Merge pull request #4443 from ye-luo/fix-spline-blocking-offset

Fix spline block offset caused by #4439
QMCPACK · Feb 7, 2023 · f437528 · f437528
2 parents b1e7aae + 7521263
commit f437528
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 11 deletions.
diff --git a/src/QMCWaveFunctions/BsplineFactory/SplineC2COMPTarget.cpp b/src/QMCWaveFunctions/BsplineFactory/SplineC2COMPTarget.cpp
@@ -160,7 +160,7 @@ void SplineC2COMPTarget<ST>::evaluateDetRatios(const VirtualParticleSet& VP,
 
         PRAGMA_OFFLOAD("omp parallel for")
         for (int index = 0; index < last - first; index++)
-          spline2offload::evaluate_v_impl_v2(spline_ptr, ix, iy, iz, index, a, b, c,
+          spline2offload::evaluate_v_impl_v2(spline_ptr, ix, iy, iz, first + index, a, b, c,
                                              offload_scratch_iat_ptr + first + index);
         const size_t first_cplx = first / 2;
         const size_t last_cplx  = omptarget::min(last / 2, orb_size);
@@ -280,7 +280,7 @@ void SplineC2COMPTarget<ST>::mw_evaluateDetRatios(const RefVectorWithLeader<SPOS
 
         PRAGMA_OFFLOAD("omp parallel for")
         for (int index = 0; index < last - first; index++)
-          spline2offload::evaluate_v_impl_v2(spline_ptr, ix, iy, iz, index, a, b, c,
+          spline2offload::evaluate_v_impl_v2(spline_ptr, ix, iy, iz, first + index, a, b, c,
                                              offload_scratch_iat_ptr + first + index);
         const size_t first_cplx = first / 2;
         const size_t last_cplx  = omptarget::min(last / 2, orb_size);
@@ -430,7 +430,7 @@ void SplineC2COMPTarget<ST>::evaluateVGL(const ParticleSet& P,
       PRAGMA_OFFLOAD("omp parallel for")
       for (int index = 0; index < last - first; index++)
       {
-        spline2offload::evaluate_vgh_impl_v2(spline_ptr, ix, iy, iz, index, a, b, c, da, db, dc, d2a, d2b, d2c,
+        spline2offload::evaluate_vgh_impl_v2(spline_ptr, ix, iy, iz, first + index, a, b, c, da, db, dc, d2a, d2b, d2c,
                                              offload_scratch_ptr + first + index, padded_size);
         const int output_index = first + index;
         offload_scratch_ptr[padded_size * SoAFields3D::LAPL + output_index] =
@@ -518,7 +518,7 @@ void SplineC2COMPTarget<ST>::evaluateVGLMultiPos(const Vector<ST, OffloadPinnedA
         PRAGMA_OFFLOAD("omp parallel for")
         for (int index = 0; index < last - first; index++)
         {
-          spline2offload::evaluate_vgh_impl_v2(spline_ptr, ix, iy, iz, index, a, b, c, da, db, dc, d2a, d2b, d2c,
+          spline2offload::evaluate_vgh_impl_v2(spline_ptr, ix, iy, iz, first + index, a, b, c, da, db, dc, d2a, d2b, d2c,
                                                offload_scratch_iw_ptr + first + index, padded_size);
           const int output_index = first + index;
           offload_scratch_iw_ptr[padded_size * SoAFields3D::LAPL + output_index] =
@@ -694,7 +694,7 @@ void SplineC2COMPTarget<ST>::mw_evaluateVGLandDetRatioGrads(const RefVectorWithL
         PRAGMA_OFFLOAD("omp parallel for")
         for (int index = 0; index < last - first; index++)
         {
-          spline2offload::evaluate_vgh_impl_v2(spline_ptr, ix, iy, iz, index, a, b, c, da, db, dc, d2a, d2b, d2c,
+          spline2offload::evaluate_vgh_impl_v2(spline_ptr, ix, iy, iz, first + index, a, b, c, da, db, dc, d2a, d2b, d2c,
                                                offload_scratch_iw_ptr + first + index, padded_size);
           const int output_index = first + index;
           offload_scratch_iw_ptr[padded_size * SoAFields3D::LAPL + output_index] =

diff --git a/src/QMCWaveFunctions/BsplineFactory/SplineC2ROMPTarget.cpp b/src/QMCWaveFunctions/BsplineFactory/SplineC2ROMPTarget.cpp
@@ -147,7 +147,7 @@ void SplineC2ROMPTarget<ST>::evaluateValue(const ParticleSet& P, const int iat,
 
         PRAGMA_OFFLOAD("omp parallel for")
         for (int index = 0; index < last - first; index++)
-          spline2offload::evaluate_v_impl_v2(spline_ptr, ix, iy, iz, index, a, b, c,
+          spline2offload::evaluate_v_impl_v2(spline_ptr, ix, iy, iz, first + index, a, b, c,
                                              offload_scratch_ptr + first + index);
         const size_t first_cplx = first / 2;
         const size_t last_cplx  = last / 2;
@@ -231,7 +231,7 @@ void SplineC2ROMPTarget<ST>::evaluateDetRatios(const VirtualParticleSet& VP,
 
         PRAGMA_OFFLOAD("omp parallel for")
         for (int index = 0; index < last - first; index++)
-          spline2offload::evaluate_v_impl_v2(spline_ptr, ix, iy, iz, index, a, b, c,
+          spline2offload::evaluate_v_impl_v2(spline_ptr, ix, iy, iz, first + index, a, b, c,
                                              offload_scratch_iat_ptr + first + index);
         const size_t first_cplx = first / 2;
         const size_t last_cplx  = last / 2;
@@ -357,7 +357,7 @@ void SplineC2ROMPTarget<ST>::mw_evaluateDetRatios(const RefVectorWithLeader<SPOS
 
         PRAGMA_OFFLOAD("omp parallel for")
         for (int index = 0; index < last - first; index++)
-          spline2offload::evaluate_v_impl_v2(spline_ptr, ix, iy, iz, index, a, b, c,
+          spline2offload::evaluate_v_impl_v2(spline_ptr, ix, iy, iz, first + index, a, b, c,
                                              offload_scratch_iat_ptr + first + index);
         const size_t first_cplx = first / 2;
         const size_t last_cplx  = last / 2;
@@ -569,7 +569,7 @@ void SplineC2ROMPTarget<ST>::evaluateVGL(const ParticleSet& P,
       PRAGMA_OFFLOAD("omp parallel for")
       for (int index = 0; index < last - first; index++)
       {
-        spline2offload::evaluate_vgh_impl_v2(spline_ptr, ix, iy, iz, index, a, b, c, da, db, dc, d2a, d2b, d2c,
+        spline2offload::evaluate_vgh_impl_v2(spline_ptr, ix, iy, iz, first + index, a, b, c, da, db, dc, d2a, d2b, d2c,
                                              offload_scratch_ptr + first + index, spline_padded_size);
         const int output_index = first + index;
         offload_scratch_ptr[spline_padded_size * SoAFields3D::LAPL + output_index] =
@@ -660,7 +660,7 @@ void SplineC2ROMPTarget<ST>::evaluateVGLMultiPos(const Vector<ST, OffloadPinnedA
         PRAGMA_OFFLOAD("omp parallel for")
         for (int index = 0; index < last - first; index++)
         {
-          spline2offload::evaluate_vgh_impl_v2(spline_ptr, ix, iy, iz, index, a, b, c, da, db, dc, d2a, d2b, d2c,
+          spline2offload::evaluate_vgh_impl_v2(spline_ptr, ix, iy, iz, first + index, a, b, c, da, db, dc, d2a, d2b, d2c,
                                                offload_scratch_iw_ptr + first + index, spline_padded_size);
           const int output_index = first + index;
           offload_scratch_iw_ptr[spline_padded_size * SoAFields3D::LAPL + output_index] =
@@ -838,7 +838,7 @@ void SplineC2ROMPTarget<ST>::mw_evaluateVGLandDetRatioGrads(const RefVectorWithL
         PRAGMA_OFFLOAD("omp parallel for")
         for (int index = 0; index < last - first; index++)
         {
-          spline2offload::evaluate_vgh_impl_v2(spline_ptr, ix, iy, iz, index, a, b, c, da, db, dc, d2a, d2b, d2c,
+          spline2offload::evaluate_vgh_impl_v2(spline_ptr, ix, iy, iz, first + index, a, b, c, da, db, dc, d2a, d2b, d2c,
                                                offload_scratch_iw_ptr + first + index, spline_padded_size);
           const int output_index = first + index;
           offload_scratch_iw_ptr[spline_padded_size * SoAFields3D::LAPL + output_index] =