[Mesa-dev] [PATCH 1/2] i965/fs: Revert "Avoid generating MOVs for assignments for expressions."
Kenneth Graunke
kenneth at whitecape.org
Sat Aug 27 03:23:27 PDT 2011
This reverts commit 53c89c67f33639afef951e178f93f4e29acc5d53, along with
the subsequent this->result = reg_undef additions it required.
Both Eric and I agree that the way he did this is really fragile; if you
forget to add this->result = reg_undef before calling accept(), it may
end up using the same register for two separate things, breaking things
in strange and mysterious ways.
The next commit will port over the new VS backend's method for solving
this problem, which is simpler, less intrusive, and still manages to
avoid MOVs in the common case.
---
src/mesa/drivers/dri/i965/brw_fs.h | 5 +-
src/mesa/drivers/dri/i965/brw_fs_visitor.cpp | 82 +++-----------------------
2 files changed, 10 insertions(+), 77 deletions(-)
diff --git a/src/mesa/drivers/dri/i965/brw_fs.h b/src/mesa/drivers/dri/i965/brw_fs.h
index 10f45f3..e22d3d3 100644
--- a/src/mesa/drivers/dri/i965/brw_fs.h
+++ b/src/mesa/drivers/dri/i965/brw_fs.h
@@ -568,10 +568,7 @@ public:
bool failed;
char *fail_msg;
- /* On entry to a visit() method, this is the storage for the
- * result. On exit, the visit() called may have changed it, in
- * which case the parent must use the new storage instead.
- */
+ /* Result of last visit() method. */
fs_reg result;
fs_reg pixel_x;
diff --git a/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp b/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp
index cdaf543..e054366 100644
--- a/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp
+++ b/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp
@@ -160,7 +160,6 @@ fs_visitor::try_emit_saturate(ir_expression *ir)
if (!sat_val)
return false;
- this->result = reg_undef;
sat_val->accept(this);
fs_reg src = this->result;
@@ -183,11 +182,7 @@ fs_visitor::visit(ir_expression *ir)
if (try_emit_saturate(ir))
return;
- /* This is where our caller would like us to put the result, if possible. */
- fs_reg saved_result_storage = this->result;
-
for (operand = 0; operand < ir->get_num_operands(); operand++) {
- this->result = reg_undef;
ir->operands[operand]->accept(this);
if (this->result.file == BAD_FILE) {
ir_print_visitor v;
@@ -205,14 +200,10 @@ fs_visitor::visit(ir_expression *ir)
assert(!ir->operands[operand]->type->is_vector());
}
- /* Inherit storage from our parent if possible, and otherwise we
- * alloc a temporary.
+ /* Storage for our result. If our result goes into an assignment, it will
+ * just get copy-propagated out, so no worries.
*/
- if (saved_result_storage.file == BAD_FILE) {
- this->result = fs_reg(this, ir->type);
- } else {
- this->result = saved_result_storage;
- }
+ this->result = fs_reg(this, ir->type);
switch (ir->operation) {
case ir_unop_logic_not:
@@ -233,9 +224,6 @@ fs_visitor::visit(ir_expression *ir)
case ir_unop_sign:
temp = fs_reg(this, ir->type);
- /* Unalias the destination. (imagine a = sign(a)) */
- this->result = fs_reg(this, ir->type);
-
emit(BRW_OPCODE_MOV, this->result, fs_reg(0.0f));
inst = emit(BRW_OPCODE_CMP, reg_null_f, op[0], fs_reg(0.0f));
@@ -514,21 +502,9 @@ fs_visitor::visit(ir_assignment *ir)
fs_inst *inst;
/* FINISHME: arrays on the lhs */
- this->result = reg_undef;
ir->lhs->accept(this);
l = this->result;
- /* If we're doing a direct assignment, an RHS expression could
- * drop its result right into our destination. Otherwise, tell it
- * not to.
