Skip to contentSkip to conversation

pinctrl: use dev_info instead of pr_info in pinctrl_select_state_locked

PATCH In review Richard Genoud 3 participants 9 messages Mar 28, 2013 – Apr 3, 2013 Pin control

lore
Share & tools
lore
#0 lore raw

PATCH 1/3 pinctrl: use dev_info instead of pr_info in pinctrl_select_state_locked

And remove superfluous brackets.

---
If it's not too late, it can be squashed with commit:
3102a76cfbf9ac4ae0cf54c7452f7ba4292a4760
"pinctrl: disable and free setting in select_state in case of error"

 drivers/pinctrl/core.c |    5 ++---
 1 files changed, 2 insertions(+), 3 deletions(-)
drivers/pinctrl/core.c +2 -3
index 9b505acc..b98a275 100644
@@ -912,9 +912,8 @@ static int pinctrl_select_state_locked(struct pinctrl *p,
 			break;
 		}
 
-		if (ret < 0) {
+		if (ret < 0)
 			goto unapply_new_state;
-		}
 	}
 
 	p->state = state;
@@ -922,7 +921,7 @@ static int pinctrl_select_state_locked(struct pinctrl *p,
 	return 0;
 
 unapply_new_state:
-	pr_info("Error applying setting, reverse things back\n");
+	dev_err(p->dev, "Error applying setting, reverse things back\n");
 
 	/*
 	 * If the loop stopped on the 1st entry, nothing has been enabled,
Signature · 1 line
1.7.2.5
#1 lore raw

PATCH 2/3 pinctrl: remove superfluous optimization in pinctrl_select_state_locked

As Stephen Warren suggested, checking first if the setting->node entry is the first in the list or not is superfluous, as it is checked again in the list_for_each_entry bellow.
So, remove it, the code will be simpler and lighter !

---
If can also be squashed whit commit:
3102a76cfbf9ac4ae0cf54c7452f7ba4292a4760
"pinctrl: disable and free setting in select_state in case of error" (But there will be a conflict on "if (old_state)" vs "FIXME comment")

 drivers/pinctrl/core.c |   10 +---------
 1 files changed, 1 insertions(+), 9 deletions(-)
drivers/pinctrl/core.c +1 -9
index b98a275..7c34937 100644
@@ -923,20 +923,12 @@ static int pinctrl_select_state_locked(struct pinctrl *p,
 unapply_new_state:
 	dev_err(p->dev, "Error applying setting, reverse things back\n");
 
-	/*
-	 * If the loop stopped on the 1st entry, nothing has been enabled,
-	 * so jump directly to the 2nd phase
-	 */
-	if (list_entry(&setting->node, typeof(*setting), node) ==
-	    list_first_entry(&state->settings, typeof(*setting), node))
-		goto reapply_old_state;
-
 	list_for_each_entry(setting2, &state->settings, node) {
 		if (&setting2->node == &setting->node)
 			break;
 		pinctrl_free_setting(true, setting2);
 	}
-reapply_old_state:
+
 	if (old_state) {
 		list_for_each_entry(setting, &old_state->settings, node) {
 			bool found = false;
Signature · 1 line
1.7.2.5
Reviewed-by on v1 2/3
#2 lore raw

Re: [PATCH 2/3] pinctrl: remove superfluous optimization in pinctrl_select_state_locked

Quoted 4 lines · Richard Genoud

As Stephen Warren suggested, checking first if the setting->node entry is the first in the list or not is superfluous, as it is checked again in the list_for_each_entry bellow.
So, remove it, the code will be simpler and lighter !

#3 lore raw

Re: [PATCH 2/3] pinctrl: remove superfluous optimization in pinctrl_select_state_locked

In reply to Richard Genoud
Quoted 11 lines · Richard Genoud

As Stephen Warren suggested, checking first if the setting->node entry is the first in the list or not is superfluous, as it is checked again in the list_for_each_entry bellow.
So, remove it, the code will be simpler and lighter !

---
If can also be squashed whit commit:
3102a76cfbf9ac4ae0cf54c7452f7ba4292a4760
"pinctrl: disable and free setting in select_state in case of error" (But there will be a conflict on "if (old_state)" vs "FIXME comment")

Applied with Stephen's Reviewed tag.

Thanks,
Linus Walleij

#4 lore raw

PATCH 3/3 pinctrl: pinctrl_select_state: set the old_state back on error

In reply to Richard Genoud

In unapply_new_state, the old state is re-applied, but p->state is not set back as it should.

---
This one can be squshed with commit:
50cf7c8ab324de348990bb028ad9ed10872d527a
pinctrl: re-enable old state in case of error in pinctrl_select_state If needed.

 drivers/pinctrl/core.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)
drivers/pinctrl/core.c +2 -0
index 7c34937..714cf74 100644
@@ -947,6 +947,8 @@ unapply_new_state:
 				pinmux_enable_setting(setting);
 		}
 	}
+
+	p->state = old_state;
 	return ret;
 }
 
Signature · 1 line
1.7.2.5
Reviewed-by on v1 3/3
#5 lore raw

Re: [PATCH 3/3] pinctrl: pinctrl_select_state: set the old_state back on error

Quoted 2 lines · Richard Genoud

In unapply_new_state, the old state is re-applied, but p->state is not set back as it should.

#6 lore raw

Re: [PATCH 3/3] pinctrl: pinctrl_select_state: set the old_state back on error

In reply to Richard Genoud
Quoted 9 lines · Richard Genoud

In unapply_new_state, the old state is re-applied, but p->state is not set back as it should.

---
This one can be squshed with commit:
50cf7c8ab324de348990bb028ad9ed10872d527a
pinctrl: re-enable old state in case of error in pinctrl_select_state If needed.

Applied with Stepgen's Review-tag.

Thanks,
Linus Walleij

#8 lore raw
In reply to Richard Genoud
Quoted 7 lines · Richard Genoud

And remove superfluous brackets.

---
If it's not too late, it can be squashed with commit:
3102a76cfbf9ac4ae0cf54c7452f7ba4292a4760
"pinctrl: disable and free setting in select_state in case of error"

Bah, applied as-is with Stephen's Review-tag.

Yours,
Linus Walleij