Skip to content

Conversation

@deven-amd
Copy link
Contributor

1

The following commit breaks the following unit-tests on ROCm

92e0315

//tensorflow/compiler/xla/service/gpu/tests:gpu_infeed_test              FAILED in 3 out of 3 in 1.8s
//tensorflow/compiler/xla/tests:local_client_execute_test_gpu            FAILED in 3 out of 3 in 13.7s
//tensorflow/compiler/xla/tests:outfeed_in_nested_computation_test_gpu   FAILED in 3 out of 3 in 2.0s
//tensorflow/compiler/xla/tests:while_test_gpu                           FAILED in 3 out of 3 in 9.9s

The cause is that part of the changes were not enabled for ROCm, and this commit fixes that.

2

The following commit breaks the following unit-test on ROCm

#47772

//tensorflow/python/kernel_tests:segment_reduction_ops_deterministic_test_gpu FAILED in 3 out of 3 in 5.3s

ROCm platform does not yet support complex datatype for the segment reduction ops, which is what leads to the failure.

This commit modifies the testcase to skip testing the complex datatype on ROCm


/cc @cheshire @chsigg

## 1

The following commit breaks the following unit-tests on ROCm

tensorflow@92e0315

```
//tensorflow/compiler/xla/service/gpu/tests:gpu_infeed_test              FAILED in 3 out of 3 in 1.8s
//tensorflow/compiler/xla/tests:local_client_execute_test_gpu            FAILED in 3 out of 3 in 13.7s
//tensorflow/compiler/xla/tests:outfeed_in_nested_computation_test_gpu   FAILED in 3 out of 3 in 2.0s
//tensorflow/compiler/xla/tests:while_test_gpu                           FAILED in 3 out of 3 in 9.9s
```

The cause is that part of the changes were not enabled for ROCm, and this commit fixes that.

## 2

The following commit breaks the following unit-test on ROCm

tensorflow#47772

```
//tensorflow/python/kernel_tests:segment_reduction_ops_deterministic_test_gpu FAILED in 3 out of 3 in 5.3s
```

ROCm platform does not yet support complex datatype for the segment reduction ops, which is what leads to the failure.

This commit modifies the testcase to skip testing the complex datatype on ROCm
@google-ml-butler google-ml-butler bot added the size:S CL Change Size: Small label Mar 19, 2021
@google-ml-butler google-ml-butler bot requested a review from joker-eph March 19, 2021 16:01
@google-cla google-cla bot added the cla: yes label Mar 19, 2021
@google-ml-butler google-ml-butler bot added kokoro:force-run Tests on submitted change ready to pull PR ready for merge process labels Mar 19, 2021
@kokoro-team kokoro-team removed the kokoro:force-run Tests on submitted change label Mar 19, 2021
@gbaned gbaned self-assigned this Mar 19, 2021
@gbaned gbaned added comp:gpu GPU related issues ready to pull PR ready for merge process and removed ready to pull PR ready for merge process labels Mar 19, 2021
@copybara-service copybara-service bot merged commit 14a8704 into tensorflow:master Mar 22, 2021
@deven-amd deven-amd deleted the google_upstream_master_rocm_fix_210319 branch April 16, 2021 18:22
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

cla: yes comp:gpu GPU related issues ready to pull PR ready for merge process size:S CL Change Size: Small

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants