On Fri, Apr 30, 2021 at 11:58:10 -0700, trix@xxxxxxxxxx wrote:
> From: Tom Rix <trix@xxxxxxxxxx>
>
> Static analysis reports this problem
> trusted-keys/trusted_tpm1.c:496:10: warning: Potential memory leak
> return ret;
> ^~~
>
> In tpm_seal() some failure handling returns directly, without
> freeing memory.
>
> Fixes: 5df16caada3f ("KEYS: trusted: Fix incorrect handling of tpm_get_random()")
> Signed-off-by: Tom Rix <trix@xxxxxxxxxx>
> ---
> security/keys/trusted-keys/trusted_tpm1.c | 8 +++++---
> 1 file changed, 5 insertions(+), 3 deletions(-)
>
> diff --git a/security/keys/trusted-keys/trusted_tpm1.c b/security/keys/trusted-keys/trusted_tpm1.c
> index 469394550801..aa108bea6739 100644
> --- a/security/keys/trusted-keys/trusted_tpm1.c
> +++ b/security/keys/trusted-keys/trusted_tpm1.c
> @@ -493,10 +493,12 @@ static int tpm_seal(struct tpm_buf *tb, uint16_t keytype,
>
> ret = tpm_get_random(chip, td->nonceodd, TPM_NONCE_SIZE);
> if (ret < 0)
> - return ret;
> + goto out;
>
> - if (ret != TPM_NONCE_SIZE)
> - return -EIO;
> + if (ret != TPM_NONCE_SIZE) {
> + ret = -EIO;
> + goto out;
> + }
>
> ordinal = htonl(TPM_ORD_SEAL);
> datsize = htonl(datalen);
I see this patch also submitted by Colin (Cc'd) in Message-Id:
<20210430113724.110746-1-colin.king@xxxxxxxxxxxxx>
To my eyes, the commit message seems a bit better over there.
--Ben
