mv SVD to main lib

Former-commit-id: d17743b27031cb30ff4c8d62cca2f0888c4646f2

RELEASE_HISTORY.txt

@@ -16,7 +16,8 @@ Version | Short description
 
 ## Version 1.2 Changes ##
 This change introduces better organization of dependencies and removes some
-deprecated/repeated functions
+deprecated/repeated functions. The 3x3 svd code and dependent functions
+(including ARAP) were absorbed into the main library.
 
 ### Deprecated/repeated functions ###
 

include/igl/Singular_Value_Decomposition_Givens_QR_Factorization_Kernel.hpp

@@ -0,0 +1,128 @@
+//#####################################################################
+// Copyright (c) 2010-2011, Eftychios Sifakis.
+//
+// Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
+//   * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
+//   * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or
+//     other materials provided with the distribution.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING,
+// BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT
+// SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+// DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+// INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+//#####################################################################
+
+//###########################################################
+// Compute the Givens half-angle, construct the Givens quaternion and the rotation sine/cosine (for the full angle)
+//###########################################################
+
+#ifdef _WIN32
+  #undef max
+  #undef min
+#endif
+
+ENABLE_SCALAR_IMPLEMENTATION(Ssh.f=SANPIVOT.f*SANPIVOT.f;)                                ENABLE_SSE_IMPLEMENTATION(Vsh=_mm_mul_ps(VANPIVOT,VANPIVOT);)                             ENABLE_AVX_IMPLEMENTATION(Vsh=_mm256_mul_ps(VANPIVOT,VANPIVOT);)
+ENABLE_SCALAR_IMPLEMENTATION(Ssh.ui=(Ssh.f>=Ssmall_number.f)?0xffffffff:0;)               ENABLE_SSE_IMPLEMENTATION(Vsh=_mm_cmpge_ps(Vsh,Vsmall_number);)                           ENABLE_AVX_IMPLEMENTATION(Vsh=_mm256_cmp_ps(Vsh,Vsmall_number, _CMP_GE_OS);) //ENABLE_AVX_IMPLEMENTATION(Vsh=_mm256_cmpge_ps(Vsh,Vsmall_number);)
+ENABLE_SCALAR_IMPLEMENTATION(Ssh.ui=Ssh.ui&SANPIVOT.ui;)                                  ENABLE_SSE_IMPLEMENTATION(Vsh=_mm_and_ps(Vsh,VANPIVOT);)                                  ENABLE_AVX_IMPLEMENTATION(Vsh=_mm256_and_ps(Vsh,VANPIVOT);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp5.f=0.;)                                                 ENABLE_SSE_IMPLEMENTATION(Vtmp5=_mm_xor_ps(Vtmp5,Vtmp5);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp5=_mm256_xor_ps(Vtmp5,Vtmp5);)
+ENABLE_SCALAR_IMPLEMENTATION(Sch.f=Stmp5.f-SAPIVOT.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vch=_mm_sub_ps(Vtmp5,VAPIVOT);)                                 ENABLE_AVX_IMPLEMENTATION(Vch=_mm256_sub_ps(Vtmp5,VAPIVOT);)
+ENABLE_SCALAR_IMPLEMENTATION(Sch.f=std::max(Sch.f,SAPIVOT.f);)                            ENABLE_SSE_IMPLEMENTATION(Vch=_mm_max_ps(Vch,VAPIVOT);)                                   ENABLE_AVX_IMPLEMENTATION(Vch=_mm256_max_ps(Vch,VAPIVOT);)
+ENABLE_SCALAR_IMPLEMENTATION(Sch.f=std::max(Sch.f,Ssmall_number.f);)                      ENABLE_SSE_IMPLEMENTATION(Vch=_mm_max_ps(Vch,Vsmall_number);)                             ENABLE_AVX_IMPLEMENTATION(Vch=_mm256_max_ps(Vch,Vsmall_number);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp5.ui=(SAPIVOT.f>=Stmp5.f)?0xffffffff:0;)                 ENABLE_SSE_IMPLEMENTATION(Vtmp5=_mm_cmpge_ps(VAPIVOT,Vtmp5);)                             ENABLE_AVX_IMPLEMENTATION(Vtmp5=_mm256_cmp_ps(VAPIVOT,Vtmp5, _CMP_GE_OS);) //ENABLE_AVX_IMPLEMENTATION(Vtmp5=_mm256_cmpge_ps(VAPIVOT,Vtmp5);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Sch.f*Sch.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vch,Vch);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vch,Vch);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Ssh.f*Ssh.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vsh,Vsh);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vsh,Vsh);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Stmp1.f+Stmp2.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_add_ps(Vtmp1,Vtmp2);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_add_ps(Vtmp1,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=rsqrt(Stmp2.f);)                                     ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_rsqrt_ps(Vtmp2);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_rsqrt_ps(Vtmp2);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp4.f=Stmp1.f*Sone_half.f;)                                ENABLE_SSE_IMPLEMENTATION(Vtmp4=_mm_mul_ps(Vtmp1,Vone_half);)                             ENABLE_AVX_IMPLEMENTATION(Vtmp4=_mm256_mul_ps(Vtmp1,Vone_half);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp3.f=Stmp1.f*Stmp4.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp3=_mm_mul_ps(Vtmp1,Vtmp4);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp3=_mm256_mul_ps(Vtmp1,Vtmp4);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp3.f=Stmp1.f*Stmp3.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp3=_mm_mul_ps(Vtmp1,Vtmp3);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp3=_mm256_mul_ps(Vtmp1,Vtmp3);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp3.f=Stmp2.f*Stmp3.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp3=_mm_mul_ps(Vtmp2,Vtmp3);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp3=_mm256_mul_ps(Vtmp2,Vtmp3);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Stmp1.f+Stmp4.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_add_ps(Vtmp1,Vtmp4);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_add_ps(Vtmp1,Vtmp4);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Stmp1.f-Stmp3.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_sub_ps(Vtmp1,Vtmp3);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_sub_ps(Vtmp1,Vtmp3);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Stmp1.f*Stmp2.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vtmp1,Vtmp2);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vtmp1,Vtmp2);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Sch.f=Sch.f+Stmp1.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vch=_mm_add_ps(Vch,Vtmp1);)                                     ENABLE_AVX_IMPLEMENTATION(Vch=_mm256_add_ps(Vch,Vtmp1);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.ui=~Stmp5.ui&Ssh.ui;)                                  ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_andnot_ps(Vtmp5,Vsh);)                                ENABLE_AVX_IMPLEMENTATION(Vtmp1=Vch;)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.ui=~Stmp5.ui&Sch.ui;)                                  ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_andnot_ps(Vtmp5,Vch);)                                ENABLE_AVX_IMPLEMENTATION(Vch=_mm256_blendv_ps(Vsh,Vch,Vtmp5);)
+ENABLE_SCALAR_IMPLEMENTATION(Sch.ui=Stmp5.ui&Sch.ui;)                                     ENABLE_SSE_IMPLEMENTATION(Vch=_mm_and_ps(Vtmp5,Vch);)                                     ENABLE_AVX_IMPLEMENTATION(Vsh=_mm256_blendv_ps(Vtmp1,Vsh,Vtmp5);)
+ENABLE_SCALAR_IMPLEMENTATION(Ssh.ui=Stmp5.ui&Ssh.ui;)                                     ENABLE_SSE_IMPLEMENTATION(Vsh=_mm_and_ps(Vtmp5,Vsh);)
+ENABLE_SCALAR_IMPLEMENTATION(Sch.ui=Sch.ui|Stmp1.ui;)                                     ENABLE_SSE_IMPLEMENTATION(Vch=_mm_or_ps(Vch,Vtmp1);)
+ENABLE_SCALAR_IMPLEMENTATION(Ssh.ui=Ssh.ui|Stmp2.ui;)                                     ENABLE_SSE_IMPLEMENTATION(Vsh=_mm_or_ps(Vsh,Vtmp2);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Sch.f*Sch.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vch,Vch);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vch,Vch);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Ssh.f*Ssh.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vsh,Vsh);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vsh,Vsh);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Stmp1.f+Stmp2.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_add_ps(Vtmp1,Vtmp2);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_add_ps(Vtmp1,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=rsqrt(Stmp2.f);)                                     ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_rsqrt_ps(Vtmp2);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_rsqrt_ps(Vtmp2);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp4.f=Stmp1.f*Sone_half.f;)                                ENABLE_SSE_IMPLEMENTATION(Vtmp4=_mm_mul_ps(Vtmp1,Vone_half);)                             ENABLE_AVX_IMPLEMENTATION(Vtmp4=_mm256_mul_ps(Vtmp1,Vone_half);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp3.f=Stmp1.f*Stmp4.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp3=_mm_mul_ps(Vtmp1,Vtmp4);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp3=_mm256_mul_ps(Vtmp1,Vtmp4);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp3.f=Stmp1.f*Stmp3.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp3=_mm_mul_ps(Vtmp1,Vtmp3);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp3=_mm256_mul_ps(Vtmp1,Vtmp3);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp3.f=Stmp2.f*Stmp3.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp3=_mm_mul_ps(Vtmp2,Vtmp3);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp3=_mm256_mul_ps(Vtmp2,Vtmp3);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Stmp1.f+Stmp4.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_add_ps(Vtmp1,Vtmp4);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_add_ps(Vtmp1,Vtmp4);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Stmp1.f-Stmp3.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_sub_ps(Vtmp1,Vtmp3);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_sub_ps(Vtmp1,Vtmp3);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Sch.f=Sch.f*Stmp1.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vch=_mm_mul_ps(Vch,Vtmp1);)                                     ENABLE_AVX_IMPLEMENTATION(Vch=_mm256_mul_ps(Vch,Vtmp1);)
+ENABLE_SCALAR_IMPLEMENTATION(Ssh.f=Ssh.f*Stmp1.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vsh=_mm_mul_ps(Vsh,Vtmp1);)                                     ENABLE_AVX_IMPLEMENTATION(Vsh=_mm256_mul_ps(Vsh,Vtmp1);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Sc.f=Sch.f*Sch.f;)                                           ENABLE_SSE_IMPLEMENTATION(Vc=_mm_mul_ps(Vch,Vch);)                                        ENABLE_AVX_IMPLEMENTATION(Vc=_mm256_mul_ps(Vch,Vch);)ENABLE_SCALAR_IMPLEMENTATION(Ss.f=Ssh.f*Ssh.f;)                                           ENABLE_SSE_IMPLEMENTATION(Vs=_mm_mul_ps(Vsh,Vsh);)                                        ENABLE_AVX_IMPLEMENTATION(Vs=_mm256_mul_ps(Vsh,Vsh);)
+ENABLE_SCALAR_IMPLEMENTATION(Sc.f=Sc.f-Ss.f;)                                             ENABLE_SSE_IMPLEMENTATION(Vc=_mm_sub_ps(Vc,Vs);)                                          ENABLE_AVX_IMPLEMENTATION(Vc=_mm256_sub_ps(Vc,Vs);)
+ENABLE_SCALAR_IMPLEMENTATION(Ss.f=Ssh.f*Sch.f;)                                           ENABLE_SSE_IMPLEMENTATION(Vs=_mm_mul_ps(Vsh,Vch);)                                        ENABLE_AVX_IMPLEMENTATION(Vs=_mm256_mul_ps(Vsh,Vch);)
+ENABLE_SCALAR_IMPLEMENTATION(Ss.f=Ss.f+Ss.f;)                                             ENABLE_SSE_IMPLEMENTATION(Vs=_mm_add_ps(Vs,Vs);)                                          ENABLE_AVX_IMPLEMENTATION(Vs=_mm256_add_ps(Vs,Vs);)
+
+//###########################################################
+// Rotate matrix A
+//###########################################################
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Ss.f*SA11.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vs,VA11);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vs,VA11);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Ss.f*SA21.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vs,VA21);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vs,VA21);)
+ENABLE_SCALAR_IMPLEMENTATION(SA11.f=Sc.f*SA11.f;)                                         ENABLE_SSE_IMPLEMENTATION(VA11=_mm_mul_ps(Vc,VA11);)                                      ENABLE_AVX_IMPLEMENTATION(VA11=_mm256_mul_ps(Vc,VA11);)
+ENABLE_SCALAR_IMPLEMENTATION(SA21.f=Sc.f*SA21.f;)                                         ENABLE_SSE_IMPLEMENTATION(VA21=_mm_mul_ps(Vc,VA21);)                                      ENABLE_AVX_IMPLEMENTATION(VA21=_mm256_mul_ps(Vc,VA21);)
+ENABLE_SCALAR_IMPLEMENTATION(SA11.f=SA11.f+Stmp2.f;)                                      ENABLE_SSE_IMPLEMENTATION(VA11=_mm_add_ps(VA11,Vtmp2);)                                   ENABLE_AVX_IMPLEMENTATION(VA11=_mm256_add_ps(VA11,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(SA21.f=SA21.f-Stmp1.f;)                                      ENABLE_SSE_IMPLEMENTATION(VA21=_mm_sub_ps(VA21,Vtmp1);)                                   ENABLE_AVX_IMPLEMENTATION(VA21=_mm256_sub_ps(VA21,Vtmp1);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Ss.f*SA12.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vs,VA12);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vs,VA12);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Ss.f*SA22.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vs,VA22);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vs,VA22);)
+ENABLE_SCALAR_IMPLEMENTATION(SA12.f=Sc.f*SA12.f;)                                         ENABLE_SSE_IMPLEMENTATION(VA12=_mm_mul_ps(Vc,VA12);)                                      ENABLE_AVX_IMPLEMENTATION(VA12=_mm256_mul_ps(Vc,VA12);)
+ENABLE_SCALAR_IMPLEMENTATION(SA22.f=Sc.f*SA22.f;)                                         ENABLE_SSE_IMPLEMENTATION(VA22=_mm_mul_ps(Vc,VA22);)                                      ENABLE_AVX_IMPLEMENTATION(VA22=_mm256_mul_ps(Vc,VA22);)
+ENABLE_SCALAR_IMPLEMENTATION(SA12.f=SA12.f+Stmp2.f;)                                      ENABLE_SSE_IMPLEMENTATION(VA12=_mm_add_ps(VA12,Vtmp2);)                                   ENABLE_AVX_IMPLEMENTATION(VA12=_mm256_add_ps(VA12,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(SA22.f=SA22.f-Stmp1.f;)                                      ENABLE_SSE_IMPLEMENTATION(VA22=_mm_sub_ps(VA22,Vtmp1);)                                   ENABLE_AVX_IMPLEMENTATION(VA22=_mm256_sub_ps(VA22,Vtmp1);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Ss.f*SA13.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vs,VA13);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vs,VA13);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Ss.f*SA23.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vs,VA23);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vs,VA23);)
+ENABLE_SCALAR_IMPLEMENTATION(SA13.f=Sc.f*SA13.f;)                                         ENABLE_SSE_IMPLEMENTATION(VA13=_mm_mul_ps(Vc,VA13);)                                      ENABLE_AVX_IMPLEMENTATION(VA13=_mm256_mul_ps(Vc,VA13);)
+ENABLE_SCALAR_IMPLEMENTATION(SA23.f=Sc.f*SA23.f;)                                         ENABLE_SSE_IMPLEMENTATION(VA23=_mm_mul_ps(Vc,VA23);)                                      ENABLE_AVX_IMPLEMENTATION(VA23=_mm256_mul_ps(Vc,VA23);)
+ENABLE_SCALAR_IMPLEMENTATION(SA13.f=SA13.f+Stmp2.f;)                                      ENABLE_SSE_IMPLEMENTATION(VA13=_mm_add_ps(VA13,Vtmp2);)                                   ENABLE_AVX_IMPLEMENTATION(VA13=_mm256_add_ps(VA13,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(SA23.f=SA23.f-Stmp1.f;)                                      ENABLE_SSE_IMPLEMENTATION(VA23=_mm_sub_ps(VA23,Vtmp1);)                                   ENABLE_AVX_IMPLEMENTATION(VA23=_mm256_sub_ps(VA23,Vtmp1);)
+
+//###########################################################
+// Update matrix U
+//###########################################################
+
+#ifdef COMPUTE_U_AS_MATRIX
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Ss.f*SU11.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vs,VU11);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vs,VU11);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Ss.f*SU12.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vs,VU12);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vs,VU12);)
+ENABLE_SCALAR_IMPLEMENTATION(SU11.f=Sc.f*SU11.f;)                                         ENABLE_SSE_IMPLEMENTATION(VU11=_mm_mul_ps(Vc,VU11);)                                      ENABLE_AVX_IMPLEMENTATION(VU11=_mm256_mul_ps(Vc,VU11);)
+ENABLE_SCALAR_IMPLEMENTATION(SU12.f=Sc.f*SU12.f;)                                         ENABLE_SSE_IMPLEMENTATION(VU12=_mm_mul_ps(Vc,VU12);)                                      ENABLE_AVX_IMPLEMENTATION(VU12=_mm256_mul_ps(Vc,VU12);)
+ENABLE_SCALAR_IMPLEMENTATION(SU11.f=SU11.f+Stmp2.f;)                                      ENABLE_SSE_IMPLEMENTATION(VU11=_mm_add_ps(VU11,Vtmp2);)                                   ENABLE_AVX_IMPLEMENTATION(VU11=_mm256_add_ps(VU11,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(SU12.f=SU12.f-Stmp1.f;)                                      ENABLE_SSE_IMPLEMENTATION(VU12=_mm_sub_ps(VU12,Vtmp1);)                                   ENABLE_AVX_IMPLEMENTATION(VU12=_mm256_sub_ps(VU12,Vtmp1);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Ss.f*SU21.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vs,VU21);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vs,VU21);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Ss.f*SU22.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vs,VU22);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vs,VU22);)
+ENABLE_SCALAR_IMPLEMENTATION(SU21.f=Sc.f*SU21.f;)                                         ENABLE_SSE_IMPLEMENTATION(VU21=_mm_mul_ps(Vc,VU21);)                                      ENABLE_AVX_IMPLEMENTATION(VU21=_mm256_mul_ps(Vc,VU21);)
+ENABLE_SCALAR_IMPLEMENTATION(SU22.f=Sc.f*SU22.f;)                                         ENABLE_SSE_IMPLEMENTATION(VU22=_mm_mul_ps(Vc,VU22);)                                      ENABLE_AVX_IMPLEMENTATION(VU22=_mm256_mul_ps(Vc,VU22);)
+ENABLE_SCALAR_IMPLEMENTATION(SU21.f=SU21.f+Stmp2.f;)                                      ENABLE_SSE_IMPLEMENTATION(VU21=_mm_add_ps(VU21,Vtmp2);)                                   ENABLE_AVX_IMPLEMENTATION(VU21=_mm256_add_ps(VU21,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(SU22.f=SU22.f-Stmp1.f;)                                      ENABLE_SSE_IMPLEMENTATION(VU22=_mm_sub_ps(VU22,Vtmp1);)                                   ENABLE_AVX_IMPLEMENTATION(VU22=_mm256_sub_ps(VU22,Vtmp1);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Ss.f*SU31.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vs,VU31);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vs,VU31);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Ss.f*SU32.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vs,VU32);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vs,VU32);)
+ENABLE_SCALAR_IMPLEMENTATION(SU31.f=Sc.f*SU31.f;)                                         ENABLE_SSE_IMPLEMENTATION(VU31=_mm_mul_ps(Vc,VU31);)                                      ENABLE_AVX_IMPLEMENTATION(VU31=_mm256_mul_ps(Vc,VU31);)
+ENABLE_SCALAR_IMPLEMENTATION(SU32.f=Sc.f*SU32.f;)                                         ENABLE_SSE_IMPLEMENTATION(VU32=_mm_mul_ps(Vc,VU32);)                                      ENABLE_AVX_IMPLEMENTATION(VU32=_mm256_mul_ps(Vc,VU32);)
+ENABLE_SCALAR_IMPLEMENTATION(SU31.f=SU31.f+Stmp2.f;)                                      ENABLE_SSE_IMPLEMENTATION(VU31=_mm_add_ps(VU31,Vtmp2);)                                   ENABLE_AVX_IMPLEMENTATION(VU31=_mm256_add_ps(VU31,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(SU32.f=SU32.f-Stmp1.f;)                                      ENABLE_SSE_IMPLEMENTATION(VU32=_mm_sub_ps(VU32,Vtmp1);)                                   ENABLE_AVX_IMPLEMENTATION(VU32=_mm256_sub_ps(VU32,Vtmp1);)
+#endif

include/igl/Singular_Value_Decomposition_Jacobi_Conjugation_Kernel.hpp

@@ -0,0 +1,118 @@
+//#####################################################################
+// Copyright (c) 2010-2011, Eftychios Sifakis.
+//
+// Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
+//   * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
+//   * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or
+//     other materials provided with the distribution.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING,
+// BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT
+// SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+// DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+// INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+//#####################################################################
+
+//###########################################################
+// Compute the Givens angle (and half-angle) 
+//###########################################################
+
+ENABLE_SCALAR_IMPLEMENTATION(Ssh.f=SS21.f*Sone_half.f;)                                   ENABLE_SSE_IMPLEMENTATION(Vsh=_mm_mul_ps(VS21,Vone_half);)                                ENABLE_AVX_IMPLEMENTATION(Vsh=_mm256_mul_ps(VS21,Vone_half);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp5.f=SS11.f-SS22.f;)                                      ENABLE_SSE_IMPLEMENTATION(Vtmp5=_mm_sub_ps(VS11,VS22);)                                   ENABLE_AVX_IMPLEMENTATION(Vtmp5=_mm256_sub_ps(VS11,VS22);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Ssh.f*Ssh.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vsh,Vsh);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vsh,Vsh);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.ui=(Stmp2.f>=Stiny_number.f)?0xffffffff:0;)            ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_cmpge_ps(Vtmp2,Vtiny_number);)                        ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_cmp_ps(Vtmp2,Vtiny_number, _CMP_GE_OS);) //ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_cmpge_ps(Vtmp2,Vtiny_number);)
+ENABLE_SCALAR_IMPLEMENTATION(Ssh.ui=Stmp1.ui&Ssh.ui;)                                     ENABLE_SSE_IMPLEMENTATION(Vsh=_mm_and_ps(Vtmp1,Vsh);)                                     ENABLE_AVX_IMPLEMENTATION(Vsh=_mm256_and_ps(Vtmp1,Vsh);)
+ENABLE_SCALAR_IMPLEMENTATION(Sch.ui=Stmp1.ui&Stmp5.ui;)                                   ENABLE_SSE_IMPLEMENTATION(Vch=_mm_and_ps(Vtmp1,Vtmp5);)                                   ENABLE_AVX_IMPLEMENTATION(Vch=_mm256_blendv_ps(Vone,Vtmp5,Vtmp1);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.ui=~Stmp1.ui&Sone.ui;)                                 ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_andnot_ps(Vtmp1,Vone);)
+ENABLE_SCALAR_IMPLEMENTATION(Sch.ui=Sch.ui|Stmp2.ui;)                                     ENABLE_SSE_IMPLEMENTATION(Vch=_mm_or_ps(Vch,Vtmp2);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Ssh.f*Ssh.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vsh,Vsh);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vsh,Vsh);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Sch.f*Sch.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vch,Vch);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vch,Vch);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp3.f=Stmp1.f+Stmp2.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp3=_mm_add_ps(Vtmp1,Vtmp2);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp3=_mm256_add_ps(Vtmp1,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp4.f=rsqrt(Stmp3.f);)                                     ENABLE_SSE_IMPLEMENTATION(Vtmp4=_mm_rsqrt_ps(Vtmp3);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp4=_mm256_rsqrt_ps(Vtmp3);)
+
+#ifdef USE_ACCURATE_RSQRT_IN_JACOBI_CONJUGATION
+ENABLE_SCALAR_IMPLEMENTATION(Ss.f=Stmp4.f*Sone_half.f;)                                   ENABLE_SSE_IMPLEMENTATION(Vs=_mm_mul_ps(Vtmp4,Vone_half);)                                ENABLE_AVX_IMPLEMENTATION(Vs=_mm256_mul_ps(Vtmp4,Vone_half);)
+ENABLE_SCALAR_IMPLEMENTATION(Sc.f=Stmp4.f*Ss.f;)                                          ENABLE_SSE_IMPLEMENTATION(Vc=_mm_mul_ps(Vtmp4,Vs);)                                       ENABLE_AVX_IMPLEMENTATION(Vc=_mm256_mul_ps(Vtmp4,Vs);)
+ENABLE_SCALAR_IMPLEMENTATION(Sc.f=Stmp4.f*Sc.f;)                                          ENABLE_SSE_IMPLEMENTATION(Vc=_mm_mul_ps(Vtmp4,Vc);)                                       ENABLE_AVX_IMPLEMENTATION(Vc=_mm256_mul_ps(Vtmp4,Vc);)
+ENABLE_SCALAR_IMPLEMENTATION(Sc.f=Stmp3.f*Sc.f;)                                          ENABLE_SSE_IMPLEMENTATION(Vc=_mm_mul_ps(Vtmp3,Vc);)                                       ENABLE_AVX_IMPLEMENTATION(Vc=_mm256_mul_ps(Vtmp3,Vc);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp4.f=Stmp4.f+Ss.f;)                                       ENABLE_SSE_IMPLEMENTATION(Vtmp4=_mm_add_ps(Vtmp4,Vs);)                                    ENABLE_AVX_IMPLEMENTATION(Vtmp4=_mm256_add_ps(Vtmp4,Vs);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp4.f=Stmp4.f-Sc.f;)                                       ENABLE_SSE_IMPLEMENTATION(Vtmp4=_mm_sub_ps(Vtmp4,Vc);)                                    ENABLE_AVX_IMPLEMENTATION(Vtmp4=_mm256_sub_ps(Vtmp4,Vc);)
+#endif
+
+ENABLE_SCALAR_IMPLEMENTATION(Ssh.f=Stmp4.f*Ssh.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vsh=_mm_mul_ps(Vtmp4,Vsh);)                                     ENABLE_AVX_IMPLEMENTATION(Vsh=_mm256_mul_ps(Vtmp4,Vsh);)
+ENABLE_SCALAR_IMPLEMENTATION(Sch.f=Stmp4.f*Sch.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vch=_mm_mul_ps(Vtmp4,Vch);)                                     ENABLE_AVX_IMPLEMENTATION(Vch=_mm256_mul_ps(Vtmp4,Vch);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Sfour_gamma_squared.f*Stmp1.f;)                      ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vfour_gamma_squared,Vtmp1);)                   ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vfour_gamma_squared,Vtmp1);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.ui=(Stmp2.f<=Stmp1.f)?0xffffffff:0;)                   ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_cmple_ps(Vtmp2,Vtmp1);)                               ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_cmp_ps(Vtmp2,Vtmp1, _CMP_LE_OS);) //ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_cmple_ps(Vtmp2,Vtmp1);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.ui=Ssine_pi_over_eight.ui&Stmp1.ui;)                   ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_and_ps(Vsine_pi_over_eight,Vtmp1);)                   ENABLE_AVX_IMPLEMENTATION(Vsh=_mm256_blendv_ps(Vsh,Vsine_pi_over_eight,Vtmp1);)
+ENABLE_SCALAR_IMPLEMENTATION(Ssh.ui=~Stmp1.ui&Ssh.ui;)                                    ENABLE_SSE_IMPLEMENTATION(Vsh=_mm_andnot_ps(Vtmp1,Vsh);)
+ENABLE_SCALAR_IMPLEMENTATION(Ssh.ui=Ssh.ui|Stmp2.ui;)                                     ENABLE_SSE_IMPLEMENTATION(Vsh=_mm_or_ps(Vsh,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.ui=Scosine_pi_over_eight.ui&Stmp1.ui;)                 ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_and_ps(Vcosine_pi_over_eight,Vtmp1);)                 ENABLE_AVX_IMPLEMENTATION(Vch=_mm256_blendv_ps(Vch,Vcosine_pi_over_eight,Vtmp1);)
+ENABLE_SCALAR_IMPLEMENTATION(Sch.ui=~Stmp1.ui&Sch.ui;)                                    ENABLE_SSE_IMPLEMENTATION(Vch=_mm_andnot_ps(Vtmp1,Vch);)
+ENABLE_SCALAR_IMPLEMENTATION(Sch.ui=Sch.ui|Stmp2.ui;)                                     ENABLE_SSE_IMPLEMENTATION(Vch=_mm_or_ps(Vch,Vtmp2);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Ssh.f*Ssh.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vsh,Vsh);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vsh,Vsh);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Sch.f*Sch.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vch,Vch);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vch,Vch);)
+ENABLE_SCALAR_IMPLEMENTATION(Sc.f=Stmp2.f-Stmp1.f;)                                       ENABLE_SSE_IMPLEMENTATION(Vc=_mm_sub_ps(Vtmp2,Vtmp1);)                                    ENABLE_AVX_IMPLEMENTATION(Vc=_mm256_sub_ps(Vtmp2,Vtmp1);)
+ENABLE_SCALAR_IMPLEMENTATION(Ss.f=Sch.f*Ssh.f;)                                           ENABLE_SSE_IMPLEMENTATION(Vs=_mm_mul_ps(Vch,Vsh);)                                        ENABLE_AVX_IMPLEMENTATION(Vs=_mm256_mul_ps(Vch,Vsh);)
+ENABLE_SCALAR_IMPLEMENTATION(Ss.f=Ss.f+Ss.f;)                                             ENABLE_SSE_IMPLEMENTATION(Vs=_mm_add_ps(Vs,Vs);)                                          ENABLE_AVX_IMPLEMENTATION(Vs=_mm256_add_ps(Vs,Vs);)
+
+//###########################################################
+// Perform the actual Givens conjugation
+//###########################################################
+
+#ifndef USE_ACCURATE_RSQRT_IN_JACOBI_CONJUGATION
+ENABLE_SCALAR_IMPLEMENTATION(Stmp3.f=Stmp1.f+Stmp2.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp3=_mm_add_ps(Vtmp1,Vtmp2);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp3=_mm256_add_ps(Vtmp1,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(SS33.f=SS33.f*Stmp3.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS33=_mm_mul_ps(VS33,Vtmp3);)                                   ENABLE_AVX_IMPLEMENTATION(VS33=_mm256_mul_ps(VS33,Vtmp3);)
+ENABLE_SCALAR_IMPLEMENTATION(SS31.f=SS31.f*Stmp3.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS31=_mm_mul_ps(VS31,Vtmp3);)                                   ENABLE_AVX_IMPLEMENTATION(VS31=_mm256_mul_ps(VS31,Vtmp3);)
+ENABLE_SCALAR_IMPLEMENTATION(SS32.f=SS32.f*Stmp3.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS32=_mm_mul_ps(VS32,Vtmp3);)                                   ENABLE_AVX_IMPLEMENTATION(VS32=_mm256_mul_ps(VS32,Vtmp3);)
+ENABLE_SCALAR_IMPLEMENTATION(SS33.f=SS33.f*Stmp3.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS33=_mm_mul_ps(VS33,Vtmp3);)                                   ENABLE_AVX_IMPLEMENTATION(VS33=_mm256_mul_ps(VS33,Vtmp3);)
+#endif
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Ss.f*SS31.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vs,VS31);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vs,VS31);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Ss.f*SS32.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vs,VS32);)                                     ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vs,VS32);)
+ENABLE_SCALAR_IMPLEMENTATION(SS31.f=Sc.f*SS31.f;)                                         ENABLE_SSE_IMPLEMENTATION(VS31=_mm_mul_ps(Vc,VS31);)                                      ENABLE_AVX_IMPLEMENTATION(VS31=_mm256_mul_ps(Vc,VS31);)
+ENABLE_SCALAR_IMPLEMENTATION(SS32.f=Sc.f*SS32.f;)                                         ENABLE_SSE_IMPLEMENTATION(VS32=_mm_mul_ps(Vc,VS32);)                                      ENABLE_AVX_IMPLEMENTATION(VS32=_mm256_mul_ps(Vc,VS32);)
+ENABLE_SCALAR_IMPLEMENTATION(SS31.f=Stmp2.f+SS31.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS31=_mm_add_ps(Vtmp2,VS31);)                                   ENABLE_AVX_IMPLEMENTATION(VS31=_mm256_add_ps(Vtmp2,VS31);)
+ENABLE_SCALAR_IMPLEMENTATION(SS32.f=SS32.f-Stmp1.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS32=_mm_sub_ps(VS32,Vtmp1);)                                   ENABLE_AVX_IMPLEMENTATION(VS32=_mm256_sub_ps(VS32,Vtmp1);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Ss.f*Ss.f;)                                          ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vs,Vs);)                                       ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vs,Vs);)         
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=SS22.f*Stmp2.f;)                                     ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(VS22,Vtmp2);)                                  ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(VS22,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp3.f=SS11.f*Stmp2.f;)                                     ENABLE_SSE_IMPLEMENTATION(Vtmp3=_mm_mul_ps(VS11,Vtmp2);)                                  ENABLE_AVX_IMPLEMENTATION(Vtmp3=_mm256_mul_ps(VS11,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp4.f=Sc.f*Sc.f;)                                          ENABLE_SSE_IMPLEMENTATION(Vtmp4=_mm_mul_ps(Vc,Vc);)                                       ENABLE_AVX_IMPLEMENTATION(Vtmp4=_mm256_mul_ps(Vc,Vc);)
+ENABLE_SCALAR_IMPLEMENTATION(SS11.f=SS11.f*Stmp4.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS11=_mm_mul_ps(VS11,Vtmp4);)                                   ENABLE_AVX_IMPLEMENTATION(VS11=_mm256_mul_ps(VS11,Vtmp4);)
+ENABLE_SCALAR_IMPLEMENTATION(SS22.f=SS22.f*Stmp4.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS22=_mm_mul_ps(VS22,Vtmp4);)                                   ENABLE_AVX_IMPLEMENTATION(VS22=_mm256_mul_ps(VS22,Vtmp4);)
+ENABLE_SCALAR_IMPLEMENTATION(SS11.f=SS11.f+Stmp1.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS11=_mm_add_ps(VS11,Vtmp1);)                                   ENABLE_AVX_IMPLEMENTATION(VS11=_mm256_add_ps(VS11,Vtmp1);)
+ENABLE_SCALAR_IMPLEMENTATION(SS22.f=SS22.f+Stmp3.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS22=_mm_add_ps(VS22,Vtmp3);)                                   ENABLE_AVX_IMPLEMENTATION(VS22=_mm256_add_ps(VS22,Vtmp3);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp4.f=Stmp4.f-Stmp2.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp4=_mm_sub_ps(Vtmp4,Vtmp2);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp4=_mm256_sub_ps(Vtmp4,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=SS21.f+SS21.f;)                                      ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_add_ps(VS21,VS21);)                                   ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_add_ps(VS21,VS21);)
+ENABLE_SCALAR_IMPLEMENTATION(SS21.f=SS21.f*Stmp4.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS21=_mm_mul_ps(VS21,Vtmp4);)                                   ENABLE_AVX_IMPLEMENTATION(VS21=_mm256_mul_ps(VS21,Vtmp4);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp4.f=Sc.f*Ss.f;)                                          ENABLE_SSE_IMPLEMENTATION(Vtmp4=_mm_mul_ps(Vc,Vs);)                                       ENABLE_AVX_IMPLEMENTATION(Vtmp4=_mm256_mul_ps(Vc,Vs);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Stmp2.f*Stmp4.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vtmp2,Vtmp4);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vtmp2,Vtmp4);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp5.f=Stmp5.f*Stmp4.f;)                                    ENABLE_SSE_IMPLEMENTATION(Vtmp5=_mm_mul_ps(Vtmp5,Vtmp4);)                                 ENABLE_AVX_IMPLEMENTATION(Vtmp5=_mm256_mul_ps(Vtmp5,Vtmp4);)
+ENABLE_SCALAR_IMPLEMENTATION(SS11.f=SS11.f+Stmp2.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS11=_mm_add_ps(VS11,Vtmp2);)                                   ENABLE_AVX_IMPLEMENTATION(VS11=_mm256_add_ps(VS11,Vtmp2);)
+ENABLE_SCALAR_IMPLEMENTATION(SS21.f=SS21.f-Stmp5.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS21=_mm_sub_ps(VS21,Vtmp5);)                                   ENABLE_AVX_IMPLEMENTATION(VS21=_mm256_sub_ps(VS21,Vtmp5);)
+ENABLE_SCALAR_IMPLEMENTATION(SS22.f=SS22.f-Stmp2.f;)                                      ENABLE_SSE_IMPLEMENTATION(VS22=_mm_sub_ps(VS22,Vtmp2);)                                   ENABLE_AVX_IMPLEMENTATION(VS22=_mm256_sub_ps(VS22,Vtmp2);)
+
+//###########################################################
+// Compute the cumulative rotation, in quaternion form
+//###########################################################
+
+ENABLE_SCALAR_IMPLEMENTATION(Stmp1.f=Ssh.f*Sqvvx.f;)                                      ENABLE_SSE_IMPLEMENTATION(Vtmp1=_mm_mul_ps(Vsh,Vqvvx);)                                   ENABLE_AVX_IMPLEMENTATION(Vtmp1=_mm256_mul_ps(Vsh,Vqvvx);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp2.f=Ssh.f*Sqvvy.f;)                                      ENABLE_SSE_IMPLEMENTATION(Vtmp2=_mm_mul_ps(Vsh,Vqvvy);)                                   ENABLE_AVX_IMPLEMENTATION(Vtmp2=_mm256_mul_ps(Vsh,Vqvvy);)
+ENABLE_SCALAR_IMPLEMENTATION(Stmp3.f=Ssh.f*Sqvvz.f;)                                      ENABLE_SSE_IMPLEMENTATION(Vtmp3=_mm_mul_ps(Vsh,Vqvvz);)                                   ENABLE_AVX_IMPLEMENTATION(Vtmp3=_mm256_mul_ps(Vsh,Vqvvz);)
+ENABLE_SCALAR_IMPLEMENTATION(Ssh.f=Ssh.f*Sqvs.f;)                                         ENABLE_SSE_IMPLEMENTATION(Vsh=_mm_mul_ps(Vsh,Vqvs);)                                      ENABLE_AVX_IMPLEMENTATION(Vsh=_mm256_mul_ps(Vsh,Vqvs);)
+
+ENABLE_SCALAR_IMPLEMENTATION(Sqvs.f=Sch.f*Sqvs.f;)                                        ENABLE_SSE_IMPLEMENTATION(Vqvs=_mm_mul_ps(Vch,Vqvs);)                                     ENABLE_AVX_IMPLEMENTATION(Vqvs=_mm256_mul_ps(Vch,Vqvs);)
+ENABLE_SCALAR_IMPLEMENTATION(Sqvvx.f=Sch.f*Sqvvx.f;)                                      ENABLE_SSE_IMPLEMENTATION(Vqvvx=_mm_mul_ps(Vch,Vqvvx);)                                   ENABLE_AVX_IMPLEMENTATION(Vqvvx=_mm256_mul_ps(Vch,Vqvvx);)
+ENABLE_SCALAR_IMPLEMENTATION(Sqvvy.f=Sch.f*Sqvvy.f;)                                      ENABLE_SSE_IMPLEMENTATION(Vqvvy=_mm_mul_ps(Vch,Vqvvy);)                                   ENABLE_AVX_IMPLEMENTATION(Vqvvy=_mm256_mul_ps(Vch,Vqvvy);)
+ENABLE_SCALAR_IMPLEMENTATION(Sqvvz.f=Sch.f*Sqvvz.f;)                                      ENABLE_SSE_IMPLEMENTATION(Vqvvz=_mm_mul_ps(Vch,Vqvvz);)                                   ENABLE_AVX_IMPLEMENTATION(Vqvvz=_mm256_mul_ps(Vch,Vqvvz);)
+
+ENABLE_SCALAR_IMPLEMENTATION(SQVVZ.f=SQVVZ.f+Ssh.f;)                                      ENABLE_SSE_IMPLEMENTATION(VQVVZ=_mm_add_ps(VQVVZ,Vsh);)                                   ENABLE_AVX_IMPLEMENTATION(VQVVZ=_mm256_add_ps(VQVVZ,Vsh);)
+ENABLE_SCALAR_IMPLEMENTATION(Sqvs.f=Sqvs.f-STMP3.f;)                                      ENABLE_SSE_IMPLEMENTATION(Vqvs=_mm_sub_ps(Vqvs,VTMP3);)                                   ENABLE_AVX_IMPLEMENTATION(Vqvs=_mm256_sub_ps(Vqvs,VTMP3);)
+ENABLE_SCALAR_IMPLEMENTATION(SQVVX.f=SQVVX.f+STMP2.f;)                                    ENABLE_SSE_IMPLEMENTATION(VQVVX=_mm_add_ps(VQVVX,VTMP2);)                                 ENABLE_AVX_IMPLEMENTATION(VQVVX=_mm256_add_ps(VQVVX,VTMP2);)
+ENABLE_SCALAR_IMPLEMENTATION(SQVVY.f=SQVVY.f-STMP1.f;)                                    ENABLE_SSE_IMPLEMENTATION(VQVVY=_mm_sub_ps(VQVVY,VTMP1);)                                 ENABLE_AVX_IMPLEMENTATION(VQVVY=_mm256_sub_ps(VQVVY,VTMP1);)

include/igl/Singular_Value_Decomposition_Kernel_Declarations.hpp

@@ -0,0 +1,137 @@
+//#####################################################################
+// Copyright (c) 2010-2011, Eftychios Sifakis.
+//
+// Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
+//   * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
+//   * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or
+//     other materials provided with the distribution.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING,
+// BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT
+// SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+// DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+// INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+//#####################################################################
+
+//###########################################################
+// Local variable declarations
+//###########################################################
+
+#ifdef PRINT_DEBUGGING_OUTPUT
+
+#ifdef USE_SSE_IMPLEMENTATION
+    float buf[4];
+    float A11,A21,A31,A12,A22,A32,A13,A23,A33;
+    float S11,S21,S31,S22,S32,S33;
+#ifdef COMPUTE_V_AS_QUATERNION
+    float QVS,QVVX,QVVY,QVVZ;
+#endif
+#ifdef COMPUTE_V_AS_MATRIX
+    float V11,V21,V31,V12,V22,V32,V13,V23,V33;
+#endif
+#ifdef COMPUTE_U_AS_QUATERNION
+    float QUS,QUVX,QUVY,QUVZ;
+#endif
+#ifdef COMPUTE_U_AS_MATRIX
+    float U11,U21,U31,U12,U22,U32,U13,U23,U33;
+#endif
+#endif
+
+#ifdef USE_AVX_IMPLEMENTATION
+    float buf[8];
+    float A11,A21,A31,A12,A22,A32,A13,A23,A33;
+    float S11,S21,S31,S22,S32,S33;
+#ifdef COMPUTE_V_AS_QUATERNION
+    float QVS,QVVX,QVVY,QVVZ;
+#endif
+#ifdef COMPUTE_V_AS_MATRIX
+    float V11,V21,V31,V12,V22,V32,V13,V23,V33;
+#endif
+#ifdef COMPUTE_U_AS_QUATERNION
+    float QUS,QUVX,QUVY,QUVZ;
+#endif
+#ifdef COMPUTE_U_AS_MATRIX
+    float U11,U21,U31,U12,U22,U32,U13,U23,U33;
+#endif
+#endif
+
+#endif
+
+const float Four_Gamma_Squared=sqrt(8.)+3.;
+const float Sine_Pi_Over_Eight=.5*sqrt(2.-sqrt(2.));
+const float Cosine_Pi_Over_Eight=.5*sqrt(2.+sqrt(2.));
+
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sfour_gamma_squared;)       ENABLE_SSE_IMPLEMENTATION(__m128 Vfour_gamma_squared;)                                    ENABLE_AVX_IMPLEMENTATION(__m256 Vfour_gamma_squared;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Ssine_pi_over_eight;)       ENABLE_SSE_IMPLEMENTATION(__m128 Vsine_pi_over_eight;)                                    ENABLE_AVX_IMPLEMENTATION(__m256 Vsine_pi_over_eight;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Scosine_pi_over_eight;)     ENABLE_SSE_IMPLEMENTATION(__m128 Vcosine_pi_over_eight;)                                  ENABLE_AVX_IMPLEMENTATION(__m256 Vcosine_pi_over_eight;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sone_half;)                 ENABLE_SSE_IMPLEMENTATION(__m128 Vone_half;)                                              ENABLE_AVX_IMPLEMENTATION(__m256 Vone_half;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sone;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vone;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vone;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Stiny_number;)              ENABLE_SSE_IMPLEMENTATION(__m128 Vtiny_number;)                                           ENABLE_AVX_IMPLEMENTATION(__m256 Vtiny_number;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Ssmall_number;)             ENABLE_SSE_IMPLEMENTATION(__m128 Vsmall_number;)                                          ENABLE_AVX_IMPLEMENTATION(__m256 Vsmall_number;)
+
+ENABLE_SCALAR_IMPLEMENTATION(Sfour_gamma_squared.f=Four_Gamma_Squared;)                   ENABLE_SSE_IMPLEMENTATION(Vfour_gamma_squared=_mm_set1_ps(Four_Gamma_Squared);)           ENABLE_AVX_IMPLEMENTATION(Vfour_gamma_squared=_mm256_set1_ps(Four_Gamma_Squared);)
+ENABLE_SCALAR_IMPLEMENTATION(Ssine_pi_over_eight.f=Sine_Pi_Over_Eight;)                   ENABLE_SSE_IMPLEMENTATION(Vsine_pi_over_eight=_mm_set1_ps(Sine_Pi_Over_Eight);)           ENABLE_AVX_IMPLEMENTATION(Vsine_pi_over_eight=_mm256_set1_ps(Sine_Pi_Over_Eight);)
+ENABLE_SCALAR_IMPLEMENTATION(Scosine_pi_over_eight.f=Cosine_Pi_Over_Eight;)               ENABLE_SSE_IMPLEMENTATION(Vcosine_pi_over_eight=_mm_set1_ps(Cosine_Pi_Over_Eight);)       ENABLE_AVX_IMPLEMENTATION(Vcosine_pi_over_eight=_mm256_set1_ps(Cosine_Pi_Over_Eight);)
+ENABLE_SCALAR_IMPLEMENTATION(Sone_half.f=.5;)                                             ENABLE_SSE_IMPLEMENTATION(Vone_half=_mm_set1_ps(.5);)                                     ENABLE_AVX_IMPLEMENTATION(Vone_half=_mm256_set1_ps(.5);)
+ENABLE_SCALAR_IMPLEMENTATION(Sone.f=1.;)                                                  ENABLE_SSE_IMPLEMENTATION(Vone=_mm_set1_ps(1.);)                                          ENABLE_AVX_IMPLEMENTATION(Vone=_mm256_set1_ps(1.);)
+ENABLE_SCALAR_IMPLEMENTATION(Stiny_number.f=1.e-20;)                                      ENABLE_SSE_IMPLEMENTATION(Vtiny_number=_mm_set1_ps(1.e-20);)                              ENABLE_AVX_IMPLEMENTATION(Vtiny_number=_mm256_set1_ps(1.e-20);)
+ENABLE_SCALAR_IMPLEMENTATION(Ssmall_number.f=1.e-12;)                                     ENABLE_SSE_IMPLEMENTATION(Vsmall_number=_mm_set1_ps(1.e-12);)                             ENABLE_AVX_IMPLEMENTATION(Vsmall_number=_mm256_set1_ps(1.e-12);)
+
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sa11;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Va11;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Va11;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sa21;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Va21;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Va21;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sa31;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Va31;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Va31;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sa12;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Va12;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Va12;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sa22;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Va22;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Va22;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sa32;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Va32;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Va32;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sa13;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Va13;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Va13;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sa23;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Va23;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Va23;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sa33;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Va33;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Va33;)
+
+#ifdef COMPUTE_V_AS_MATRIX
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sv11;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vv11;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vv11;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sv21;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vv21;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vv21;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sv31;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vv31;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vv31;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sv12;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vv12;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vv12;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sv22;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vv22;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vv22;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sv32;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vv32;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vv32;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sv13;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vv13;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vv13;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sv23;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vv23;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vv23;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sv33;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vv33;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vv33;)
+#endif
+
+#ifdef COMPUTE_V_AS_QUATERNION
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sqvs;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vqvs;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vqvs;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sqvvx;)                     ENABLE_SSE_IMPLEMENTATION(__m128 Vqvvx;)                                                  ENABLE_AVX_IMPLEMENTATION(__m256 Vqvvx;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sqvvy;)                     ENABLE_SSE_IMPLEMENTATION(__m128 Vqvvy;)                                                  ENABLE_AVX_IMPLEMENTATION(__m256 Vqvvy;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sqvvz;)                     ENABLE_SSE_IMPLEMENTATION(__m128 Vqvvz;)                                                  ENABLE_AVX_IMPLEMENTATION(__m256 Vqvvz;)
+#endif
+
+#ifdef COMPUTE_U_AS_MATRIX
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Su11;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vu11;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vu11;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Su21;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vu21;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vu21;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Su31;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vu31;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vu31;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Su12;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vu12;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vu12;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Su22;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vu22;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vu22;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Su32;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vu32;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vu32;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Su13;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vu13;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vu13;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Su23;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vu23;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vu23;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Su33;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vu33;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vu33;)
+#endif
+
+#ifdef COMPUTE_U_AS_QUATERNION
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Squs;)                      ENABLE_SSE_IMPLEMENTATION(__m128 Vqus;)                                                   ENABLE_AVX_IMPLEMENTATION(__m256 Vqus;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Squvx;)                     ENABLE_SSE_IMPLEMENTATION(__m128 Vquvx;)                                                  ENABLE_AVX_IMPLEMENTATION(__m256 Vquvx;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Squvy;)                     ENABLE_SSE_IMPLEMENTATION(__m128 Vquvy;)                                                  ENABLE_AVX_IMPLEMENTATION(__m256 Vquvy;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Squvz;)                     ENABLE_SSE_IMPLEMENTATION(__m128 Vquvz;)                                                  ENABLE_AVX_IMPLEMENTATION(__m256 Vquvz;)
+#endif
+
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sc;)                        ENABLE_SSE_IMPLEMENTATION(__m128 Vc;)                                                     ENABLE_AVX_IMPLEMENTATION(__m256 Vc;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Ss;)                        ENABLE_SSE_IMPLEMENTATION(__m128 Vs;)                                                     ENABLE_AVX_IMPLEMENTATION(__m256 Vs;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Sch;)                       ENABLE_SSE_IMPLEMENTATION(__m128 Vch;)                                                    ENABLE_AVX_IMPLEMENTATION(__m256 Vch;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Ssh;)                       ENABLE_SSE_IMPLEMENTATION(__m128 Vsh;)                                                    ENABLE_AVX_IMPLEMENTATION(__m256 Vsh;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Stmp1;)                     ENABLE_SSE_IMPLEMENTATION(__m128 Vtmp1;)                                                  ENABLE_AVX_IMPLEMENTATION(__m256 Vtmp1;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Stmp2;)                     ENABLE_SSE_IMPLEMENTATION(__m128 Vtmp2;)                                                  ENABLE_AVX_IMPLEMENTATION(__m256 Vtmp2;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Stmp3;)                     ENABLE_SSE_IMPLEMENTATION(__m128 Vtmp3;)                                                  ENABLE_AVX_IMPLEMENTATION(__m256 Vtmp3;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Stmp4;)                     ENABLE_SSE_IMPLEMENTATION(__m128 Vtmp4;)                                                  ENABLE_AVX_IMPLEMENTATION(__m256 Vtmp4;)
+ENABLE_SCALAR_IMPLEMENTATION(union {float f;unsigned int ui;} Stmp5;)                     ENABLE_SSE_IMPLEMENTATION(__m128 Vtmp5;)                                                  ENABLE_AVX_IMPLEMENTATION(__m256 Vtmp5;)

include/igl/Singular_Value_Decomposition_Main_Kernel_Body.hpp.REMOVED.git-id

@@ -0,0 +1 @@
+e8898a8aa8e920ab7c2cd0bfaac5acc2e4991daf

include/igl/Singular_Value_Decomposition_Preamble.hpp

@@ -0,0 +1,67 @@
+//#####################################################################
+// Copyright (c) 2010-2011, Eftychios Sifakis.
+//
+// Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
+//   * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
+//   * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or
+//     other materials provided with the distribution.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING,
+// BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT
+// SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+// DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+// INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+//#####################################################################
+
+#ifdef PRINT_DEBUGGING_OUTPUT
+#include <iomanip>
+#include <iostream>
+#endif
+
+#ifdef USE_SCALAR_IMPLEMENTATION
+#define ENABLE_SCALAR_IMPLEMENTATION(X) X
+#else
+#define ENABLE_SCALAR_IMPLEMENTATION(X)
+#endif
+
+#ifdef USE_SSE_IMPLEMENTATION
+#define ENABLE_SSE_IMPLEMENTATION(X) X
+#else
+#define ENABLE_SSE_IMPLEMENTATION(X)
+#endif
+
+#ifdef USE_AVX_IMPLEMENTATION
+#include <immintrin.h>
+#define ENABLE_AVX_IMPLEMENTATION(X) X
+#else
+// Stefan: removed include. Why does it import MMX instructions, shouldn't this be under the #ifdef USE_SSE_IMPLEMENTATION above?
+//#include <xmmintrin.h>
+#define ENABLE_AVX_IMPLEMENTATION(X)
+#endif
+
+#ifdef USE_SCALAR_IMPLEMENTATION
+// Alec: Why is this using sse intrinsics if it's supposed to be the scalar
+// implementation?
+#ifdef __SSE__
+#include <mmintrin.h>
+// Changed to inline
+inline float rsqrt(const float f)
+{
+    float buf[4];
+    buf[0]=f;
+    __m128 v=_mm_loadu_ps(buf);
+    v=_mm_rsqrt_ss(v);
+    _mm_storeu_ps(buf,v);
+    return buf[0];
+}
+#else
+#include <cmath>
+inline float rsqrt(const float f)
+{
+  return 1./sqrtf(f);
+}
+#endif
+#endif
+
+

