Skip to content

Fix timing with TBB and a minor GCC bug #2296

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

Merged
ProfFan merged 1 commit into from
Nov 4, 2025
Merged

Fix timing with TBB and a minor GCC bug #2296

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
57 changes: 43 additions & 14 deletions examples/TimeTBB.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,11 +15,13 @@
* @date November 6, 2013
*/

#include <gtsam/global_includes.h>
#include <gtsam/base/Matrix.h>
#include <gtsam/global_includes.h>

#include <map>
#include <algorithm>
#include <iostream>
#include <map>
#include <random>

using namespace std;
using namespace gtsam;
Expand Down Expand Up @@ -53,18 +55,30 @@ struct WorkerWithoutAllocation
{
vector<double>& results;

WorkerWithoutAllocation(vector<double>& results) : results(results) {}
WorkerWithoutAllocation(vector<double>& results)
: results(results), gen(rd()), dis(-1.0, 1.0) {}

// Copy constructor for TBB
WorkerWithoutAllocation(const WorkerWithoutAllocation& other)
: results(other.results), gen(rd()), dis(-1.0, 1.0) {}

void operator()(const tbb::blocked_range<size_t>& r) const
{
for(size_t i = r.begin(); i != r.end(); ++i)
{
FixedMatrix m1 = FixedMatrix::Random();
FixedMatrix m2 = FixedMatrix::Random();
FixedMatrix m1;
FixedMatrix m2;
std::generate(m1.data(), m1.data() + m1.size(),
[&]() { return dis(gen); });
std::generate(m2.data(), m2.data() + m2.size(),
[&]() { return dis(gen); });
FixedMatrix prod = m1 * m2;
results[i] = prod.norm();
}
}
std::random_device rd;
mutable std::minstd_rand gen;
mutable std::uniform_real_distribution<double> dis;
};

/* ************************************************************************* */
Expand All @@ -86,9 +100,11 @@ map<int, double> testWithoutMemoryAllocation(int num_threads)
tbb::tick_count t0 = tbb::tick_count::now();

// Run parallel code (as a task group) inside of task arena
arena.execute([&]{
tg.run_and_wait([&]{
tbb::parallel_for(tbb::blocked_range<size_t>(0, numberOfProblems), WorkerWithoutAllocation(results));
arena.execute([&] {
tg.run_and_wait([&] {
tbb::parallel_for(
tbb::blocked_range<size_t>(0, numberOfProblems, grainSize),
WorkerWithoutAllocation(results));
});
});

Expand All @@ -105,7 +121,12 @@ struct WorkerWithAllocation
{
vector<double>& results;

WorkerWithAllocation(vector<double>& results) : results(results) {}
WorkerWithAllocation(vector<double>& results)
: results(results), gen(rd()), dis(-1.0, 1.0) {}

// Copy constructor for TBB
WorkerWithAllocation(const WorkerWithAllocation& other)
: results(other.results), gen(rd()), dis(-1.0, 1.0) {}

void operator()(const tbb::blocked_range<size_t>& r) const
{
Expand All @@ -119,8 +140,10 @@ struct WorkerWithAllocation
double *proddata = allocator.allocate(problemSize * problemSize);
Eigen::Map<Matrix> prod(proddata, problemSize, problemSize);

m1 = Eigen::Matrix4d::Random(problemSize, problemSize);
m2 = Eigen::Matrix4d::Random(problemSize, problemSize);
std::generate(m1.data(), m1.data() + m1.size(),
[&]() { return dis(gen); });
std::generate(m2.data(), m2.data() + m2.size(),
[&]() { return dis(gen); });
prod = m1 * m2;
results[i] = prod.norm();

Expand All @@ -129,6 +152,10 @@ struct WorkerWithAllocation
allocator.deallocate(proddata, problemSize * problemSize);
}
}

std::random_device rd;
mutable std::minstd_rand gen;
mutable std::uniform_real_distribution<double> dis;
};

/* ************************************************************************* */
Expand All @@ -150,9 +177,11 @@ map<int, double> testWithMemoryAllocation(int num_threads)
tbb::tick_count t0 = tbb::tick_count::now();

// Run parallel code (as a task group) inside of task arena
arena.execute([&]{
tg.run_and_wait([&]{
tbb::parallel_for(tbb::blocked_range<size_t>(0, numberOfProblems), WorkerWithAllocation(results));
arena.execute([&] {
tg.run_and_wait([&] {
tbb::parallel_for(
tbb::blocked_range<size_t>(0, numberOfProblems, grainSize),
WorkerWithAllocation(results));
});
});

Expand Down
6 changes: 6 additions & 0 deletions gtsam/geometry/Gal3.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,12 @@
* refer to the aforementioned paper.
*/

// GCC bug workaround
#if defined(__GNUC__) && __GNUC__ == 15
#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
#endif

#include <gtsam/base/Matrix.h>
#include <gtsam/base/numericalDerivative.h>
#include <gtsam/geometry/Event.h>
Expand Down
Loading