[PATCH][next] dm ioctl: rate limit a couple of ioctl based error messages

[PATCH][next] dm ioctl: rate limit a couple of ioctl based error messages

[Colin Ian King]

It is possible to spam the kernel log with a misbehaving user process that
is passing incorrect dm ioctls to /dev/mapper/control. Use a rate limit
on these error messages to reduce the noise.

Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
---
 drivers/md/dm-ioctl.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
index f299ff393a6a..d42eac944eb5 100644
--- a/drivers/md/dm-ioctl.c
+++ b/drivers/md/dm-ioctl.c
@@ -1912,7 +1912,7 @@ static int check_version(unsigned int cmd, struct dm_ioctl __user *user,
 
 	if ((kernel_params->version[0] != DM_VERSION_MAJOR) ||
 	    (kernel_params->version[1] > DM_VERSION_MINOR)) {
-		DMERR("ioctl interface mismatch: kernel(%u.%u.%u), user(%u.%u.%u), cmd(%d)",
+		DMERR_LIMIT("ioctl interface mismatch: kernel(%u.%u.%u), user(%u.%u.%u), cmd(%d)",
 		      DM_VERSION_MAJOR, DM_VERSION_MINOR,
 		      DM_VERSION_PATCHLEVEL,
 		      kernel_params->version[0],
@@ -1961,7 +1961,7 @@ static int copy_params(struct dm_ioctl __user *user, struct dm_ioctl *param_kern
 
 	if (unlikely(param_kernel->data_size < minimum_data_size) ||
 	    unlikely(param_kernel->data_size > DM_MAX_TARGETS * DM_MAX_TARGET_PARAMS)) {
-		DMERR("Invalid data size in the ioctl structure: %u",
+		DMERR_LIMIT("Invalid data size in the ioctl structure: %u",
 		      param_kernel->data_size);
 		return -EINVAL;
 	}
-- 
2.39.5

Re: [PATCH][next] dm ioctl: rate limit a couple of ioctl based error messages

[Mikulas Patocka]

Hi

I'd like to ask - have you experienced these errors? What was the process 
that triggered them? Was it some robot that tests syscalls with random 
parameters?

Mikulas


On Thu, 24 Oct 2024, Colin Ian King wrote:

> It is possible to spam the kernel log with a misbehaving user process that
> is passing incorrect dm ioctls to /dev/mapper/control. Use a rate limit
> on these error messages to reduce the noise.
> 
> Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
> ---
>  drivers/md/dm-ioctl.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
> index f299ff393a6a..d42eac944eb5 100644
> --- a/drivers/md/dm-ioctl.c
> +++ b/drivers/md/dm-ioctl.c
> @@ -1912,7 +1912,7 @@ static int check_version(unsigned int cmd, struct dm_ioctl __user *user,
>  
>  	if ((kernel_params->version[0] != DM_VERSION_MAJOR) ||
>  	    (kernel_params->version[1] > DM_VERSION_MINOR)) {
> -		DMERR("ioctl interface mismatch: kernel(%u.%u.%u), user(%u.%u.%u), cmd(%d)",
> +		DMERR_LIMIT("ioctl interface mismatch: kernel(%u.%u.%u), user(%u.%u.%u), cmd(%d)",
>  		      DM_VERSION_MAJOR, DM_VERSION_MINOR,
>  		      DM_VERSION_PATCHLEVEL,
>  		      kernel_params->version[0],
> @@ -1961,7 +1961,7 @@ static int copy_params(struct dm_ioctl __user *user, struct dm_ioctl *param_kern
>  
>  	if (unlikely(param_kernel->data_size < minimum_data_size) ||
>  	    unlikely(param_kernel->data_size > DM_MAX_TARGETS * DM_MAX_TARGET_PARAMS)) {
> -		DMERR("Invalid data size in the ioctl structure: %u",
> +		DMERR_LIMIT("Invalid data size in the ioctl structure: %u",
>  		      param_kernel->data_size);
>  		return -EINVAL;
>  	}
> -- 
> 2.39.5
> 

Re: [PATCH][next] dm ioctl: rate limit a couple of ioctl based error messages

[Colin King (gmail)]

On 24/10/2024 15:01, Mikulas Patocka wrote:
> Hi
> 
> I'd like to ask - have you experienced these errors? What was the process
> that triggered them? Was it some robot that tests syscalls with random
> parameters?

