[PATCH 05/18] Add ft_first_node().

Clean up some of the open-coded data structure references by providing a
function to return a pointer to the first node in the tree.  This is only
used in high-level functions trying to access the beginning of the tree,
not in low-level code that is actually manipulating the data structure.

Signed-off-by: Scott Wood <redacted>
---
 arch/powerpc/boot/flatdevtree.c |   13 +++++++++----
 1 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/arch/powerpc/boot/flatdevtree.c b/arch/powerpc/boot/flatdevtree.c
index c76c194..78e179f 100644
--- a/arch/powerpc/boot/flatdevtree.c
+++ b/arch/powerpc/boot/flatdevtree.c

@@ -29,6 +29,11 @@
 
 #define _ALIGN(x, al)	(((x) + (al) - 1) & ~((al) - 1))
 
+static char *ft_first_node(struct ft_cxt *cxt)
+{
+	return cxt->rgn[FT_STRUCT].start;
+}
+
 /* Routines for keeping node ptrs returned by ft_find_device current */
 /* First entry not used b/c it would return 0 and be taken as NULL/error */
 static void *ft_node_add(struct ft_cxt *cxt, char *node)

@@ -590,7 +595,7 @@ int ft_add_rsvmap(struct ft_cxt *cxt, u6
 
 void ft_begin_tree(struct ft_cxt *cxt)
 {
-	cxt->p = cxt->rgn[FT_STRUCT].start;
+	cxt->p = ft_first_node(cxt);
 }
 
 void ft_end_tree(struct ft_cxt *cxt)

@@ -636,7 +641,7 @@ void *ft_find_device(struct ft_cxt *cxt,
 	/* require absolute path */
 	if (srch_path[0] != '/')
 		return NULL;
-	node = ft_find_descendent(cxt, cxt->rgn[FT_STRUCT].start, srch_path);
+	node = ft_find_descendent(cxt, ft_first_node(cxt), srch_path);
 	return ft_node_add(cxt, node);
 }
 

@@ -717,7 +722,7 @@ void *ft_get_parent(struct ft_cxt *cxt,
 			return cxt->genealogy[d > 0 ? d - 1 : 0];
 
 	/* have to do it the hard way... */
-	p = cxt->rgn[FT_STRUCT].start;
+	p = ft_first_node(cxt);
 	d = 0;
 	while ((p = ft_next(cxt, p, &atom)) != NULL) {
 		switch (atom.tag) {

@@ -855,7 +860,7 @@ void *ft_create_node(struct ft_cxt *cxt,
 	char *p, *next;
 	int depth = 0;
 
-	p = cxt->rgn[FT_STRUCT].start;
+	p = ft_first_node(cxt);
 	while ((next = ft_next(cxt, p, &atom)) != NULL) {
 		switch (atom.tag) {
 		case OF_DT_BEGIN_NODE:

-- 
1.4.4

+static char *ft_first_node(struct ft_cxt *cxt)
+{
+	return cxt->rgn[FT_STRUCT].start;
+}

Probably should be ft_root_node() instead?  Looks
fine otherwise (obviously ;-) )


Segher

Jerry Van Baren wrote:

Have you considered replacing flatdevtree.c with David Gibson's fdtlib? 
 The interface is a lot cleaner, which likely carries through to the 
underlying code.
<http://thread.gmane.org/gmane.linux.ports.ppc64.devel/12771/focus=12877>

I tried pulling from the git tree, but since I'm using the http protocol 
(our firewall unfortunately won't allow the git protocol to go through), 
all I get is:

Cannot get remote repository information.
Perhaps git-update-server-info needs to be run there?

-Scott

Scott Wood wrote:

Clean up some of the open-coded data structure references by providing a
function to return a pointer to the first node in the tree.  This is only
used in high-level functions trying to access the beginning of the tree,
not in low-level code that is actually manipulating the data structure.

Signed-off-by: Scott Wood <redacted>
---
 arch/powerpc/boot/flatdevtree.c |   13 +++++++++----
 1 files changed, 9 insertions(+), 4 deletions(-)

Hi Scott,

Have you considered replacing flatdevtree.c with David Gibson's fdtlib? 
  The interface is a lot cleaner, which likely carries through to the 
underlying code.
<http://thread.gmane.org/gmane.linux.ports.ppc64.devel/12771/focus=12877>

Best regards,
gvb

On Wed, Jan 24, 2007 at 03:59:18PM -0600, Scott Wood wrote:

Jerry Van Baren wrote:

quoted

Have you considered replacing flatdevtree.c with David Gibson's fdtlib? 
 The interface is a lot cleaner, which likely carries through to the 
underlying code.

In any case I intend to replace the flatdevtree.c code with libfdt at
some point in the future (not too far away, I hope).  I'm delaying on
a couple of things at present: first, sorting out IBM-internal
bureaucratic hurdles, second letting the API settle a bit so that I'm
pretty confident it won't want changing.

quoted

<http://thread.gmane.org/gmane.linux.ports.ppc64.devel/12771/focus=12877>

I tried pulling from the git tree, but since I'm using the http protocol 
(our firewall unfortunately won't allow the git protocol to go through), 
all I get is:

Cannot get remote repository information.
Perhaps git-update-server-info needs to be run there?

Um.. I'm running git-update-server-info, but the repository wasn't
anywhere http visible.  Just moved things around, now try:
	git clone http://ozlabs.org/~dgibson/libfdt.git

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson