intel: factor out dispatch PS enabling logic
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Reviewed-by: Ivan Briano <ivan.briano@intel.com> Tested-by: Mark Janes <markjanes@swizzler.org> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20169>
This commit is contained in:
@@ -984,6 +984,57 @@ brw_fs_simd_width_for_ksp(unsigned ksp_idx, bool simd8_enabled,
|
||||
}
|
||||
}
|
||||
|
||||
static inline void
|
||||
brw_fs_get_dispatch_enables(const struct intel_device_info *devinfo,
|
||||
const struct brw_wm_prog_data *prog_data,
|
||||
unsigned rasterization_samples,
|
||||
bool *enable_8,
|
||||
bool *enable_16,
|
||||
bool *enable_32)
|
||||
{
|
||||
assert(rasterization_samples != 0);
|
||||
|
||||
*enable_8 = prog_data->dispatch_8;
|
||||
*enable_16 = prog_data->dispatch_16;
|
||||
*enable_32 = prog_data->dispatch_32;
|
||||
|
||||
if (prog_data->persample_dispatch) {
|
||||
/* Starting with SandyBridge (where we first get MSAA), the different
|
||||
* pixel dispatch combinations are grouped into classifications A
|
||||
* through F (SNB PRM Vol. 2 Part 1 Section 7.7.1). On most hardware
|
||||
* generations, the only configurations supporting persample dispatch
|
||||
* are those in which only one dispatch width is enabled.
|
||||
*
|
||||
* The Gfx12 hardware spec has a similar dispatch grouping table, but
|
||||
* the following conflicting restriction applies (from the page on
|
||||
* "Structure_3DSTATE_PS_BODY"), so we need to keep the SIMD16 shader:
|
||||
*
|
||||
* "SIMD32 may only be enabled if SIMD16 or (dual)SIMD8 is also
|
||||
* enabled."
|
||||
*/
|
||||
if (*enable_32 || *enable_16)
|
||||
*enable_8 = false;
|
||||
if (devinfo->ver < 12 && *enable_32)
|
||||
*enable_16 = false;
|
||||
}
|
||||
|
||||
/* The docs for 3DSTATE_PS::32 Pixel Dispatch Enable say:
|
||||
*
|
||||
* "When NUM_MULTISAMPLES = 16 or FORCE_SAMPLE_COUNT = 16,
|
||||
* SIMD32 Dispatch must not be enabled for PER_PIXEL dispatch
|
||||
* mode."
|
||||
*
|
||||
* 16x MSAA only exists on Gfx9+, so we can skip this on Gfx8.
|
||||
*/
|
||||
if (devinfo->ver >= 9 && rasterization_samples == 16 &&
|
||||
!prog_data->persample_dispatch) {
|
||||
assert(*enable_8 || *enable_16);
|
||||
*enable_32 = false;
|
||||
}
|
||||
|
||||
assert(*enable_8 || *enable_16 || *enable_32);
|
||||
}
|
||||
|
||||
#define brw_wm_state_simd_width_for_ksp(wm_state, ksp_idx) \
|
||||
brw_fs_simd_width_for_ksp((ksp_idx), (wm_state)._8PixelDispatchEnable, \
|
||||
(wm_state)._16PixelDispatchEnable, \
|
||||
|
@@ -7607,26 +7607,6 @@ brw_compile_fs(const struct brw_compiler *compiler,
|
||||
}
|
||||
}
|
||||
|
||||
if (prog_data->persample_dispatch) {
|
||||
/* Starting with SandyBridge (where we first get MSAA), the different
|
||||
* pixel dispatch combinations are grouped into classifications A
|
||||
* through F (SNB PRM Vol. 2 Part 1 Section 7.7.1). On most hardware
|
||||
* generations, the only configurations supporting persample dispatch
|
||||
* are those in which only one dispatch width is enabled.
|
||||
*
|
||||
* The Gfx12 hardware spec has a similar dispatch grouping table, but
|
||||
* the following conflicting restriction applies (from the page on
|
||||
* "Structure_3DSTATE_PS_BODY"), so we need to keep the SIMD16 shader:
|
||||
*
|
||||
* "SIMD32 may only be enabled if SIMD16 or (dual)SIMD8 is also
|
||||
* enabled."
|
||||
*/
|
||||
if (simd32_cfg || simd16_cfg)
|
||||
simd8_cfg = NULL;
|
||||
if (simd32_cfg && devinfo->ver < 12)
|
||||
simd16_cfg = NULL;
|
||||
}
|
||||
|
||||
fs_generator g(compiler, params->log_data, mem_ctx, &prog_data->base,
|
||||
v8->runtime_check_aads_emit, MESA_SHADER_FRAGMENT);
|
||||
|
||||
|
Reference in New Issue
Block a user