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 for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20211212212117.GA5737@mail.hallyn.com>
Date:   Sun, 12 Dec 2021 15:21:17 -0600
From:   "Serge E. Hallyn" <serge@...lyn.com>
To:     Jianglei Nie <niejianglei2021@....com>
Cc:     jejb@...ux.ibm.com, jarkko@...nel.org, zohar@...ux.ibm.com,
        dhowells@...hat.com, jmorris@...ei.org, serge@...lyn.com,
        linux-integrity@...r.kernel.org, keyrings@...r.kernel.org,
        linux-security-module@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] security:trusted_tpm2: Fix memory leak in
 tpm2_key_encode()

On Sun, Dec 12, 2021 at 09:54:03PM +0800, Jianglei Nie wrote:
> Line 36 (#1) allocates a memory chunk for scratch by kmalloc(), but
> it is never freed through the function, which will lead to a memory
> leak.
> 
> We should kfree() scratch before the function returns (#2, #3 and #4).
> 
> 31 static int tpm2_key_encode(struct trusted_key_payload *payload,
> 32			   struct trusted_key_options *options,
> 33			   u8 *src, u32 len)
> 34 {
> 36	u8 *scratch = kmalloc(SCRATCH_SIZE, GFP_KERNEL);
>       	// #1: kmalloc space
> 37	u8 *work = scratch, *work1;
> 50	if (!scratch)
> 51		return -ENOMEM;
> 
> 56	if (options->blobauth_len == 0) {
> 60		if (WARN(IS_ERR(w), "BUG: Boolean failed to encode"))
> 61			return PTR_ERR(w); // #2: missing kfree
> 63	}
> 
> 71	if (WARN(work - scratch + pub_len + priv_len + 14 > SCRATCH_SIZE,
> 72		 "BUG: scratch buffer is too small"))
> 73		return -EINVAL; // #3: missing kfree
> 
>   	// #4: missing kfree: scratch is never used afterwards.
> 82	if (WARN(IS_ERR(work1), "BUG: ASN.1 encoder failed"))
> 83		return PTR_ERR(work1);
> 
> 85	return work1 - payload->blob;
> 86 }
> 
> Signed-off-by: Jianglei Nie <niejianglei2021@....com>

I don't know that we need to keep the line by line recap in
the full git log, but it def looks correct:

Reviewed-by: Serge Hallyn <serge@...lyn.com>

thanks,
-serge

> ---
>  security/keys/trusted-keys/trusted_tpm2.c | 9 +++++++--
>  1 file changed, 7 insertions(+), 2 deletions(-)
> 
> diff --git a/security/keys/trusted-keys/trusted_tpm2.c b/security/keys/trusted-keys/trusted_tpm2.c
> index 0165da386289..3408a74c855f 100644
> --- a/security/keys/trusted-keys/trusted_tpm2.c
> +++ b/security/keys/trusted-keys/trusted_tpm2.c
> @@ -57,8 +57,10 @@ static int tpm2_key_encode(struct trusted_key_payload *payload,
>  		unsigned char bool[3], *w = bool;
>  		/* tag 0 is emptyAuth */
>  		w = asn1_encode_boolean(w, w + sizeof(bool), true);
> -		if (WARN(IS_ERR(w), "BUG: Boolean failed to encode"))
> +		if (WARN(IS_ERR(w), "BUG: Boolean failed to encode")) {
> +			kfree(scratch);
>  			return PTR_ERR(w);
> +		}
>  		work = asn1_encode_tag(work, end_work, 0, bool, w - bool);
>  	}
>  
> @@ -69,9 +71,12 @@ static int tpm2_key_encode(struct trusted_key_payload *payload,
>  	 * trigger, so if it does there's something nefarious going on
>  	 */
>  	if (WARN(work - scratch + pub_len + priv_len + 14 > SCRATCH_SIZE,
> -		 "BUG: scratch buffer is too small"))
> +		 "BUG: scratch buffer is too small")) {
> +		kfree(scratch);
>  		return -EINVAL;
> +	}
>  
> +	kfree(scratch);
>  	work = asn1_encode_integer(work, end_work, options->keyhandle);
>  	work = asn1_encode_octet_string(work, end_work, pub, pub_len);
>  	work = asn1_encode_octet_string(work, end_work, priv, priv_len);
> -- 
> 2.25.1

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