I hit this when running stress-ng's dev test, this exercises various 
ioctls for race conditions.  It's not a major issue, just a corner case 
found when I was running some stress tests.

Colin

> 
> Mikulas
> 
> 
> On Thu, 24 Oct 2024, Colin Ian King wrote:
> 
>> It is possible to spam the kernel log with a misbehaving user process that
>> is passing incorrect dm ioctls to /dev/mapper/control. Use a rate limit
>> on these error messages to reduce the noise.
>>
>> Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
>> ---
>>   drivers/md/dm-ioctl.c | 4 ++--
>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
>> index f299ff393a6a..d42eac944eb5 100644
>> --- a/drivers/md/dm-ioctl.c
>> +++ b/drivers/md/dm-ioctl.c
>> @@ -1912,7 +1912,7 @@ static int check_version(unsigned int cmd, struct dm_ioctl __user *user,
>>   
>>   	if ((kernel_params->version[0] != DM_VERSION_MAJOR) ||
>>   	    (kernel_params->version[1] > DM_VERSION_MINOR)) {
>> -		DMERR("ioctl interface mismatch: kernel(%u.%u.%u), user(%u.%u.%u), cmd(%d)",
>> +		DMERR_LIMIT("ioctl interface mismatch: kernel(%u.%u.%u), user(%u.%u.%u), cmd(%d)",
>>   		      DM_VERSION_MAJOR, DM_VERSION_MINOR,
>>   		      DM_VERSION_PATCHLEVEL,
>>   		      kernel_params->version[0],
>> @@ -1961,7 +1961,7 @@ static int copy_params(struct dm_ioctl __user *user, struct dm_ioctl *param_kern
>>   
>>   	if (unlikely(param_kernel->data_size < minimum_data_size) ||
>>   	    unlikely(param_kernel->data_size > DM_MAX_TARGETS * DM_MAX_TARGET_PARAMS)) {
>> -		DMERR("Invalid data size in the ioctl structure: %u",
>> +		DMERR_LIMIT("Invalid data size in the ioctl structure: %u",
>>   		      param_kernel->data_size);
>>   		return -EINVAL;
>>   	}
>> -- 
>> 2.39.5
>>
> 

Re: [PATCH][next] dm ioctl: rate limit a couple of ioctl based error messages

[Mike Snitzer]

On Thu, Oct 24, 2024 at 12:04:31PM +0100, Colin Ian King wrote:
> It is possible to spam the kernel log with a misbehaving user process that
> is passing incorrect dm ioctls to /dev/mapper/control. Use a rate limit
> on these error messages to reduce the noise.
> 
> Signed-off-by: Colin Ian King <colin.i.king@gmail.com>

Mikulas, please update the header to include the context that
stress-ng is what Colin was using when he experienced excessive
warnings.

Thanks.

Acked-by: Mike Snitzer <snitzer@kernel.org>

Re: [PATCH][next] dm ioctl: rate limit a couple of ioctl based error messages

[Mikulas Patocka]

On Tue, 29 Oct 2024, Mike Snitzer wrote:

> On Thu, Oct 24, 2024 at 12:04:31PM +0100, Colin Ian King wrote:
> > It is possible to spam the kernel log with a misbehaving user process that
> > is passing incorrect dm ioctls to /dev/mapper/control. Use a rate limit
> > on these error messages to reduce the noise.
> > 
> > Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
> 
> Mikulas, please update the header to include the context that
> stress-ng is what Colin was using when he experienced excessive
> warnings.
> 
> Thanks.
> 
> Acked-by: Mike Snitzer <snitzer@kernel.org>

OK, done.

Mikulas