Rodrigo, when you're ready, can you pull in Art's requests?
On Fri, Jun 20, 2014 at 03:56:08PM +0000, Runyan, Arthur J wrote:
> You give me too much credit. I just gave you an explanation of what the
> hardware does, then you ran with it.
>
> On Thu, Jun 19, 2014 at 12:06:13PM -0700, Ben Widawsky wrote:
> + DRM_INFO("Reducing the compressed framebuffer size. This may
> lead to increased power. Try to increase stolen memory size if available in
> BIOS.\n");
>
> I prefer "This may lead to less power savings than a non-reduced size." since
> FBC is still going to save power.
Got it.
>
> dpfc_ctl = DPFC_CTL_PLANE(intel_crtc->plane);
> if (drm_format_plane_cpp(fb->pixel_format, 0) == 2)
> + dev_priv->fbc.threshold++;
> +
> + switch (dev_priv->fbc.threshold) {
> + case 4:
> + dpfc_ctl |= DPFC_CTL_LIMIT_4X;
> + break;
> + case 2:
> dpfc_ctl |= DPFC_CTL_LIMIT_2X;
> - else
> + break;
> + case 1:
> dpfc_ctl |= DPFC_CTL_LIMIT_1X;
> + break;
> + }
>
> I Am Not A Coder, but at a glance it looks like the ++ could lead to
> undefined case 3 when you want case 4.
>
Thanks for your feedback. 3 is an invalid value, but a default case
should be added here. I'd do a BUG_ON, but that's strictly forbidden.
--
Ben Widawsky, Intel Open Source Technology Center
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