- */
- if (ir->condition ||
- !(ir->lhs->type->is_scalar() ||
- (ir->lhs->type->is_vector() &&
- ir->write_mask == (1 << ir->lhs->type->vector_elements) - 1))) {
- this->result = reg_undef;
- }
-
ir->rhs->accept(this);
r = this->result;
@@ -543,13 +519,9 @@ fs_visitor::visit(ir_assignment *ir)
ir->lhs->type->is_vector()) {
for (int i = 0; i < ir->lhs->type->vector_elements; i++) {
if (ir->write_mask & (1 << i)) {
- if (ir->condition) {
- inst = emit(BRW_OPCODE_MOV, l, r);
+ inst = emit(BRW_OPCODE_MOV, l, r);
+ if (ir->condition)
inst->predicated = true;
- } else if (!l.equals(&r)) {
- inst = emit(BRW_OPCODE_MOV, l, r);
- }
-
r.reg_offset++;
}
l.reg_offset++;
@@ -590,19 +562,16 @@ fs_visitor::emit_texture_gen4(ir_texture *ir, fs_reg dst, fs_reg coordinate,
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen), fs_reg(0.0f));
mlen++;
} else if (ir->op == ir_txb) {
- this->result = reg_undef;
ir->lod_info.bias->accept(this);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen), this->result);
mlen++;
} else {
assert(ir->op == ir_txl);
- this->result = reg_undef;
ir->lod_info.lod->accept(this);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen), this->result);
mlen++;
}
- this->result = reg_undef;
ir->shadow_comparitor->accept(this);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen), this->result);
mlen++;
@@ -617,11 +586,9 @@ fs_visitor::emit_texture_gen4(ir_texture *ir, fs_reg dst, fs_reg coordinate,
/* gen4's SIMD8 sampler always has the slots for u,v,r present. */
mlen += 3;
} else if (ir->op == ir_txd) {
- this->result = reg_undef;
ir->lod_info.grad.dPdx->accept(this);
fs_reg dPdx = this->result;
- this->result = reg_undef;
ir->lod_info.grad.dPdy->accept(this);
fs_reg dPdy = this->result;
@@ -660,7 +627,6 @@ fs_visitor::emit_texture_gen4(ir_texture *ir, fs_reg dst, fs_reg coordinate,
} else if (ir->op == ir_txs) {
/* There's no SIMD8 resinfo message on Gen4. Use SIMD16 instead. */
simd16 = true;
- this->result = reg_undef;
ir->lod_info.lod->accept(this);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen, BRW_REGISTER_TYPE_UD), this->result);
mlen += 2;
@@ -684,12 +650,10 @@ fs_visitor::emit_texture_gen4(ir_texture *ir, fs_reg dst, fs_reg coordinate,
mlen += 6;
if (ir->op == ir_txb) {
- this->result = reg_undef;
ir->lod_info.bias->accept(this);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen), this->result);
mlen++;
} else {
- this->result = reg_undef;
ir->lod_info.lod->accept(this);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen), this->result);
mlen++;
@@ -789,7 +753,6 @@ fs_visitor::emit_texture_gen5(ir_texture *ir, fs_reg dst, fs_reg coordinate,
if (ir->shadow_comparitor && ir->op != ir_txd) {
mlen = MAX2(mlen, header_present + 4 * reg_width);
- this->result = reg_undef;
ir->shadow_comparitor->accept(this);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen), this->result);
mlen += reg_width;
@@ -801,7 +764,6 @@ fs_visitor::emit_texture_gen5(ir_texture *ir, fs_reg dst, fs_reg coordinate,
inst = emit(FS_OPCODE_TEX, dst);
break;
case ir_txb:
- this->result = reg_undef;
ir->lod_info.bias->accept(this);
mlen = MAX2(mlen, header_present + 4 * reg_width);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen), this->result);
@@ -811,7 +773,6 @@ fs_visitor::emit_texture_gen5(ir_texture *ir, fs_reg dst, fs_reg coordinate,
break;
case ir_txl:
- this->result = reg_undef;
ir->lod_info.lod->accept(this);
mlen = MAX2(mlen, header_present + 4 * reg_width);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen), this->result);
@@ -820,11 +781,9 @@ fs_visitor::emit_texture_gen5(ir_texture *ir, fs_reg dst, fs_reg coordinate,
inst = emit(FS_OPCODE_TXL, dst);
break;
case ir_txd: {
- this->result = reg_undef;
ir->lod_info.grad.dPdx->accept(this);
fs_reg dPdx = this->result;
- this->result = reg_undef;
ir->lod_info.grad.dPdy->accept(this);
fs_reg dPdy = this->result;
@@ -853,7 +812,6 @@ fs_visitor::emit_texture_gen5(ir_texture *ir, fs_reg dst, fs_reg coordinate,
break;
}
case ir_txs:
- this->result = reg_undef;
ir->lod_info.lod->accept(this);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen, BRW_REGISTER_TYPE_UD), this->result);
mlen += reg_width;
@@ -893,7 +851,6 @@ fs_visitor::emit_texture_gen7(ir_texture *ir, fs_reg dst, fs_reg coordinate,
}
if (ir->shadow_comparitor && ir->op != ir_txd) {
- this->result = reg_undef;
ir->shadow_comparitor->accept(this);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen), this->result);
mlen += reg_width;
@@ -904,13 +861,11 @@ fs_visitor::emit_texture_gen7(ir_texture *ir, fs_reg dst, fs_reg coordinate,
case ir_tex:
break;
case ir_txb:
