Merge branch 'master' into tmp-usd-alab-v2-T100452tmp-usd-alab-v2-T100452

6 files changed, 145 insertions, 53 deletions

diff --git a/intern/cycles/util/CMakeLists.txt b/intern/cycles/util/CMakeLists.txt
index 81a7607baab..997d574a3b0 100644
--- a/intern/cycles/util/CMakeLists.txt
+++ b/intern/cycles/util/CMakeLists.txt
@@ -3,7 +3,6 @@
 
 set(INC
   ..
-  ../../glew-mx
 )
 
 set(INC_SYS
@@ -149,6 +148,4 @@ endif()
 include_directories(${INC})
 include_directories(SYSTEM ${INC_SYS})
 
-add_definitions(${GL_DEFINITIONS})
-
 cycles_add_library(cycles_util "${LIB}" ${SRC} ${SRC_HEADERS})
diff --git a/intern/cycles/util/defines.h b/intern/cycles/util/defines.h
index c7118ca09c9..1969529eff0 100644
--- a/intern/cycles/util/defines.h
+++ b/intern/cycles/util/defines.h
@@ -89,46 +89,6 @@
 #  define UNLIKELY(x) (x)
 #endif
 
-#if defined(__GNUC__) || defined(__clang__)
-#  if defined(__cplusplus)
-/* Some magic to be sure we don't have reference in the type. */
-template<typename T> static inline T decltype_helper(T x)
-{
-  return x;
-}
-#    define TYPEOF(x) decltype(decltype_helper(x))
-#  else
-#    define TYPEOF(x) typeof(x)
-#  endif
-#endif
-
-/* Causes warning:
- * incompatible types when assigning to type 'Foo' from type 'Bar'
- * ... the compiler optimizes away the temp var */
-#ifdef __GNUC__
-#  define CHECK_TYPE(var, type) \
-    { \
-      TYPEOF(var) * __tmp; \
-      __tmp = (type *)NULL; \
-      (void)__tmp; \
-    } \
-    (void)0
-
-#  define CHECK_TYPE_PAIR(var_a, var_b) \
-    { \
-      TYPEOF(var_a) * __tmp; \
-      __tmp = (typeof(var_b) *)NULL; \
-      (void)__tmp; \
-    } \
-    (void)0
-#else
-#  define CHECK_TYPE(var, type)
-#  define CHECK_TYPE_PAIR(var_a, var_b)
-#endif
-
-/* can be used in simple macros */
-#define CHECK_TYPE_INLINE(val, type) ((void)(((type)0) != (val)))
-
 #ifndef __KERNEL_GPU__
 #  include <cassert>
 #  define util_assert(statement) assert(statement)
diff --git a/intern/cycles/util/hash.h b/intern/cycles/util/hash.h
index 081b33025d8..4f83f331229 100644
--- a/intern/cycles/util/hash.h
+++ b/intern/cycles/util/hash.h
@@ -4,10 +4,28 @@
 #ifndef __UTIL_HASH_H__
 #define __UTIL_HASH_H__
 
+#include "util/math.h"
 #include "util/types.h"
 
 CCL_NAMESPACE_BEGIN
 
+/* [0, uint_max] -> [0.0, 1.0) */
+ccl_device_forceinline float uint_to_float_excl(uint n)
+{
+  // Note: we divide by 4294967808 instead of 2^32 because the latter
+  // leads to a [0.0, 1.0] mapping instead of [0.0, 1.0) due to floating
+  // point rounding error. 4294967808 unfortunately leaves (precisely)
+  // one unused ulp between the max number this outputs and 1.0, but
+  // that's the best you can do with this construction.
+  return (float)n * (1.0f / 4294967808.0f);
+}
+
+/* [0, uint_max] -> [0.0, 1.0] */
+ccl_device_forceinline float uint_to_float_incl(uint n)
+{
+  return (float)n * (1.0f / (float)0xFFFFFFFFu);
+}
+
 /* ***** Jenkins Lookup3 Hash Functions ***** */
 
 /* Source: http://burtleburtle.net/bob/c/lookup3.c */
@@ -116,22 +134,22 @@ ccl_device_inline uint hash_uint4(uint kx, uint ky, uint kz, uint kw)
 
 ccl_device_inline float hash_uint_to_float(uint kx)
 {
-  return (float)hash_uint(kx) / (float)0xFFFFFFFFu;
+  return uint_to_float_incl(hash_uint(kx));
 }
 
 ccl_device_inline float hash_uint2_to_float(uint kx, uint ky)
 {
-  return (float)hash_uint2(kx, ky) / (float)0xFFFFFFFFu;
+  return uint_to_float_incl(hash_uint2(kx, ky));
 }
 
 ccl_device_inline float hash_uint3_to_float(uint kx, uint ky, uint kz)
 {
-  return (float)hash_uint3(kx, ky, kz) / (float)0xFFFFFFFFu;
+  return uint_to_float_incl(hash_uint3(kx, ky, kz));
 }
 
 ccl_device_inline float hash_uint4_to_float(uint kx, uint ky, uint kz, uint kw)
 {
-  return (float)hash_uint4(kx, ky, kz, kw) / (float)0xFFFFFFFFu;
+  return uint_to_float_incl(hash_uint4(kx, ky, kz, kw));
 }
 
 /* Hashing float or float[234] into a float in the range [0, 1]. */
@@ -359,6 +377,123 @@ ccl_device_inline avxi hash_avxi4(avxi kx, avxi ky, avxi kz, avxi kw)
 
 #endif
 
+/* ***** Hash Prospector Hash Functions *****
+ *
+ * These are based on the high-quality 32-bit hash/mixing functions from
+ * https://github.com/skeeto/hash-prospector
+ */
+
+ccl_device_inline uint hash_hp_uint(uint i)
+{
+  // The actual mixing function from Hash Prospector.
+  i ^= i >> 16;
+  i *= 0x21f0aaad;
+  i ^= i >> 15;
+  i *= 0xd35a2d97;
+  i ^= i >> 15;
+
+  // The xor is just to make input zero not map to output zero.
+  // The number is randomly selected and isn't special.
+  return i ^ 0xe6fe3beb;
+}
+
+/* Seedable version of hash_hp_uint() above. */
+ccl_device_inline uint hash_hp_seeded_uint(uint i, uint seed)
+{
+  // Manipulate the seed so it doesn't interact poorly with n when they
+  // are both e.g. incrementing.  This isn't fool-proof, but is good
+  // enough for practical use.
+  seed ^= seed << 19;
+
+  return hash_hp_uint(i ^ seed);
+}
+
+/* Outputs [0.0, 1.0). */
+ccl_device_inline float hash_hp_float(uint i)
+{
+  return uint_to_float_excl(hash_hp_uint(i));
+}
+
+/* Outputs [0.0, 1.0). */
+ccl_device_inline float hash_hp_seeded_float(uint i, uint seed)
+{
+  return uint_to_float_excl(hash_hp_seeded_uint(i, seed));
+}
+
+/* ***** Modified Wang Hash Functions *****
+ *
+ * These are based on a bespoke modified version of the Wang hash, and
+ * can serve as a faster hash when quality isn't critical.
+ *
+ * The original Wang hash is documented here:
+ * https://www.burtleburtle.net/bob/hash/integer.html
+ */
+
+ccl_device_inline uint hash_wang_seeded_uint(uint i, uint seed)
+{
+  i = (i ^ 61) ^ seed;
+  i += i << 3;
+  i ^= i >> 4;
+  i *= 0x27d4eb2d;
+  return i;
+}
+
+/* Outputs [0.0, 1.0). */
+ccl_device_inline float hash_wang_seeded_float(uint i, uint seed)
+{
+  return uint_to_float_excl(hash_wang_seeded_uint(i, seed));
+}
+
+/* ***** Index Shuffling Hash Function *****
+ *
+ * This function takes an index, the length of the thing the index points
+ * into, and returns a shuffled index.  For example, if you pass indices
+ * 0 through 19 to this function with a length parameter of 20, it will
+ * return the indices in a shuffled order with no repeats.  Indices
+ * larger than the length parameter will simply repeat the same shuffled
+ * pattern over and over.
+ *
+ * This is useful for iterating over an array in random shuffled order
+ * without having to shuffle the array itself.
+ *
+ * Passing different seeds results in different random shuffles.
+ *
+ * This function runs in average O(1) time.
+ *
+ * See https://andrew-helmer.github.io/permute/ for details on how this
+ * works.
+ */
+ccl_device_inline uint hash_shuffle_uint(uint i, uint length, uint seed)
+{
+  i = i % length;
+  uint mask = (1 << (32 - count_leading_zeros(length - 1))) - 1;
+
+  do {
+    i ^= seed;
+    i *= 0xe170893d;
+    i ^= seed >> 16;
+    i ^= (i & mask) >> 4;
+    i ^= seed >> 8;
+    i *= 0x0929eb3f;
+    i ^= seed >> 23;
+    i ^= (i & mask) >> 1;
+    i *= 1 | seed >> 27;
+    i *= 0x6935fa69;
+    i ^= (i & mask) >> 11;
+    i *= 0x74dcb303;
+    i ^= (i & mask) >> 2;
+    i *= 0x9e501cc3;
+    i ^= (i & mask) >> 2;
+    i *= 0xc860a3df;
+    i &= mask;
+    i ^= i >> 5;
+  } while (i >= length);
+
+  return i;
+}
+
+/* ********** */
+
 #ifndef __KERNEL_GPU__
 static inline uint hash_string(const char *str)
 {
diff --git a/intern/cycles/util/math.h b/intern/cycles/util/math.h
index 0585dcc8ad5..0905b3ec5c9 100644
--- a/intern/cycles/util/math.h
+++ b/intern/cycles/util/math.h
@@ -886,16 +886,16 @@ ccl_device_inline float2 map_to_tube(const float3 co)
 
 ccl_device_inline float2 map_to_sphere(const float3 co)
 {
-  float l = len(co);
+  float l = dot(co, co);
   float u, v;
   if (l > 0.0f) {
     if (UNLIKELY(co.x == 0.0f && co.y == 0.0f)) {
       u = 0.0f; /* Otherwise domain error. */
     }
     else {
-      u = (1.0f - atan2f(co.x, co.y) / M_PI_F) / 2.0f;
+      u = (0.5f - atan2f(co.x, co.y) * M_1_2PI_F);
     }
-    v = 1.0f - safe_acosf(co.z / l) / M_PI_F;
+    v = 1.0f - safe_acosf(co.z / sqrtf(l)) * M_1_PI_F;
   }
   else {
     u = v = 0.0f;
diff --git a/intern/cycles/util/opengl.h b/intern/cycles/util/opengl.h
index 090deb861c4..fefee4ec022 100644
--- a/intern/cycles/util/opengl.h
+++ b/intern/cycles/util/opengl.h
@@ -7,6 +7,6 @@
 /* OpenGL header includes, used everywhere we use OpenGL, to deal with
  * platform differences in one central place. */
 
-#include <GL/glew.h>
+#include <epoxy/gl.h>
 
 #endif /* __UTIL_OPENGL_H__ */
diff --git a/intern/cycles/util/version.h b/intern/cycles/util/version.h
index 8260dd4f55d..fb02c3eaeb7 100644
--- a/intern/cycles/util/version.h
+++ b/intern/cycles/util/version.h
@@ -9,7 +9,7 @@
 CCL_NAMESPACE_BEGIN
 
 #define CYCLES_VERSION_MAJOR 3
-#define CYCLES_VERSION_MINOR 2
+#define CYCLES_VERSION_MINOR 4
 #define CYCLES_VERSION_PATCH 0
 
 #define CYCLES_MAKE_VERSION_STRING2(a, b, c) #a "." #b "." #c
@@ -19,7 +19,7 @@ CCL_NAMESPACE_BEGIN
 
 /* Blender libraries version compatible with this version */
 
-#define CYCLES_BLENDER_LIBRARIES_VERSION 3.1
+#define CYCLES_BLENDER_LIBRARIES_VERSION 3.3
 
 CCL_NAMESPACE_END