[STF] stackable stf resources by caugonnet · Pull Request #2674 · NVIDIA/cccl

caugonnet · 2024-10-31T09:00:20Z

Description

This introduces helper methods to improve how we nest contexts to better leverage CUDA Graphs

Checklist

New or existing tests cover these changes.
The documentation is up to date with these changes.

copy-pr-bot · 2024-10-31T09:00:24Z

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

caugonnet · 2024-10-31T09:02:49Z

/ok to test

caugonnet · 2024-10-31T14:49:26Z

/ok to test

github-actions · 2024-10-31T15:13:37Z

🟩 CI finished in 23m 04s: Pass: 100%/54 | Total: 4h 36m | Avg: 5m 06s | Max: 17m 44s | Hits: 89%/224

🟩 cudax: Pass: 100%/54 | Total: 4h 36m | Avg: 5m 06s | Max: 17m 44s | Hits: 89%/224

🟩 cpu
  🟩 amd64              Pass: 100%/50  | Total:  4h 20m | Avg:  5m 13s | Max: 17m 44s | Hits:  89%/224   
  🟩 arm64              Pass: 100%/4   | Total: 15m 16s | Avg:  3m 49s | Max:  4m 54s
🟩 ctk
  🟩 12.0               Pass: 100%/19  | Total:  1h 40m | Avg:  5m 16s | Max: 17m 32s | Hits:  89%/112   
  🟩 12.5               Pass: 100%/2   | Total:  9m 25s | Avg:  4m 42s | Max:  4m 47s
  🟩 12.6               Pass: 100%/33  | Total:  2h 46m | Avg:  5m 03s | Max: 17m 44s | Hits:  89%/112   
🟩 cudacxx
  🟩 nvcc12.0           Pass: 100%/19  | Total:  1h 40m | Avg:  5m 16s | Max: 17m 32s | Hits:  89%/112   
  🟩 nvcc12.5           Pass: 100%/2   | Total:  9m 25s | Avg:  4m 42s | Max:  4m 47s
  🟩 nvcc12.6           Pass: 100%/33  | Total:  2h 46m | Avg:  5m 03s | Max: 17m 44s | Hits:  89%/112   
🟩 cudacxx_family
  🟩 nvcc               Pass: 100%/54  | Total:  4h 36m | Avg:  5m 06s | Max: 17m 44s | Hits:  89%/224   
🟩 cxx
  🟩 Clang9             Pass: 100%/2   | Total:  7m 21s | Avg:  3m 40s | Max:  4m 02s
  🟩 Clang10            Pass: 100%/2   | Total:  6m 59s | Avg:  3m 29s | Max:  3m 41s
  🟩 Clang11            Pass: 100%/4   | Total: 13m 28s | Avg:  3m 22s | Max:  3m 34s
  🟩 Clang12            Pass: 100%/4   | Total: 13m 16s | Avg:  3m 19s | Max:  3m 27s
  🟩 Clang13            Pass: 100%/4   | Total: 13m 11s | Avg:  3m 17s | Max:  3m 23s
  🟩 Clang14            Pass: 100%/4   | Total: 27m 40s | Avg:  6m 55s | Max: 17m 27s
  🟩 Clang15            Pass: 100%/2   | Total:  7m 01s | Avg:  3m 30s | Max:  3m 37s
  🟩 Clang16            Pass: 100%/4   | Total: 13m 48s | Avg:  3m 27s | Max:  3m 48s
  🟩 Clang17            Pass: 100%/2   | Total:  7m 19s | Avg:  3m 39s | Max:  3m 42s
  🟩 Clang18            Pass: 100%/2   | Total: 19m 21s | Avg:  9m 40s | Max: 15m 57s
  🟩 GCC9               Pass: 100%/2   | Total:  7m 29s | Avg:  3m 44s | Max:  3m 56s
  🟩 GCC10              Pass: 100%/4   | Total: 15m 40s | Avg:  3m 55s | Max:  4m 13s
  🟩 GCC11              Pass: 100%/4   | Total: 14m 55s | Avg:  3m 43s | Max:  3m 53s
  🟩 GCC12              Pass: 100%/7   | Total:  1h 07m | Avg:  9m 34s | Max: 17m 44s
  🟩 GCC13              Pass: 100%/3   | Total: 12m 09s | Avg:  4m 03s | Max:  4m 54s
  🟩 MSVC14.36          Pass: 100%/1   | Total: 10m 04s | Avg: 10m 04s | Max: 10m 04s | Hits:  89%/112   
  🟩 MSVC14.39          Pass: 100%/1   | Total: 10m 06s | Avg: 10m 06s | Max: 10m 06s | Hits:  89%/112   
  🟩 NVHPC24.7          Pass: 100%/2   | Total:  9m 25s | Avg:  4m 42s | Max:  4m 47s
🟩 cxx_family
  🟩 Clang              Pass: 100%/30  | Total:  2h 09m | Avg:  4m 18s | Max: 17m 27s
  🟩 GCC                Pass: 100%/20  | Total:  1h 57m | Avg:  5m 51s | Max: 17m 44s
  🟩 MSVC               Pass: 100%/2   | Total: 20m 10s | Avg: 10m 05s | Max: 10m 06s | Hits:  89%/224   
  🟩 NVHPC              Pass: 100%/2   | Total:  9m 25s | Avg:  4m 42s | Max:  4m 47s
🟩 gpu
  🟩 v100               Pass: 100%/54  | Total:  4h 36m | Avg:  5m 06s | Max: 17m 44s | Hits:  89%/224   
🟩 jobs
  🟩 Build              Pass: 100%/49  | Total:  3h 10m | Avg:  3m 53s | Max: 10m 06s | Hits:  89%/224   
  🟩 Test               Pass: 100%/5   | Total:  1h 25m | Avg: 17m 04s | Max: 17m 44s
🟩 sm
  🟩 90                 Pass: 100%/1   | Total:  3m 15s | Avg:  3m 15s | Max:  3m 15s
  🟩 90a                Pass: 100%/1   | Total:  3m 22s | Avg:  3m 22s | Max:  3m 22s
🟩 std
  🟩 17                 Pass: 100%/29  | Total:  2h 12m | Avg:  4m 33s | Max: 17m 32s
  🟩 20                 Pass: 100%/25  | Total:  2h 24m | Avg:  5m 45s | Max: 17m 44s | Hits:  89%/224

👃 Inspect Changes

Modifications in project?

	Project
	CCCL Infrastructure
	libcu++
	CUB
	Thrust
+/-	CUDA Experimental
	python
	CCCL C Parallel Library
	Catch2Helper

Modifications in project or dependencies?

	Project
	CCCL Infrastructure
	libcu++
	CUB
	Thrust
+/-	CUDA Experimental
	python
	CCCL C Parallel Library
	Catch2Helper

🏃‍ Runner counts (total jobs: 54)

#	Runner
43	`linux-amd64-cpu16`
5	`linux-amd64-gpu-v100-latest-1`
4	`linux-arm64-cpu16`
2	`windows-amd64-cpu16`

caugonnet · 2024-11-04T10:29:10Z

/ok to test

github-actions · 2024-11-04T11:42:50Z

🟩 CI finished in 1h 12m: Pass: 100%/54 | Total: 4h 28m | Avg: 4m 58s | Max: 23m 23s | Hits: 89%/224

🟩 cudax: Pass: 100%/54 | Total: 4h 28m | Avg: 4m 58s | Max: 23m 23s | Hits: 89%/224

🟩 cpu
  🟩 amd64              Pass: 100%/50  | Total:  4h 15m | Avg:  5m 06s | Max: 23m 23s | Hits:  89%/224   
  🟩 arm64              Pass: 100%/4   | Total: 13m 38s | Avg:  3m 24s | Max:  4m 29s
🟩 ctk
  🟩 12.0               Pass: 100%/19  | Total:  1h 35m | Avg:  5m 02s | Max: 22m 16s | Hits:  89%/112   
  🟩 12.5               Pass: 100%/2   | Total:  9m 39s | Avg:  4m 49s | Max:  4m 55s
  🟩 12.6               Pass: 100%/33  | Total:  2h 43m | Avg:  4m 56s | Max: 23m 23s | Hits:  89%/112   
🟩 cudacxx
  🟩 nvcc12.0           Pass: 100%/19  | Total:  1h 35m | Avg:  5m 02s | Max: 22m 16s | Hits:  89%/112   
  🟩 nvcc12.5           Pass: 100%/2   | Total:  9m 39s | Avg:  4m 49s | Max:  4m 55s
  🟩 nvcc12.6           Pass: 100%/33  | Total:  2h 43m | Avg:  4m 56s | Max: 23m 23s | Hits:  89%/112   
🟩 cudacxx_family
  🟩 nvcc               Pass: 100%/54  | Total:  4h 28m | Avg:  4m 58s | Max: 23m 23s | Hits:  89%/224   
🟩 cxx
  🟩 Clang9             Pass: 100%/2   | Total:  6m 47s | Avg:  3m 23s | Max:  3m 29s
  🟩 Clang10            Pass: 100%/2   | Total:  7m 03s | Avg:  3m 31s | Max:  3m 42s
  🟩 Clang11            Pass: 100%/4   | Total: 12m 35s | Avg:  3m 08s | Max:  3m 26s
  🟩 Clang12            Pass: 100%/4   | Total: 12m 22s | Avg:  3m 05s | Max:  3m 12s
  🟩 Clang13            Pass: 100%/4   | Total: 12m 54s | Avg:  3m 13s | Max:  3m 25s
  🟩 Clang14            Pass: 100%/4   | Total: 27m 30s | Avg:  6m 52s | Max: 17m 45s
  🟩 Clang15            Pass: 100%/2   | Total:  6m 52s | Avg:  3m 26s | Max:  3m 37s
  🟩 Clang16            Pass: 100%/4   | Total: 13m 29s | Avg:  3m 22s | Max:  3m 40s
  🟩 Clang17            Pass: 100%/2   | Total:  7m 10s | Avg:  3m 35s | Max:  3m 38s
  🟩 Clang18            Pass: 100%/2   | Total: 20m 58s | Avg: 10m 29s | Max: 17m 48s
  🟩 GCC9               Pass: 100%/2   | Total:  6m 10s | Avg:  3m 05s | Max:  3m 17s
  🟩 GCC10              Pass: 100%/4   | Total: 12m 43s | Avg:  3m 10s | Max:  3m 21s
  🟩 GCC11              Pass: 100%/4   | Total: 12m 17s | Avg:  3m 04s | Max:  3m 12s
  🟩 GCC12              Pass: 100%/7   | Total:  1h 16m | Avg: 10m 58s | Max: 23m 23s
  🟩 GCC13              Pass: 100%/3   | Total: 10m 06s | Avg:  3m 22s | Max:  4m 29s
  🟩 MSVC14.36          Pass: 100%/1   | Total:  6m 56s | Avg:  6m 56s | Max:  6m 56s | Hits:  89%/112   
  🟩 MSVC14.39          Pass: 100%/1   | Total:  6m 30s | Avg:  6m 30s | Max:  6m 30s | Hits:  89%/112   
  🟩 NVHPC24.7          Pass: 100%/2   | Total:  9m 39s | Avg:  4m 49s | Max:  4m 55s
🟩 cxx_family
  🟩 Clang              Pass: 100%/30  | Total:  2h 07m | Avg:  4m 15s | Max: 17m 48s
  🟩 GCC                Pass: 100%/20  | Total:  1h 58m | Avg:  5m 54s | Max: 23m 23s
  🟩 MSVC               Pass: 100%/2   | Total: 13m 26s | Avg:  6m 43s | Max:  6m 56s | Hits:  89%/224   
  🟩 NVHPC              Pass: 100%/2   | Total:  9m 39s | Avg:  4m 49s | Max:  4m 55s
🟩 gpu
  🟩 v100               Pass: 100%/54  | Total:  4h 28m | Avg:  4m 58s | Max: 23m 23s | Hits:  89%/224   
🟩 jobs
  🟩 Build              Pass: 100%/49  | Total:  2h 49m | Avg:  3m 27s | Max:  6m 56s | Hits:  89%/224   
  🟩 Test               Pass: 100%/5   | Total:  1h 39m | Avg: 19m 56s | Max: 23m 23s
🟩 sm
  🟩 90                 Pass: 100%/1   | Total:  2m 46s | Avg:  2m 46s | Max:  2m 46s
  🟩 90a                Pass: 100%/1   | Total:  2m 45s | Avg:  2m 45s | Max:  2m 45s
🟩 std
  🟩 17                 Pass: 100%/29  | Total:  2h 14m | Avg:  4m 39s | Max: 23m 23s
  🟩 20                 Pass: 100%/25  | Total:  2h 13m | Avg:  5m 21s | Max: 18m 29s | Hits:  89%/224

