I had simply copied these values from another driver when adding initial
compute support to iris. The actual hardware limit is UINT32_MAX (see
the GPGPU_WALKER/COMPUTE_WALKER ThreadGroupID{X,Y,Z}Dimension fields).
Thanks to Karol Herbst for noticing the unnecessarily low limit.
References: https://gitlab.freedesktop.org/mesa/mesa/-/issues/7676
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Tested-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19826>
RET((uint64_t []) { 3 });
case PIPE_COMPUTE_CAP_MAX_GRID_SIZE:
- RET(((uint64_t []) { 65535, 65535, 65535 }));
+ RET(((uint64_t []) { UINT32_MAX, UINT32_MAX, UINT32_MAX }));
case PIPE_COMPUTE_CAP_MAX_BLOCK_SIZE:
/* MaxComputeWorkGroupSize[0..2] */