Re: [PATCH v2] mm: Add kvfree_sensitive() for freeing sensitive data objects

[Waiman Long]

On 4/6/20 4:00 PM, Matthew Wilcox wrote:
> On Mon, Apr 06, 2020 at 02:58:27PM -0400, Waiman Long wrote:
>> +/**
>> + * kvfree_sensitive - free a data object containing sensitive information
>> + * @addr - address of the data object to be freed
>> + * @len  - length of the data object
> Did you try building this with W=1?  I believe this is incorrect kerneldoc.
> It should be @addr: and @len:
>
> Also, it reads better in the htmldocs if you capitalise the first letter
> of each sentence and finish with a full stop.
>
You are right. I use the wrong delimiter here. I just send out a v3
patch to fix that. Thanks for noticing it.


>> @@ -914,7 +911,7 @@ long keyctl_read_key(key_serial_t keyid, char __user *buffer, size_t buflen)
>>  		 */
>>  		if (ret > key_data_len) {
>>  			if (unlikely(key_data))
>> -				__kvzfree(key_data, key_data_len);
>> +				kvfree_sensitive(key_data, key_data_len);
> I'd drop the test of key_data here.
>
I would like to keep the unlikely tag here to emphaize the fact that
this path should not be taken. I have comments up a few lines to talk
about it, though it didn't show up in the diff.

Cheers,
Longman