Hi Jiri,

thank you for your comments.


Jiri Svoboda <[email protected]> wrote:

> Hi Miroslav,
> 
> good catch. It looks like GPT uses a weird Microsoft-specific encoding of 
> GUID/UUID, whereas we use the encoding per RFC 4122. The other case where we 
> call uuid_encode() is in ext4 superblock, where it is appropriate since ext4 
> encodes UUIDs according to RFC 4122 (in network byte order).
> 
> I never thought alternate encodings would exist. In light of this, it might 
> make more sense for the internal representation to be actually structured, at 
> least for the first three dash-delimited fields (and maybe for all of them).

Do you mean something like:

#define _UUID_NODE_LEN 6

typedef struct {
        uint32_t time_low;
        uint16_t time_mid;
        uint16_t time_hi_and_version;
        uint8_t clock_seq_hi_and_reserved;
        uint8_t clock_seq_low;
        uint8_t node[_UUID_NODE_LEN];
} uuid_t;

> Your changes look mostly OK, although I am not a big fan of the helper 
> function (which couldn't exist anyway if we changed the internal 
> representation to structured).  Some of the inline comments incorrectly start 
> with a double asterisk, e.g. /** time_low */ even though they are not 
> documentation comments.

Sorry, I thought you used the /** ... */ style for normal comments as well.

I will try to put something together.

--
mc

_______________________________________________
HelenOS-devel mailing list
[email protected]
http://lists.modry.cz/listinfo/helenos-devel

Reply via email to