On Thu, Jul 21, 2011 at 12:15 PM, Dave Airlie <airl...@gmail.com> wrote: > From: Dave Airlie <airl...@redhat.com> > > I've been hacking around trying to make this test work and discovered > writing some bits up at bit 16 of the exported word seems to act like > some sort of mask. However I've no idea what this is or why this fixes > this test. However it breaks the stencilreaddraw test in mesa-demos, which > means its not entirely correct. My current trend is to blame some float vs > int mixup somewhere, otherwise I'm pretty lost. > > Maybe AMD guys could take a look when they resurface.
shader export ARRAY_BASE for EXPORT_POS: 60 is position 61 is misc vector 62, 63 are clip distance vectors The use of the values exported in 61-63 are controlled by PA_CL_VS_OUT_CNTL: VS_OUT_MISC_VEC_ENA - enables the use of all fields in export 61 USE_VTX_POINT_SIZE - point size in the X channel of export 61 USE_VTX_EDGE_FLAG - edge flag in the Y channel of export 61 USE_VTX_RENDER_TARGET_INDX - render target index in the Z channel of export 61 USE_VTX_VIEWPORT_INDX - viewport index in the W channel of export 61 USE_VTX_KILL_FLAG - kill flag in the Z channel of export 61 (mutually exclusive from render target index) VS_OUT_CCDIST0_VEC_ENA/VS_OUT_CCDIST1_VEC_ENA - enable clip distance vectors shader export ARRAY_BASE for EXPORT_PIXEL: 0-7 CB targets 61 computed Z The use of the values exported in the computed Z vector are controlled by DB_SHADER_CONTROL: Z_EXPORT_ENABLE - Z as a float in RED STENCIL_REF_EXPORT_ENABLE - stencil ref as int in GREEN COVERAGE_TO_MASK_ENABLE - alpha to mask in ALPHA MASK_EXPORT_ENABLE - pixel sample mask in BLUE DB_SOURCE_FORMAT - export control restrictions Alex _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
