[PATCH] mei: remove dev_err message on an unsupported ioctl

[PATCH] mei: remove dev_err message on an unsupported ioctl

[Colin King]

From: Colin Ian King <colin.king@canonical.com>

Currently the driver spams the kernel log on unsupported ioctls which is
unnecessary as the ioctl returns -ENOIOCTLCMD to indicate this anyway.
I suspect this was originally for debugging purposes but it really is not
required so remove it.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Cc: stable <stable@vger.kernel.org>
---
 drivers/misc/mei/main.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/misc/mei/main.c b/drivers/misc/mei/main.c
index 758dc73602d5..7bb013644aeb 100644
--- a/drivers/misc/mei/main.c
+++ b/drivers/misc/mei/main.c
@@ -507,7 +507,6 @@ static long mei_ioctl(struct file *file, unsigned int cmd, unsigned long data)
 		break;
 
 	default:
-		dev_err(dev->dev, ": unsupported ioctl %d.\n", cmd);
 		rets = -ENOIOCTLCMD;
 	}
 
-- 
2.15.1

RE: [PATCH] mei: remove dev_err message on an unsupported ioctl

[Winkler, Tomas]

> 
> From: Colin Ian King <colin.king@canonical.com>
> 
> Currently the driver spams the kernel log on unsupported ioctls which is
> unnecessary as the ioctl returns -ENOIOCTLCMD to indicate this anyway.
> I suspect this was originally for debugging purposes but it really is not
> required so remove it.
>

This is rather strange as most of the legimit applications are using one IOCTL from kernel 3.0
Do you have any reference where this call is originated from, frankly this is the first time I got such complain.

In any case it would be maybe better to change it a warn once call.
 
Thanks
Tomas

> Signed-off-by: Colin Ian King <colin.king@canonical.com>
> ---
>  drivers/misc/mei/main.c | 1 -
>  1 file changed, 1 deletion(-)
> 
> diff --git a/drivers/misc/mei/main.c b/drivers/misc/mei/main.c index
> 758dc73602d5..7bb013644aeb 100644
> --- a/drivers/misc/mei/main.c
> +++ b/drivers/misc/mei/main.c
> @@ -507,7 +507,6 @@ static long mei_ioctl(struct file *file, unsigned int
> cmd, unsigned long data)
>  		break;
> 
>  	default:
> -		dev_err(dev->dev, ": unsupported ioctl %d.\n", cmd);
>  		rets = -ENOIOCTLCMD;
>  	}
> 
> --
> 2.15.1

Re: [PATCH] mei: remove dev_err message on an unsupported ioctl

[Colin Ian King]

On 27/02/18 17:26, Winkler, Tomas wrote:
>>
>> From: Colin Ian King <colin.king@canonical.com>
>>
>> Currently the driver spams the kernel log on unsupported ioctls which is
>> unnecessary as the ioctl returns -ENOIOCTLCMD to indicate this anyway.
>> I suspect this was originally for debugging purposes but it really is not
>> required so remove it.
>>
> 
> This is rather strange as most of the legimit applications are using one IOCTL from kernel 3.0
> Do you have any reference where this call is originated from, frankly this is the first time I got such complain.

I hit this while stressing the kernel with my stress-ng kernel stressing
tool; this was not using the legitimate ioctls (to see if things
explode) and I got a rather full kernel log. So it was being abused
somewhat ;-)

> 
> In any case it would be maybe better to change it a warn once call.

Probably so, I can fix that up if that is preferred.

Colin

>  
> Thanks
> Tomas
> 
>> Signed-off-by: Colin Ian King <colin.king@canonical.com>
>> ---
>>  drivers/misc/mei/main.c | 1 -
>>  1 file changed, 1 deletion(-)
>>
>> diff --git a/drivers/misc/mei/main.c b/drivers/misc/mei/main.c index
>> 758dc73602d5..7bb013644aeb 100644
>> --- a/drivers/misc/mei/main.c
>> +++ b/drivers/misc/mei/main.c
>> @@ -507,7 +507,6 @@ static long mei_ioctl(struct file *file, unsigned int
>> cmd, unsigned long data)
>>  		break;
>>
>>  	default:
>> -		dev_err(dev->dev, ": unsupported ioctl %d.\n", cmd);
>>  		rets = -ENOIOCTLCMD;
>>  	}
>>
>> --
>> 2.15.1
> 

RE: [PATCH] mei: remove dev_err message on an unsupported ioctl

[Winkler, Tomas]

> On 27/02/18 17:26, Winkler, Tomas wrote:
> >>
> >> From: Colin Ian King <colin.king@canonical.com>
> >>
> >> Currently the driver spams the kernel log on unsupported ioctls which
> >> is unnecessary as the ioctl returns -ENOIOCTLCMD to indicate this
> anyway.
> >> I suspect this was originally for debugging purposes but it really is
> >> not required so remove it.
> >>
> >
> > This is rather strange as most of the legimit applications are using
> > one IOCTL from kernel 3.0 Do you have any reference where this call is
> originated from, frankly this is the first time I got such complain.
> 
> I hit this while stressing the kernel with my stress-ng kernel stressing tool;
> this was not using the legitimate ioctls (to see if things
> explode) and I got a rather full kernel log. So it was being abused somewhat
> ;-)

Okay, thanks for the info.

Re: [PATCH] mei: remove dev_err message on an unsupported ioctl

[Greg Kroah-Hartman]

On Tue, Feb 27, 2018 at 05:26:22PM +0000, Winkler, Tomas wrote:
> > 
> > From: Colin Ian King <colin.king@canonical.com>
> > 
> > Currently the driver spams the kernel log on unsupported ioctls which is
> > unnecessary as the ioctl returns -ENOIOCTLCMD to indicate this anyway.
> > I suspect this was originally for debugging purposes but it really is not
> > required so remove it.
> >
> 
> This is rather strange as most of the legimit applications are using one IOCTL from kernel 3.0
> Do you have any reference where this call is originated from, frankly this is the first time I got such complain.
> 
> In any case it would be maybe better to change it a warn once call.

Just drop the message, it should not be this easy for userspace to spam
the kernel log.  The patch looks fine to me.

thanks,

greg k-h

RE: [PATCH] mei: remove dev_err message on an unsupported ioctl

[Winkler, Tomas]

> 
> On Tue, Feb 27, 2018 at 05:26:22PM +0000, Winkler, Tomas wrote:
> > >
> > > From: Colin Ian King <colin.king@canonical.com>
> > >
> > > Currently the driver spams the kernel log on unsupported ioctls
> > > which is unnecessary as the ioctl returns -ENOIOCTLCMD to indicate this
> anyway.
> > > I suspect this was originally for debugging purposes but it really
> > > is not required so remove it.
> > >
> >
> > This is rather strange as most of the legimit applications are using
> > one IOCTL from kernel 3.0 Do you have any reference where this call is
> originated from, frankly this is the first time I got such complain.
> >
> > In any case it would be maybe better to change it a warn once call.
> 
> Just drop the message, it should not be this easy for userspace to spam the
> kernel log.  The patch looks fine to me.
> 
Fair enough, add my ack too.

Still I'm curious if this comes from the field.

Thanks
Tomas

Re: [PATCH] mei: remove dev_err message on an unsupported ioctl

[Greg Kroah-Hartman]

On Tue, Feb 27, 2018 at 09:44:41PM +0000, Winkler, Tomas wrote:
> > 
> > On Tue, Feb 27, 2018 at 05:26:22PM +0000, Winkler, Tomas wrote:
> > > >
> > > > From: Colin Ian King <colin.king@canonical.com>
> > > >
> > > > Currently the driver spams the kernel log on unsupported ioctls
> > > > which is unnecessary as the ioctl returns -ENOIOCTLCMD to indicate this
> > anyway.
> > > > I suspect this was originally for debugging purposes but it really
> > > > is not required so remove it.
> > > >
> > >
> > > This is rather strange as most of the legimit applications are using
> > > one IOCTL from kernel 3.0 Do you have any reference where this call is
> > originated from, frankly this is the first time I got such complain.
> > >
> > > In any case it would be maybe better to change it a warn once call.
> > 
> > Just drop the message, it should not be this easy for userspace to spam the
> > kernel log.  The patch looks fine to me.
> > 
> Fair enough, add my ack too.
> 
> Still I'm curious if this comes from the field.

fuzzers are "in the field" :)

RE: [PATCH] mei: remove dev_err message on an unsupported ioctl

[Winkler, Tomas]

> 
> On Tue, Feb 27, 2018 at 09:44:41PM +0000, Winkler, Tomas wrote:
> > >
> > > On Tue, Feb 27, 2018 at 05:26:22PM +0000, Winkler, Tomas wrote:
> > > > >
> > > > > From: Colin Ian King <colin.king@canonical.com>
> > > > >
> > > > > Currently the driver spams the kernel log on unsupported ioctls
> > > > > which is unnecessary as the ioctl returns -ENOIOCTLCMD to
> > > > > indicate this
> > > anyway.
> > > > > I suspect this was originally for debugging purposes but it
> > > > > really is not required so remove it.
> > > > >
> > > >
> > > > This is rather strange as most of the legimit applications are
> > > > using one IOCTL from kernel 3.0 Do you have any reference where
> > > > this call is
> > > originated from, frankly this is the first time I got such complain.
> > > >
> > > > In any case it would be maybe better to change it a warn once call.
> > >
> > > Just drop the message, it should not be this easy for userspace to
> > > spam the kernel log.  The patch looks fine to me.
> > >
> > Fair enough, add my ack too.
> >
> > Still I'm curious if this comes from the field.
> 
> fuzzers are "in the field" :)

It's definitely fuzzers season.

Tomas