Subject: [PATCH v2] gpio: just plain warning when nonexisting gpio requested

From: Enrico Weigelt <[email protected]>

When trying to export an nonexisting gpio ID, the kernel prints out a
big warning w/ stacktrace, sounding like a huge problem. In fact it's
a pretty normal situation, like file or device not found.

So, just print a more relaxed warning instead.

changes v2: drop defining pr_fmt()

Signed-off-by: Enrico Weigelt <[email protected]>
---
drivers/gpio/gpiolib.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
index 089ddcaa9bc6..6639e7b77d6c 100644
--- a/drivers/gpio/gpiolib.c
+++ b/drivers/gpio/gpiolib.c
@@ -1,4 +1,5 @@
// SPDX-License-Identifier: GPL-2.0
+
#include <linux/bitmap.h>
#include <linux/kernel.h>
#include <linux/module.h>
@@ -119,7 +120,7 @@ struct gpio_desc *gpio_to_desc(unsigned gpio)
spin_unlock_irqrestore(&gpio_lock, flags);

if (!gpio_is_valid(gpio))
- WARN(1, "invalid GPIO %d\n", gpio);
+ pr_warn("invalid GPIO %d\n", gpio);

return NULL;
}
--
2.11.0


2020-12-05 22:32:04

by Linus Walleij

[permalink] [raw]
Subject: Re: [PATCH v2] gpio: just plain warning when nonexisting gpio requested

On Wed, Dec 2, 2020 at 2:37 PM Enrico Weigelt, metux IT consult
<[email protected]> wrote:

> From: Enrico Weigelt <[email protected]>
>
> When trying to export an nonexisting gpio ID, the kernel prints out a
> big warning w/ stacktrace, sounding like a huge problem. In fact it's
> a pretty normal situation, like file or device not found.
>
> So, just print a more relaxed warning instead.
>
> changes v2: drop defining pr_fmt()
>
> Signed-off-by: Enrico Weigelt <[email protected]>

Hm hm. It would mainly happen if someone was using the sysfs
ABI right? I don't mind if sysfs feels scary to use.

But I see it can happen for other code using the old GPIO
symbols so OK I applied it.

Yours,
Linus Walleij