Skip to content

Commit

Permalink
change nRowsPack naming back
Browse files Browse the repository at this point in the history
  • Loading branch information
@rickb80
rickb80 committed Jul 5, 2024
1 parent 93028d9 commit ee68649
Showing 1 changed file with 20 additions and 20 deletions.
40 changes: 20 additions & 20 deletions src/starkpil/chelpers_steps.hpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,7 @@

class CHelpersSteps {
public:
uint64_t nRowsPack = 4;
uint64_t nrowsPack = 4;
uint64_t nCols;
vector<uint64_t> nColsStages;
vector<uint64_t> nColsStagesAcc;
Expand Down Expand Up @@ -61,7 +61,7 @@ class CHelpersSteps {
nCols = nColsStagesAcc[11] + 6;
}

inline virtual void storePolinomials(StarkInfo &starkInfo, StepsParams &params, __m256i *bufferT_, uint8_t* storePol, uint64_t row, uint64_t nRowsPack, uint64_t domainExtended) {
inline virtual void storePolinomials(StarkInfo &starkInfo, StepsParams &params, __m256i *bufferT_, uint8_t* storePol, uint64_t row, uint64_t nrowsPack, uint64_t domainExtended) {
if(domainExtended) {
// Store either polinomial f or polinomial q
for(uint64_t k = 0; k < nColsStages[10]; ++k) {
Expand Down Expand Up @@ -90,8 +90,8 @@ class CHelpersSteps {
}
}

inline virtual void loadPolinomials(StarkInfo &starkInfo, StepsParams &params, __m256i *bufferT_, uint64_t row, uint64_t stage, uint64_t nRowsPack, uint64_t domainExtended) {
Goldilocks::Element bufferT[2*nRowsPack];
inline virtual void loadPolinomials(StarkInfo &starkInfo, StepsParams &params, __m256i *bufferT_, uint64_t row, uint64_t stage, uint64_t nrowsPack, uint64_t domainExtended) {
Goldilocks::Element bufferT[2*nrowsPack];
ConstantPolsStarks *constPols = domainExtended ? params.pConstPols2ns : params.pConstPols;
Polinomial &x = domainExtended ? params.x_2ns : params.x_n;
uint64_t domainSize = domainExtended ? 1 << starkInfo.starkStruct.nBitsExt : 1 << starkInfo.starkStruct.nBits;
Expand All @@ -100,20 +100,20 @@ class CHelpersSteps {
std::vector<uint64_t> nextStrides = {0, nextStride};
for(uint64_t k = 0; k < starkInfo.nConstants; ++k) {
for(uint64_t o = 0; o < 2; ++o) {
for(uint64_t j = 0; j < nRowsPack; ++j) {
for(uint64_t j = 0; j < nrowsPack; ++j) {
uint64_t l = (row + j + nextStrides[o]) % domainSize;
bufferT[nRowsPack*o + j] = ((Goldilocks::Element *)constPols->address())[l * starkInfo.nConstants + k];
bufferT[nrowsPack*o + j] = ((Goldilocks::Element *)constPols->address())[l * starkInfo.nConstants + k];
}
Goldilocks::load_avx(bufferT_[nColsStagesAcc[5*o] + k], &bufferT[nRowsPack*o]);
Goldilocks::load_avx(bufferT_[nColsStagesAcc[5*o] + k], &bufferT[nrowsPack*o]);
}
}

// Load x and Zi
for(uint64_t j = 0; j < nRowsPack; ++j) {
for(uint64_t j = 0; j < nrowsPack; ++j) {
bufferT[j] = x[row + j][0];
}
Goldilocks::load_avx(bufferT_[starkInfo.nConstants], &bufferT[0]);
for(uint64_t j = 0; j < nRowsPack; ++j) {
for(uint64_t j = 0; j < nrowsPack; ++j) {
bufferT[j] = params.zi[row + j][0];
}

