Skip to contentSkip to conversation

of: overlay: Miscellaneous improvements

PATCH Applied Geert Uytterhoeven 3 participants 7 messages Jul 15, 2022 – Jul 19, 2022 Device tree

lore
Share & tools
lore
#0 lore raw

cover of: overlay: Miscellaneous improvements

	Hi,

While performing the long-overdue rebase of my topic/overlays branch[1] on top of the overlay rework in v5.19-rc1, I identified a few areas for improvement in the upstream code.

Thanks for your comments!

[1] https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/log/?h=topic/overlays

Geert Uytterhoeven (2):
of: overlay: Move devicetree_corrupt() check up
of: overlay: Simplify of_overlay_fdt_apply() tail

 drivers/of/overlay.c | 20 +++++++-------------
 1 file changed, 7 insertions(+), 13 deletions(-)
Signature · 12 lines
2.25.1

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds
#1 lore raw

PATCH 1/2 of: overlay: Move devicetree_corrupt() check up

There is no point in doing several preparatory steps in of_overlay_fdt_apply(), only to see of_overlay_apply() return early because of a corrupt device tree.

Move the check for a corrupt device tree from of_overlay_apply() to of_overlay_fdt_apply(), to check for this as early as possible.

---
 drivers/of/overlay.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)
drivers/of/overlay.c +5 -6
index 4044ddcb02c60a58..84a8d402009cb3b2 100644
@@ -903,12 +903,6 @@ static int of_overlay_apply(struct overlay_changeset *ovcs)
 {
 	int ret = 0, ret_revert, ret_tmp;
 
-	if (devicetree_corrupt()) {
-		pr_err("devicetree state suspect, refuse to apply overlay\n");
-		ret = -EBUSY;
-		goto out;
-	}
-
 	ret = of_resolve_phandles(ovcs->overlay_root);
 	if (ret)
 		goto out;
@@ -983,6 +977,11 @@ int of_overlay_fdt_apply(const void *overlay_fdt, u32 overlay_fdt_size,
 
 	*ret_ovcs_id = 0;
 
+	if (devicetree_corrupt()) {
+		pr_err("devicetree state suspect, refuse to apply overlay\n");
+		return -EBUSY;
+	}
+
 	if (overlay_fdt_size < sizeof(struct fdt_header) ||
 	    fdt_check_header(overlay_fdt)) {
 		pr_err("Invalid overlay_fdt header\n");
Signature · 1 line
2.25.1
maintainer Applied
#2 lore raw

Re: [PATCH 1/2] of: overlay: Move devicetree_corrupt() check up

Quoted 4 lines · Geert Uytterhoeven

There is no point in doing several preparatory steps in of_overlay_fdt_apply(), only to see of_overlay_apply() return early because of a corrupt device tree.

Quoted 8 lines

Move the check for a corrupt device tree from of_overlay_apply() to of_overlay_fdt_apply(), to check for this as early as possible.

---
 drivers/of/overlay.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

Applied, thanks!

#3 lore raw

PATCH 2/2 of: overlay: Simplify of_overlay_fdt_apply() tail

In reply to Geert Uytterhoeven

It does not hurt to fill in the changeset id while the mutex is still held. After doing so, the function tails for the success and failure cases become identical, so they can be unified.

---
 drivers/of/overlay.c | 9 ++-------
 1 file changed, 2 insertions(+), 7 deletions(-)
drivers/of/overlay.c +2 -7
index 84a8d402009cb3b2..bd8ff4df723da217 100644
@@ -1043,20 +1043,15 @@ int of_overlay_fdt_apply(const void *overlay_fdt, u32 overlay_fdt_size,
 	 * goto err_free_ovcs.  Instead, the caller of of_overlay_fdt_apply()
 	 * can call of_overlay_remove();
 	 */
-
-	mutex_unlock(&of_mutex);
-	of_overlay_mutex_unlock();
-
 	*ret_ovcs_id = ovcs->id;
-
-	return ret;
+	goto out_unlock;
 
 err_free_ovcs:
 	free_overlay_changeset(ovcs);
 
+out_unlock:
 	mutex_unlock(&of_mutex);
 	of_overlay_mutex_unlock();
-
 	return ret;
 }
 EXPORT_SYMBOL_GPL(of_overlay_fdt_apply);
Signature · 1 line
2.25.1
maintainer Applied
#4 lore raw

Re: [PATCH 2/2] of: overlay: Simplify of_overlay_fdt_apply() tail

Quoted 9 lines · Geert Uytterhoeven

It does not hurt to fill in the changeset id while the mutex is still held. After doing so, the function tails for the success and failure cases become identical, so they can be unified.

---
 drivers/of/overlay.c | 9 ++-------
 1 file changed, 2 insertions(+), 7 deletions(-)

Applied, thanks!

#5 lore raw
In reply to Geert Uytterhoeven
Quoted 8 lines · Geert Uytterhoeven
	Hi,

While performing the long-overdue rebase of my topic/overlays branch[1] on top of the overlay rework in v5.19-rc1, I identified a few areas for improvement in the upstream code.

Thanks for your comments!

Quoted 9 lines

[1] https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/log/?h=topic/overlays

Geert Uytterhoeven (2):
of: overlay: Move devicetree_corrupt() check up
of: overlay: Simplify of_overlay_fdt_apply() tail

 drivers/of/overlay.c | 20 +++++++-------------
 1 file changed, 7 insertions(+), 13 deletions(-)

The patches look good, based on a visual inspection. I'll build the kernel and test the patches tomorrow morning.

-Frank

Reviewed-by Tested-by on the v1 series
#6 lore raw
In reply to Geert Uytterhoeven
Quoted 8 lines · Geert Uytterhoeven
	Hi,

While performing the long-overdue rebase of my topic/overlays branch[1] on top of the overlay rework in v5.19-rc1, I identified a few areas for improvement in the upstream code.

Thanks for your comments!

Quoted 9 lines

[1] https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/log/?h=topic/overlays

Geert Uytterhoeven (2):
of: overlay: Move devicetree_corrupt() check up
of: overlay: Simplify of_overlay_fdt_apply() tail

 drivers/of/overlay.c | 20 +++++++-------------
 1 file changed, 7 insertions(+), 13 deletions(-)

For the entire series: