[Bf-blender-cvs] [6ef435bb29f] tmp-macs-arm-cycles: BLI: add BLI_simd.h header to wrap SSE includes
Brecht Van Lommel
noreply at git.blender.org
Mon Feb 15 19:57:29 CET 2021
Commit: 6ef435bb29fc5eef3be159792220830d948f3e1c
Author: Brecht Van Lommel
Date: Sun Feb 14 14:20:51 2021 +0100
Branches: tmp-macs-arm-cycles
https://developer.blender.org/rB6ef435bb29fc5eef3be159792220830d948f3e1c
BLI: add BLI_simd.h header to wrap SSE includes
In preparation of adding Neon support.
Ref T78710
===================================================================
M source/blender/blenkernel/intern/lattice_deform.c
A source/blender/blenlib/BLI_simd.h
M source/blender/blenlib/intern/math_base_inline.c
M source/blender/blenlib/intern/math_color_inline.c
M source/blender/blenlib/intern/math_geom.c
M source/blender/blenlib/intern/math_matrix.c
M source/blender/compositor/operations/COM_BlurBaseOperation.cpp
M source/blender/compositor/operations/COM_BlurBaseOperation.h
M source/blender/compositor/operations/COM_GaussianXBlurOperation.cpp
M source/blender/compositor/operations/COM_GaussianXBlurOperation.h
M source/blender/compositor/operations/COM_GaussianYBlurOperation.cpp
M source/blender/compositor/operations/COM_GaussianYBlurOperation.h
M source/blender/modifiers/intern/MOD_meshdeform.c
===================================================================
diff --git a/source/blender/blenkernel/intern/lattice_deform.c b/source/blender/blenkernel/intern/lattice_deform.c
index 81c2f492f7e..2651042939f 100644
--- a/source/blender/blenkernel/intern/lattice_deform.c
+++ b/source/blender/blenkernel/intern/lattice_deform.c
@@ -31,6 +31,7 @@
#include "MEM_guardedalloc.h"
#include "BLI_math.h"
+#include "BLI_simd.h"
#include "BLI_task.h"
#include "BLI_utildefines.h"
@@ -49,10 +50,6 @@
#include "BKE_deform.h"
-#ifdef __SSE2__
-# include <emmintrin.h>
-#endif
-
/* -------------------------------------------------------------------- */
/** \name Lattice Deform API
* \{ */
@@ -171,7 +168,7 @@ void BKE_lattice_deform_data_eval_co(LatticeDeformData *lattice_deform_data,
/* vgroup influence */
float co_prev[4] = {0}, weight_blend = 0.0f;
copy_v3_v3(co_prev, co);
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
__m128 co_vec = _mm_loadu_ps(co_prev);
#endif
@@ -232,7 +229,7 @@ void BKE_lattice_deform_data_eval_co(LatticeDeformData *lattice_deform_data,
u = v * tu[uu - ui + 1];
idx_u = CLAMPIS(uu, 0, idx_u_max);
const int idx = idx_w + idx_v + idx_u;
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
{
__m128 weight_vec = _mm_set1_ps(u);
/* We need to address special case for last item to avoid accessing invalid memory. */
@@ -256,7 +253,7 @@ void BKE_lattice_deform_data_eval_co(LatticeDeformData *lattice_deform_data,
}
}
}
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
{
copy_v3_v3(co, (float *)&co_vec);
}
diff --git a/source/blender/blenlib/BLI_simd.h b/source/blender/blenlib/BLI_simd.h
new file mode 100644
index 00000000000..1518b6c1de2
--- /dev/null
+++ b/source/blender/blenlib/BLI_simd.h
@@ -0,0 +1,28 @@
+/*
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ */
+
+#pragma once
+
+/** \file
+ * \ingroup bli
+ *
+ * SIMD instruction support.
+ */
+
+#if defined(__SSE2__)
+# include <emmintrin.h>
+# define BLI_HAVE_SSE2
+#endif
diff --git a/source/blender/blenlib/intern/math_base_inline.c b/source/blender/blenlib/intern/math_base_inline.c
index 28aa81e5858..39945960e68 100644
--- a/source/blender/blenlib/intern/math_base_inline.c
+++ b/source/blender/blenlib/intern/math_base_inline.c
@@ -31,11 +31,8 @@
#include <stdio.h>
#include <stdlib.h>
-#ifdef __SSE2__
-# include <emmintrin.h>
-#endif
-
#include "BLI_math_base.h"
+#include "BLI_simd.h"
#ifdef __cplusplus
extern "C" {
@@ -685,10 +682,10 @@ MINLINE int integer_digits_i(const int i)
/* Internal helpers for SSE2 implementation.
*
- * NOTE: Are to be called ONLY from inside `#ifdef __SSE2__` !!!
+ * NOTE: Are to be called ONLY from inside `#ifdef BLI_HAVE_SSE2` !!!
*/
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
/* Calculate initial guess for arg^exp based on float representation
* This method gives a constant bias, which can be easily compensated by
@@ -769,7 +766,7 @@ MALWAYS_INLINE __m128 _bli_math_blend_sse(const __m128 mask, const __m128 a, con
return _mm_or_ps(_mm_and_ps(mask, a), _mm_andnot_ps(mask, b));
}
-#endif /* __SSE2__ */
+#endif /* BLI_HAVE_SSE2 */
/* Low level conversion functions */
MINLINE unsigned char unit_float_to_uchar_clamp(float val)
diff --git a/source/blender/blenlib/intern/math_color_inline.c b/source/blender/blenlib/intern/math_color_inline.c
index 26828cb8717..4c50c1c7af8 100644
--- a/source/blender/blenlib/intern/math_color_inline.c
+++ b/source/blender/blenlib/intern/math_color_inline.c
@@ -34,7 +34,7 @@
/******************************** Color Space ********************************/
-# ifdef __SSE2__
+# ifdef BLI_HAVE_SSE2
MALWAYS_INLINE __m128 srgb_to_linearrgb_v4_simd(const __m128 c)
{
@@ -75,7 +75,7 @@ MINLINE void linearrgb_to_srgb_v3_v3(float srgb[3], const float linear[3])
srgb[2] = r[2];
}
-# else /* __SSE2__ */
+# else /* BLI_HAVE_SSE2 */
MINLINE void srgb_to_linearrgb_v3_v3(float linear[3], const float srgb[3])
{
@@ -90,7 +90,7 @@ MINLINE void linearrgb_to_srgb_v3_v3(float srgb[3], const float linear[3])
srgb[1] = linearrgb_to_srgb(linear[1]);
srgb[2] = linearrgb_to_srgb(linear[2]);
}
-# endif /* __SSE2__ */
+# endif /* BLI_HAVE_SSE2 */
MINLINE void srgb_to_linearrgb_v4(float linear[4], const float srgb[4])
{
diff --git a/source/blender/blenlib/intern/math_geom.c b/source/blender/blenlib/intern/math_geom.c
index 5d78bb0b901..b7e94e6c512 100644
--- a/source/blender/blenlib/intern/math_geom.c
+++ b/source/blender/blenlib/intern/math_geom.c
@@ -5907,7 +5907,7 @@ static float ff_quad_form_factor(float *p, float *n, float *q0, float *q1, float
#if 0
-# include <xmmintrin.h>
+# include "BLI_simd.h"
static __m128 sse_approx_acos(__m128 x)
{
diff --git a/source/blender/blenlib/intern/math_matrix.c b/source/blender/blenlib/intern/math_matrix.c
index 6b5efc3f8c4..b460d75d77f 100644
--- a/source/blender/blenlib/intern/math_matrix.c
+++ b/source/blender/blenlib/intern/math_matrix.c
@@ -277,7 +277,7 @@ void mul_m4_m4m4_uniq(float R[4][4], const float A[4][4], const float B[4][4])
BLI_assert(!ELEM(R, A, B));
/* matrix product: R[j][k] = A[j][i] . B[i][k] */
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
__m128 A0 = _mm_loadu_ps(A[0]);
__m128 A1 = _mm_loadu_ps(A[1]);
__m128 A2 = _mm_loadu_ps(A[2]);
diff --git a/source/blender/compositor/operations/COM_BlurBaseOperation.cpp b/source/blender/compositor/operations/COM_BlurBaseOperation.cpp
index 3fe154c397e..612a71037f7 100644
--- a/source/blender/compositor/operations/COM_BlurBaseOperation.cpp
+++ b/source/blender/compositor/operations/COM_BlurBaseOperation.cpp
@@ -88,7 +88,7 @@ float *BlurBaseOperation::make_gausstab(float rad, int size)
return gausstab;
}
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
__m128 *BlurBaseOperation::convert_gausstab_sse(const float *gausstab, int size)
{
int n = 2 * size + 1;
diff --git a/source/blender/compositor/operations/COM_BlurBaseOperation.h b/source/blender/compositor/operations/COM_BlurBaseOperation.h
index c452b2e4ea1..56dacc96710 100644
--- a/source/blender/compositor/operations/COM_BlurBaseOperation.h
+++ b/source/blender/compositor/operations/COM_BlurBaseOperation.h
@@ -23,16 +23,14 @@
#define MAX_GAUSSTAB_RADIUS 30000
-#ifdef __SSE2__
-# include <emmintrin.h>
-#endif
+#include "BLI_simd.h"
class BlurBaseOperation : public NodeOperation, public QualityStepHelper {
private:
protected:
BlurBaseOperation(DataType data_type);
float *make_gausstab(float rad, int size);
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
__m128 *convert_gausstab_sse(const float *gausstab, int size);
#endif
float *make_dist_fac_inverse(float rad, int size, int falloff);
diff --git a/source/blender/compositor/operations/COM_GaussianXBlurOperation.cpp b/source/blender/compositor/operations/COM_GaussianXBlurOperation.cpp
index e08d30e5ddf..90333f7dd79 100644
--- a/source/blender/compositor/operations/COM_GaussianXBlurOperation.cpp
+++ b/source/blender/compositor/operations/COM_GaussianXBlurOperation.cpp
@@ -26,7 +26,7 @@
GaussianXBlurOperation::GaussianXBlurOperation() : BlurBaseOperation(COM_DT_COLOR)
{
this->m_gausstab = nullptr;
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
this->m_gausstab_sse = nullptr;
#endif
this->m_filtersize = 0;
@@ -55,7 +55,7 @@ void GaussianXBlurOperation::initExecution()
/* TODO(sergey): De-duplicate with the case below and Y blur. */
this->m_gausstab = BlurBaseOperation::make_gausstab(rad, m_filtersize);
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
this->m_gausstab_sse = BlurBaseOperation::convert_gausstab_sse(this->m_gausstab, m_filtersize);
#endif
}
@@ -70,7 +70,7 @@ void GaussianXBlurOperation::updateGauss()
m_filtersize = min_ii(ceil(rad), MAX_GAUSSTAB_RADIUS);
this->m_gausstab = BlurBaseOperation::make_gausstab(rad, m_filtersize);
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
this->m_gausstab_sse = BlurBaseOperation::convert_gausstab_sse(this->m_gausstab, m_filtersize);
#endif
}
@@ -95,7 +95,7 @@ void GaussianXBlurOperation::executePixel(float output[4], int x, int y, void *d
int offsetadd = getOffsetAdd();
int bufferindex = ((xmin - bufferstartx) * 4) + ((ymin - bufferstarty) * 4 * bufferwidth);
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
__m128 accum_r = _mm_load_ps(color_accum);
for (int nx = xmin, index = (xmin - x) + this->m_filtersize; nx < xmax;
nx += step, index += step) {
@@ -162,7 +162,7 @@ void GaussianXBlurOperation::deinitExecution()
MEM_freeN(this->m_gausstab);
this->m_gausstab = nullptr;
}
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
if (this->m_gausstab_sse) {
MEM_freeN(this->m_gausstab_sse);
this->m_gausstab_sse = nullptr;
diff --git a/source/blender/compositor/operations/COM_GaussianXBlurOperation.h b/source/blender/compositor/operations/COM_GaussianXBlurOperation.h
index 9348c05f906..b2bcd79e716 100644
--- a/source/blender/compositor/operations/COM_GaussianXBlurOperation.h
+++ b/source/blender/compositor/operations/COM_GaussianXBlurOperation.h
@@ -24,7 +24,7 @@
class GaussianXBlurOperation : public BlurBaseOperation {
private:
float *m_gausstab;
-#ifdef __SSE2__
+#ifdef BLI_HAVE_SSE2
__m128 *m_gausstab_sse;
#endif
int m_filtersize;
diff --git a/source/blender/compositor/operations/COM_GaussianYBlurOperation.cpp b/source/blender/compositor/operations/COM_GaussianYBlurOperation.cpp
index 7710b065ccd..c5b3cf24239 100644
--- a/source/blender/compositor/operations/COM_GaussianYBlurOperation.cpp
+++ b/source/blender/compositor/operations/COM_GaussianYBlurOperation.cpp
@@ -26,7 +26,7 @@
@@ Diff output truncated at 10240 characters. @@
More information about the Bf-blender-cvs
mailing list