👃 Inspect Changes

Modifications in project?

	Project
	CCCL Infrastructure
	libcu++
	CUB
	Thrust
+/-	CUDA Experimental
	python
	CCCL C Parallel Library
	Catch2Helper

Modifications in project or dependencies?

	Project
	CCCL Infrastructure
	libcu++
	CUB
	Thrust
+/-	CUDA Experimental
	python
	CCCL C Parallel Library
	Catch2Helper

🏃‍ Runner counts (total jobs: 54)

#	Runner
43	`linux-amd64-cpu16`
5	`linux-amd64-gpu-v100-latest-1`
4	`linux-arm64-cpu16`
2	`windows-amd64-cpu16`

cudax/examples/stf/binary_fhe_stackable.cu

cudax/include/cuda/experimental/__stf/utility/stackable_ctx.cuh

cudax/include/cuda/experimental/stf.cuh

caugonnet · 2025-01-14T10:37:09Z

/ok to test

github-actions · 2025-01-14T11:20:50Z

🟩 CI finished in 40m 53s: Pass: 100%/20 | Total: 3h 17m | Avg: 9m 53s | Max: 24m 35s | Hits: 582%/312

🟩 cudax: Pass: 100%/20 | Total: 3h 17m | Avg: 9m 53s | Max: 24m 35s | Hits: 582%/312

🟩 cpu
  🟩 amd64              Pass: 100%/16  | Total:  2h 45m | Avg: 10m 19s | Max: 24m 35s | Hits: 582%/312   
  🟩 arm64              Pass: 100%/4   | Total: 32m 37s | Avg:  8m 09s | Max:  8m 52s
🟩 ctk
  🟩 12.0               Pass: 100%/1   | Total: 11m 33s | Avg: 11m 33s | Max: 11m 33s | Hits: 582%/156   
  🟩 12.5               Pass: 100%/2   | Total: 11m 00s | Avg:  5m 30s | Max:  5m 42s
  🟩 12.6               Pass: 100%/17  | Total:  2h 55m | Avg: 10m 18s | Max: 24m 35s | Hits: 582%/156   
🟩 cudacxx
  🟩 nvcc12.0           Pass: 100%/1   | Total: 11m 33s | Avg: 11m 33s | Max: 11m 33s | Hits: 582%/156   
  🟩 nvcc12.5           Pass: 100%/2   | Total: 11m 00s | Avg:  5m 30s | Max:  5m 42s
  🟩 nvcc12.6           Pass: 100%/17  | Total:  2h 55m | Avg: 10m 18s | Max: 24m 35s | Hits: 582%/156   
🟩 cudacxx_family
  🟩 nvcc               Pass: 100%/20  | Total:  3h 17m | Avg:  9m 53s | Max: 24m 35s | Hits: 582%/312   
🟩 cxx
  🟩 Clang14            Pass: 100%/1   | Total:  9m 26s | Avg:  9m 26s | Max:  9m 26s
  🟩 Clang15            Pass: 100%/1   | Total:  9m 50s | Avg:  9m 50s | Max:  9m 50s
  🟩 Clang16            Pass: 100%/1   | Total:  9m 15s | Avg:  9m 15s | Max:  9m 15s
  🟩 Clang17            Pass: 100%/1   | Total:  9m 56s | Avg:  9m 56s | Max:  9m 56s
  🟩 Clang18            Pass: 100%/4   | Total: 41m 39s | Avg: 10m 24s | Max: 16m 15s
  🟩 GCC10              Pass: 100%/1   | Total:  9m 42s | Avg:  9m 42s | Max:  9m 42s
  🟩 GCC11              Pass: 100%/1   | Total:  9m 13s | Avg:  9m 13s | Max:  9m 13s
  🟩 GCC12              Pass: 100%/2   | Total: 34m 07s | Avg: 17m 03s | Max: 24m 35s
  🟩 GCC13              Pass: 100%/4   | Total: 30m 57s | Avg:  7m 44s | Max:  8m 52s
  🟩 MSVC14.36          Pass: 100%/1   | Total: 11m 33s | Avg: 11m 33s | Max: 11m 33s | Hits: 582%/156   
  🟩 MSVC14.39          Pass: 100%/1   | Total: 11m 06s | Avg: 11m 06s | Max: 11m 06s | Hits: 582%/156   
  🟩 NVHPC24.7          Pass: 100%/2   | Total: 11m 00s | Avg:  5m 30s | Max:  5m 42s
🟩 cxx_family
  🟩 Clang              Pass: 100%/8   | Total:  1h 20m | Avg: 10m 00s | Max: 16m 15s
  🟩 GCC                Pass: 100%/8   | Total:  1h 23m | Avg: 10m 29s | Max: 24m 35s
  🟩 MSVC               Pass: 100%/2   | Total: 22m 39s | Avg: 11m 19s | Max: 11m 33s | Hits: 582%/312   
  🟩 NVHPC              Pass: 100%/2   | Total: 11m 00s | Avg:  5m 30s | Max:  5m 42s
🟩 gpu
  🟩 v100               Pass: 100%/20  | Total:  3h 17m | Avg:  9m 53s | Max: 24m 35s | Hits: 582%/312   
🟩 jobs
  🟩 Build              Pass: 100%/18  | Total:  2h 36m | Avg:  8m 43s | Max: 11m 33s | Hits: 582%/312   
  🟩 Test               Pass: 100%/2   | Total: 40m 50s | Avg: 20m 25s | Max: 24m 35s
🟩 sm
  🟩 90                 Pass: 100%/1   | Total:  6m 57s | Avg:  6m 57s | Max:  6m 57s
  🟩 90a                Pass: 100%/1   | Total:  7m 24s | Avg:  7m 24s | Max:  7m 24s
🟩 std
  🟩 17                 Pass: 100%/4   | Total: 28m 13s | Avg:  7m 03s | Max:  7m 50s
  🟩 20                 Pass: 100%/16  | Total:  2h 49m | Avg: 10m 35s | Max: 24m 35s | Hits: 582%/312

👃 Inspect Changes

Modifications in project?

	Project
	CCCL Infrastructure
	libcu++
	CUB
	Thrust
+/-	CUDA Experimental
	python
	CCCL C Parallel Library
	Catch2Helper

Modifications in project or dependencies?

	Project
	CCCL Infrastructure
	libcu++
	CUB
	Thrust
+/-	CUDA Experimental
	python
	CCCL C Parallel Library
	Catch2Helper

🏃‍ Runner counts (total jobs: 20)

#	Runner
12	`linux-amd64-cpu16`
4	`linux-arm64-cpu16`
2	`windows-amd64-cpu16`
2	`linux-amd64-gpu-v100-latest-1`

caugonnet · 2025-01-15T07:50:21Z

/ok to test

github-actions · 2025-01-15T08:33:28Z

🟩 CI finished in 42m 04s: Pass: 100%/20 | Total: 4h 07m | Avg: 12m 22s | Max: 22m 12s | Hits: 582%/312

🟩 cudax: Pass: 100%/20 | Total: 4h 07m | Avg: 12m 22s | Max: 22m 12s | Hits: 582%/312

🟩 cpu
  🟩 amd64              Pass: 100%/16  | Total:  3h 22m | Avg: 12m 39s | Max: 22m 12s | Hits: 582%/312   
  🟩 arm64              Pass: 100%/4   | Total: 44m 46s | Avg: 11m 11s | Max: 11m 49s
🟩 ctk
  🟩 12.0               Pass: 100%/1   | Total: 10m 50s | Avg: 10m 50s | Max: 10m 50s | Hits: 582%/156   
  🟩 12.5               Pass: 100%/2   | Total: 11m 41s | Avg:  5m 50s | Max:  5m 51s
  🟩 12.6               Pass: 100%/17  | Total:  3h 44m | Avg: 13m 13s | Max: 22m 12s | Hits: 582%/156   
🟩 cudacxx
  🟩 nvcc12.0           Pass: 100%/1   | Total: 10m 50s | Avg: 10m 50s | Max: 10m 50s | Hits: 582%/156   
  🟩 nvcc12.5           Pass: 100%/2   | Total: 11m 41s | Avg:  5m 50s | Max:  5m 51s
  🟩 nvcc12.6           Pass: 100%/17  | Total:  3h 44m | Avg: 13m 13s | Max: 22m 12s | Hits: 582%/156   
🟩 cudacxx_family
  🟩 nvcc               Pass: 100%/20  | Total:  4h 07m | Avg: 12m 22s | Max: 22m 12s | Hits: 582%/312   
🟩 cxx
  🟩 Clang14            Pass: 100%/1   | Total: 12m 12s | Avg: 12m 12s | Max: 12m 12s
  🟩 Clang15            Pass: 100%/1   | Total: 13m 19s | Avg: 13m 19s | Max: 13m 19s
  🟩 Clang16            Pass: 100%/1   | Total: 13m 09s | Avg: 13m 09s | Max: 13m 09s
  🟩 Clang17            Pass: 100%/1   | Total: 13m 12s | Avg: 13m 12s | Max: 13m 12s
  🟩 Clang18            Pass: 100%/4   | Total: 54m 39s | Avg: 13m 39s | Max: 18m 11s
  🟩 GCC10              Pass: 100%/1   | Total: 13m 50s | Avg: 13m 50s | Max: 13m 50s
  🟩 GCC11              Pass: 100%/1   | Total: 14m 10s | Avg: 14m 10s | Max: 14m 10s
  🟩 GCC12              Pass: 100%/2   | Total: 36m 46s | Avg: 18m 23s | Max: 22m 12s
  🟩 GCC13              Pass: 100%/4   | Total: 42m 15s | Avg: 10m 33s | Max: 11m 49s
  🟩 MSVC14.36          Pass: 100%/1   | Total: 10m 50s | Avg: 10m 50s | Max: 10m 50s | Hits: 582%/156   
  🟩 MSVC14.39          Pass: 100%/1   | Total: 11m 18s | Avg: 11m 18s | Max: 11m 18s | Hits: 582%/156   
  🟩 NVHPC24.7          Pass: 100%/2   | Total: 11m 41s | Avg:  5m 50s | Max:  5m 51s
🟩 cxx_family
  🟩 Clang              Pass: 100%/8   | Total:  1h 46m | Avg: 13m 18s | Max: 18m 11s
  🟩 GCC                Pass: 100%/8   | Total:  1h 47m | Avg: 13m 22s | Max: 22m 12s
  🟩 MSVC               Pass: 100%/2   | Total: 22m 08s | Avg: 11m 04s | Max: 11m 18s | Hits: 582%/312   
  🟩 NVHPC              Pass: 100%/2   | Total: 11m 41s | Avg:  5m 50s | Max:  5m 51s
🟩 gpu
  🟩 v100               Pass: 100%/20  | Total:  4h 07m | Avg: 12m 22s | Max: 22m 12s | Hits: 582%/312   
🟩 jobs
  🟩 Build              Pass: 100%/18  | Total:  3h 26m | Avg: 11m 29s | Max: 14m 34s | Hits: 582%/312   
  🟩 Test               Pass: 100%/2   | Total: 40m 23s | Avg: 20m 11s | Max: 22m 12s
🟩 sm
  🟩 90                 Pass: 100%/1   | Total:  8m 53s | Avg:  8m 53s | Max:  8m 53s
  🟩 90a                Pass: 100%/1   | Total: 10m 48s | Avg: 10m 48s | Max: 10m 48s
🟩 std
  🟩 17                 Pass: 100%/4   | Total: 36m 10s | Avg:  9m 02s | Max: 10m 45s
  🟩 20                 Pass: 100%/16  | Total:  3h 31m | Avg: 13m 11s | Max: 22m 12s | Hits: 582%/312

👃 Inspect Changes

Modifications in project?