include/igl/svd3x3/arap.cpp → include/igl/arap.cpp

@@ -6,18 +6,18 @@
 // v. 2.0. If a copy of the MPL was not distributed with this file, You can
 // obtain one at http://mozilla.org/MPL/2.0/.
 #include "arap.h"
-#include <igl/colon.h>
-#include <igl/cotmatrix.h>
-#include <igl/massmatrix.h>
-#include <igl/group_sum_matrix.h>
-#include <igl/covariance_scatter_matrix.h>
-#include <igl/speye.h>
-#include <igl/mode.h>
-#include <igl/project_isometrically_to_plane.h>
-#include <igl/slice.h>
-#include <igl/arap_rhs.h>
-#include <igl/repdiag.h>
-#include <igl/columnize.h>
+#include "colon.h"
+#include "cotmatrix.h"
+#include "massmatrix.h"
+#include "group_sum_matrix.h"
+#include "covariance_scatter_matrix.h"
+#include "speye.h"
+#include "mode.h"
+#include "project_isometrically_to_plane.h"
+#include "slice.h"
+#include "arap_rhs.h"
+#include "repdiag.h"
+#include "columnize.h"
 #include "fit_rotations.h"
 #include <cassert>
 #include <iostream>

include/igl/svd3x3/arap.h → include/igl/arap.h

@@ -7,9 +7,9 @@
 // obtain one at http://mozilla.org/MPL/2.0/.
 #ifndef IGL_ARAP_H
 #define IGL_ARAP_H
-#include <igl/igl_inline.h>
-#include <igl/min_quad_with_fixed.h>
-#include <igl/ARAPEnergyType.h>
+#include "igl_inline.h"
+#include "min_quad_with_fixed.h"
+#include "ARAPEnergyType.h"
 #include <Eigen/Core>
 #include <Eigen/Sparse>
 

include/igl/svd3x3/arap_dof.cpp → include/igl/arap_dof.cpp

@@ -7,30 +7,30 @@
 // obtain one at http://mozilla.org/MPL/2.0/.
 #include "arap_dof.h"
 
-#include <igl/cotmatrix.h>
-#include <igl/massmatrix.h>
-#include <igl/speye.h>
-#include <igl/repdiag.h>
-#include <igl/repmat.h>
-#include <igl/slice.h>
-#include <igl/colon.h>
-#include <igl/full.h>
-#include <igl/is_sparse.h>
-#include <igl/mode.h>
-#include <igl/is_symmetric.h>
-#include <igl/group_sum_matrix.h>
-#include <igl/arap_rhs.h>
-#include <igl/covariance_scatter_matrix.h>
-#include <igl/svd3x3/fit_rotations.h>
-
-#include <igl/verbose.h>
-#include <igl/print_ijv.h>
-
-#include <igl/get_seconds_hires.h>
+#include "cotmatrix.h"
+#include "massmatrix.h"
+#include "speye.h"
+#include "repdiag.h"
+#include "repmat.h"
+#include "slice.h"
+#include "colon.h"
+#include "full.h"
+#include "is_sparse.h"
+#include "mode.h"
+#include "is_symmetric.h"
+#include "group_sum_matrix.h"
+#include "arap_rhs.h"
+#include "covariance_scatter_matrix.h"
+#include "fit_rotations.h"
+
+#include "verbose.h"
+#include "print_ijv.h"
+
+#include "get_seconds_hires.h"
 //#include "MKLEigenInterface.h"
-#include <igl/min_quad_dense.h>
-#include <igl/get_seconds.h>
-#include <igl/columnize.h>
+#include "min_quad_dense.h"
+#include "get_seconds.h"
+#include "columnize.h"
 
 // defined if no early exit is supported, i.e., always take a fixed number of iterations
 #define IGL_ARAP_DOF_FIXED_ITERATIONS_COUNT

include/igl/svd3x3/arap_dof.h → include/igl/arap_dof.h

@@ -7,11 +7,11 @@
 // obtain one at http://mozilla.org/MPL/2.0/.
 #ifndef IGL_ARAP_ENERGY_TYPE_DOF_H
 #define IGL_ARAP_ENERGY_TYPE_DOF_H
-#include <igl/igl_inline.h>
+#include "igl_inline.h"
 
 #include <Eigen/Dense>
 #include <Eigen/Sparse>
-#include <igl/ARAPEnergyType.h>
+#include "ARAPEnergyType.h"
 #include <vector>
 
 namespace igl

include/igl/svd3x3/fit_rotations.cpp → include/igl/fit_rotations.cpp

@@ -7,12 +7,12 @@
 // obtain one at http://mozilla.org/MPL/2.0/.
 #include "fit_rotations.h"
 #include "polar_svd3x3.h"
-#include <igl/repmat.h>
-#include <igl/verbose.h>
-#include <igl/polar_dec.h>
-#include <igl/polar_svd.h>
-#include <igl/matlab_format.h>
-#include <igl/C_STR.h>
+#include "repmat.h"
+#include "verbose.h"
+#include "polar_dec.h"
+#include "polar_svd.h"
+#include "matlab_format.h"
+#include "C_STR.h"
 #include <iostream>
 
 template <typename DerivedS, typename DerivedD>

include/igl/svd3x3/fit_rotations.h → include/igl/fit_rotations.h

@@ -7,7 +7,7 @@
 // obtain one at http://mozilla.org/MPL/2.0/.
 #ifndef IGL_FIT_ROTATIONS_H
 #define IGL_FIT_ROTATIONS_H
-#include <igl/igl_inline.h>
+#include "igl_inline.h"
 #include <Eigen/Core>
 
 namespace igl

include/igl/svd3x3/polar_svd3x3.h → include/igl/polar_svd3x3.h

@@ -8,7 +8,7 @@
 #ifndef IGL_POLAR_SVD3X3_H
 #define IGL_POLAR_SVD3X3_H
 #include <Eigen/Core>
-#include <igl/igl_inline.h>
+#include "igl_inline.h"
 namespace igl
 {
   // Computes the closest rotation to input matrix A using specialized 3x3 SVD

include/igl/readOBJ.cpp

@@ -323,7 +323,5 @@ IGL_INLINE bool igl::readOBJ(
 // Explicit template specialization
 // generated by autoexplicit.sh
 template bool igl::readOBJ<Eigen::Matrix<double, -1, -1, 0, -1, -1>, Eigen::Matrix<int, -1, -1, 0, -1, -1> >(std::basic_string<char, std::char_traits<char>, std::allocator<char> >, Eigen::PlainObjectBase<Eigen::Matrix<double, -1, -1, 0, -1, -1> >&, Eigen::PlainObjectBase<Eigen::Matrix<int, -1, -1, 0, -1, -1> >&);
-template bool igl::readOBJ<Eigen::Matrix<double, -1, 3, 1, -1, 3>, Eigen::Matrix<unsigned int, -1, -1, 1, -1, -1>, Eigen::Matrix<double, -1, 2, 1, -1, 2> >(std::basic_string<char, std::char_traits<char>, std::allocator<char> >, Eigen::PlainObjectBase<Eigen::Matrix<double, -1, 3, 1, -1, 3> >&, Eigen::PlainObjectBase<Eigen::Matrix<unsigned int, -1, -1, 1, -1, -1> >&, Eigen::PlainObjectBase<Eigen::Matrix<double, -1, 3, 1, -1, 3> >&, Eigen::PlainObjectBase<Eigen::Matrix<unsigned int, -1, -1, 1, -1, -1> >&, Eigen::PlainObjectBase<Eigen::Matrix<double, -1, 2, 1, -1, 2> >&, Eigen::PlainObjectBase<Eigen::Matrix<unsigned int, -1, -1, 1, -1, -1> >&);
-template bool igl::readOBJ<Eigen::Matrix<double, -1, -1, 0, -1, -1>, Eigen::Matrix<int, -1, -1, 0, -1, -1>, Eigen::Matrix<double, -1, -1, 0, -1, -1> >(std::basic_string<char, std::char_traits<char>, std::allocator<char> >, Eigen::PlainObjectBase<Eigen::Matrix<double, -1, -1, 0, -1, -1> >&, Eigen::PlainObjectBase<Eigen::Matrix<int, -1, -1, 0, -1, -1> >&, Eigen::PlainObjectBase<Eigen::Matrix<double, -1, -1, 0, -1, -1> >&, Eigen::PlainObjectBase<Eigen::Matrix<int, -1, -1, 0, -1, -1> >&, Eigen::PlainObjectBase<Eigen::Matrix<double, -1, -1, 0, -1, -1> >&, Eigen::PlainObjectBase<Eigen::Matrix<int, -1, -1, 0, -1, -1> >&);
 template bool igl::readOBJ<Eigen::Matrix<double, -1, 3, 0, -1, 3>, Eigen::Matrix<int, -1, 3, 0, -1, 3> >(std::string, Eigen::PlainObjectBase<Eigen::Matrix<double, -1, 3, 0, -1, 3> >&, Eigen::PlainObjectBase<Eigen::Matrix<int, -1, 3, 0, -1, 3> >&);
 #endif

include/igl/svd3x3/svd3x3.cpp → include/igl/svd3x3.cpp

@@ -15,7 +15,7 @@
 #undef USE_AVX_IMPLEMENTATION
 #define COMPUTE_U_AS_MATRIX
 #define COMPUTE_V_AS_MATRIX
-#include <Singular_Value_Decomposition_Preamble.hpp>
+#include "Singular_Value_Decomposition_Preamble.hpp"
 
 #pragma runtime_checks( "u", off )  // disable runtime asserts on xor eax,eax type of stuff (doesn't always work, disable explicitly in compiler settings)
 template<typename T>
@@ -23,7 +23,7 @@ IGL_INLINE void igl::svd3x3(const Eigen::Matrix<T, 3, 3>& A, Eigen::Matrix<T, 3,
 {
   // this code only supports the scalar version (otherwise we'd need to pass arrays of matrices)  
 
-#include <Singular_Value_Decomposition_Kernel_Declarations.hpp>
+#include "Singular_Value_Decomposition_Kernel_Declarations.hpp"
 
   ENABLE_SCALAR_IMPLEMENTATION(Sa11.f=A(0,0);)                                      ENABLE_SSE_IMPLEMENTATION(Va11=_mm_loadu_ps(a11);)                                  ENABLE_AVX_IMPLEMENTATION(Va11=_mm256_loadu_ps(a11);)
     ENABLE_SCALAR_IMPLEMENTATION(Sa21.f=A(1,0);)                                      ENABLE_SSE_IMPLEMENTATION(Va21=_mm_loadu_ps(a21);)                                  ENABLE_AVX_IMPLEMENTATION(Va21=_mm256_loadu_ps(a21);)
@@ -35,7 +35,7 @@ IGL_INLINE void igl::svd3x3(const Eigen::Matrix<T, 3, 3>& A, Eigen::Matrix<T, 3,
     ENABLE_SCALAR_IMPLEMENTATION(Sa23.f=A(1,2);)                                      ENABLE_SSE_IMPLEMENTATION(Va23=_mm_loadu_ps(a23);)                                  ENABLE_AVX_IMPLEMENTATION(Va23=_mm256_loadu_ps(a23);)
     ENABLE_SCALAR_IMPLEMENTATION(Sa33.f=A(2,2);)                                      ENABLE_SSE_IMPLEMENTATION(Va33=_mm_loadu_ps(a33);)                                  ENABLE_AVX_IMPLEMENTATION(Va33=_mm256_loadu_ps(a33);)
 
-#include <Singular_Value_Decomposition_Main_Kernel_Body.hpp>
+#include "Singular_Value_Decomposition_Main_Kernel_Body.hpp"
 
     ENABLE_SCALAR_IMPLEMENTATION(U(0,0)=Su11.f;)                                      ENABLE_SSE_IMPLEMENTATION(_mm_storeu_ps(u11,Vu11);)                                 ENABLE_AVX_IMPLEMENTATION(_mm256_storeu_ps(u11,Vu11);)
     ENABLE_SCALAR_IMPLEMENTATION(U(1,0)=Su21.f;)                                      ENABLE_SSE_IMPLEMENTATION(_mm_storeu_ps(u21,Vu21);)                                 ENABLE_AVX_IMPLEMENTATION(_mm256_storeu_ps(u21,Vu21);)

include/igl/svd3x3/svd3x3.h → include/igl/svd3x3.h

@@ -7,7 +7,7 @@
 // obtain one at http://mozilla.org/MPL/2.0/.
 #ifndef IGL_SVD3X3_H
 #define IGL_SVD3X3_H
-#include <igl/igl_inline.h>
+#include "igl_inline.h"
 #include <Eigen/Dense>
 
 namespace igl

include/igl/svd3x3/svd3x3_avx.h → include/igl/svd3x3_avx.h

@@ -7,7 +7,7 @@
 // obtain one at http://mozilla.org/MPL/2.0/.
 #ifndef IGL_SVD3X3_AVX_H
 #define IGL_SVD3X3_AVX_H
-#include <igl/igl_inline.h>
+#include "igl_inline.h"
 #include <Eigen/Dense>
 
 namespace igl

include/igl/svd3x3/svd3x3_sse.h → include/igl/svd3x3_sse.h

@@ -7,7 +7,7 @@
 // obtain one at http://mozilla.org/MPL/2.0/.
 #ifndef IGL_SVD3X3_SSE_H
 #define IGL_SVD3X3_SSE_H
-#include <igl/igl_inline.h>
+#include "igl_inline.h"
 #include <Eigen/Dense>
 
 namespace igl

optional/CMakeLists.txt

@@ -89,12 +89,6 @@ else(PNG_FOUND)
   message(STATUS "PNG not found")
 endif(PNG_FOUND)
 
-## Check for SVD3X3
-find_package(SVD3X3 QUIET)
-if(SVD3X3_FOUND)
-  include_directories( ${SVD3X3_INCLUDE_DIR})
-endif(SVD3X3_FOUND)
-
 ## Check for TETGEN
 find_package(TETGEN QUIET)
 if(TETGEN_FOUND)
@@ -235,14 +229,6 @@ if (PNG_FOUND)
   add_library(iglpng STATIC ${SOURCES_PNG})
 endif (PNG_FOUND)
 
-#### Compile the svd3x3 part
-if (SVD3X3_FOUND)
-  file(GLOB SOURCES_SVD3X3
-    "${PROJECT_SOURCE_DIR}/../include/igl/svd3x3/*.cpp"
-  )
-  add_library(iglsvd3x3 STATIC ${SOURCES_SVD3X3})
-endif (SVD3X3_FOUND)
-
 #### Compile the tetgen part
 if (TETGEN_FOUND)
   file(GLOB SOURCES_TETGEN

tutorial/cmake/FindLIBIGL.cmake

@@ -115,13 +115,6 @@ if(LIBIGL_USE_STATIC_LIBRARY)
     message(FATAL_ERROR "could NOT find matlab")
   endif(MATLAB_FOUND)
 
-  FIND_LIBRARY( LIBIGLSVD3X3_LIBRARY NAMES iglsvd3x3 PATHS ${LIBIGL_LIB_DIRS})
-  if(NOT LIBIGLSVD3X3_LIBRARY)
-    set(LIBIGL_FOUND FALSE)
-    message(FATAL_ERROR "could NOT find libiglsvd3x3")
-  endif(NOT LIBIGLSVD3X3_LIBRARY)
-  set(LIBIGL_LIBRARIES ${LIBIGL_LIBRARIES}  ${LIBIGLSVD3X3_LIBRARY})
-
   FIND_LIBRARY( LIBIGLTETGEN_LIBRARY NAMES igltetgen PATHS ${LIBIGL_LIB_DIRS})
   if(NOT LIBIGLTETGEN_LIBRARY)
     set(LIBIGL_FOUND FALSE)

tutorial/cmake/FindSVD3X3.cmake

@@ -1,28 +0,0 @@
-# - Try to find the SVD3X3 library
-# Once done this will define
-#
-#  SVD3X3_FOUND - system has SVD3X3
-#  SVD3X3_INCLUDE_DIR - the SVD3X3 include directory
-
-FIND_PATH(SVD3X3_INCLUDE_DIR Singular_Value_Decomposition_Givens_QR_Factorization_Kernel.hpp
-   /usr/include
-   /usr/local/include
-   ${PROJECT_SOURCE_DIR}/../libigl/external/Singular_Value_Decomposition/
-   ${PROJECT_SOURCE_DIR}/../../external/Singular_Value_Decomposition/
-   NO_DEFAULT_PATH
-)
-
-SET(SVD3X3_FOUND "NO")
-IF (SVD3X3_INCLUDE_DIR)
-	SET(SVD3X3_FOUND "YES")
-ENDIF (SVD3X3_INCLUDE_DIR)
-
-if(SVD3X3_INCLUDE_DIR)
-   message(STATUS "Found SVD3X3: ${SVD3X3_INCLUDE_DIR}")
-else(SVD3X3_INCLUDE_DIR)
-  if (NOT SVD3X3_FIND_QUIETLY)
-   message(FATAL_ERROR "could NOT find SVD3X3")
- endif(NOT SVD3X3_FIND_QUIETLY)
-endif(SVD3X3_INCLUDE_DIR)
-
-MARK_AS_ADVANCED(SVD3X3_INCLUDE_DIR SVD3X3_LIBRARIES)