[Intel-gfx] [PATCH] drm/i915:fix the bug that testdisplay can't work with 32 bits depth (bug # 40011)

Lan, Hai hai.lan at intel.com
Sun Aug 14 00:49:13 CEST 2011


Hi Keith
Thanks for your explanation. So bug 40011 can be treated as not-a-bug.

Hai Lan

> -----Original Message-----
> From: Keith Packard [mailto:keithp at keithp.com]
> Sent: Sunday, August 14, 2011 1:47 AM
> To: Lan, Hai; intel-gfx at lists.freedesktop.org
> Cc: Lan, Hai
> Subject: Re: [PATCH] drm/i915:fix the bug that testdisplay can't work with 32
> bits depth (bug # 40011)
> 
> On Sat, 13 Aug 2011 23:22:14 -0400, Hai Lan <hai.lan at intel.com> wrote:
> 
> >  	case 32:
> >  		if (fb->depth == 24)
> >  			dspcntr |= DISPPLANE_32BPP_NO_ALPHA;
> > +		else if (fb->depth == 32)
> > +			dspcntr |= DISPPLANE_32BPP_NO_ALPHA;
> >  		else if (fb->depth == 30)
> >  			dspcntr |= DISPPLANE_32BPP_30BIT_NO_ALPHA;
> >  		else
> 
> There is no hardware support for depth 32 output -- the hardware can only
> do 24-bits (8-8-8-8) XRGB or 30-bits (2-10-10-10) XRGB. Passing depth 32 to
> this function is in fact an error which the kernel now correctly reports.
> 
> This patch simply assumes that the application programmer created a depth
> 24 frame buffer and then mistakenly told the kernel that it was depth 32.
> 
> --
> keith.packard at intel.com



More information about the Intel-gfx mailing list