Re: [PATCH v9 6/6] tpm: pass an array of tpm_extend_digest structures to tpm_pcr_extend()

[Roberto Sassu <roberto.sassu@huawei.com>]

On 2/1/2019 8:15 PM, Mimi Zohar wrote:
> Hi Roberto,
> 
> Sorry for the delayed review.  A few comments inline below, minor
> suggestions.
> 
>> diff --git a/security/integrity/ima/ima.h b/security/integrity/ima/ima.h
>> index cc12f3449a72..e6b2dcb0846a 100644
>> --- a/security/integrity/ima/ima.h
>> +++ b/security/integrity/ima/ima.h
>> @@ -56,6 +56,7 @@ extern int ima_policy_flag;
>>   extern int ima_hash_algo;
>>   extern int ima_appraise;
>>   extern struct tpm_chip *ima_tpm_chip;
>> +extern struct tpm_digest *digests;
>>   
>>   /* IMA event related data */
>>   struct ima_event_data {
>> diff --git a/security/integrity/ima/ima_init.c b/security/integrity/ima/ima_init.c
>> index 6bb42a9c5e47..296a965b11ef 100644
>> --- a/security/integrity/ima/ima_init.c
>> +++ b/security/integrity/ima/ima_init.c
>> @@ -27,6 +27,7 @@
>>   /* name for boot aggregate entry */
>>   static const char boot_aggregate_name[] = "boot_aggregate";
>>   struct tpm_chip *ima_tpm_chip;
>> +struct tpm_digest *digests;
> 
> "digests" is used in the new ima_init_digests() and in
> ima_pcr_extend().  It's nice that the initialization routines are
> grouped together here in ima_init.c, but wouldn't it better to define
> "digests" in ima_queued.c, where it is currently being used?
>   "digests" could then be defined as static.
> 
>>   
>>   /* Add the boot aggregate to the IMA measurement list and extend
>>    * the PCR register.
>> @@ -104,6 +105,24 @@ void __init ima_load_x509(void)
>>   }
>>   #endif
>>   
>> +int __init ima_init_digests(void)
>> +{
>> +	int i;
>> +
>> +	if (!ima_tpm_chip)
>> +		return 0;
>> +
>> +	digests = kcalloc(ima_tpm_chip->nr_allocated_banks, sizeof(*digests),
>> +			  GFP_NOFS);
>> +	if (!digests)
>> +		return -ENOMEM;
>> +
>> +	for (i = 0; i < ima_tpm_chip->nr_allocated_banks; i++)
>> +		digests[i].alg_id = ima_tpm_chip->allocated_banks[i].alg_id;
>> +
>> +	return 0;
>> +}
>> +
>>   int __init ima_init(void)
>>   {
>>   	int rc;
>> @@ -125,6 +144,9 @@ int __init ima_init(void)
>>   
>>   	ima_load_kexec_buffer();
>>   
>> +	rc = ima_init_digests();
> 
> Ok. Getting the tpm chip is at the beginning of this function.
>   Deferring allocating "digests" to here, avoids having to free memory
> on failure.

If I understood the code correctly, ima_init() does not undo the actions
of previous functions. For example, if ima_init_template() fails,
ima_shash_tfm is not freed.

Probably, this can be improved later.

Roberto


> ima_load_kexec_buffer() restores prior measurements, but doesn't
> extend the TPM.  For anyone reading the code, a short comment above
> ima_load_kexec_buffer() would make sense.
> 
> Mimi
>     
>> +	if (rc != 0)
>> +		return rc;
>>   	rc = ima_add_boot_aggregate();	/* boot aggregate must be first entry */
>>   	if (rc != 0)
>>   		return rc;
>> diff --git a/security/integrity/ima/ima_queue.c b/security/integrity/ima/ima_queue.c
>> index 0e41dc1df1d4..719e88ca58f6 100644
>> --- a/security/integrity/ima/ima_queue.c
>> +++ b/security/integrity/ima/ima_queue.c
>> @@ -140,11 +140,15 @@ unsigned long ima_get_binary_runtime_size(void)
>>   static int ima_pcr_extend(const u8 *hash, int pcr)
>>   {
>>   	int result = 0;
>> +	int i;
>>   
>>   	if (!ima_tpm_chip)
>>   		return result;
>>   
>> -	result = tpm_pcr_extend(ima_tpm_chip, pcr, hash);
>> +	for (i = 0; i < ima_tpm_chip->nr_allocated_banks; i++)
>> +		memcpy(digests[i].digest, hash, TPM_DIGEST_SIZE);
>> +
>> +	result = tpm_pcr_extend(ima_tpm_chip, pcr, digests);
>>   	if (result != 0)
>>   		pr_err("Error Communicating to TPM chip, result: %d\n", result);
>>   	return result;
>>
> 

-- 
HUAWEI TECHNOLOGIES Duesseldorf GmbH, HRB 56063
Managing Director: Bo PENG, Jian LI, Yanli SHI