i965 implements this with a single (multiple destination) instruction, ADDC. Emitting ADDC directly from uaddCarry() would be ideal, but our optimization passes don't know how to copy with expressions with side-effects.
Radeon has an ADDC_UINT instruction that only generates the carry bit. I've chosen to go this route and implement uaddCarry() by doing the addition and the carry operations separately. --- src/glsl/builtin_functions.cpp | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/src/glsl/builtin_functions.cpp b/src/glsl/builtin_functions.cpp index 31f7489..7da9142 100644 --- a/src/glsl/builtin_functions.cpp +++ b/src/glsl/builtin_functions.cpp @@ -512,6 +512,7 @@ private: B1(fma) B2(ldexp) B2(frexp) + B1(uaddCarry) #undef B0 #undef B1 #undef B2 @@ -1838,6 +1839,12 @@ builtin_builder::create_builtins() _frexp(glsl_type::vec3_type, glsl_type::ivec3_type), _frexp(glsl_type::vec4_type, glsl_type::ivec4_type), NULL); + add_function("uaddCarry", + _uaddCarry(glsl_type::uint_type), + _uaddCarry(glsl_type::uvec2_type), + _uaddCarry(glsl_type::uvec3_type), + _uaddCarry(glsl_type::uvec4_type), + NULL); #undef F #undef FI #undef FIU @@ -3584,6 +3591,20 @@ builtin_builder::_frexp(const glsl_type *x_type, const glsl_type *exp_type) return sig; } + +ir_function_signature * +builtin_builder::_uaddCarry(const glsl_type *type) +{ + ir_variable *x = in_var(type, "x"); + ir_variable *y = in_var(type, "y"); + ir_variable *carry = out_var(type, "carry"); + MAKE_SIG(type, gpu_shader5, 3, x, y, carry); + + body.emit(assign(carry, ir_builder::carry(x, y))); + body.emit(ret(add(x, y))); + + return sig; +} /** @} */ /******************************************************************************/ -- 1.8.3.2 _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev