[TRTLLM-13696][test] Part1: Add CPU only CI stage

[tongyuantongyu]

Summary by CodeRabbit

• New Features

  • Added support for running selected tests on CPU-only environments, including new CPU Generic stage coverage for x86 and ARM.
  • Introduced a new cpu_only test marker and related filtering so CPU-focused tests can be isolated more easily.

• Bug Fixes

  • Improved handling when no GPU is available across runtime checks, profiling, and test utilities.
  • Updated test orchestration to better handle CPU-only and generic setups, reducing failures from missing CUDA/NVIDIA tooling.

Description

This PR adds new CPU only CI stages and moved eligible tests to run without GPU. Also made a handful of fixes so TensorRT LLM can import properly without GPU.

Test Coverage

New CPU stages.

PR Checklist

Please review the following before submitting your PR:

• PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.

• PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.

• Test cases are provided for new code paths (see test instructions)

• If PR introduces API changes, an appropriate PR label is added - either api-compatible or api-breaking. For api-breaking, include BREAKING in the PR title.

• Any new dependencies have been scanned for license and vulnerabilities

• CODEOWNERS updated if ownership changes

• Documentation updated as needed

• Update tava architecture diagram if there is a significant design change in PR.

• The reviewers assigned automatically/manually are appropriate for the PR.

• Please check this after reviewing the above items as appropriate for this PR.

GitHub Bot Help

To see a list of available CI bot commands, please comment /bot help.

[tongyuantongyu]

/bot run --stages "CPU-Generic-x86-1, CPU-Generic-arm-1"

[tongyuantongyu]

/bot run --stage-list "CPU-Generic-x86-1, CPU-Generic-arm-1"

[tensorrt-cicd]

PR_Github #56883 Bot args parsing error: usage: /bot [-h]
{run,kill,skip,submit,reviewers,reuse-pipeline,reuse-review} ...
/bot: error: unrecognized arguments: --stages CPU-Generic-x86-1, CPU-Generic-arm-1

Link to invocation

[tensorrt-cicd]

PR_Github #56886 [ run ] triggered by Bot. Commit: 54190f6 Link to invocation

[tongyuantongyu]

/bot run --stage-list "CPU-Generic-x86-1, CPU-Generic-arm-1"

[tensorrt-cicd]

PR_Github #56886 [ run ] completed with state FAILURE. Commit: 54190f6
/LLM/main/L0_MergeRequest_PR pipeline #45694 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[tensorrt-cicd]

PR_Github #56898 [ run ] triggered by Bot. Commit: 9e7e885 Link to invocation

[tensorrt-cicd]

PR_Github #56898 [ run ] completed with state FAILURE. Commit: 9e7e885
/LLM/main/L0_MergeRequest_PR pipeline #45707 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[tongyuantongyu]

/bot run --stage-list "CPU-Generic-x86-1, CPU-Generic-arm-1" --disable-fail-fast

[tensorrt-cicd]

PR_Github #57075 [ run ] triggered by Bot. Commit: 9e7e885 Link to invocation

[tensorrt-cicd]

PR_Github #57075 [ run ] completed with state SUCCESS. Commit: 9e7e885
/LLM/main/L0_MergeRequest_PR pipeline #45865 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[tongyuantongyu]

/bot run --stage-list "CPU-Generic-x86-1, CPU-Generic-arm-1" --disable-fail-fast

[tongyuantongyu]

/bot run --stage-list "CPU-Generic-x86-1, CPU-Generic-arm-1" --disable-fail-fast

[tensorrt-cicd]

PR_Github #57326 [ run ] triggered by Bot. Commit: 5d73a0c Link to invocation

[tensorrt-cicd]

PR_Github #57326 [ run ] completed with state FAILURE. Commit: 5d73a0c
/LLM/main/L0_MergeRequest_PR pipeline #46084 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[tongyuantongyu]

/bot run --stage-list "CPU-Generic-x86-1, CPU-Generic-arm-1" --disable-fail-fast

[tongyuantongyu]

/bot run --stage-list "CPU-Generic-x86-1, CPU-Generic-arm-1" --disable-fail-fast

[tensorrt-cicd]

PR_Github #57405 [ run ] triggered by Bot. Commit: 224f08d Link to invocation

[tongyuantongyu]

/bot run --stage-list "CPU-Generic-x86-1, CPU-Generic-arm-1" --disable-fail-fast

[tensorrt-cicd]

PR_Github #57405 [ run ] completed with state FAILURE. Commit: 224f08d
/LLM/main/L0_MergeRequest_PR pipeline #46149 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[tensorrt-cicd]

PR_Github #57428 [ run ] triggered by Bot. Commit: b9f517b Link to invocation

[tensorrt-cicd]

PR_Github #57428 [ run ] completed with state FAILURE. Commit: b9f517b
/LLM/main/L0_MergeRequest_PR pipeline #46169 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[tensorrt-cicd]

PR_Github #58424 [ run ] completed with state SUCCESS. Commit: 7df51e1
/LLM/main/L0_MergeRequest_PR pipeline #47041 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[tongyuantongyu]

/bot run

[tensorrt-cicd]

PR_Github #58567 [ run ] triggered by Bot. Commit: e11b696 Link to invocation

[tensorrt-cicd]

PR_Github #58567 [ run ] completed with state SUCCESS. Commit: e11b696
/LLM/main/L0_MergeRequest_PR pipeline #47162 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[tongyuantongyu]

/bot run --stage-list "B300-PyTorch-1"

[tongyuantongyu]

/bot run --stage-list "B300-PyTorch-1"

[tensorrt-cicd]

PR_Github #58659 [ run ] triggered by Bot. Commit: db2bb50 Link to invocation

[tensorrt-cicd]

PR_Github #58659 [ run ] completed with state SUCCESS. Commit: db2bb50
/LLM/main/L0_MergeRequest_PR pipeline #47245 (Partly Tested) completed with status: 'SUCCESS'

CI Report

Link to invocation

[tongyuantongyu]

/bot run

[tensorrt-cicd]

PR_Github #58805 [ run ] triggered by Bot. Commit: db2bb50 Link to invocation

[tensorrt-cicd]

PR_Github #58805 [ run ] completed with state DISABLED
Pipeline is freezed and top-1 instance is under maintenance.

Link to invocation

[chzblych]

/bot run

[tensorrt-cicd]

PR_Github #58825 [ run ] triggered by Bot. Commit: db2bb50 Link to invocation

[tensorrt-cicd]

PR_Github #58825 [ run ] completed with state FAILURE. Commit: db2bb50
/LLM/main/L0_MergeRequest_PR pipeline #47383 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[tongyuantongyu]

/bot run

[tensorrt-cicd]

PR_Github #58865 [ run ] triggered by Bot. Commit: 7fc5195 Link to invocation

[tensorrt-cicd]

PR_Github #58865 [ run ] completed with state SUCCESS. Commit: 7fc5195
/LLM/main/L0_MergeRequest_PR pipeline #47412 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[tongyuantongyu]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #58905 [ run ] triggered by Bot. Commit: 7fc5195 Link to invocation

[tensorrt-cicd]

PR_Github #58905 [ run ] completed with state SUCCESS. Commit: 7fc5195
/LLM/main/L0_MergeRequest_PR pipeline #47446 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[tongyuantongyu]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #58978 [ run ] triggered by Bot. Commit: 7fc5195 Link to invocation

[tensorrt-cicd]

PR_Github #58978 [ run ] completed with state SUCCESS. Commit: 7fc5195
/LLM/main/L0_MergeRequest_PR pipeline #47509 completed with status: 'SUCCESS'

CI Report

Link to invocation