[ROCm] set ROCM_ARCH to gfx900 and gfx906 for CI builds

[jeffdaily]

This change adds the arch settings for caffe2 builds, fixes some typos,
and clarifies that this setting applies to both CircleCI and Jenkins.

[jeffdaily]

Corresponding to pytorch/ossci-job-dsl#83, this will reduce build times for ROCm CI (Jenkins) jobs.

[facebook-github-bot]

@malfet has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

[dr-ci]

💊 CI failures summary and remediations

As of commit c44eed6 (more details on the Dr. CI page):

---

• 3/3 failures possibly* introduced in this PR
  • 1/3 non-CircleCI failure(s)

---

🕵️ 1 new failure recognized by patterns

The following CI failures do not appear to be due to upstream breakages:

pytorch_linux_backward_compatibility_check_test (1/1)

Step: "Run tests" (full log | diagnosis details | 🔁 rerun)

Nov 10 18:32:48 The PR is introducing backward incompatible changes to the operator library. Please contact PyTorch team to confirm whether this change is wanted or not.

Nov 10 18:32:48 processing existing schema:  __getstate__(__torch__.torch.classes.xnnpack.LinearOpContext _0) -> ((Tensor, Tensor?, Scalar?, Scalar?) _0) 
Nov 10 18:32:48 processing existing schema:  __setstate__(__torch__.torch.classes.xnnpack.LinearOpContext _0, (Tensor, Tensor?, Scalar?, Scalar?) _1) -> (None _0) 
Nov 10 18:32:48 processing existing schema:  __getstate__(__torch__.torch.classes.xnnpack.Conv2dOpContext _0) -> ((Tensor, Tensor?, int[], int[], int[], int, Scalar?, Scalar?) _0) 
Nov 10 18:32:48 processing existing schema:  __setstate__(__torch__.torch.classes.xnnpack.Conv2dOpContext _0, (Tensor, Tensor?, int[], int[], int[], int, Scalar?, Scalar?) _1) -> (None _0) 
Nov 10 18:32:48 processing existing schema:  __getstate__(__torch__.torch.classes.xnnpack.TransposeConv2dOpContext _0) -> ((Tensor, Tensor?, int[], int[], int[], int[], int, Scalar?, Scalar?) _0) 
Nov 10 18:32:48 processing existing schema:  __setstate__(__torch__.torch.classes.xnnpack.TransposeConv2dOpContext _0, (Tensor, Tensor?, int[], int[], int[], int[], int, Scalar?, Scalar?) _1) -> (None _0) 
Nov 10 18:32:48 processing existing schema:  __init__(__torch__.torch.classes._nnapi.Compilation _0) -> (None _0) 
Nov 10 18:32:48 processing existing schema:  init(__torch__.torch.classes._nnapi.Compilation _0, Tensor _1, Tensor[] _2) -> (None _0) 
Nov 10 18:32:48 processing existing schema:  run(__torch__.torch.classes._nnapi.Compilation _0, Tensor[] _1, Tensor[] _2) -> (None _0) 
Nov 10 18:32:48 processing existing schema:  __init__(__torch__.torch.classes.dist_rpc.WorkerInfo _0, str _1, int _2) -> (None _0) 
Nov 10 18:32:48 The PR is introducing backward incompatible changes to the operator library. Please contact PyTorch team to confirm whether this change is wanted or not.  
Nov 10 18:32:48  
Nov 10 18:32:48 Broken ops: [ 
Nov 10 18:32:48 	aten::_foreach_log(Tensor[] tensors) -> (Tensor[]) 
Nov 10 18:32:48 	aten::_foreach_round(Tensor[] tensors) -> (Tensor[]) 
Nov 10 18:32:48 	aten::_foreach_sinh(Tensor[] tensors) -> (Tensor[]) 
Nov 10 18:32:48 	aten::_foreach_lgamma_(Tensor[] self) -> () 
Nov 10 18:32:48 	aten::_foreach_lgamma(Tensor[] tensors) -> (Tensor[]) 
Nov 10 18:32:48 	aten::_foreach_log10(Tensor[] tensors) -> (Tensor[]) 
Nov 10 18:32:48 	aten::_foreach_round_(Tensor[] self) -> () 
Nov 10 18:32:48 	aten::_foreach_sin(Tensor[] tensors) -> (Tensor[]) 

---

1 job timed out:

• pytorch_linux_xenial_py3_clang5_asan_test2

---

ci.pytorch.org: 1 failed

• Failed: pr/caffe2-pytorch-linux-bionic-rocm3.8-py3.6-test

---

This comment was automatically generated by Dr. CI (expand for details).

Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions on the GitHub issue tracker or post in the (internal) Dr. CI Users group.

See how this bot performed.

This comment has been revised 3 times.

[jeffdaily]

CI failures not due to this PR.

[facebook-github-bot]

@malfet merged this pull request in 7691cf1.

[facebook-github-bot]

@malfet merged this pull request in 7691cf1.