intel/compiler: Zero out the header for texel fetch

It looks like even if we pass the header not present in the sampler descriptor,
it's not helping with the correct behavior of texelFetch.

Experiment on real HW shows that if we just zero out the header and include it
in the message, it helps with the correct behavior. I'm not sure if there is a
valid HW workaround for this one.

We can skip masking the sampler message header bits 4:0 but masking them out
doesn't hurt in this case.

Increasing number of parameter impact sampler performance, For example,
a sample message using 5 parameters will not be able to sustain the same
throughput as a sample message with only 4 valid parameters. We should
look out for any perf impact with respect to texel fetch.

This patch fixes ~3k tests involving texelFetch instruction on Xe3+

Signed-off-by: Sagar Ghuge <sagar.ghuge@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33562>

src/intel/compiler/brw_lower_logical_sends.cpp

@@ -706,7 +706,8 @@ emit_load_payload_with_padding(const brw_builder &bld, const brw_reg &dst,
 }
 
 static bool
-shader_opcode_needs_header(opcode op)
+shader_opcode_needs_header(opcode op,
+                           const struct intel_device_info *devinfo)
 {
    switch (op) {
    case SHADER_OPCODE_TG4_LOGICAL:
@@ -718,6 +719,9 @@ shader_opcode_needs_header(opcode op)
    case SHADER_OPCODE_TG4_IMPLICIT_LOD_LOGICAL:
    case SHADER_OPCODE_SAMPLEINFO_LOGICAL:
       return true;
+   case SHADER_OPCODE_TXF_LOGICAL:
+      /* Xe3 HW does not seem to work unless we force a header. */
+      return devinfo->ver >= 30;
    default:
       break;
    }
@@ -764,7 +768,7 @@ lower_sampler_logical_send(const brw_builder &bld, brw_inst *inst,
    assert((surface.file == BAD_FILE) != (surface_handle.file == BAD_FILE));
    assert((sampler.file == BAD_FILE) != (sampler_handle.file == BAD_FILE));
 
-   if (shader_opcode_needs_header(op) || inst->offset != 0 ||
+   if (shader_opcode_needs_header(op, devinfo) || inst->offset != 0 ||
        sampler_handle.file != BAD_FILE ||
        is_high_sampler(devinfo, sampler) ||
        residency) {