Bump version to 0.3.0, update documentation

CHANGELOG.md

@@ -1,5 +1,29 @@
 # CHANGELOG
 
+## [0.3.0] - 2022-07-08
+
+- Add support for FP16 storage type, see `wp.float16`
+- Add support for per-dimension byte strides, see `wp.array.strides`
+- Add support for passing Python classes as kernel arguments, see `@wp.struct` decorator
+- Add additional bounds checks for builtin matrix types
+- Add additional floating point checks, see `wp.config.verify_fp`
+- Add interleaved user source with generated code to aid debugging
+- Add generalized GPU marching cubes implementation, see `wp.MarchingCubes` class 
+- Add additional scalar*matrix vector operators
+- Add support for retrieving a single row from builtin types, e.g.: `r = m33[i]`
+- Add  `wp.log2()` and `wp.log10()` builtins
+- Add support for quickly instancing `wp.sim.ModelBuilder` objects to improve env. creation performance for RL
+- Remove custom CUB version and improve compatability with CUDA 11.7
+- Fix to preserve external user-gradients when calling `wp.Tape.zero()`
+- Fix to only allocate gradient of a Torch tensor if `requires_grad=True`
+- Fix for ray-cast precision in edge case on GPU (watertightness issue)
+- Fix for kernel hot-reload when definition changes
+- Fix for NVCC warnings on Linux
+- Fix for generated function names when kernels are defined as class functions 
+- Fix for reload of generated CPU kernel code on Linux
+- Fix for example scripts to ouput USD at 60 timecodes per-second (better Kit compatability)
+
+
 ## [0.2.3] - 2022-06-13
 
 - Fix for incorrect 4d array bounds checking

VERSION.md

@@ -1 +1 @@
-0.2.3
+0.3.0

docs/_build/html/_sources/modules/functions.rst.txt

@@ -16,6 +16,7 @@ Scalar Types
 .. autoclass:: uint32
 .. autoclass:: int64
 .. autoclass:: uint64
+.. autoclass:: float16
 .. autoclass:: float32
 .. autoclass:: float64
 Vector Types
@@ -154,6 +155,16 @@ Scalar Math
    Return the natural log (base-e) of x, where x is positive.
 
 
+.. function:: log2(x: float32) -> float
+
+   Return the natural log (base-2) of x, where x is positive.
+
+
+.. function:: log10(x: float32) -> float
+
+   Return the natural log (base-10) of x, where x is positive.
+
+
 .. function:: exp(x: float32) -> float
 
    Return base-e exponential, e^x.
@@ -733,6 +744,36 @@ Utility
    Atomically subtract ``value`` onto the array at location given by indices.
 
 
+.. function:: index(a: vec2, i: int32) -> float
+
+
+.. function:: index(a: vec3, i: int32) -> float
+
+
+.. function:: index(a: vec4, i: int32) -> float
+
+
+.. function:: index(a: quat, i: int32) -> float
+
+
+.. function:: index(a: mat22, i: int32) -> vec2
+
+
+.. function:: index(a: mat22, i: int32, j: int32) -> float
+
+
+.. function:: index(a: mat33, i: int32) -> vec3
+
+
+.. function:: index(a: mat33, i: int32, j: int32) -> float
+
+
+.. function:: index(a: mat44, i: int32) -> vec4
+
+
+.. function:: index(a: mat44, i: int32, j: int32) -> float
+
+
 .. function:: expect_eq(arg1: int8, arg2: int8) -> None
 
    Prints an error to stdout if arg1 and arg2 are not equal
@@ -773,6 +814,11 @@ Utility
    Prints an error to stdout if arg1 and arg2 are not equal
 
 
+.. function:: expect_eq(arg1: float16, arg2: float16) -> None
+
+   Prints an error to stdout if arg1 and arg2 are not equal
+
+
 .. function:: expect_eq(arg1: float32, arg2: float32) -> None
 
    Prints an error to stdout if arg1 and arg2 are not equal
@@ -833,6 +879,11 @@ Utility
    Prints an error to stdout if arg1 and arg2 are not equal
 
 
+.. function:: lerp(a: float16, b: float16, t: float32) -> float16
+
+   Linearly interpolate two values a and b using factor t, computed as ``a*(1-t) + b*t``
+
+
 .. function:: lerp(a: float32, b: float32, t: float32) -> float32
 
    Linearly interpolate two values a and b using factor t, computed as ``a*(1-t) + b*t``
@@ -1240,6 +1291,15 @@ Operators
 .. function:: mul(x: float32, y: quat) -> quat
 
 
+.. function:: mul(x: float32, y: mat22) -> mat22
+
+
+.. function:: mul(x: float32, y: mat33) -> mat33
+
+
+.. function:: mul(x: float32, y: mat44) -> mat44
+
+
 .. function:: mul(x: vec2, y: float32) -> vec2
 
 

docs/_build/html/index.html

@@ -70,7 +70,7 @@
 
 
               <div class="version">
-                0.2.3
+                0.3.0
               </div>
 
 

docs/_build/html/modules/introduction.html

@@ -71,7 +71,7 @@
 
 
               <div class="version">
-                0.2.3
+                0.3.0
               </div>
 
 

docs/_build/html/modules/runtime.html

@@ -71,7 +71,7 @@
 
 
               <div class="version">
-                0.2.3
+                0.3.0
               </div>
 
 
@@ -451,7 +451,7 @@ <h3>Multi-dimensional arrays<a class="headerlink" href="#multi-dimensional-array
 
 <dl class="py class">
 <dt class="sig sig-object py" id="warp.array">
-<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">warp.</span></span><span class="sig-name descname"><span class="pre">array</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">data</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">dtype</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">shape</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">length</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">0</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">ptr</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">capacity</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">0</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">device</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">copy</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">owner</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">ndim</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">requires_grad</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">False</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#warp.array" title="Permalink to this definition">¶</a></dt>
+<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">warp.</span></span><span class="sig-name descname"><span class="pre">array</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">data</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">dtype</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">shape</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">strides</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">length</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">0</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">ptr</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">capacity</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">0</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">device</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">copy</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">owner</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">ndim</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">requires_grad</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">False</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#warp.array" title="Permalink to this definition">¶</a></dt>
 <dd><p>Constructs a new Warp array object from existing data.</p>
 <p>When the <code class="docutils literal notranslate"><span class="pre">data</span></code> argument is a valid list, tuple, or ndarray the array will be constructed from this object’s data.
 For objects that are not stored sequentially in memory (e.g.: a list), then the data will first
@@ -466,6 +466,7 @@ <h3>Multi-dimensional arrays<a class="headerlink" href="#multi-dimensional-array
 <li><p><strong>data</strong> (<em>Union</em><em>[</em><a class="reference external" href="https://docs.python.org/3/library/stdtypes.html#list" title="(in Python v3.10)"><em>list</em></a><em>, </em><a class="reference external" href="https://docs.python.org/3/library/stdtypes.html#tuple" title="(in Python v3.10)"><em>tuple</em></a><em>, </em><em>ndarray</em><em>]</em>) – </p></li>
 <li><p><strong>dtype</strong> (<em>Union</em>) – One of the built-in types, e.g.: <a class="reference internal" href="functions.html#id39" title="warp.mat33"><code class="xref py py-class docutils literal notranslate"><span class="pre">warp.mat33</span></code></a>, if dtype is None and data an ndarray then it will be inferred from the array data type</p></li>
 <li><p><strong>shape</strong> (<em>Tuple</em>) – Dimensions of the array</p></li>
+<li><p><strong>strides</strong> (<em>Tuple</em>) – Number of bytes in each dimension between successive elements of the array</p></li>
 <li><p><strong>length</strong> (<a class="reference external" href="https://docs.python.org/3/library/functions.html#int" title="(in Python v3.10)"><em>int</em></a>) – Number of elements (rows) of the data type (deprecated, users should use <cite>shape</cite> argument)</p></li>
 <li><p><strong>ptr</strong> (<a class="reference internal" href="functions.html#warp.uint64" title="warp.uint64"><em>uint64</em></a>) – Address of an external memory address to alias (data should be None)</p></li>
 <li><p><strong>capacity</strong> (<a class="reference external" href="https://docs.python.org/3/library/functions.html#int" title="(in Python v3.10)"><em>int</em></a>) – Maximum size in bytes of the ptr allocation (data should be None)</p></li>

docs/_build/html/search.html

@@ -72,7 +72,7 @@
 
 
               <div class="version">
-                0.2.3
+                0.3.0
               </div>
 
 

docs/conf.py

@@ -21,7 +21,7 @@
 copyright = '2022, NVIDIA'
 author = 'NVIDIA'
 
-version = '0.2.3'
+version = '0.3.0'
 
 # -- General configuration ---------------------------------------------------
 

docs/modules/functions.rst

@@ -16,6 +16,7 @@ Scalar Types
 .. autoclass:: uint32
 .. autoclass:: int64
 .. autoclass:: uint64
+.. autoclass:: float16
 .. autoclass:: float32
 .. autoclass:: float64
 Vector Types
@@ -154,6 +155,16 @@ Scalar Math
    Return the natural log (base-e) of x, where x is positive.
 
 
+.. function:: log2(x: float32) -> float
+
+   Return the natural log (base-2) of x, where x is positive.
+
+
+.. function:: log10(x: float32) -> float
+
+   Return the natural log (base-10) of x, where x is positive.
+
+
 .. function:: exp(x: float32) -> float
 
    Return base-e exponential, e^x.
@@ -733,6 +744,36 @@ Utility
    Atomically subtract ``value`` onto the array at location given by indices.
 
 
+.. function:: index(a: vec2, i: int32) -> float
+
+
+.. function:: index(a: vec3, i: int32) -> float
+
+
+.. function:: index(a: vec4, i: int32) -> float
+
+
+.. function:: index(a: quat, i: int32) -> float
+
+
+.. function:: index(a: mat22, i: int32) -> vec2
+
+
+.. function:: index(a: mat22, i: int32, j: int32) -> float
+
+
+.. function:: index(a: mat33, i: int32) -> vec3
+
+
+.. function:: index(a: mat33, i: int32, j: int32) -> float
+
+
+.. function:: index(a: mat44, i: int32) -> vec4
+
+
+.. function:: index(a: mat44, i: int32, j: int32) -> float
+
+
 .. function:: expect_eq(arg1: int8, arg2: int8) -> None
 
    Prints an error to stdout if arg1 and arg2 are not equal
@@ -773,6 +814,11 @@ Utility
    Prints an error to stdout if arg1 and arg2 are not equal
 
 
+.. function:: expect_eq(arg1: float16, arg2: float16) -> None
+
+   Prints an error to stdout if arg1 and arg2 are not equal
+
+
 .. function:: expect_eq(arg1: float32, arg2: float32) -> None
 
    Prints an error to stdout if arg1 and arg2 are not equal
@@ -833,6 +879,11 @@ Utility
    Prints an error to stdout if arg1 and arg2 are not equal
 
 
+.. function:: lerp(a: float16, b: float16, t: float32) -> float16
+
+   Linearly interpolate two values a and b using factor t, computed as ``a*(1-t) + b*t``
+
+
 .. function:: lerp(a: float32, b: float32, t: float32) -> float32
 
    Linearly interpolate two values a and b using factor t, computed as ``a*(1-t) + b*t``
@@ -1240,6 +1291,15 @@ Operators
 .. function:: mul(x: float32, y: quat) -> quat
 
 
+.. function:: mul(x: float32, y: mat22) -> mat22
+
+
+.. function:: mul(x: float32, y: mat33) -> mat33
+
+
+.. function:: mul(x: float32, y: mat44) -> mat44
+
+
 .. function:: mul(x: vec2, y: float32) -> vec2
 
 

exts/omni.warp/config/extension.toml

@@ -1,7 +1,7 @@
 
 [package]
 # Semantic Versioning is used: https://semver.org/
-version = "0.2.3"
+version = "0.3.0"
 authors = ["Miles Macklin"]
 title = "Warp"
 description="A Python framework for high-performance simulation and graphics programming"

exts/omni.warp/docs/CHANGELOG.md

@@ -1,5 +1,29 @@
 # CHANGELOG
 
+## [0.3.0] - 2022-07-08
+
+- Add support for FP16 storage type, see `wp.float16`
+- Add support for per-dimension byte strides, see `wp.array.strides`
+- Add support for passing Python classes as kernel arguments, see `@wp.struct` decorator
+- Add additional bounds checks for builtin matrix types
+- Add additional floating point checks, see `wp.config.verify_fp`
+- Add interleaved user source with generated code to aid debugging
+- Add generalized GPU marching cubes implementation, see `wp.MarchingCubes` class 
+- Add additional scalar*matrix vector operators
+- Add support for retrieving a single row from builtin types, e.g.: `r = m33[i]`
+- Add  `wp.log2()` and `wp.log10()` builtins
+- Add support for quickly instancing `wp.sim.ModelBuilder` objects to improve env. creation performance for RL
+- Remove custom CUB version and improve compatability with CUDA 11.7
+- Fix to preserve external user-gradients when calling `wp.Tape.zero()`
+- Fix to only allocate gradient of a Torch tensor if `requires_grad=True`
+- Fix for ray-cast precision in edge case on GPU (watertightness issue)
+- Fix for kernel hot-reload when definition changes
+- Fix for NVCC warnings on Linux
+- Fix for generated function names when kernels are defined as class functions 
+- Fix for reload of generated CPU kernel code on Linux
+- Fix for example scripts to ouput USD at 60 timecodes per-second (better Kit compatability)
+
+
 ## [0.2.3] - 2022-06-13
 
 - Fix for incorrect 4d array bounds checking

setup.py

@@ -2,7 +2,7 @@
 
 setuptools.setup(
     name="warp-lang",
-    version="0.2.3",
+    version="0.3.0",
     author="NVIDIA",
     author_email="[email protected]",
     description="A Python framework for high-performance simulation and graphics programming",

warp/config.py

@@ -7,7 +7,7 @@
 
 import os
 
-version = "0.2.3"
+version = "0.3.0"
 
 cuda_path = None        # path to local CUDA toolchain, if None at init time warp will attempt to find the SDK using CUDA_PATH env var
 

warp/native/builtin.h

@@ -960,6 +960,13 @@ CUDA_CALLABLE inline T lerp(const T& a, const T& b, float t)
     return a*(1.0-t) + b*t;
 }
 
+template <>
+CUDA_CALLABLE inline float16 lerp(const float16& a, const float16& b, float t)
+{
+    return float(a)*(1.0-t) + float(b)*t;
+}
+
+
 template <typename T>
 CUDA_CALLABLE inline void adj_lerp(const T& a, const T& b, float t, T& adj_a, T& adj_b, float& adj_t, const T& adj_ret)
 {