Expand All @@ -123,30 +123,30 @@ class CHelpersSteps {
if(stage < s) break;
for(uint64_t k = 0; k < nColsStages[s]; ++k) {
for(uint64_t o = 0; o < 2; ++o) {
for(uint64_t j = 0; j < nRowsPack; ++j) {
for(uint64_t j = 0; j < nrowsPack; ++j) {
uint64_t l = (row + j + nextStrides[o]) % domainSize;
bufferT[nRowsPack*o + j] = params.pols[offsetsStages[s] + l * nColsStages[s] + k];
bufferT[nrowsPack*o + j] = params.pols[offsetsStages[s] + l * nColsStages[s] + k];
}
Goldilocks::load_avx(bufferT_[nColsStagesAcc[5*o + s] + k], &bufferT[nRowsPack*o]);
Goldilocks::load_avx(bufferT_[nColsStagesAcc[5*o + s] + k], &bufferT[nrowsPack*o]);
}
}
}

if(stage == 5) {
for(uint64_t k = 0; k < nColsStages[nStages + 1]; ++k) {
for(uint64_t o = 0; o < 2; ++o) {
for(uint64_t j = 0; j < nRowsPack; ++j) {
for(uint64_t j = 0; j < nrowsPack; ++j) {
uint64_t l = (row + j + nextStrides[o]) % domainSize;
bufferT[nRowsPack*o + j] = params.pols[offsetsStages[nStages + 1] + l * nColsStages[nStages + 1] + k];
bufferT[nrowsPack*o + j] = params.pols[offsetsStages[nStages + 1] + l * nColsStages[nStages + 1] + k];
}
Goldilocks::load_avx(bufferT_[nColsStagesAcc[5*o + nStages + 1] + k], &bufferT[nRowsPack*o]);
Goldilocks::load_avx(bufferT_[nColsStagesAcc[5*o + nStages + 1] + k], &bufferT[nrowsPack*o]);
}
}

// Load xDivXSubXi & xDivXSubWXi
for(uint64_t d = 0; d < 2; ++d) {
for(uint64_t i = 0; i < FIELD_EXTENSION; ++i) {
for(uint64_t j = 0; j < nRowsPack; ++j) {
for(uint64_t j = 0; j < nrowsPack; ++j) {
bufferT[j] = params.xDivXSubXi[d*domainSize + row + j][i];
}
Goldilocks::load_avx(bufferT_[nColsStagesAcc[11] + FIELD_EXTENSION*d + i], &bufferT[0]);
Expand All @@ -156,7 +156,7 @@ class CHelpersSteps {
}

virtual void calculateExpressions(StarkInfo &starkInfo, StepsParams &params, ParserArgs &parserArgs, ParserParams &parserParams) {
assert(nRowsPack == 4);
assert(nrowsPack == 4);
bool domainExtended = parserParams.stage > 3 ? true : false;
uint64_t domainSize = domainExtended ? 1 << starkInfo.starkStruct.nBitsExt : 1 << starkInfo.starkStruct.nBits;
uint8_t *ops = &parserArgs.ops[parserParams.opsOffset];
Expand Down Expand Up @@ -199,14 +199,14 @@ class CHelpersSteps {
}

#pragma omp parallel for
for (uint64_t i = 0; i < domainSize; i+= nRowsPack) {
for (uint64_t i = 0; i < domainSize; i+= nrowsPack) {
uint64_t i_args = 0;

__m256i bufferT_[2*nCols];
__m256i tmp1[parserParams.nTemp1];
Goldilocks3::Element_avx tmp3[parserParams.nTemp3];

loadPolinomials(starkInfo, params, bufferT_, i, parserParams.stage, nRowsPack, domainExtended);
loadPolinomials(starkInfo, params, bufferT_, i, parserParams.stage, nrowsPack, domainExtended);

for (uint64_t kk = 0; kk < parserParams.nOps; ++kk) {
switch (ops[kk]) {
Expand Down Expand Up @@ -696,7 +696,7 @@ class CHelpersSteps {
}
}
}
storePolinomials(starkInfo, params, bufferT_, storePol, i, nRowsPack, domainExtended);
storePolinomials(starkInfo, params, bufferT_, storePol, i, nrowsPack, domainExtended);
if (i_args != parserParams.nArgs) std::cout << " " << i_args << " - " << parserParams.nArgs << std::endl;
assert(i_args == parserParams.nArgs);
}
Expand Down

0 comments on commit ee68649

Please sign in to comment.