OpenHarmony/third_party_mesa3d
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[915] Use a quartic term to improve the accuracy of SIN results.

Browse Source 
This is described in the link in the comment, and is the same technique that
r300 uses.
This commit is contained in:
Eric Anholt
2008-02-06 15:38:16 -08:00
parent d98abcbef0
commit 2551a5ee80
1 changed files with 54 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files

77
src/mesa/drivers/dri/i915/i915_fragprog.c
View File

@@ -43,11 +43,19 @@
#include "i915_context.h" #include "i915_context.h"
#include "i915_program.h" #include "i915_program.h"
static const GLfloat sin_quad_constants[4] = { static const GLfloat sin_quad_constants[2][4] = {
4.0, {
-4.0, 2.0,
2.0, -1.0,
-1.0 .5,
0.0
},
{
4.0,
-4.0,
1.0 / (2.0 * M_PI),
.2225
}
}; };
static const GLfloat sin_constants[4] = { 1.0, static const GLfloat sin_constants[4] = { 1.0,
@@ -341,7 +349,7 @@ upload_program(struct i915_fragment_program *p)
while (1) { while (1) {
GLuint src0, src1, src2, flags; GLuint src0, src1, src2, flags;
GLuint tmp = 0, consts = 0; GLuint tmp = 0, consts0 = 0, consts1 = 0;
switch (inst->Opcode) { switch (inst->Opcode) {
case OPCODE_ABS: case OPCODE_ABS:
@@ -690,15 +698,16 @@ upload_program(struct i915_fragment_program *p)
case OPCODE_SIN: case OPCODE_SIN:
src0 = src_vector(p, &inst->SrcReg[0], program); src0 = src_vector(p, &inst->SrcReg[0], program);
tmp = i915_get_utemp(p); tmp = i915_get_utemp(p);
consts = i915_emit_const4fv(p, sin_quad_constants); consts0 = i915_emit_const4fv(p, sin_quad_constants[0]);
consts1 = i915_emit_const4fv(p, sin_quad_constants[1]);
/* Reduce range from repeating about [-pi,pi] to [-1,1] */ /* Reduce range from repeating about [-pi,pi] to [-1,1] */
i915_emit_arith(p, i915_emit_arith(p,
A0_MAD, A0_MAD,
tmp, A0_DEST_CHANNEL_X, 0, tmp, A0_DEST_CHANNEL_X, 0,
src0, src0,
i915_emit_const1f(p, 1.0 / (2.0 * M_PI)), swizzle(consts1, Z, ZERO, ZERO, ZERO), /* 1/(2pi) */
i915_emit_const1f(p, .5)); swizzle(consts0, Z, ZERO, ZERO, ZERO)); /* .5 */
i915_emit_arith(p, A0_FRC, tmp, A0_DEST_CHANNEL_X, 0, tmp, 0, 0); i915_emit_arith(p, A0_FRC, tmp, A0_DEST_CHANNEL_X, 0, tmp, 0, 0);
@@ -706,19 +715,15 @@ upload_program(struct i915_fragment_program *p)
A0_MAD, A0_MAD,
tmp, A0_DEST_CHANNEL_X, 0, tmp, A0_DEST_CHANNEL_X, 0,
tmp, tmp,
swizzle(consts, Z, ZERO, ZERO, ZERO), /* 2 */ swizzle(consts0, X, ZERO, ZERO, ZERO), /* 2 */
swizzle(consts, W, ZERO, ZERO, ZERO)); /* -1 */ swizzle(consts0, Y, ZERO, ZERO, ZERO)); /* -1 */
/* Compute sin using a quadratic. While it has increased total /* Compute sin using a quadratic and quartic. It gives continuity
* error over the range, it does give continuity that the 4-component * that repeating the Taylor series lacks every 2*pi, and has
* Taylor series lacks when repeating the range due to its * reduced error.
* sin(PI) != 0 behavior.
* *
* The idea was described at: * The idea was described at:
* http://www.devmaster.net/forums/showthread.php?t=5784 * http://www.devmaster.net/forums/showthread.php?t=5784
*
* If we're concerned about the error of this approximation, we should
* probably incorporate a second pass to include a x**4 factor.
*/ */
/* tmp.y = abs(tmp.x); {x, abs(x), 0, 0} */ /* tmp.y = abs(tmp.x); {x, abs(x), 0, 0} */
@@ -737,15 +742,41 @@ upload_program(struct i915_fragment_program *p)
tmp, tmp,
0); 0);
/* result = tmp.xy DP sin_quad_constants.xy */ /* tmp.x = tmp.xy DP sin_quad_constants[2].xy */
i915_emit_arith(p, i915_emit_arith(p,
A0_DP3, A0_DP3,
tmp, A0_DEST_CHANNEL_X, 0,
tmp,
swizzle(consts1, X, Y, ZERO, ZERO),
0);
/* tmp.x now contains a first approximation (y). Now, weight it
* against tmp.y**2 to get closer.
*/
i915_emit_arith(p,
A0_MAX,
tmp, A0_DEST_CHANNEL_Y, 0,
swizzle(tmp, ZERO, X, ZERO, ZERO),
negate(swizzle(tmp, ZERO, X, ZERO, ZERO), 0, 1, 0, 0),
0);
/* tmp.y = tmp.x * tmp.y - tmp.x; {y, y * abs(y) - y, 0, 0} */
i915_emit_arith(p,
A0_MAD,
tmp, A0_DEST_CHANNEL_Y, 0,
swizzle(tmp, ZERO, X, ZERO, ZERO),
swizzle(tmp, ZERO, Y, ZERO, ZERO),
negate(swizzle(tmp, ZERO, X, ZERO, ZERO), 0, 1, 0, 0));
/* result = .2225 * tmp.y + tmp.x =.2225(y * abs(y) - y) + y= */
i915_emit_arith(p,
A0_MAD,
get_result_vector(p, inst), get_result_vector(p, inst),
get_result_flags(inst), 0, get_result_flags(inst), 0,
tmp, swizzle(consts1, W, W, W, W),
swizzle(i915_emit_const4fv(p, sin_quad_constants), swizzle(tmp, Y, Y, Y, Y),
X, Y, ZERO, ZERO), swizzle(tmp, X, X, X, X));
0);
break; break;
case OPCODE_SLT: case OPCODE_SLT: