Re: [PATCH v2 08/11] crypto: testmgr - check err on akcipher maxsize

[Tudor Ambarus]

On 23.05.2017 07:08, Herbert Xu wrote:

On Wed, May 17, 2017 at 06:00:35PM +0300, Tudor Ambarus wrote:

crypto_akcipher_maxsize() returns minimum length for output buffer
or error code if key hasn't been set.

Signed-off-by: Tudor Ambarus 
---
  crypto/testmgr.c | 5 +
  1 file changed, 5 insertions(+)

diff --git a/crypto/testmgr.c b/crypto/testmgr.c
index 6f5f3ed..87a4abd 100644
--- a/crypto/testmgr.c
+++ b/crypto/testmgr.c
@@ -2150,6 +2150,11 @@ static int test_akcipher_one(struct crypto_akcipher *tfm,
  
  	err = -ENOMEM;

out_len_max = crypto_akcipher_maxsize(tfm);


Can this call be reached without a setkey or with a failed setkey?


As of now, this call is reached only after a successful setkey.
If some user call it before setkey we will end up in a NULL dereference.
I tend to keep the error checking.

Thanks,
ta

Re: [PATCH v2 08/11] crypto: testmgr - check err on akcipher maxsize

[Herbert Xu]

On Wed, May 17, 2017 at 06:00:35PM +0300, Tudor Ambarus wrote:
> crypto_akcipher_maxsize() returns minimum length for output buffer
> or error code if key hasn't been set.
> 
> Signed-off-by: Tudor Ambarus 
> ---
>  crypto/testmgr.c | 5 +
>  1 file changed, 5 insertions(+)
> 
> diff --git a/crypto/testmgr.c b/crypto/testmgr.c
> index 6f5f3ed..87a4abd 100644
> --- a/crypto/testmgr.c
> +++ b/crypto/testmgr.c
> @@ -2150,6 +2150,11 @@ static int test_akcipher_one(struct crypto_akcipher 
> *tfm,
>  
>   err = -ENOMEM;
>   out_len_max = crypto_akcipher_maxsize(tfm);

Can this call be reached without a setkey or with a failed setkey?
If not we should not check for errors.

Cheers,
-- 
Email: Herbert Xu 
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt