KonstantinSeurer
/
mesa
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

glsl: Optimize pow(2, x) --> exp2(x). 

On Haswell, POW takes 24 cycles, while EXP2 only takes 14.  Plus, using
POW requires putting 2.0 in a register, while EXP2 doesn't.

I believe that EXP2 will be faster than POW on basically all GPUs, so
it makes sense to optimize it.

Looking at the savage2 subset of shader-db:
total instructions in shared programs: 113225 -> 113179 (-0.04%)
instructions in affected programs:     2139 -> 2093 (-2.15%)
instances of 'math pow':               795 -> 749 (-6.14%)
instances of 'math exp':               389 -> 435 (11.8%)

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
This commit is contained in:
Kenneth Graunke 2014-01-05 22:57:01 -08:00
parent 5e3fd6a9db
commit 847bc36a38
1 changed files with 11 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
src/glsl/opt_algebraic.cpp
View File

@ -87,6 +87,12 @@ is_vec_one(ir_constant *ir)
return (ir == NULL) ? false : ir->is_one();
}
static inline bool
is_vec_two(ir_constant *ir)
{
return (ir == NULL) ? false : ir->is_value(2.0, 2);
}
static inline bool
is_vec_negative_one(ir_constant *ir)
{
@ -420,6 +426,11 @@ ir_algebraic_visitor::handle_expression(ir_expression *ir)
/* 1^x == 1 */
if (is_vec_one(op_const[0]))
return op_const[0];
/* pow(2,x) == exp2(x) */
if (is_vec_two(op_const[0]))
return expr(ir_unop_exp2, ir->operands[1]);
break;
case ir_unop_rcp: