>
-#include <linux/kconfig.h>
 #include <linux/of.h>
 #include <linux/property.h>
 #include <linux/phy.h>
@@ -517,7 +517,6 @@ EXPORT_SYMBOL_GPL(fwnode_property_read_string);
 int fwnode_property_match_string(const struct fwnode_handle *fwnode,
 	const char *propname, const char *string)
 {
-	const char **values;
 	int nval, ret;
 
 	nval = fwnode_property_string_array_count(fwnode, propname);
@@ -527,20 +526,18 @@ int fwnode_property_match_string(const struct fwnode_handle *fwnode,
 	if (nval == 0)
 		return -ENODATA;
 
-	values = kcalloc(nval, sizeof(*values), GFP_KERNEL);
+	const char **values __free(kfree) = kcalloc(nval, sizeof(*values), GFP_KERNEL);
 	if (!values)
 		return -ENOMEM;
 
 	ret = fwnode_property_read_string_array(fwnode, propname, values, nval);
 	if (ret < 0)
-		goto out_free;
+		return ret;
 
 	ret = match_string(values, nval, string);
 	if (ret < 0)
-		ret = -ENODATA;
+		return -ENODATA;
 
-out_free:
-	kfree(values);
 	return ret;
 }
 EXPORT_SYMBOL_GPL(fwnode_property_match_string);
@@ -1128,8 +1125,9 @@ struct fwnode_handle *
 fwnode_graph_get_next_endpoint(const struct fwnode_handle *fwnode,
 			       struct fwnode_handle *prev)
 {
-	struct fwnode_handle *ep, *port_parent = NULL;
 	const struct fwnode_handle *parent;
+	struct fwnode_handle *port_parent __free(fwnode_handle) = NULL;
+	struct fwnode_handle *ep;
 
 	/*
 	 * If this function is in a loop and the previous iteration returned
@@ -1147,13 +1145,9 @@ fwnode_graph_get_next_endpoint(const struct fwnode_handle *fwnode,
 
 	ep = fwnode_call_ptr_op(parent, graph_get_next_endpoint, prev);
 	if (ep)
-		goto out_put_port_parent;
+		return ep;
 
-	ep = fwnode_graph_get_next_endpoint(parent->secondary, NULL);
-
-out_put_port_parent:
-	fwnode_handle_put(port_parent);
-	return ep;
+	return fwnode_graph_get_next_endpoint(parent->secondary, NULL);
 }
 EXPORT_SYMBOL_GPL(fwnode_graph_get_next_endpoint);
 
Signature · 1 line
2.50.1
#4 lore raw

Re: [PATCH v4 2/3] device property: Refactor to use RAII approach

Quoted 58 lines · Andy Shevchenko

In a couple of functions code can be made cleaner with help of __free() macro. Refactor these to use RAII approach.

---
 drivers/base/property.c | 22 ++++++++--------------
 1 file changed, 8 insertions(+), 14 deletions(-)
drivers/base/property.c +4 -7
index f7b30d9c8716..808e8a90c125 100644
@@ -7,10 +7,10 @@
  *          Mika Westerberg <[email protected]>
  */
 
+#include <linux/cleanup.h>
 #include <linux/device.h>
 #include <linux/err.h>
 #include <linux/export.h>
-#include <linux/kconfig.h>
 #include <linux/of.h>
 #include <linux/property.h>
 #include <linux/phy.h>
@@ -517,7 +517,6 @@ EXPORT_SYMBOL_GPL(fwnode_property_read_string);
 int fwnode_property_match_string(const struct fwnode_handle *fwnode,
        const char *propname, const char *string)
 {
-       const char **values;
        int nval, ret;
 
        nval = fwnode_property_string_array_count(fwnode, propname);
@@ -527,20 +526,18 @@ int fwnode_property_match_string(const struct fwnode_handle *fwnode,
        if (nval == 0)
                return -ENODATA;
 
-       values = kcalloc(nval, sizeof(*values), GFP_KERNEL);
+       const char **values __free(kfree) = kcalloc(nval, sizeof(*values), GFP_KERNEL);
        if (!values)
                return -ENOMEM;
 
        ret = fwnode_property_read_string_array(fwnode, propname, values, nval);
        if (ret < 0)
-               goto out_free;
+               return ret;
 
        ret = match_string(values, nval, string);
        if (ret < 0)
-               ret = -ENODATA;
+               return -ENODATA;
 
-out_free:
-       kfree(values);
        return ret;
 }
 EXPORT_SYMBOL_GPL(fwnode_property_match_string);

drivers/base/property.c struct fwnode_handle * · line 1129

Quoted 6 lines
drivers/base/property.c +1 -1
@@ -1128,8 +1125,9 @@ struct fwnode_handle *
 fwnode_graph_get_next_endpoint(const struct fwnode_handle *fwnode,
                               struct fwnode_handle *prev)
 {
-       struct fwnode_handle *ep, *port_parent = NULL;
        const struct fwnode_handle *parent;
+       struct fwnode_handle *port_parent __free(fwnode_handle) = NULL;

The thing on the right-hand side of the assignment should be a constructor, shouldn't it?

Quoted 23 lines
drivers/base/property.c +3 -7
+       struct fwnode_handle *ep;
 
        /*
         * If this function is in a loop and the previous iteration returned
@@ -1147,13 +1145,9 @@ fwnode_graph_get_next_endpoint(const struct fwnode_handle *fwnode,
 
        ep = fwnode_call_ptr_op(parent, graph_get_next_endpoint, prev);
        if (ep)
-               goto out_put_port_parent;
+               return ep;
 
-       ep = fwnode_graph_get_next_endpoint(parent->secondary, NULL);
-
-out_put_port_parent:
-       fwnode_handle_put(port_parent);
-       return ep;
+       return fwnode_graph_get_next_endpoint(parent->secondary, NULL);
 }
 EXPORT_SYMBOL_GPL(fwnode_graph_get_next_endpoint);
 
--

2.50.1

author reviewer
#5 lore raw

Re: [PATCH v4 2/3] device property: Refactor to use RAII approach

Quoted 4 lines · Rafael J. Wysocki
Quoted 2 lines · Andy Shevchenko

In a couple of functions code can be made cleaner with help of __free() macro. Refactor these to use RAII approach.

...

Quoted 9 lines

drivers/base/property.c struct fwnode_handle * · line 1129

Quoted 6 lines

fwnode_graph_get_next_endpoint(const struct fwnode_handle *fwnode,
struct fwnode_handle *prev)
{

drivers/base/property.c +1 -1
-       struct fwnode_handle *ep, *port_parent = NULL;
        const struct fwnode_handle *parent;
+       struct fwnode_handle *port_parent __free(fwnode_handle) = NULL;

The thing on the right-hand side of the assignment should be a constructor, shouldn't it?

Nope. It's a recommendation to do a such, but in this case it's hard to achieve (since the constructor is inside the conditional scope) and actually is not needed as we don't mix up with other RAII calls, such a guard()().

Signature · 2 lines
With Best Regards,
Andy Shevchenko
author reviewer
#6 lore raw

PATCH v4 3/3 device property: add test cases for fwnode_for_each_child_node()

In reply to Andy Shevchenko

From: Xu Yang <[email protected]>

Add test cases for fwnode_for_each_child_node() API.

Test command:

$ ./tools/testing/kunit/kunit.py run property-entry
---
 drivers/base/test/Kconfig               |   1 +
 drivers/base/test/property-entry-test.c | 136 ++++++++++++++++++++++++
 2 files changed, 137 insertions(+)
drivers/base/test/Kconfig +1 -0
index 2756870615cc..95fc42e91564 100644
@@ -17,4 +17,5 @@ config DM_KUNIT_TEST
 config DRIVER_PE_KUNIT_TEST
 	tristate "KUnit Tests for property entry API" if !KUNIT_ALL_TESTS
 	depends on KUNIT
+	select OF
 	default KUNIT_ALL_TESTS
drivers/base/test/property-entry-test.c +136 -0

154 lines not on this page · raw message

Signature · 1 line
2.50.1
maintainer
#7 lore raw

Re: [PATCH v4 3/3] device property: add test cases for fwnode_for_each_child_node()

(Cc: OF)

Quoted 9 lines · Andy Shevchenko

From: Xu Yang <[email protected]>

Add test cases for fwnode_for_each_child_node() API.

