glsl: Use array deref for access to vector components

We've assumed that we could lower per-component vector access from

  vec[i] = scalar

to

  vec = ir_triop_vector_insert(vec, scalar, i)

but with SSBOs (and compute shader SLM and tesselation outputs) this is
no longer valid. If a vector is "externally visible", multiple threads
can write independent components simultaneously. With lowering to
ir_triop_vector_insert, each thread read the entire vector, changes one
component, then writes out the entire vector. This is racy.

Instead of generating a ir_binop_vector_extract when we see v[i], we
generate ir_dereference_array. We then add a lowering pass to lower the
ir_dereference_array to ir_binop_vector_extract for rvalues and for to
vector_insert for lvalues in a separate lowering pass.

The resulting IR is the same as before, but we now have a window between
ast->ir conversion and the lowering pass where v[i] appears in the IR as
an array deref. This lets us run lowering passes that lower the vector
access to I/O (eg for SSBO load/store) before we lower the per-component
access to full vector writes.

Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Signed-off-by: Kristian Høgsberg Kristensen <krh@bitplanet.net>
This commit is contained in:
Kristian Høgsberg Kristensen
2015-11-04 14:58:54 -08:00
parent 60dd5287ff
commit 96b22fb080
10 changed files with 138 additions and 68 deletions
+6 -18
View File
@@ -256,18 +256,10 @@ verify_parameter_modes(_mesa_glsl_parse_state *state,
actual->variable_referenced()->name);
return false;
} else if (!actual->is_lvalue()) {
/* Even though ir_binop_vector_extract is not an l-value, let it
* slop through. generate_call will handle it correctly.
*/
ir_expression *const expr = ((ir_rvalue *) actual)->as_expression();
if (expr == NULL
|| expr->operation != ir_binop_vector_extract
|| !expr->operands[0]->is_lvalue()) {
_mesa_glsl_error(&loc, state,
"function parameter '%s %s' is not an lvalue",
mode, formal->name);
return false;
}
_mesa_glsl_error(&loc, state,
"function parameter '%s %s' is not an lvalue",
mode, formal->name);
return false;
}
}
@@ -376,12 +368,8 @@ fix_parameter(void *mem_ctx, ir_rvalue *actual, const glsl_type *formal_type,
ir_rvalue *lhs = actual;
if (expr != NULL && expr->operation == ir_binop_vector_extract) {
rhs = new(mem_ctx) ir_expression(ir_triop_vector_insert,
expr->operands[0]->type,
expr->operands[0]->clone(mem_ctx, NULL),
rhs,
expr->operands[1]->clone(mem_ctx, NULL));
lhs = expr->operands[0]->clone(mem_ctx, NULL);
lhs == new(mem_ctx) ir_dereference_array(expr->operands[0]->clone(mem_ctx, NULL),
expr->operands[1]->clone(mem_ctx, NULL));
}
ir_assignment *const assignment_2 = new(mem_ctx) ir_assignment(lhs, rhs);