i965: Add FS backend for builtin gl_SampleMask[]
V2:
- Update comments
- Add a special backend instructions to compute sample_mask.
- Add a new variable uses_omask in brw_wm_prog_data.
V3:
- Make changes to support simd16 mode.
- Delete redundant AND instruction and handle the register
stride in FS backend instruction.
Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Paul Berry <stereotype441@gmail.com>
diff --git a/src/mesa/drivers/dri/i965/brw_context.h b/src/mesa/drivers/dri/i965/brw_context.h
index 665807c..bec4d6b 100644
--- a/src/mesa/drivers/dri/i965/brw_context.h
+++ b/src/mesa/drivers/dri/i965/brw_context.h
@@ -384,6 +384,7 @@
GLuint nr_pull_params;
bool dual_src_blend;
bool uses_pos_offset;
+ bool uses_omask;
uint32_t prog_offset_16;
/**
diff --git a/src/mesa/drivers/dri/i965/brw_defines.h b/src/mesa/drivers/dri/i965/brw_defines.h
index e03ac58..bfea88a 100644
--- a/src/mesa/drivers/dri/i965/brw_defines.h
+++ b/src/mesa/drivers/dri/i965/brw_defines.h
@@ -794,6 +794,7 @@
FS_OPCODE_VARYING_PULL_CONSTANT_LOAD_GEN7,
FS_OPCODE_MOV_DISPATCH_TO_FLAGS,
FS_OPCODE_DISCARD_JUMP,
+ FS_OPCODE_SET_OMASK,
FS_OPCODE_SET_SAMPLE_ID,
FS_OPCODE_SET_SIMD4X2_OFFSET,
FS_OPCODE_PACK_HALF_2x16_SPLIT,
diff --git a/src/mesa/drivers/dri/i965/brw_fs.h b/src/mesa/drivers/dri/i965/brw_fs.h
index 0ed8c05..e9bf52f 100644
--- a/src/mesa/drivers/dri/i965/brw_fs.h
+++ b/src/mesa/drivers/dri/i965/brw_fs.h
@@ -443,6 +443,7 @@
struct hash_table *variable_ht;
fs_reg frag_depth;
+ fs_reg sample_mask;
fs_reg outputs[BRW_MAX_DRAW_BUFFERS];
unsigned output_components[BRW_MAX_DRAW_BUFFERS];
fs_reg dual_src_output;
@@ -549,6 +550,10 @@
struct brw_reg offset);
void generate_mov_dispatch_to_flags(fs_inst *inst);
+ void generate_set_omask(fs_inst *inst,
+ struct brw_reg dst,
+ struct brw_reg sample_mask);
+
void generate_set_sample_id(fs_inst *inst,
struct brw_reg dst,
struct brw_reg src0,
diff --git a/src/mesa/drivers/dri/i965/brw_fs_generator.cpp b/src/mesa/drivers/dri/i965/brw_fs_generator.cpp
index 2553098..d67c4e4 100644
--- a/src/mesa/drivers/dri/i965/brw_fs_generator.cpp
+++ b/src/mesa/drivers/dri/i965/brw_fs_generator.cpp
@@ -1068,6 +1068,44 @@
brw_pop_insn_state(p);
}
+/* Sets vstride=16, width=8, hstride=2 or vstride=0, width=1, hstride=0
+ * (when mask is passed as a uniform) of register mask before moving it
+ * to register dst.
+ */
+void
+fs_generator::generate_set_omask(fs_inst *inst,
+ struct brw_reg dst,
+ struct brw_reg mask)
+{
+ bool stride_8_8_1 =
+ (mask.vstride == BRW_VERTICAL_STRIDE_8 &&
+ mask.width == BRW_WIDTH_8 &&
+ mask.hstride == BRW_HORIZONTAL_STRIDE_1);
+
+ bool stride_0_1_0 =
+ (mask.vstride == BRW_VERTICAL_STRIDE_0 &&
+ mask.width == BRW_WIDTH_1 &&
+ mask.hstride == BRW_HORIZONTAL_STRIDE_0);
+
+ assert(stride_8_8_1 || stride_0_1_0);
+ assert(dst.type == BRW_REGISTER_TYPE_UW);
+
+ if (dispatch_width == 16)
+ dst = vec16(dst);
+ brw_push_insn_state(p);
+ brw_set_compression_control(p, BRW_COMPRESSION_NONE);
+ brw_set_mask_control(p, BRW_MASK_DISABLE);
+
+ if (stride_8_8_1) {
+ brw_MOV(p, dst, stride(retype(brw_vec1_reg(mask.file, mask.nr, 0),
+ dst.type), 16, 8, 2));
+ } else if (stride_0_1_0) {
+ brw_MOV(p, dst, stride(retype(brw_vec1_reg(mask.file, mask.nr, 0),
+ dst.type), 0, 1, 0));
+ }
+ brw_pop_insn_state(p);
+}
+
/* Sets vstride=1, width=4, hstride=0 of register src1 during
* the ADD instruction.
*/
@@ -1666,6 +1704,10 @@
generate_set_simd4x2_offset(inst, dst, src[0]);
break;
+ case FS_OPCODE_SET_OMASK:
+ generate_set_omask(inst, dst, src[0]);
+ break;
+
case FS_OPCODE_SET_SAMPLE_ID:
generate_set_sample_id(inst, dst, src[0], src[1]);
break;
diff --git a/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp b/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp
index 0993637..dd60671 100644
--- a/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp
+++ b/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp
@@ -82,6 +82,8 @@
}
} else if (ir->location == FRAG_RESULT_DEPTH) {
this->frag_depth = *reg;
+ } else if (ir->location == FRAG_RESULT_SAMPLE_MASK) {
+ this->sample_mask = *reg;
} else {
/* gl_FragData or a user-defined FS output */
assert(ir->location >= FRAG_RESULT_DATA0 &&
@@ -2526,6 +2528,16 @@
pop_force_uncompressed();
}
+ c->prog_data.uses_omask =
+ fp->Base.OutputsWritten & BITFIELD64_BIT(FRAG_RESULT_SAMPLE_MASK);
+ if(c->prog_data.uses_omask) {
+ this->current_annotation = "FB write oMask";
+ assert(this->sample_mask.file != BAD_FILE);
+ /* Hand over gl_SampleMask. Only lower 16 bits are relevant. */
+ emit(FS_OPCODE_SET_OMASK, fs_reg(MRF, nr, BRW_REGISTER_TYPE_UW), this->sample_mask);
+ nr += 1;
+ }
+
/* Reserve space for color. It'll be filled in per MRT below. */
int color_mrf = nr;
nr += 4 * reg_width;