Add a missing mutex lock in the SchurEliminator. This
was lost somewhere along in the BLAS based refactoring.

Change-Id: I90b94fa9c3a8ea1b900a18f76ef6a7d0dbf24318

internal/ceres/schur_eliminator_impl.h

diff --git a/internal/ceres/schur_eliminator_impl.h b/internal/ceres/schur_eliminator_impl.h
index 5afb2f2..050b1cf 100644
--- a/internal/ceres/schur_eliminator_impl.h
+++ b/internal/ceres/schur_eliminator_impl.h
@@ -676,6 +676,7 @@
                                          &row_stride, &col_stride);
       if (cell_info != NULL) {
         // block += b1.transpose() * b2;
+        CeresMutexLock l(&cell_info->m);
         MatrixTransposeMatrixMultiply
             <kRowBlockSize, kFBlockSize, kRowBlockSize, kFBlockSize, 1>(
                 row_values + row.cells[i].position, row.block.size, block1_size,