- this->result = reg_undef;
ir->lod_info.bias->accept(this);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen), this->result);
mlen += reg_width;
break;
case ir_txl:
- this->result = reg_undef;
ir->lod_info.lod->accept(this);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen), this->result);
mlen += reg_width;
@@ -919,11 +874,9 @@ fs_visitor::emit_texture_gen7(ir_texture *ir, fs_reg dst, fs_reg coordinate,
if (c->dispatch_width == 16)
fail("Gen7 does not support sample_d/sample_d_c in SIMD16 mode.");
- this->result = reg_undef;
ir->lod_info.grad.dPdx->accept(this);
fs_reg dPdx = this->result;
- this->result = reg_undef;
ir->lod_info.grad.dPdy->accept(this);
fs_reg dPdy = this->result;
@@ -949,7 +902,6 @@ fs_visitor::emit_texture_gen7(ir_texture *ir, fs_reg dst, fs_reg coordinate,
break;
}
case ir_txs:
- this->result = reg_undef;
ir->lod_info.lod->accept(this);
emit(BRW_OPCODE_MOV, fs_reg(MRF, base_mrf + mlen, BRW_REGISTER_TYPE_UD), this->result);
mlen += reg_width;
@@ -1013,7 +965,6 @@ fs_visitor::visit(ir_texture *ir)
return swizzle_result(ir, fs_reg(0.0f), sampler);
}
- this->result = reg_undef;
if (ir->coordinate)
ir->coordinate->accept(this);
fs_reg coordinate = this->result;
@@ -1127,7 +1078,6 @@ fs_visitor::visit(ir_texture *ir)
if (hw_compare_supported) {
inst->shadow_compare = true;
} else {
- this->result = reg_undef;
ir->shadow_comparitor->accept(this);
fs_reg ref = this->result;
@@ -1206,7 +1156,6 @@ fs_visitor::swizzle_result(ir_texture *ir, fs_reg orig_val, int sampler)
void
fs_visitor::visit(ir_swizzle *ir)
{
- this->result = reg_undef;
ir->val->accept(this);
fs_reg val = this->result;
@@ -1269,7 +1218,6 @@ fs_visitor::visit(ir_constant *ir)
const unsigned size = type_size(ir->type->fields.array);
for (unsigned i = 0; i < ir->type->length; i++) {
- this->result = reg_undef;
ir->array_elements[i]->accept(this);
fs_reg src_reg = this->result;
@@ -1285,7 +1233,6 @@ fs_visitor::visit(ir_constant *ir)
ir_instruction *const field = (ir_instruction *) node;
const unsigned size = type_size(field->type);
- this->result = reg_undef;
field->accept(this);
fs_reg src_reg = this->result;
@@ -1336,7 +1283,6 @@ fs_visitor::emit_bool_to_cond_code(ir_rvalue *ir)
for (unsigned int i = 0; i < expr->get_num_operands(); i++) {
assert(expr->operands[i]->type->is_scalar());
- this->result = reg_undef;
expr->operands[i]->accept(this);
op[i] = this->result;
}
@@ -1401,7 +1347,6 @@ fs_visitor::emit_bool_to_cond_code(ir_rvalue *ir)
return;
}
- this->result = reg_undef;
ir->accept(this);
if (intel->gen >= 6) {
@@ -1431,7 +1376,6 @@ fs_visitor::emit_if_gen6(ir_if *ir)
for (unsigned int i = 0; i < expr->get_num_operands(); i++) {
assert(expr->operands[i]->type->is_scalar());
- this->result = reg_undef;
expr->operands[i]->accept(this);
op[i] = this->result;
}
@@ -1493,7 +1437,6 @@ fs_visitor::emit_if_gen6(ir_if *ir)
return;
}
- this->result = reg_undef;
ir->condition->accept(this);
fs_inst *inst = emit(BRW_OPCODE_IF, reg_null_d, this->result, fs_reg(0));
@@ -1526,7 +1469,7 @@ fs_visitor::visit(ir_if *ir)
foreach_list(node, &ir->then_instructions) {
ir_instruction *ir = (ir_instruction *)node;
this->base_ir = ir;
- this->result = reg_undef;
+
ir->accept(this);
}
@@ -1536,7 +1479,7 @@ fs_visitor::visit(ir_if *ir)
foreach_list(node, &ir->else_instructions) {
ir_instruction *ir = (ir_instruction *)node;
this->base_ir = ir;
- this->result = reg_undef;
+
ir->accept(this);
}
}
@@ -1559,14 +1502,10 @@ fs_visitor::visit(ir_loop *ir)
counter = *(variable_storage(ir->counter));
if (ir->from) {
- this->result = counter;
-
this->base_ir = ir->from;
- this->result = counter;
ir->from->accept(this);
- if (!this->result.equals(&counter))
- emit(BRW_OPCODE_MOV, counter, this->result);
+ emit(BRW_OPCODE_MOV, counter, this->result);
}
}
@@ -1574,7 +1513,6 @@ fs_visitor::visit(ir_loop *ir)
if (ir->to) {
this->base_ir = ir->to;
- this->result = reg_undef;
ir->to->accept(this);
fs_inst *inst = emit(BRW_OPCODE_CMP, reg_null_cmp, counter, this->result);
@@ -1588,13 +1526,11 @@ fs_visitor::visit(ir_loop *ir)
ir_instruction *ir = (ir_instruction *)node;
this->base_ir = ir;
- this->result = reg_undef;
ir->accept(this);
}
if (ir->increment) {
this->base_ir = ir->increment;
- this->result = reg_undef;
ir->increment->accept(this);
emit(BRW_OPCODE_ADD, counter, counter, this->result);
}
@@ -1644,7 +1580,7 @@ fs_visitor::visit(ir_function *ir)
foreach_list(node, &sig->body) {
ir_instruction *ir = (ir_instruction *)node;
this->base_ir = ir;
- this->result = reg_undef;
+
ir->accept(this);
}
}
--
1.7.6
More information about the mesa-dev
mailing list