Re: [Xen-devel] [PATCH 1/2] efi/boot: fix set_color function

[Jan Beulich <jbeulich@xxxxxxxx>]

On 04.10.2019 12:54, Igor Druzhinin wrote:
> On 04/10/2019 11:34, Jan Beulich wrote:
>> On 03.10.2019 15:49, Igor Druzhinin wrote:
>>> - 0 is a possible and allowed value for a color mask accroding to
>>>   UEFI Spec 2.6 (11.9) especially for reserved mask
>>
>> Hmm, looking at 2.8 (where it's section 12.9, which in turn is why
>> section titles would be more helpful in such references) I can't
>> see the case being mentioned explicitly. I can accept that
>> ReservedMask might be zero, but then I'd prefer to handle that
>> case in the caller, rather than allowing zero also for the three
>> colors.
> 
> "If a bit is set in RedMask, GreenMask, or BlueMask then those bits of
> the pixel represent the corresponding color." - "If a bit is set..."
> implies it might not be set.

This talks about the function of individual bits. There's nothing said
about not bit at all being set for a particular color.

> Nothing prevents mask for the colors be 0 as well.

I wouldn't read it like this, no. I'm fine imply such for the reserved
field, but I'd rather consider it a broken mode if one of the colors
has no way of representing at all. In particular

"The color intensities must increase as the color values for a each
 color mask increase with a minimum intensity of all bits in a color
 mask clear to a maximum intensity of all bits in a color mask set."

suggests to me that there can't be zero bits set, or else there'd be
no difference between minimum and maximum intensity. Also, while
mathematically it makes sense for "all bits" to include the case of
zero of them, it doesn't (to me at least) in day-to-day use of the
language.

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel