Re: [PATCH v3 1/4] HID: hid-picolcd*: Convert sprintf/scnprintf to sysfs_emit/sysfs_emit_at

[Christophe JAILLET]

Le 18/03/2024 à 02:28, Li Zhijian a écrit :
> Per filesystems/sysfs.rst, show() should only use sysfs_emit()
> or sysfs_emit_at() when formatting the value to be returned to user space.
>
> coccinelle complains that there are still a couple of functions that use
> snprintf(). Convert them to sysfs_emit().
>
> scnprintf() will be converted as weel if they have.
>
> Generally, this patch is generated by
> make coccicheck M=<path/to/file> MODE=patch \
> COCCI=scripts/coccinelle/api/device_attr_show.cocci
>
> No functional change intended
>
> CC: "Bruno Prémont" <bonbons@xxxxxxxxxxxxxxxxx>
> CC: Jiri Kosina <jikos@xxxxxxxxxx>
> CC: Benjamin Tissoires <benjamin.tissoires@xxxxxxxxxx>
> CC: linux-input@xxxxxxxxxxxxxxx
> Signed-off-by: Li Zhijian <lizhijian@xxxxxxxxxxx>
> ---
> V3:
>     Covert more file(drivers/hid/hid-picolcd_fb.c) as suggested by Bruno
>
> This is a part of the work "Fix coccicheck device_attr_show warnings"[1]
> Split them per subsystem so that the maintainer can review it easily
> [1] https://lore.kernel.org/lkml/20240116041129.3937800-1-lizhijian@xxxxxxxxxxx/
> Signed-off-by: Li Zhijian <lizhijian@xxxxxxxxxxx>
> ---
>   drivers/hid/hid-picolcd_core.c | 6 +++---
>   drivers/hid/hid-picolcd_fb.c   | 4 ++--
>   2 files changed, 5 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/hid/hid-picolcd_core.c b/drivers/hid/hid-picolcd_core.c
> index bbda231a7ce3..fa46fb6eab3f 100644
> --- a/drivers/hid/hid-picolcd_core.c
> +++ b/drivers/hid/hid-picolcd_core.c
> @@ -256,9 +256,9 @@ static ssize_t picolcd_operation_mode_show(struct device *dev,
>   	struct picolcd_data *data = dev_get_drvdata(dev);
>   
>   	if (data->status & PICOLCD_BOOTLOADER)
> -		return snprintf(buf, PAGE_SIZE, "[bootloader] lcd\n");
> +		return sysfs_emit(buf, "[bootloader] lcd\n");
>   	else
> -		return snprintf(buf, PAGE_SIZE, "bootloader [lcd]\n");
> +		return sysfs_emit(buf, "bootloader [lcd]\n");
>   }
>   
>   static ssize_t picolcd_operation_mode_store(struct device *dev,
> @@ -301,7 +301,7 @@ static ssize_t picolcd_operation_mode_delay_show(struct device *dev,
>   {
>   	struct picolcd_data *data = dev_get_drvdata(dev);
>   
> -	return snprintf(buf, PAGE_SIZE, "%hu\n", data->opmode_delay);
> +	return sysfs_emit(buf, "%hu\n", data->opmode_delay);
>   }
>   
>   static ssize_t picolcd_operation_mode_delay_store(struct device *dev,
> diff --git a/drivers/hid/hid-picolcd_fb.c b/drivers/hid/hid-picolcd_fb.c
> index d7dddd99d325..369c78d70e66 100644
> --- a/drivers/hid/hid-picolcd_fb.c
> +++ b/drivers/hid/hid-picolcd_fb.c
> @@ -424,9 +424,9 @@ static ssize_t picolcd_fb_update_rate_show(struct device *dev,

Hi,

just above we have:
	for (i = 1; i <= PICOLCDFB_UPDATE_RATE_LIMIT; i++)

>   		if (ret >= PAGE_SIZE)
>   			break;

and PICOLCDFB_UPDATE_RATE_LIMIT is 10, so it is not possible to have ret 
>= PAGE_SIZE. Should it happen, sysfs_emit_at() handles it.
So, this test can also be removed, IMHO.

CJ

>   		else if (i == fb_update_rate)
> -			ret += scnprintf(buf+ret, PAGE_SIZE-ret, "[%u] ", i);
> +			ret += sysfs_emit_at(buf, ret, "[%u] ", i);
>   		else
> -			ret += scnprintf(buf+ret, PAGE_SIZE-ret, "%u ", i);
> +			ret += sysfs_emit_at(buf, ret, "%u ", i);
>   	if (ret > 0)
>   		buf[min(ret, (size_t)PAGE_SIZE)-1] = '\n';
>   	return ret;