Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Remove reinterpret_casts from util/linpack.H #1409

Closed
wants to merge 2 commits into from
Closed

Remove reinterpret_casts from util/linpack.H #1409

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
2bb3dd6
Remove reinterpret_casts from util/linpack.H
yut23 Dec 14, 2023
82df5b7
Fix off-by-one error
yut23 Dec 14, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
55 changes: 24 additions & 31 deletions util/linpack.H
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,25 +10,21 @@ template <int num_eqs>
AMREX_GPU_HOST_DEVICE AMREX_FORCE_INLINE
void dgesl (RArray2D& a1, IArray1D& pivot1, RArray1D& b1)
{
auto const& a = reinterpret_cast<Array2D<Real, 0, num_eqs-1, 0, num_eqs-1>&>(a1);
auto const& pivot = reinterpret_cast<Array1D<short, 0, num_eqs-1>&>(pivot1);
auto& b = reinterpret_cast<Array1D<Real, 0, num_eqs-1>&>(b1);

int nm1 = num_eqs - 1;

// solve a * x = b
// first solve l * y = b
if (nm1 >= 1) {
for (int k = 0; k < nm1; ++k) {
int l = pivot(k) - 1;
Real t = b(l);
int l = pivot1(k+1) - 1;
Real t = b1(l+1);
if (l != k) {
b(l) = b(k);
b(k) = t;
b1(l+1) = b1(k+1);
b1(k+1) = t;
}

for (int j = k+1; j < num_eqs; ++j) {
b(j) += t * a(j,k);
b1(j+1) += t * a1(j+1,k+1);
}
}
}
Expand All @@ -37,10 +33,10 @@ void dgesl (RArray2D& a1, IArray1D& pivot1, RArray1D& b1)
for (int kb = 0; kb < num_eqs; ++kb) {

int k = num_eqs - kb - 1;
b(k) = b(k) / a(k,k);
Real t = -b(k);
b1(k+1) = b1(k+1) / a1(k+1,k+1);
Real t = -b1(k+1);
for (int j = 0; j < k; ++j) {
b(j) += t * a(j,k);
b1(j+1) += t * a1(j+1,k+1);
}
}

Expand All @@ -52,9 +48,6 @@ template <int num_eqs>
AMREX_GPU_HOST_DEVICE AMREX_FORCE_INLINE
void dgefa (RArray2D& a1, IArray1D& pivot1, int& info)
{
auto& a = reinterpret_cast<Array2D<Real, 0, num_eqs-1, 0, num_eqs-1>&>(a1);
auto& pivot = reinterpret_cast<Array1D<short, 0, num_eqs-1>&>(pivot1);

// dgefa factors a matrix by gaussian elimination.
// a is returned in the form a = l * u where
// l is a product of permutation and unit lower
Expand All @@ -73,41 +66,41 @@ void dgefa (RArray2D& a1, IArray1D& pivot1, int& info)

// find l = pivot index
int l = k;
Real dmax = std::abs(a(k,k));
Real dmax = std::abs(a1(k+1,k+1));
for (int i = k+1; i < num_eqs; ++i) {
if (std::abs(a(i,k)) > dmax) {
if (std::abs(a1(i+1,k+1)) > dmax) {
l = i;
dmax = std::abs(a(i,k));
dmax = std::abs(a1(i+1,k+1));
}
}

pivot(k) = l + 1;
pivot1(k+1) = l + 1;

// zero pivot implies this column already triangularized
if (a(l,k) != 0.0e0_rt) {
if (a1(l+1,k+1) != 0.0e0_rt) {

// interchange if necessary
if (l != k) {
t = a(l,k);
a(l,k) = a(k,k);
a(k,k) = t;
t = a1(l+1,k+1);
a1(l+1,k+1) = a1(k+1,k+1);
a1(k+1,k+1) = t;
}

// compute multipliers
t = -1.0e0_rt / a(k,k);
t = -1.0e0_rt / a1(k+1,k+1);
for (int j = k+1; j < num_eqs; ++j) {
a(j,k) *= t;
a1(j+1,k+1) *= t;
}

// row elimination with column indexing
for (int j = k+1; j < num_eqs; ++j) {
t = a(l,j);
t = a1(l+1,j+1);
if (l != k) {
a(l,j) = a(k,j);
a(k,j) = t;
a1(l+1,j+1) = a1(k+1,j+1);
a1(k+1,j+1) = t;
}
for (int i = k+1; i < num_eqs; ++i) {
a(i,j) += t * a(i,k);
a1(i+1,j+1) += t * a1(i+1,k+1);
}
}
}
Expand All @@ -121,9 +114,9 @@ void dgefa (RArray2D& a1, IArray1D& pivot1, int& info)

}

pivot(num_eqs-1) = num_eqs;
pivot1(num_eqs) = num_eqs;

if (a(num_eqs-1,num_eqs-1) == 0.0e0_rt) {
if (a1(num_eqs,num_eqs) == 0.0e0_rt) {
info = num_eqs;
}

Expand Down