Fix sdot_ bug for runtime F2C symbol conflicts by using cblas where available

Author: vlasenkov

THGeneral.h.in

@@ -14,6 +14,7 @@
 #cmakedefine USE_BLAS
 #cmakedefine USE_LAPACK
 #cmakedefine BLAS_F2C
+#cmakedefine BLAS_USE_CBLAS_DOT
 
 #ifdef __cplusplus
 # define TH_EXTERNC extern "C"

cmake/FindBLAS.cmake

@@ -274,6 +274,22 @@ int main() {
   ELSE (BLAS_F2C_DOUBLE_WORKS AND NOT BLAS_F2C_FLOAT_WORKS)
     SET(BLAS_F2C FALSE)
   ENDIF (BLAS_F2C_DOUBLE_WORKS AND NOT BLAS_F2C_FLOAT_WORKS)
+  CHECK_C_SOURCE_RUNS("
+#include <stdlib.h>
+#include <stdio.h>
+float x[4] = { 1, 2, 3, 4 };
+float y[4] = { .1, .01, .001, .0001 };
+extern float cblas_sdot();
+int main() {
+  int i;
+  double r = cblas_sdot(4, x, 1, y, 1);
+  exit((float)r != (float).1234);
+}" BLAS_USE_CBLAS_DOT )
+  IF (BLAS_USE_CBLAS_DOT)
+    SET(BLAS_USE_CBLAS_DOT TRUE)
+  ELSE (BLAS_USE_CBLAS_DOT)
+    SET(BLAS_USE_CBLAS_DOT FALSE)
+  ENDIF (BLAS_USE_CBLAS_DOT)
 ENDIF(BLAS_LIBRARIES)
 
 # epilogue

generic/THBlas.c

@@ -18,7 +18,18 @@ TH_EXTERNC void scopy_(int *n, float *x, int *incx, float *y, int *incy);
 TH_EXTERNC void daxpy_(int *n, double *a, double *x, int *incx, double *y, int *incy);
 TH_EXTERNC void saxpy_(int *n, float *a, float *x, int *incx, float *y, int *incy);
 TH_EXTERNC double ddot_(int *n, double *x, int *incx, double *y, int *incy);
+#ifdef BLAS_USE_CBLAS_DOT
+TH_EXTERNC float cblas_sdot(const int n, const float *x, const int incx, const float *y, const int incy);
+#ifndef THBlas_C_sdot_
+#define THBlas_C_sdot_
+inline ffloat sdot_(const int *n, const float *x, const int *incx, const float *y, const int *incy)
+{
+  return cblas_sdot(*n, x, *incx, y, *incy);
+}
+#endif
+#else
 TH_EXTERNC ffloat sdot_(int *n, float *x, int *incx, float *y, int *incy);
+#endif
 TH_EXTERNC void dgemv_(char *trans, int *m, int *n, double *alpha, double *a, int *lda, double *x, int *incx, double *beta, double *y, int *incy);
 TH_EXTERNC void sgemv_(char *trans, int *m, int *n, float *alpha, float *a, int *lda, float *x, int *incx, float *beta, float *y, int *incy);
 TH_EXTERNC void dger_(int *m, int *n, double *alpha, double *x, int *incx, double *y, int *incy, double *a, int *lda);