2017-11-22 08:08:34

by Corentin LABBE

Subject: [PATCH] crypto: arm64/aes - do not call crypto_unregister_skcipher twice on error

When a cipher fail to register in aes_init(), the error path go thought
aes_exit() then crypto_unregister_skciphers().
Since aes_exit calls also crypto_unregister_skcipher, this trigger a
refcount_t: underflow; use-after-free.

Signed-off-by: Corentin Labbe <[email protected]>
---
arch/arm64/crypto/aes-glue.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/arch/arm64/crypto/aes-glue.c b/arch/arm64/crypto/aes-glue.c
index 998ba519a026..9e42ec96243e 100644
--- a/arch/arm64/crypto/aes-glue.c
+++ b/arch/arm64/crypto/aes-glue.c
@@ -664,7 +664,10 @@ static int __init aes_init(void)
return 0;

unregister_simds:
- aes_exit();
+ for (i = 0; i < ARRAY_SIZE(aes_simd_algs); i++)
+ if (aes_simd_algs[i])
+ simd_skcipher_free(aes_simd_algs[i]);
+ crypto_unregister_shashes(mac_algs, ARRAY_SIZE(mac_algs));
unregister_ciphers:
crypto_unregister_skciphers(aes_algs, ARRAY_SIZE(aes_algs));
return err;
--
2.13.6

2017-11-22 08:55:14

by Ard Biesheuvel

Subject: Re: [PATCH] crypto: arm64/aes - do not call crypto_unregister_skcipher twice on error

Hello Corentin,

On 22 November 2017 at 08:08, Corentin Labbe <[email protected]> wrote:
> When a cipher fail

fails

> to register in aes_init(), the error path go thought

goes through

> aes_exit() then crypto_unregister_skciphers().
> Since aes_exit calls also crypto_unregister_skcipher, this trigger a

triggers

> refcount_t: underflow; use-after-free.
>
> Signed-off-by: Corentin Labbe <[email protected]>
> ---
> arch/arm64/crypto/aes-glue.c | 5 ++++-
> 1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/arch/arm64/crypto/aes-glue.c b/arch/arm64/crypto/aes-glue.c
> index 998ba519a026..9e42ec96243e 100644
> --- a/arch/arm64/crypto/aes-glue.c
> +++ b/arch/arm64/crypto/aes-glue.c
> @@ -664,7 +664,10 @@ static int __init aes_init(void)
> return 0;
>
> unregister_simds:
> - aes_exit();
> + for (i = 0; i < ARRAY_SIZE(aes_simd_algs); i++)
> + if (aes_simd_algs[i])
> + simd_skcipher_free(aes_simd_algs[i]);
> + crypto_unregister_shashes(mac_algs, ARRAY_SIZE(mac_algs));
> unregister_ciphers:
> crypto_unregister_skciphers(aes_algs, ARRAY_SIZE(aes_algs));
> return err;
> --
> 2.13.6
>
>


Would this also fix it?

diff --git a/arch/arm64/crypto/aes-glue.c b/arch/arm64/crypto/aes-glue.c
index 998ba519a026..2fa850e86aa8 100644
--- a/arch/arm64/crypto/aes-glue.c
+++ b/arch/arm64/crypto/aes-glue.c
@@ -665,6 +665,7 @@ static int __init aes_init(void)

unregister_simds:
aes_exit();
+ return err;
unregister_ciphers:
crypto_unregister_skciphers(aes_algs, ARRAY_SIZE(aes_algs));
return err;

2017-11-24 04:58:53

by Herbert Xu

Subject: Re: [PATCH] crypto: arm64/aes - do not call crypto_unregister_skcipher twice on error

On Wed, Nov 22, 2017 at 08:55:14AM +0000, Ard Biesheuvel wrote:
>
> Would this also fix it?

Looks good. Could you resubmit with a sign-off?

Thanks,
--
Email: Herbert Xu <[email protected]>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

2017-11-24 05:52:37

by Corentin LABBE

Subject: Re: [PATCH] crypto: arm64/aes - do not call crypto_unregister_skcipher twice on error

On Wed, Nov 22, 2017 at 08:55:14AM +0000, Ard Biesheuvel wrote:
> Hello Corentin,
>
> On 22 November 2017 at 08:08, Corentin Labbe <[email protected]> wrote:
> > When a cipher fail
>
> fails
>
> > to register in aes_init(), the error path go thought
>
> goes through
>
> > aes_exit() then crypto_unregister_skciphers().
> > Since aes_exit calls also crypto_unregister_skcipher, this trigger a
>
> triggers
>
> > refcount_t: underflow; use-after-free.
> >
> > Signed-off-by: Corentin Labbe <[email protected]>
> > ---
> > arch/arm64/crypto/aes-glue.c | 5 ++++-
> > 1 file changed, 4 insertions(+), 1 deletion(-)
> >
> > diff --git a/arch/arm64/crypto/aes-glue.c b/arch/arm64/crypto/aes-glue.c
> > index 998ba519a026..9e42ec96243e 100644
> > --- a/arch/arm64/crypto/aes-glue.c
> > +++ b/arch/arm64/crypto/aes-glue.c
> > @@ -664,7 +664,10 @@ static int __init aes_init(void)
> > return 0;
> >
> > unregister_simds:
> > - aes_exit();
> > + for (i = 0; i < ARRAY_SIZE(aes_simd_algs); i++)
> > + if (aes_simd_algs[i])
> > + simd_skcipher_free(aes_simd_algs[i]);
> > + crypto_unregister_shashes(mac_algs, ARRAY_SIZE(mac_algs));
> > unregister_ciphers:
> > crypto_unregister_skciphers(aes_algs, ARRAY_SIZE(aes_algs));
> > return err;
> > --
> > 2.13.6
> >
> >
>
>
> Would this also fix it?
>
> diff --git a/arch/arm64/crypto/aes-glue.c b/arch/arm64/crypto/aes-glue.c
> index 998ba519a026..2fa850e86aa8 100644
> --- a/arch/arm64/crypto/aes-glue.c
> +++ b/arch/arm64/crypto/aes-glue.c
> @@ -665,6 +665,7 @@ static int __init aes_init(void)
>
> unregister_simds:
> aes_exit();
> + return err;
> unregister_ciphers:
> crypto_unregister_skciphers(aes_algs, ARRAY_SIZE(aes_algs));
> return err;

Yes it is better.

I will send a v2 today.

Regards