Test command:

$ ./tools/testing/kunit/kunit.py run property-entry

This patch broke the build [1], hence I dropped it from the queue (but kept the first two patches of this series).

of_node_init() can't be used from modules, either it needs EXPORT_SYMBOL_GPL() itself, or of_node_ktype needs to be exported for this kunit test to work.

[1] https://lore.kernel.org/all/[email protected]/

author reviewer
#8 lore raw

Re: [PATCH v4 3/3] device property: add test cases for fwnode_for_each_child_node()

Quoted 7 lines · Danilo Krummrich
Quoted 6 lines · Andy Shevchenko

Add test cases for fwnode_for_each_child_node() API.

Test command:

$ ./tools/testing/kunit/kunit.py run property-entry
Quoted 8 lines
Quoted 2 lines

This patch broke the build [1], hence I dropped it from the queue (but kept the first two patches of this series).

of_node_init() can't be used from modules, either it needs EXPORT_SYMBOL_GPL() itself, or of_node_ktype needs to be exported for this kunit test to work.

Oh, thanks for pointing out.

Quoted 1 line

[1] https://lore.kernel.org/all/[email protected]/

Signature · 2 lines
With Best Regards,
Andy Shevchenko
#9 lore raw

Re: [PATCH v4 3/3] device property: add test cases for fwnode_for_each_child_node()

Hi Danilo and Andy,

Quoted 11 lines · Andy Shevchenko
Quoted 10 lines · Danilo Krummrich
Quoted 8 lines · Andy Shevchenko

Add test cases for fwnode_for_each_child_node() API.

Test command:

$ ./tools/testing/kunit/kunit.py run property-entry
Quoted 7 lines
Quoted 5 lines

This patch broke the build [1], hence I dropped it from the queue (but kept the first two patches of this series).

of_node_init() can't be used from modules, either it needs EXPORT_SYMBOL_GPL() itself, or of_node_ktype needs to be exported for this kunit test to work.

Oh, thanks for pointing out.

Fix patches are sent:
https://lore.kernel.org/imx/[email protected]/

Thanks,
Xu Yang

#10 lore raw
In reply to Andy Shevchenko

Hi Greg and Danilo,

Quoted 15 lines · Andy Shevchenko

This series fixes two issues in the fwnode child iteration logic when a secondary fwnode is present.

The first issue is a refcount imbalance in software_node_get_next_child(). When a software node is used as a secondary fwnode, the iteration code may incorrectly decrement the refcount of child nodes that do not belong to the software node hierarchy. This results in refcount underflow and possible use-after-free.

The second issue is an infinite loop in fwnode_for_each_child_node(), caused by improper handling of iteration state across primary and secondary fwnodes. When iterating over children from both primary and secondary fwnodes, the code may incorrectly resume iteration from the primary fwnode even when the current child belongs to the secondary, leading to repeated traversal and a loop.

Quoted 3 lines

Both issues are triggered when mixing different fwnode types through the secondary mechanism, and stem from incorrect assumptions about ownership and traversal context of child nodes.

Gentle ping on this series.

Thanks,
Xu Yang

Quoted 31 lines

Changes in v4:
- amended the fix and test case (Andy)
- added patch 2 to align other implementations with RAII approach (Andy)
- tested on Intel Galileo board for which the initial code was developed (Andy)
- Link to v3: https://patch.msgid.link/[email protected]

Changes in v3:
- remove software node patch
- add a kunit test case suggested by Andy Shevchenko
- Link to v2: https://patch.msgid.link/[email protected]

Changes in v2:
- use __free() to cleanup parent fwnode
- Link to v1: https://lore.kernel.org/r/[email protected]

Andy Shevchenko (1):
device property: Refactor to use RAII approach

Xu Yang (2):
device property: fix infinite loop in fwnode_for_each_child_node() device property: add test cases for fwnode_for_each_child_node()

 drivers/base/property.c                 |  41 ++++---
 drivers/base/test/Kconfig               |   1 +
 drivers/base/test/property-entry-test.c | 136 ++++++++++++++++++++++++
 3 files changed, 161 insertions(+), 17 deletions(-)
Signature · 1 line
2.50.1