On 31/12/15 16:14, Sean M. Pappalardo wrote:
> I asked in #intel-gfx on freenode, the Intel graphics driver development
> channel and they suggested trying the i915g driver instead since it
> doesn't report an error in the same situation. They said the root
> problem is that "the shader is too long (instruction count) for the
> hardware" and that "the graphics hardware doesn't match the capablities
> that GL requires (and the driver is lying so that it mostly works.)
> There is a flag to disable the lie. But certainly try i915g, llvmpipe,
> or convince Qt to use a different back-end."
>
Sean,
That is so far over my head I need a space suit.
Mark B
;-)
On 31/12/15 16:14, Sean M. Pappalardo wrote:
> I asked in #intel-gfx on freenode, the Intel graphics driver development
> channel and they suggested trying the i915g driver instead since it
> doesn't report an error in the same situation. They said the root
> problem is that "the shader is too long (instruction count) for the
> hardware" and that "the graphics hardware doesn't match the capablities
> that GL requires (and the driver is lying so that it mostly works.)
> There is a flag to disable the lie. But certainly try i915g, llvmpipe,
> or convince Qt to use a different back-end."
>