[dm-devel] [PATCH v2] dm raid: fix parse_raid_params() variable range issue

[Heinz Mauelshagen]

On 03/22/2018 08:41 PM, Mike Snitzer wrote:
> On Thu, Mar 22 2018 at  1:21pm -0400,
> Heinz Mauelshagen <heinzm at redhat.com> wrote:
>
>> This v2 addresses Mikulas' point about the variable range and folds in
>> "[PATCH] dm raid: use __within_range() more in parse_raid_params()":
>>
>> parse_raid_parames() compared variable "int value" with
>> INT_MAX to prevent overflow of mddev variables set.
>>
>> Change type to "long long value".
> Can you elaborate on the risk/issue that is being fixed here?

Fix addresses a coverity finding supporting the full,
positive range of the "struct mddev" int members
set here.  I.e. the "int" cast is compared with INT_MAX.

>
> User specifying a value that overflows an int?

No, kstroint() catches that.

>
> (also: see below for inline comment about last hunk)

See below...

>> Whilst on it, use __within_range() throughout and
>> add a sync min/max rate check.
>>
>> Signed-off-by: Heinz Mauelshagen <heinzm at redhat.com>
>> ---
>>   drivers/md/dm-raid.c | 16 +++++++++++-----
>>   1 file changed, 11 insertions(+), 5 deletions(-)
>>
>> diff --git a/drivers/md/dm-raid.c b/drivers/md/dm-raid.c
>> index c1d1034ff7b7..c0e3d2aa9346 100644
>> --- a/drivers/md/dm-raid.c
>> +++ b/drivers/md/dm-raid.c
>> @@ -1141,7 +1141,7 @@ static int validate_raid_redundancy(struct raid_set *rs)
>>   static int parse_raid_params(struct raid_set *rs, struct dm_arg_set *as,
>>   			     unsigned int num_raid_params)
>>   {
>> -	int value, raid10_format = ALGORITHM_RAID10_DEFAULT;
>> +	long long value, raid10_format = ALGORITHM_RAID10_DEFAULT;
>>   	unsigned int raid10_copies = 2;
>>   	unsigned int i, write_mostly = 0;
>>   	unsigned int region_size = 0;
>> @@ -1153,7 +1153,7 @@ static int parse_raid_params(struct raid_set *rs, struct dm_arg_set *as,
>>   	arg = dm_shift_arg(as);
>>   	num_raid_params--; /* Account for chunk_size argument */
>>   
>> -	if (kstrtoint(arg, 10, &value) < 0) {
>> +	if (kstrtoll(arg, 10, &value) < 0) {
>>   		rs->ti->error = "Bad numerical argument given for chunk_size";
>>   		return -EINVAL;
>>   	}
>> @@ -1315,7 +1315,7 @@ static int parse_raid_params(struct raid_set *rs, struct dm_arg_set *as,
>>   		/*
>>   		 * Parameters with number values from here on.
>>   		 */
>> -		if (kstrtoint(arg, 10, &value) < 0) {
>> +		if (kstrtoll(arg, 10, &value) < 0) {
>>   			rs->ti->error = "Bad numerical argument given in raid params";
>>   			return -EINVAL;
>>   		}
>> @@ -1430,7 +1430,7 @@ static int parse_raid_params(struct raid_set *rs, struct dm_arg_set *as,
>>   				rs->ti->error = "Only one min_recovery_rate argument pair allowed";
>>   				return -EINVAL;
>>   			}
>> -			if (value > INT_MAX) {
>> +			if (!__within_range(value, 0, INT_MAX)) {
>>   				rs->ti->error = "min_recovery_rate out of range";
>>   				return -EINVAL;
>>   			}
>> @@ -1440,7 +1440,7 @@ static int parse_raid_params(struct raid_set *rs, struct dm_arg_set *as,
>>   				rs->ti->error = "Only one max_recovery_rate argument pair allowed";
>>   				return -EINVAL;
>>   			}
>> -			if (value > INT_MAX) {
>> +			if (!__within_range(value, 0, INT_MAX)) {
>>   				rs->ti->error = "max_recovery_rate out of range";
>>   				return -EINVAL;
>>   			}
>> @@ -1472,6 +1472,12 @@ static int parse_raid_params(struct raid_set *rs, struct dm_arg_set *as,
>>   		}
>>   	}
>>   
>> +	if (rs->md.sync_speed_max &&
>> +	    rs->md.sync_speed_max < rs->md.sync_speed_min) {
>> +		rs->ti->error = "sync speed max smaller than min";
>> +		return -EINVAL;
>> +	}
>> +
>>   	if (test_bit(__CTR_FLAG_SYNC, &rs->ctr_flags) &&
>>   	    test_bit(__CTR_FLAG_NOSYNC, &rs->ctr_flags)) {
>>   		rs->ti->error = "sync and nosync are mutually exclusive";
>> -- 
>> 2.14.3
>>
> Isn't this last hunk unrelated?

No, once using __within_range to ensure positive values
for sync min/max, this hunk ensures that those are sane
if both are set.

Heinz

>
> --
> dm-devel mailing list
> dm-devel at redhat.com
> https://www.redhat.com/mailman/listinfo/dm-devel