linux-kernel - Re: [PATCH v2 1/2] KEYS: Don't write out to userspace while holding key semaphore

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 13 Mar 2020 09:29:47 -0400
From:   Waiman Long <longman@...hat.com>
To:     Jarkko Sakkinen <jarkko.sakkinen@...ux.intel.com>
Cc:     David Howells <dhowells@...hat.com>,
        James Morris <jmorris@...ei.org>,
        "Serge E. Hallyn" <serge@...lyn.com>,
        Mimi Zohar <zohar@...ux.ibm.com>, keyrings@...r.kernel.org,
        linux-kernel@...r.kernel.org,
        linux-security-module@...r.kernel.org,
        linux-integrity@...r.kernel.org,
        Sumit Garg <sumit.garg@...aro.org>,
        Jerry Snitselaar <jsnitsel@...hat.com>,
        Roberto Sassu <roberto.sassu@...wei.com>,
        Eric Biggers <ebiggers@...gle.com>,
        Chris von Recklinghausen <crecklin@...hat.com>
Subject: Re: [PATCH v2 1/2] KEYS: Don't write out to userspace while holding
 key semaphore

On 3/12/20 9:04 PM, Jarkko Sakkinen wrote:
> On Sun, Mar 08, 2020 at 01:04:09PM -0400, Waiman Long wrote:
>> +		/*
>> +		 * Read methods will just return the required length
>> +		 * without any copying if the provided length isn't big
>> +		 * enough.
>> +		 */
>> +		if ((ret > 0) && (ret <= buflen) && buffer &&
>> +		    copy_to_user(buffer, tmpbuf, ret))
>> +			ret = -EFAULT;
> Please, reorg and remove redundant parentheses:
>
> /*
>  * Read methods will just return the required length
>  * without any copying if the provided length isn't big
>  * enough.
>  */
> if (ret > 0 && ret <= buflen) {
> 	if (buffer && copy_to_user(buffer, tmpbuf, ret))
> 		ret = -EFAULT;
> }
>
> Now the comment is attached to the exact right thing. The previous
> organization is a pain to look at when backtracking commits for
> whatever reason in the future.
Yes, I can reorganize the code.
> I'm also wondering, would it be possible to rework the code in a way
> that you don't have check whether buffer is valid on a constant basis?

One way to do that is to extract the down_read/up_read block into a
helper function and then have 2 separate paths - one for length
retrieval and another one for reading the key. I think that will make
the code a bit easier easier to read.

Thanks,
Longman