| /external/gemmlowp/meta/generators/ |
| zip_Nx8_neon.py | 69 def GenerateLoadAggregateStore(emitter, lanes, output_address, alignment):
85 emitter.DereferenceIncrement(output_address, 64))
89 output_address):
164 emitter.DereferenceIncrement(output_address, 64))
167 def GenerateAggregatorReduction(emitter, registers, lanes, output_address,
211 emitter.Dereference(output_address, None))
213 emitter.EmitVStore('1.32', low, emitter.Dereference(output_address, 64))
216 emitter.DereferenceIncrement(output_address, 64))
218 emitter.Dereference(output_address, None))
221 emitter.DereferenceIncrement(output_address, 64)
[all...] |
| streams_common.py | 88 aggregators, output_address):
92 emitter, registers, aggregators, output_address,
100 aggregators, output_address):
104 emitter, registers, aggregators, output_address,
112 output_address, multiplicative_sum_offset,
137 emitter.Dereference(output_address,
251 output_address = registers.MapOutputParameter('out')
273 output_address)
283 stride, output_address)
287 self.asm_emitter, registers, aggregators, output_address)
[all...] |
| transform_kernels_common.py | 58 output_address):
74 emitter.EmitVStoreAE(8, elements, load, output_address, None)
75 emitter.EmitPld(output_address)
104 output_address):
143 emitter.EmitVStoreAE(32, elements, load, output_address, None)
144 emitter.EmitPld(output_address)
173 output_address):
212 emitter.EmitVStoreAE(8, elements, load, output_address, None)
213 emitter.EmitPld(output_address)
254 output_address)
[all...] |
| /external/tensorflow/tensorflow/compiler/xla/service/gpu/ |
| ir_emitter.h | 146 // sequentially consistent memory model. `output_address` and `source_address`
148 // atomicAdd(output_address, *source_address).
150 const HloComputation& nested_computation, llvm::Value* output_address,
189 llvm::Value* output_address,
196 llvm::Value* output_address,
204 // sequentially consistent memory model. `output_address` and `source_address`
206 // atomicAdd(output_address, *source_address).
|
| ir_emitter.cc | 179 const HloComputation& computation, llvm::Value* output_address,
199 {output_address, source},
200 {output_address->getType()}, &ir_builder_);
205 ir_builder_.CreateAtomicRMW(llvm::AtomicRMWInst::Add, output_address,
219 ir_builder_.CreateAtomicRMW(opcode, output_address, source,
230 ir_builder_.CreateAtomicRMW(opcode, output_address, source,
240 // 1. Reads the value from the memory pointed to by output_address and
246 // pointed to by output_address, and compares the value with old_output. If
259 // atomic binary operation. We mask off the last two bits of the output_address
274 // atomic_address = output_address & ((int64)(-4))
[all...] |
| ir_emitter_unnested.cc | [all...] |
| /prebuilts/gcc/linux-x86/host/x86_64-w64-mingw32-4.8/lib/gcc/x86_64-w64-mingw32/4.8.3/plugin/include/ |
| output.h | 111 extern void output_address (rtx);
|
| /toolchain/binutils/binutils-2.27/gold/ |
| ehframe.cc | 194 typename elfcpp::Elf_types<size>::Elf_Addr output_address; local
195 output_address = this->address();
203 p->first - output_address);
205 p->second - output_address);
[all...] |
| aarch64.cc | 2289 AArch64_address output_address; local
2384 AArch64_address output_address; local
[all...] |
| arm.cc | [all...] |
| output.cc | 1063 offset = os->output_address(relobj, lsi, addend);
1087 address = os->output_address(relobj, this->shndx_, address);
3077 Output_section::output_address(const Relobj* object, unsigned int shndx, function in class:gold::Output_section
[all...] |
| mips.cc | 1555 output_address() const function in class:__anon4450::Mips16_stub_section
[all...] |
| object.h | [all...] |
| object.cc | [all...] |
| output.h | [all...] |
| symtab.cc | [all...] |
| /external/v8/src/ |
| deoptimizer.cc | 949 Address output_address = local
1189 Address output_address = local
2143 Address output_address = local
2177 Address output_address = local
[all...] |
| /external/tensorflow/tensorflow/compiler/xla/service/cpu/ |
| ir_emitter.cc | [all...] |
