[Spice-devel] [spice-common PATCH 3/7 v2] ppc: Added supprt for bigendian color byte order

[Lukas Venhoda]

Hi,
> In the subject line "support" (typo)
> "machines"

Can those be fixed before pushing? Or would you prefer me to resend the patch?

> Why do we get both on BE machines, and only one byte order on LE
> machines ? surface_create is called sometimes to create surfaces using
> the 'native' byte order, and other times it's called for surfaces coming
> from the network, so using LE byte-order?

It is my understanding, that pixman_utils converts everything to LE, even on BE machines.
Example being, if we change colors to the new macros in spice_surface_format_to_pixman(),
the colors change back to the "wrong" colors.

That being sad, pixman_utils is not called every time, so sometimes the image will be in
RGBA order and sometimes in ABGR.

I tryed looking more into the issue, but leaving it as is worked best.

Seeing, as I didn't want to accidentally break stuff I didn't know how worked,
I just changed what was necessary and left the rest as is. (Including jpeg for example)

Hope that answers your questions.

Lukas
----- Original Message -----
> From: "Christophe Fergeau" <cfergeau at redhat.com>
> To: "Lukas Venhoda" <lvenhoda at redhat.com>
> Cc: spice-devel at lists.freedesktop.org
> Sent: Wednesday, July 8, 2015 5:35:41 PM
> Subject: Re: [Spice-devel] [spice-common PATCH 3/7 v2] ppc: Added supprt for bigendian color byte order
> 
> In the subject line "support" (typo)
> 
> On Thu, Jul 02, 2015 at 04:01:40PM +0200, Lukas Venhoda wrote:
> > On LE machine, color order will always be ARGB.
> > On BE mcahines both ARGB and BGRA states are correct.
> 
> "machines"
> 
> Why do we get both on BE machines, and only one byte order on LE
> machines ? surface_create is called sometimes to create surfaces using
> the 'native' byte order, and other times it's called for surfaces coming
> from the network, so using LE byte-order?
> 
> Christophe
>