drivers/pinctrl/core.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
When applying a pinctrl state, -EPROBE_DEFER may be returned if
dependencies are not ready and the consumer will retry probing.
This is normal probe ordering behaviour and not a real error.
However, pinctrl core currently logs:
"Error applying setting, reverse things back"
even when the return value is -EPROBE_DEFER, resulting in noisy
boot-time error messages.
Replace dev_err() with dev_err_probe() to handle -EPROBE_DEFER
consistently and suppress error logging for deferred probes.
No functional change intended.
Signed-off-by: Michal Piekos <michal.piekos@mmpsystems.pl>
---
This series contains a single patch to suppress spurious
boot-time error messages when applying pinctrl states during
normal probe deferral.
Currently, pinctrl core logs an error even when
-EPROBE_DEFER is returned, which is expected probe-ordering
behaviour.
Switch to dev_err_probe() so deferred probes do not emit
misleading error logs.
No functional change intended.
---
drivers/pinctrl/core.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/pinctrl/core.c b/drivers/pinctrl/core.c
index b5e97689589f..2edc9bdad183 100644
--- a/drivers/pinctrl/core.c
+++ b/drivers/pinctrl/core.c
@@ -1350,7 +1350,8 @@ static int pinctrl_commit_state(struct pinctrl *p, struct pinctrl_state *state)
goto restore_old_state;
unapply_new_state:
- dev_err(p->dev, "Error applying setting, reverse things back\n");
+ dev_err_probe(p->dev, ret,
+ "Error applying setting, reverse things back\n");
/*
* All we can do here is pinmux_disable_setting.
---
base-commit: eb71ab2bf72260054677e348498ba995a057c463
change-id: 20260301-pinctrl-dev-err-probe-00d80f8202a6
Best regards,
--
Michal Piekos <michal.piekos@mmpsystems.pl>
On Sun, Mar 1, 2026 at 5:46 PM Michal Piekos <michal.piekos@mmpsystems.pl> wrote: > When applying a pinctrl state, -EPROBE_DEFER may be returned if > dependencies are not ready and the consumer will retry probing. > > This is normal probe ordering behaviour and not a real error. > However, pinctrl core currently logs: > > "Error applying setting, reverse things back" > > even when the return value is -EPROBE_DEFER, resulting in noisy > boot-time error messages. > > Replace dev_err() with dev_err_probe() to handle -EPROBE_DEFER > consistently and suppress error logging for deferred probes. > > No functional change intended. > > Signed-off-by: Michal Piekos <michal.piekos@mmpsystems.pl> > --- > This series contains a single patch to suppress spurious > boot-time error messages when applying pinctrl states during > normal probe deferral. > > Currently, pinctrl core logs an error even when > -EPROBE_DEFER is returned, which is expected probe-ordering > behaviour. > > Switch to dev_err_probe() so deferred probes do not emit > misleading error logs. Patch applied. Mysterious overly-verbose commit log have traces of AI spice. It is mandatory to mention in the commit message if an AI tool was used to generate the patch, was it? In that case tell me so I can augment the commit log. Yours, Linus Walleij
On Tue, Mar 03, 2026 at 08:32:39AM +0100, Linus Walleij wrote: > On Sun, Mar 1, 2026 at 5:46 PM Michal Piekos > <michal.piekos@mmpsystems.pl> wrote: > > > When applying a pinctrl state, -EPROBE_DEFER may be returned if > > dependencies are not ready and the consumer will retry probing. > > > > This is normal probe ordering behaviour and not a real error. > > However, pinctrl core currently logs: > > > > "Error applying setting, reverse things back" > > > > even when the return value is -EPROBE_DEFER, resulting in noisy > > boot-time error messages. > > > > Replace dev_err() with dev_err_probe() to handle -EPROBE_DEFER > > consistently and suppress error logging for deferred probes. > > > > No functional change intended. > > > > Signed-off-by: Michal Piekos <michal.piekos@mmpsystems.pl> > > --- > > This series contains a single patch to suppress spurious > > boot-time error messages when applying pinctrl states during > > normal probe deferral. > > > > Currently, pinctrl core logs an error even when > > -EPROBE_DEFER is returned, which is expected probe-ordering > > behaviour. > > > > Switch to dev_err_probe() so deferred probes do not emit > > misleading error logs. > > Patch applied. > > Mysterious overly-verbose commit log have traces of AI > spice. > > It is mandatory to mention in the commit message if an AI > tool was used to generate the patch, was it? In that > case tell me so I can augment the commit log. > > Yours, > Linus Walleij AI was used in patch message preparation. AI was not used for patch code. Thank you Michal Piekos
© 2016 - 2026 Red Hat, Inc.