Bugfix in BLR COLWISE.

@liuyangzhuan
pghysels · Jul 19, 2024 · e7f2c85 · e7f2c85
1 parent cf768ce
commit e7f2c85
Show file tree

Hide file tree

Showing 2 changed files with 22 additions and 26 deletions.
diff --git a/src/BLR/BLRMatrix.cpp b/src/BLR/BLRMatrix.cpp
@@ -1403,7 +1403,8 @@ namespace strumpack {
       B11.piv_.resize(B11.rows());
       auto rb = B11.rowblocks();
       auto rb2 = B22.rowblocks();
-      const std::size_t CP = 1; // code below is now simplified for CP == 1
+      const std::size_t CP = 1;
+      // code below is now simplified for CP == 1
 #pragma omp parallel if(!omp_in_parallel())
 #pragma omp single nowait
       {
@@ -1412,7 +1413,8 @@ namespace strumpack {
           B21.fill_col(0., i, CP);
           blockcol(i, true, CP);
           for (std::size_t k=0; k<i; k++) {
-            if (admissible(k, i)) B11.compress_tile(k, i, opts);
+            if (admissible(k, i))
+              B11.compress_tile(k, i, opts);
             std::vector<int> tpiv
               (B11.piv_.begin()+B11.tileroff(k),
                B11.piv_.begin()+B11.tileroff(k+1));
@@ -1449,16 +1451,6 @@ namespace strumpack {
           B12.fill_col(0., i, CP);
           B22.fill_col(0., i, CP);
           blockcol(i, false, CP);
-#pragma omp taskloop
-          for (std::size_t k=0; k<rb; k++) {
-            B12.compress_tile(k, i, opts);
-            std::vector<int> tpiv
-              (B11.piv_.begin()+B11.tileroff(k),
-               B11.piv_.begin()+B11.tileroff(k+1));
-            B12.tile(k, i).laswp(tpiv, true);
-            trsm(Side::L, UpLo::L, Trans::N, Diag::U,
-                 scalar_t(1.), B11.tile(k, k), B12.tile(k, i));
-          }
           for (std::size_t k=0; k<rb; k++) {
 #pragma omp taskloop
             for (std::size_t lk=k+1; lk<rb+rb2; lk++)
@@ -1471,6 +1463,16 @@ namespace strumpack {
                      B21.tile(lk-rb, k), B12.tile(k, i), scalar_t(1.),
                      B22.tile_dense(lk-rb, i).D());
           }
+#pragma omp taskloop
+          for (std::size_t k=0; k<rb; k++) {
+            B12.compress_tile(k, i, opts);
+            std::vector<int> tpiv
+              (B11.piv_.begin()+B11.tileroff(k),
+               B11.piv_.begin()+B11.tileroff(k+1));
+            B12.tile(k, i).laswp(tpiv, true);
+            trsm(Side::L, UpLo::L, Trans::N, Diag::U,
+                 scalar_t(1.), B11.tile(k, k), B12.tile(k, i));
+          }
 #pragma omp taskloop
           for (std::size_t k=0; k<rb2; k++)
             if (i != k)

diff --git a/src/sparse/fronts/FrontBLR.cpp b/src/sparse/fronts/FrontBLR.cpp
@@ -100,29 +100,23 @@ namespace strumpack {
             F12blr_(e.r, e.c) = e.v;
     }
     if (part) {
+      auto lo = F11blr_.tilecoff(i);
+      auto hi = F11blr_.tilecoff(std::min(i+CP, F11blr_.colblocks()));
       if (lchild_)
         lchild_->extend_add_to_blr_col
-          (F11blr_, F12blr_, F21blr_, F22blr_, this, F11blr_.tilecoff(i),
-           F11blr_.tilecoff(std::min(i+CP, F11blr_.colblocks())),
-           task_depth, opts);
+          (F11blr_, F12blr_, F21blr_, F22blr_, this, lo, hi, task_depth, opts);
       if (rchild_)
         rchild_->extend_add_to_blr_col
-          (F11blr_, F12blr_, F21blr_, F22blr_, this, F11blr_.tilecoff(i),
-           F11blr_.tilecoff(std::min(i+CP, F11blr_.colblocks())),
-           task_depth, opts);
+          (F11blr_, F12blr_, F21blr_, F22blr_, this, lo, hi, task_depth, opts);
     } else {
+      auto lo = F22blr_.tilecoff(i) + dim_sep();
+      auto hi = F22blr_.tilecoff(std::min(i+CP, F22blr_.colblocks())) + dim_sep();
       if (lchild_)
         lchild_->extend_add_to_blr_col
-          (F11blr_, F12blr_, F21blr_, F22blr_, this,
-           F22blr_.tilecoff(i) + dim_sep(),
-           F22blr_.tilecoff(std::min(i+CP, F22blr_.colblocks())) + dim_sep(),
-           task_depth, opts);
+          (F11blr_, F12blr_, F21blr_, F22blr_, this, lo, hi, task_depth, opts);
       if (rchild_)
         rchild_->extend_add_to_blr_col
-          (F11blr_, F12blr_, F21blr_, F22blr_, this,
-           F22blr_.tilecoff(i) + dim_sep(),
-           F22blr_.tilecoff(std::min(i+CP, F22blr_.colblocks())) + dim_sep(),
-           task_depth, opts);
+          (F11blr_, F12blr_, F21blr_, F22blr_, this, lo, hi, task_depth, opts);
     }
   }