You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.

hybrid_execution_context.cc 2.8 kB

4 years ago
4 years ago
4 years ago
4 years ago
1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465666768697071727374757677787980818283848586878889
  1. /**
  2. * Copyright 2019-2020 Huawei Technologies Co., Ltd
  3. *
  4. * Licensed under the Apache License, Version 2.0 (the "License");
  5. * you may not use this file except in compliance with the License.
  6. * You may obtain a copy of the License at
  7. *
  8. * http://www.apache.org/licenses/LICENSE-2.0
  9. *
  10. * Unless required by applicable law or agreed to in writing, software
  11. * distributed under the License is distributed on an "AS IS" BASIS,
  12. * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  13. * See the License for the specific language governing permissions and
  14. * limitations under the License.
  15. */
  16. #include "hybrid_execution_context.h"
  17. #include <atomic>
  18. namespace ge {
  19. namespace hybrid {
  20. namespace {
  21. const uint32_t kEndOfSequence = 0x0704000a;
  22. const uint32_t kEndOfSequenceNew = 507005;
  23. const int32_t kModelAbortNormal = 0x0704000e;
  24. const int32_t kModelAbortNormalNew = 507024;
  25. std::atomic_ulong context_id_gen {};
  26. } // namespace
  27. GraphExecutionContext::GraphExecutionContext() {
  28. context_id = context_id_gen++;
  29. }
  30. void GraphExecutionContext::SetErrorCode(Status error_code) {
  31. std::lock_guard<std::mutex> lk(mu);
  32. this->status = error_code;
  33. }
  34. Status GraphExecutionContext::GetStatus() const {
  35. std::lock_guard<std::mutex> lk(mu);
  36. return this->status;
  37. }
  38. Status GraphExecutionContext::Synchronize(rtStream_t rt_stream) {
  39. auto rt_ret = rtStreamSynchronize(rt_stream);
  40. if (rt_ret == RT_ERROR_NONE) {
  41. return SUCCESS;
  42. }
  43. if (rt_ret == kEndOfSequence || rt_ret == kEndOfSequenceNew) {
  44. GELOGI("Got end of sequence");
  45. is_eos_ = true;
  46. return END_OF_SEQUENCE;
  47. }
  48. if (rt_ret == kModelAbortNormal || rt_ret == kModelAbortNormalNew) {
  49. GELOGI("The model with multiple datasets aborts normally");
  50. return SUCCESS;
  51. }
  52. GELOGE(RT_FAILED, "[Invoke][rtStreamSynchronize] failed, ret = %d", rt_ret);
  53. REPORT_CALL_ERROR("E19999", "invoke rtStreamSynchronize failed, ret = %d", rt_ret);
  54. return RT_FAILED;
  55. }
  56. Status GraphExecutionContext::DumpExceptionInfo(const std::vector<rtExceptionInfo> &exception_infos) {
  57. if (exception_infos.empty()) {
  58. GELOGI("[Dump][ExceptionInfo] Exception info is null.");
  59. return SUCCESS;
  60. }
  61. GELOGI("[Dump][ExceptionInfo] Start to search dynamic op info and to dump.");
  62. if (exception_dumper.DumpExceptionInfo(exception_infos) != SUCCESS) {
  63. GELOGE(FAILED, "[Dump][Exception] Dump dynamic op exception info failed.");
  64. return FAILED;
  65. }
  66. GELOGI("[Dump][ExceptionInfo] Start to search static op info and to dump.");
  67. for (const auto &iter : davinci_model) {
  68. if (iter != nullptr) {
  69. if (iter->DumpExceptionInfo(exception_infos) != SUCCESS) {
  70. GELOGE(FAILED, "[Dump][ExceptionInfo] Dump static op exception info failed.");
  71. return FAILED;
  72. }
  73. }
  74. }
  75. return SUCCESS;
  76. }
  77. } // namespace hybrid
  78. } // namespace ge

图引擎模块(GE)是MindSpore的一个子模块,其代码由C++实现,位于前端模块ME和底层硬件之间,起到承接作用。图引擎模块以ME下发的图作为输入,然后进行一系列的深度图优化操作,最后输出一张可以在底层硬件上高效运行的图。GE针对昇腾AI处理器的硬件结构特点,做了特定的优化工作,以此来充分发挥出昇腾AI处理器的强大算力。在进行模型训练/推理时,GE会被自动调用而用户并不感知。GE主要由GE API和GE Core两部分组成,详细的架构图如下所示