Re: [PATCH v3 3/6] tpm: introduce tpm_pcr_bank_info structure with digest_size from TPM

[Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>]

On Wed, Jun 21, 2017 at 04:29:38PM +0200, Roberto Sassu wrote:
> This patch introduces the new structure tpm_pcr_bank_info to store
> information regarding PCR banks. The next patch will replace the array of
> TPM algorithms IDs with an array of the new structure.

You should be able to express what you are doing without referring to
the "next patch" as there is no well defined order between commits
before they are in the mainline.

In this you might considering merging couple of commits if they are
so closely coupled.

> tpm_pcr_bank_info contains the TPM algorithm ID, the digest size and,
> optionally, the corresponding crypto ID, if a mapping exists. These

There is probably not the one and only algorithm ID and digest. If I
interperet this sentence to the word I would except either zero or one
mappings in total.

> information will be used by IMA to calculate the digest of an event
> and to provide measurements logs to userspace applications. The new
> structure has been defined in include/linux/tpm.h, as it will be passed
> to functions outside the TPM driver.
> 
> The purpose of this patch is to fix a serious issue in tpm2_pcr_extend():
> if the mapping between a TPM algorithm and a crypto algorithm is not
> defined, the PCR bank with the unknown algorithm is not extended.

Don't really understand what you are trying to say here as this patch
contains zero changes to tpm2_pcr_extend(). You shoud be carefuly when
using strong words such as "serious". tpm2_pcr_extend() works in well
defined way.

> This gives the opportunity to an attacker to reply to remote attestation
> requests with a list of fake measurements. Instead, the digest size
> is retrieved from the output buffer of a PCR read, without relying
> on the crypto subsystem.

This last paragraph is total nonsense.

I went through the whole commit message and did not find anything that
would explain why you make the code changes. Seriously, if you don't
know what you are doing it doesn't help if you explain alot of random
stuff

Maybe one way to explain part of the changes would be simply that the
mapping is needed because kernel subsystems communicate with the crypto
subsystem IDs. Just a one suggestion.

> Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>


> ---
>  drivers/char/tpm/tpm.h      | 11 -----------
>  drivers/char/tpm/tpm2-cmd.c | 30 ++++++++++++++++++++++++++++++
>  include/linux/tpm.h         | 19 +++++++++++++++++++
>  3 files changed, 49 insertions(+), 11 deletions(-)
> 
> diff --git a/drivers/char/tpm/tpm.h b/drivers/char/tpm/tpm.h
> index 1df0521..62c600d 100644
> --- a/drivers/char/tpm/tpm.h
> +++ b/drivers/char/tpm/tpm.h
> @@ -98,17 +98,6 @@ enum tpm2_return_codes {
>  	TPM2_RC_REFERENCE_H0	= 0x0910,
>  };
>  
> -enum tpm2_algorithms {
> -	TPM2_ALG_ERROR		= 0x0000,
> -	TPM2_ALG_SHA1		= 0x0004,
> -	TPM2_ALG_KEYEDHASH	= 0x0008,
> -	TPM2_ALG_SHA256		= 0x000B,
> -	TPM2_ALG_SHA384		= 0x000C,
> -	TPM2_ALG_SHA512		= 0x000D,
> -	TPM2_ALG_NULL		= 0x0010,
> -	TPM2_ALG_SM3_256	= 0x0012,
> -};
> -
>  enum tpm2_command_codes {
>  	TPM2_CC_FIRST		= 0x011F,
>  	TPM2_CC_SELF_TEST	= 0x0143,
> diff --git a/drivers/char/tpm/tpm2-cmd.c b/drivers/char/tpm/tpm2-cmd.c
> index 6a9fe0d..74a68ea 100644
> --- a/drivers/char/tpm/tpm2-cmd.c
> +++ b/drivers/char/tpm/tpm2-cmd.c
> @@ -992,6 +992,36 @@ int tpm2_probe(struct tpm_chip *chip)
>  }
>  EXPORT_SYMBOL_GPL(tpm2_probe);
>  
> +static int tpm2_init_pcr_bank_info(struct tpm_chip *chip, u16 alg_id,
> +				   struct tpm_pcr_bank_info *active_bank)

Generates a compiler warning (unused function).

> +{
> +	struct tpm_buf buf;
> +	struct tpm2_pcr_read_out *pcrread_out;
> +	int rc = 0;
> +	int i;
> +
> +	active_bank->alg_id = alg_id;
> +
> +	rc = tpm2_pcr_read_tpm_buf(chip, 0, alg_id, &buf, NULL);
> +	if (rc)
> +		goto out;
> +
> +	pcrread_out = (struct tpm2_pcr_read_out *)&buf.data[TPM_HEADER_SIZE];
> +
> +	active_bank->digest_size = be16_to_cpu(pcrread_out->digest_size);
> +	active_bank->crypto_id = HASH_ALGO__LAST;
> +
> +	for (i = 0; i < ARRAY_SIZE(tpm2_hash_map); i++) {
> +		if (active_bank->alg_id != tpm2_hash_map[i].tpm_id)
> +			continue;
> +
> +		active_bank->crypto_id = tpm2_hash_map[i].crypto_id;
> +	}
> +out:
> +	tpm_buf_destroy(&buf);
> +	return rc;
> +}
> +
>  struct tpm2_pcr_selection {
>  	__be16  hash_alg;
>  	u8  size_of_select;
> diff --git a/include/linux/tpm.h b/include/linux/tpm.h
> index 5a090f5..ff06738 100644
> --- a/include/linux/tpm.h
> +++ b/include/linux/tpm.h
> @@ -22,6 +22,8 @@
>  #ifndef __LINUX_TPM_H__
>  #define __LINUX_TPM_H__
>  
> +#include <crypto/hash_info.h>
> +
>  #define TPM_DIGEST_SIZE 20	/* Max TPM v1.2 PCR size */
>  
>  /*
> @@ -37,6 +39,17 @@ enum TPM_OPS_FLAGS {
>  	TPM_OPS_AUTO_STARTUP = BIT(0),
>  };
>  
> +enum tpm2_algorithms {
> +	TPM2_ALG_ERROR		= 0x0000,
> +	TPM2_ALG_SHA1		= 0x0004,
> +	TPM2_ALG_KEYEDHASH	= 0x0008,
> +	TPM2_ALG_SHA256		= 0x000B,
> +	TPM2_ALG_SHA384		= 0x000C,
> +	TPM2_ALG_SHA512		= 0x000D,
> +	TPM2_ALG_NULL		= 0x0010,
> +	TPM2_ALG_SM3_256	= 0x0012,
> +};
> +
>  struct tpm_class_ops {
>  	unsigned int flags;
>  	const u8 req_complete_mask;
> @@ -52,6 +65,12 @@ struct tpm_class_ops {
>  	void (*relinquish_locality)(struct tpm_chip *chip, int loc);
>  };
>  
> +struct tpm_pcr_bank_info {
> +	enum tpm2_algorithms alg_id;
> +	enum hash_algo crypto_id;
> +	u32 digest_size;
> +};
> +
>  #if defined(CONFIG_TCG_TPM) || defined(CONFIG_TCG_TPM_MODULE)
>  
>  extern int tpm_is_tpm2(u32 chip_num);
> -- 
> 2.9.3
> 

The code changes does not "live for its own".

/Jarkko