On Fri, 6 Mar 2015 10:52:53, Eric Botcazou wrote:
>
>> I know it because strict_volatile_bitfield_p checks this:
>> /* Check for cases of unaligned fields that must be split. */
>> if (bitnum % BITS_PER_UNIT + bitsize> modesize
>>
>> Therefore, if bitsize == modesize, we know that bitnum % BITS_PER_UNIT
>> must be zero.
>
> Isn't that precisely the condition you're trying to relax in the other change?
>
Yes, but modesize is a mutliple of BITS_PER_UNIT,
therefore:
bitnum % modesize + bitsize> modesize
implies
bitnum % BITS_PER_UNIT + bitsize> modesize
I agree that the comments are still out of sync.
Thanks
Bernd.
