Re: [PATCH 1/4] xen/lib: Export additional sha256 functions

[Jan Beulich <jbeulich@suse.com>]

On 06.05.2025 16:05, Andrew Cooper wrote:
> On 06/05/2025 2:56 pm, Frediano Ziglio wrote:
>> diff --git a/xen/include/xen/sha2.h b/xen/include/xen/sha2.h
>> index 47d97fbf01..ea8bad67e4 100644
>> --- a/xen/include/xen/sha2.h
>> +++ b/xen/include/xen/sha2.h
>> @@ -9,6 +9,16 @@
>>  
>>  #define SHA2_256_DIGEST_SIZE 32
>>  
>> +struct sha2_256_state {
>> +    uint32_t state[SHA2_256_DIGEST_SIZE / sizeof(uint32_t)];
>> +    uint8_t buf[64];
>> +    size_t count; /* Byte count. */
>> +};
>> +
>> +void sha2_256_init(struct sha2_256_state *s);
>> +void sha2_256_update(struct sha2_256_state *s, const void *msg,
>> +                     size_t len);
>> +void sha2_256_final(struct sha2_256_state *s, void *_dst);
>>  void sha2_256_digest(uint8_t digest[SHA2_256_DIGEST_SIZE],
>>                       const void *msg, size_t len);
> 
> sha2_256_digest() is unlike the others as it holds sha2_256_state
> internally.  I'd suggest having all of the additions below this point,
> which group them more nicely.
> 
> Can fix on commit.  Otherwise LGTM.

I notice this was committed, but isn't this introducing new Misra
violations (extern functions without external callers)?

Jan