Merge pull request #1026 from arrayfire/hotfixes-3.1.2

Hotfixes 3.1.2

CMakeModules/Version.cmake

@@ -3,7 +3,7 @@
 #
 SET(AF_VERSION_MAJOR "3")
 SET(AF_VERSION_MINOR "1")
-SET(AF_VERSION_PATCH "1")
+SET(AF_VERSION_PATCH "2")
 
 SET(AF_VERSION "${AF_VERSION_MAJOR}.${AF_VERSION_MINOR}.${AF_VERSION_PATCH}")
 SET(AF_API_VERSION_CURRENT ${AF_VERSION_MAJOR}${AF_VERSION_MINOR})

CMakeModules/build_forge.cmake

@@ -22,7 +22,7 @@ ENDIF()
 ExternalProject_Add(
     forge-ext
     GIT_REPOSITORY https://github.com/arrayfire/forge.git
-    GIT_TAG af3.1
+    GIT_TAG af3.1.2
     PREFIX "${prefix}"
     INSTALL_DIR "${prefix}"
     UPDATE_COMMAND ""

CMakeModules/osx_install/cpu_scripts/postinstall

@@ -3,7 +3,7 @@
 set -e
 set -o pipefail
 
-err_file=/tmp/AFInstaller.err
+err_file=/tmp/AFInstallerCPU.err
 brew=/usr/local/bin/brew
 
 if [ ! -f $brew ]; then
@@ -20,5 +20,5 @@ if [ -z $user ]; then
     exit 1
 fi
 
-su $user -c "$brew tap homebrew/versions"
-su $user -c "$brew install fftw glfw3 fontconfig"
+su $user -c "$brew tap homebrew/versions" 2> $err_file
+su $user -c "$brew install fftw glfw3 fontconfig" 2> $err_file

CMakeModules/osx_install/cuda_scripts/postinstall

@@ -19,5 +19,5 @@ if [ -z $user ]; then
     exit 1
 fi
 
-su $user -c "$brew tap homebrew/versions"
-su $user -c "$brew install glfw3 fontconfig"
+su $user -c "$brew tap homebrew/versions" 2> $err_file
+su $user -c "$brew install glfw3 fontconfig" 2> $err_file

CMakeModules/osx_install/opencl_scripts/postinstall

@@ -19,5 +19,5 @@ if [ -z $user ]; then
     exit 1
 fi
 
-su $user -c "$brew tap homebrew/versions"
-su $user -c "$brew install glfw3 fontconfig"
+su $user -c "$brew tap homebrew/versions" 2> $err_file
+su $user -c "$brew install glfw3 fontconfig" 2> $err_file

docs/CMakeLists.txt

@@ -22,7 +22,7 @@ ADD_CUSTOM_TARGET(docs
     ALL
     COMMAND ${DOXYGEN_EXECUTABLE} ${AF_DOCS_CONFIG_OUT}
     COMMAND cmake -E copy_directory ${ASSETS_DIR} ${CMAKE_CURRENT_BINARY_DIR}
-    COMMAND cmake -E remove_directory ${CMAKE_CURRENT_BINARY_DIR}/.git
+    COMMAND cmake -E remove ${CMAKE_CURRENT_BINARY_DIR}/.git
     WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}
     COMMENT "Generating Documentation"
     VERBATIM)

docs/details/array.dox

@@ -3,7 +3,7 @@
 \addtogroup arrayfire_func
 @{
 \defgroup array_mem_operator_paren operator()
-\ingroup index_mat 
+\ingroup index_mat
 
 \brief Gets a reference to a set of elements
 
@@ -92,7 +92,7 @@ Adds and assigns values
 \defgroup array_mem_operator_minus_eq operator-=
 \ingroup index_mat
 
-\brief Substracts and assigns the value(s) of val to the elements of the
+\brief Subtracts and assigns the value(s) of val to the elements of the
 af::array
 
 Substracts and assigns values
@@ -116,14 +116,5 @@ Multiplies and assigns values
 
 Divides and assigns values
 
-===============================================================================
-
-\defgroup array_mem_operator_plus operator+
-\ingroup arith_mat
-
-\brief Divides and assigns the value(s) of val to the elements of the af::array
-
-Divides and assigns values
-
 @}
 */

docs/pages/README.md

@@ -135,24 +135,22 @@ If you wish to cite ArrayFire in an academic publication, please use the
 following reference:
 
 Formatted:
-```
-Yalamanchili, P., Arshad, U., Mohammed, Z., Garigipati, P., Entschev, P.,
-Kloppenborg, B., Malcolm, James and Melonakos, J. (2015).
-ArrayFire - A high performance software library for parallel computing with an
-easy-to-use API. Atlanta: AccelerEyes. Retrieved from https://github.com/arrayfire/arrayfire
-```
+
+    Yalamanchili, P., Arshad, U., Mohammed, Z., Garigipati, P., Entschev, P.,
+    Kloppenborg, B., Malcolm, James and Melonakos, J. (2015).
+    ArrayFire - A high performance software library for parallel computing with an
+    easy-to-use API. Atlanta: AccelerEyes. Retrieved from https://github.com/arrayfire/arrayfire
 
 BibTeX:
-```bibtex
-@misc{Yalamanchili2015,
-abstract = {ArrayFire is a high performance software library for parallel computing with an easy-to-use API. Its array based function set makes parallel programming simple. ArrayFire's multiple backends (CUDA, OpenCL and native CPU) make it platform independent and highly portable. A few lines of code in ArrayFire can replace dozens of lines of parallel computing code, saving you valuable time and lowering development costs.},
-address = {Atlanta},
-author = {Yalamanchili, Pavan and Arshad, Umar and Mohammed, Zakiuddin and Garigipati, Pradeep and Entschev, Peter and Kloppenborg, Brian and Malcolm, James and Melonakos, John},
-publisher = {AccelerEyes},
-title = {{ArrayFire - A high performance software library for parallel computing with an easy-to-use API}},
-url = {https://github.com/arrayfire/arrayfire},
-year = {2015}
-}
-```
+
+    @misc{Yalamanchili2015,
+    abstract = {ArrayFire is a high performance software library for parallel computing with an easy-to-use API. Its array based function set makes parallel programming simple. ArrayFire's multiple backends (CUDA, OpenCL and native CPU) make it platform independent and highly portable. A few lines of code in ArrayFire can replace dozens of lines of parallel computing code, saving you valuable time and lowering development costs.},
+    address = {Atlanta},
+    author = {Yalamanchili, Pavan and Arshad, Umar and Mohammed, Zakiuddin and Garigipati, Pradeep and Entschev, Peter and Kloppenborg, Brian and Malcolm, James and Melonakos, John},
+    publisher = {AccelerEyes},
+    title = {{ArrayFire - A high performance software library for parallel computing with an easy-to-use API}},
+    url = {https://github.com/arrayfire/arrayfire},
+    year = {2015}
+    }
 
 ArrayFire development is funded by ArrayFire LLC and several third parties, please see the list of <a href="https://github.com/arrayfire/arrayfire/blob/master/ACKNOWLEDGEMENTS.md">acknowledgements</a>.

docs/pages/release_notes.md

@@ -1,6 +1,30 @@
 Release Notes {#releasenotes}
 ==============
 
+v3.1.2
+==============
+
+Bug Fixes
+---------
+
+* Fixed [bug](https://github.com/arrayfire/arrayfire/commit/4698f12) in assign that was causing test to fail
+* Fixed bug in convolve. Frequency condition now depends on kernel size only
+* Fixed [bug](https://github.com/arrayfire/arrayfire/issues/1005) in indexed reductions for complex type in OpenCL backend
+* Fixed [bug](https://github.com/arrayfire/arrayfire/issues/1006) in kernel name generation in ireduce for OpenCL backend
+* Fixed non-linear to linear indices in ireduce
+* Fixed [bug](https://github.com/arrayfire/arrayfire/issues/1011) in reductions for small arrays
+* Fixed [bug](https://github.com/arrayfire/arrayfire/issues/1010) in histogram for indexed arrays
+* Fixed [compiler error](https://github.com/arrayfire/arrayfire/issues/1015) CPUID for non-compliant devices
+* Fixed [failing tests](https://github.com/arrayfire/arrayfire/issues/1008) on i386 platforms
+* Add missing AFAPI
+
+Other
+-------
+
+* Documentation: Added missing examples and other corrections
+* Documentation: Fixed warnings in documentation building
+* Installers: Send error messages to log file in OSX Installer
+
 v3.1.1
 ==============
 

docs/pages/using_on_linux.md

@@ -133,7 +133,7 @@ for the CPU, OpenCL and CUDA backends repsectively).
 Here is a minimial example MakeFile which uses ArrayFire's CPU backend:
 
     LIBS=-lafcpu
-    LIB_PATHS=/usr/lib
+    LIB_PATHS=-L/usr/lib
     INCLUDES=-I/usr/include
     CC=g++ $(COMPILER_OPTIONS)
     COMPILER_OPTIONS=-std=c++11 -g

docs/pages/using_on_osx.md

@@ -81,7 +81,7 @@ for the CPU, OpenCL and CUDA backends repsectively).
 Here is a minimial example MakeFile which uses ArrayFire's CPU backend:
 
     LIBS=-lafcpu
-    LIB_PATHS=/usr/lib
+    LIB_PATHS=-L/usr/lib
     INCLUDES=-I/usr/include
     CC=g++ $(COMPILER_OPTIONS)
     COMPILER_OPTIONS=-std=c++11 -g

include/af/defines.h

@@ -50,8 +50,8 @@
     typedef long long   dim_t;
 #endif
 
-#if defined(__arm__) || defined(_M_ARM) || defined(_M_ARMT) || defined(_ARM) || defined(_TARGET_ARCH_ARM) || defined(__TARGET_ARCH_THUMB) || defined(__thumb__)
-#define ARM_ARCH
+#if defined(__x86_64__) || defined(_M_X64) || defined(__i386__) || defined(_M_IX86) || defined(_WIN64)
+#define USE_CPUID
 #endif
 
 #include <stdlib.h>

include/af/dim4.hpp

@@ -53,9 +53,9 @@ class AFAPI dim4
     const   dim_t* get() const   { return dims; }
 };
 
-dim4 operator+(const dim4& first, const dim4& second);
-dim4 operator-(const dim4& first, const dim4& second);
-dim4 operator*(const dim4& first, const dim4& second);
+AFAPI dim4 operator+(const dim4& first, const dim4& second);
+AFAPI dim4 operator-(const dim4& first, const dim4& second);
+AFAPI dim4 operator*(const dim4& first, const dim4& second);
 
 static inline
 std::ostream&

include/arrayfire.h

@@ -39,7 +39,7 @@
       @defgroup data_mat Functions to create arrays.
       constant, random, range, etc.
 
-      @defgroup index_mat Indexing operation on arrays
+      @defgroup index_mat Assignment & Indexing operation on arrays
       Access sub regions of an array object
 
       @defgroup manip_mat Move and Reorder array content
@@ -234,36 +234,50 @@
 */
 
 /**
-\example helloworld.cpp
-\example pi.cpp
-\example integer.cpp
-\example rainfall.cpp
-\example vectorize.cpp
-\example black_scholes_options.cpp
-\example monte_carlo_options.cpp
-\example heston_model.cpp
+\example matching.cpp
+\example fast.cpp
 \example harris.cpp
-\example kmeans.cpp
-\example knn.cpp
-\example bagging.cpp
-\example naive_bayes.cpp
+\example susan.cpp
+\example logistic_regression.cpp
+\example rbm.cpp
 \example perceptron.cpp
 \example neural_network.cpp
-\example rbm.cpp
+\example bagging.cpp
+\example naive_bayes.cpp
 \example deep_belief_net.cpp
-\example logistic_regression.cpp
+\example kmeans.cpp
+\example softmax_regression.cpp
+\example knn.cpp
+\example monte_carlo_options.cpp
+\example heston_model.cpp
+\example black_scholes_options.cpp
+\example blas.cpp
+\example fft.cpp
+\example pi.cpp
+\example svd.cpp
+\example cholesky.cpp
+\example qr.cpp
+\example lu.cpp
 \example conway.cpp
-\example conway_pretty.cpp
-\example fractal.cpp
 \example histogram.cpp
+\example fractal.cpp
 \example plot2d.cpp
-\example brain_segmentation.cpp
-\example image_demo.cpp
-\example image_editing.cpp
+\example conway_pretty.cpp
+\example helloworld.cpp
+\example vectorize.cpp
+\example integer.cpp
+\example convolve.cpp
+\example rainfall.cpp
 \example morphing.cpp
-\example optical_flow.cpp
+\example image_demo.cpp
+\example brain_segmentation.cpp
 \example pyramids.cpp
+\example binary_thresholding.cpp
+\example optical_flow.cpp
+\example adaptive_thresholding.cpp
+\example image_editing.cpp
 \example edge.cpp
+\example filters.cpp
 */
 
 #include "af/compatible.h"

src/api/c/assign.cpp

@@ -231,7 +231,7 @@ af_err af_assign_gen(af_array *out,
         ARG_ASSERT(2, (ndims == 1) || (ndims == (dim_t)lInfo.ndims()));
 
         if (ndims == 1 && ndims != (dim_t)lInfo.ndims()) {
-            af_array tmp_in, tmp_out;
+            af_array tmp_in = 0, tmp_out = 0;
             AF_CHECK(af_flat(&tmp_in, lhs));
             AF_CHECK(af_assign_gen(&tmp_out, tmp_in, ndims, indexs, rhs_));
             AF_CHECK(af_moddims(out, tmp_out, lInfo.ndims(), lInfo.dims().get()));

src/api/c/convolve.cpp

@@ -146,12 +146,12 @@ bool isFreqDomain(const af_array &signal, const af_array filter, af_conv_domain
 
     if (identifyBatchKind<baseDim>(sdims, fdims) == CONVOLVE_BATCH_DIFF) return true;
 
-    int batch = 1;
+    int kbatch = 1;
     for(int i = 3; i >= baseDim; i--) {
-        batch *= std::max(fdims[i], sdims[i]);
+        kbatch *= fdims[i];
     }
 
-    if (batch >= 10) return true;
+    if (kbatch >= 10) return true;
 
     if (baseDim == 1) {
         if (fdims[0] > 128) return true;
@@ -177,44 +177,46 @@ af_err af_convolve1(af_array *out, const af_array signal, const af_array filter,
     try {
         if (isFreqDomain<1>(signal, filter, domain))
             return af_fft_convolve1(out, signal, filter, mode);
-    } CATCHALL;
 
-    if (mode == AF_CONV_EXPAND)
+        if (mode == AF_CONV_EXPAND)
             return convolve<1, true >(out, signal, filter);
-    else
-        return convolve<1, false>(out, signal, filter);
+        else
+            return convolve<1, false>(out, signal, filter);
+    } CATCHALL;
 }
 
 af_err af_convolve2(af_array *out, const af_array signal, const af_array filter, const af_conv_mode mode, af_conv_domain domain)
 {
     try {
         if (isFreqDomain<2>(signal, filter, domain))
             return af_fft_convolve2(out, signal, filter, mode);
-    } CATCHALL;
 
-    if (mode == AF_CONV_EXPAND)
-        return convolve<2, true >(out, signal, filter);
-    else
-        return convolve<2, false>(out, signal, filter);
+        if (mode == AF_CONV_EXPAND)
+            return convolve<2, true >(out, signal, filter);
+        else
+            return convolve<2, false>(out, signal, filter);
+    } CATCHALL;
 }
 
 af_err af_convolve3(af_array *out, const af_array signal, const af_array filter, const af_conv_mode mode, af_conv_domain domain)
 {
     try {
         if (isFreqDomain<3>(signal, filter, domain))
             return af_fft_convolve3(out, signal, filter, mode);
-    } CATCHALL;
 
-    if (mode == AF_CONV_EXPAND)
-        return convolve<3, true >(out, signal, filter);
-    else
-        return convolve<3, false>(out, signal, filter);
+        if (mode == AF_CONV_EXPAND)
+            return convolve<3, true >(out, signal, filter);
+        else
+            return convolve<3, false>(out, signal, filter);
+    } CATCHALL;
 }
 
 af_err af_convolve2_sep(af_array *out, const af_array signal, const af_array col_filter, const af_array row_filter, const af_conv_mode mode)
 {
-    if (mode == AF_CONV_EXPAND)
-        return convolve2_sep<true >(out, signal, col_filter, row_filter);
-    else
-        return convolve2_sep<false>(out, signal, col_filter, row_filter);
+    try {
+        if (mode == AF_CONV_EXPAND)
+            return convolve2_sep<true >(out, signal, col_filter, row_filter);
+        else
+            return convolve2_sep<false>(out, signal, col_filter, row_filter);
+    } CATCHALL;
 }