On 17.02.2008 05:02, Peter Stuge wrote:
> On Sun, Feb 17, 2008 at 12:34:24AM +0100, Carl-Daniel Hailfinger wrote:
>
>> +enum compalgo {
>> + none = 0,
>> + lzma = 1,
>> + nrv2b = 2,
>> + zeroes = 3,
>> + /* invalid should always be the last entry. */
>> + invalid
>> +};
>>
>
> It's fairly common to uppercase these. Do we want that too?
>
This was just a straight copy from util/lar, but I agree. We use these
enums like #defines, so uppercasing seems indeed the best way to bring
this hint to the developer.
>> + const char *algoname = algo_name[(archive->compression >= invalid)
>> + ? invalid : archive->compression];
>>
>
> Not sure about invalid. I would be less worried if invalid was set to
> some explicit value, e.g. -1.
We use archive->compression as array index. The line above is there to
make sure we do not dereference a random pointer outside the array. That
also rules out a negative value for invalid because accessing arrays
with negative indices is real scary EVIL.
> The current proposal means we can never remove algorithms.
>
How so? We have to keep the number reserved if we ever abandon a
compression algorithm, but that's it.
Regards,
Carl-Daniel
--
http://www.hailfinger.org/
--
coreboot mailing list
[email protected]
http://www.coreboot.org/mailman/listinfo/coreboot
