I've pushed this. But I do wonder if we should just set LINE_WIDTH_SEPARATE to always false and not worry about this on any GPU. It's not like we really need the two line widths.
On Fri, Apr 20, 2018 at 6:32 PM, Rhys Perry <pendingchao...@gmail.com> wrote: > Signed-off-by: Rhys Perry <pendingchao...@gmail.com> > --- > src/gallium/drivers/nouveau/nvc0/nvc0_state.c | 4 +++- > 1 file changed, 3 insertions(+), 1 deletion(-) > > diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_state.c > b/src/gallium/drivers/nouveau/nvc0/nvc0_state.c > index 99d45a238a..1977a04832 100644 > --- a/src/gallium/drivers/nouveau/nvc0/nvc0_state.c > +++ b/src/gallium/drivers/nouveau/nvc0/nvc0_state.c > @@ -233,7 +233,9 @@ nvc0_rasterizer_state_create(struct pipe_context *pipe, > SB_IMMED_3D(so, MULTISAMPLE_ENABLE, cso->multisample); > > SB_IMMED_3D(so, LINE_SMOOTH_ENABLE, cso->line_smooth); > - if (cso->line_smooth || cso->multisample) > + /* On GM20x+, LINE_WIDTH_SMOOTH controls both aliased and smooth > + * rendering and LINE_WIDTH_ALIASED seems to be ignored */ > + if (cso->line_smooth || cso->multisample || class_3d>=GM200_3D_CLASS) > SB_BEGIN_3D(so, LINE_WIDTH_SMOOTH, 1); > else > SB_BEGIN_3D(so, LINE_WIDTH_ALIASED, 1); > -- > 2.14.3 > > _______________________________________________ > mesa-dev mailing list > mesa-dev@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/mesa-dev _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev
