Skip to content

perf: eliminate temporary allocations in Philox rounds #1982

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
leejet merged 1 commit into from
Sep 16, 2026
Merged

perf: eliminate temporary allocations in Philox rounds #1982

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
21 changes: 7 additions & 14 deletions src/core/rng_philox.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,13 +19,6 @@ class PhiloxRNG : public RNG {
float two_pow32_inv = 2.3283064e-10f;
float two_pow32_inv_2pi = 2.3283064e-10f * 6.2831855f;

std::vector<uint32_t> uint32(uint64_t x) {
std::vector<uint32_t> result(2);
result[0] = static_cast<uint32_t>(x & 0xFFFFFFFF);
result[1] = static_cast<uint32_t>(x >> 32);
return result;
}

std::vector<std::vector<uint32_t>> uint32(const std::vector<uint64_t>& x) {
uint32_t N = (uint32_t)x.size();
std::vector<std::vector<uint32_t>> result(2, std::vector<uint32_t>(N));
Expand All @@ -43,13 +36,13 @@ class PhiloxRNG : public RNG {
const std::vector<std::vector<uint32_t>>& key) {
uint32_t N = (uint32_t)counter[0].size();
for (uint32_t i = 0; i < N; i++) {
std::vector<uint32_t> v1 = uint32(static_cast<uint64_t>(counter[0][i]) * static_cast<uint64_t>(philox_m[0]));
std::vector<uint32_t> v2 = uint32(static_cast<uint64_t>(counter[2][i]) * static_cast<uint64_t>(philox_m[1]));
const uint64_t v1 = static_cast<uint64_t>(counter[0][i]) * static_cast<uint64_t>(philox_m[0]);
const uint64_t v2 = static_cast<uint64_t>(counter[2][i]) * static_cast<uint64_t>(philox_m[1]);

counter[0][i] = v2[1] ^ counter[1][i] ^ key[0][i];
counter[1][i] = v2[0];
counter[2][i] = v1[1] ^ counter[3][i] ^ key[1][i];
counter[3][i] = v1[0];
counter[0][i] = static_cast<uint32_t>(v2 >> 32) ^ counter[1][i] ^ key[0][i];
counter[1][i] = static_cast<uint32_t>(v2);
counter[2][i] = static_cast<uint32_t>(v1 >> 32) ^ counter[3][i] ^ key[1][i];
counter[3][i] = static_cast<uint32_t>(v1);
}
}

Expand Down Expand Up @@ -126,4 +119,4 @@ class PhiloxRNG : public RNG {
}
};

#endif // __SD_CORE_RNG_PHILOX_HPP__
#endif // __SD_CORE_RNG_PHILOX_HPP__
Loading