Thread (19 messages) 19 messages, 2 authors, 2014-05-23

Re: [PATCH 2/3] of: Make of_find_node_by_path() handle /aliases

From: Grant Likely <hidden>
Date: 2014-05-21 16:09:43
Also in: lkml
Subsystem: open firmware and flattened device tree, the rest · Maintainers: Rob Herring, Saravana Kannan, Linus Torvalds

On Tue, 20 May 2014 19:55:45 -0700, Frank Rowand [off-list ref] wrote:
On 5/13/2014 7:58 AM, Grant Likely wrote:
quoted
Make of_find_node_by_path() handle aliases as prefixes. To make this
quoted
work the name search is refactored to search by path component instead
quoted
of by full string. This should be a more efficient search, and it makes
quoted
it possible to start a search at a subnode of a tree.
quoted
quoted
Signed-off-by: David Daney <redacted>
quoted
Signed-off-by: Pantelis Antoniou <redacted>
quoted
[grant.likely: Rework to not require allocating at runtime]
quoted
Acked-by: Rob Herring <robh@kernel.org>
quoted
Signed-off-by: Grant Likely <redacted>
quoted
---
quoted
 drivers/of/base.c | 60 +++++++++++++++++++++++++++++++++++++++++++++++++++----
quoted
 1 file changed, 56 insertions(+), 4 deletions(-)


Was this patch created against a tree that has modifications to device tree

locking?
  I get a hang due to deadlock when I apply it.  Patch to verify the
cause is below.
Ummm... I may have forgotten to enable CONFIG_LOCKDEP when testing. Try
the following fix, and can you please pass me that brown paper bag.

g.

commit 35e9c5ae6c3d0b5eb91579f397d8e1ecb95ee711
Author: Grant Likely [off-list ref]
Date:   Thu May 22 01:04:17 2014 +0900

    dt: Create unlocked version of for_each_child_of_node()
    
    When iterating over nodes, sometimes it needs to be done when the DT
    lock is already held. This patch makes an unlocked version of the
    for_each_child_of_node() macro.
    
    Signed-off-by: Grant Likely [off-list ref]
diff --git a/drivers/of/base.c b/drivers/of/base.c
index 8900d378c07e..c05a143b6a70 100644
--- a/drivers/of/base.c
+++ b/drivers/of/base.c
@@ -695,6 +695,22 @@ struct device_node *of_get_next_parent(struct device_node *node)
 }
 EXPORT_SYMBOL(of_get_next_parent);
 
+#define __for_each_child_of_node(parent, child) \
+	for (child = __of_get_next_child(parent, NULL); child != NULL; \
+	     child = __of_get_next_child(parent, child))
+static struct device_node *__of_get_next_child(const struct device_node *node,
+						struct device_node *prev)
+{
+	struct device_node *next;
+
+	next = prev ? prev->sibling : node->child;
+	for (; next; next = next->sibling)
+		if (of_node_get(next))
+			break;
+	of_node_put(prev);
+	return next;
+}
+
 /**
  *	of_get_next_child - Iterate a node childs
  *	@node:	parent node
@@ -710,11 +726,7 @@ struct device_node *of_get_next_child(const struct device_node *node,
 	unsigned long flags;
 
 	raw_spin_lock_irqsave(&devtree_lock, flags);
-	next = prev ? prev->sibling : node->child;
-	for (; next; next = next->sibling)
-		if (of_node_get(next))
-			break;
-	of_node_put(prev);
+	next = __of_get_next_child(node, prev);
 	raw_spin_unlock_irqrestore(&devtree_lock, flags);
 	return next;
 }
@@ -780,7 +792,7 @@ static struct device_node *__of_find_node_by_path(struct device_node *parent,
 	if (!len)
 		return parent;
 
-	for_each_child_of_node(parent, child) {
+	__for_each_child_of_node(parent, child) {
 		const char *name = strrchr(child->full_name, '/');
 		if (WARN(!name, "malformed device_node %s\n", child->full_name))
 			continue;
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help