	Project
	CCCL Infrastructure
	libcu++
	CUB
	Thrust
+/-	CUDA Experimental
	python
	CCCL C Parallel Library
	Catch2Helper

Modifications in project or dependencies?

	Project
	CCCL Infrastructure
	libcu++
	CUB
	Thrust
+/-	CUDA Experimental
	python
	CCCL C Parallel Library
	Catch2Helper

🏃‍ Runner counts (total jobs: 20)

#	Runner
12	`linux-amd64-cpu16`
4	`linux-arm64-cpu16`
2	`windows-amd64-cpu16`
2	`linux-amd64-gpu-v100-latest-1`

cudax/examples/stf/binary_fhe_stackable.cu

cudax/include/cuda/experimental/__stf/utility/stackable_ctx.cuh

caugonnet · 2025-01-22T09:33:13Z

cudax/include/cuda/experimental/__stf/stackable/stackable_ctx.cuh

+ * @brief This class defines a context that behaves as a context which can have nested subcontexts (implemented as local
+ * CUDA graphs)
+ */
+class stackable_ctx


We need a == operator too

cudax/include/cuda/experimental/__stf/utility/stackable_ctx.cuh

cudax/examples/stf/binary_fhe_stackable.cu

cudax/include/cuda/experimental/__stf/internal/logical_data.cuh

caugonnet · 2025-01-29T07:56:35Z

cudax/test/stf/local_stf/stackable2.cu

+
+    ctx.pop();
+  }
+


TODO check results.

caugonnet · 2025-01-30T10:12:29Z

/ok to test

github-actions · 2025-01-30T10:52:45Z

🟨 CI finished in 39m 08s: Pass: 85%/20 | Total: 4h 10m | Avg: 12m 30s | Max: 17m 58s | Hits: 388%/522

🟨 cudax: Pass: 85%/20 | Total: 4h 10m | Avg: 12m 30s | Max: 17m 58s | Hits: 388%/522

🔍 cpu: amd64 🔍
  🔍 amd64              Pass:  81%/16  | Total:  3h 21m | Avg: 12m 34s | Max: 17m 58s | Hits: 388%/522   
  🟩 arm64              Pass: 100%/4   | Total: 49m 07s | Avg: 12m 16s | Max: 13m 08s
🔍 ctk: 12.6 🔍
  🟩 12.0               Pass: 100%/1   | Total:  9m 06s | Avg:  9m 06s | Max:  9m 06s | Hits: 388%/261   
  🟩 12.5               Pass: 100%/2   | Total: 14m 30s | Avg:  7m 15s | Max:  7m 17s
  🔍 12.6               Pass:  82%/17  | Total:  3h 46m | Avg: 13m 20s | Max: 17m 58s | Hits: 388%/261   
🔍 cudacxx: nvcc12.6 🔍
  🟩 nvcc12.0           Pass: 100%/1   | Total:  9m 06s | Avg:  9m 06s | Max:  9m 06s | Hits: 388%/261   
  🟩 nvcc12.5           Pass: 100%/2   | Total: 14m 30s | Avg:  7m 15s | Max:  7m 17s
  🔍 nvcc12.6           Pass:  82%/17  | Total:  3h 46m | Avg: 13m 20s | Max: 17m 58s | Hits: 388%/261   
🔍 std: 20 🔍
  🟩 17                 Pass: 100%/4   | Total: 41m 03s | Avg: 10m 15s | Max: 11m 56s
  🔍 20                 Pass:  81%/16  | Total:  3h 29m | Avg: 13m 04s | Max: 17m 58s | Hits: 388%/522   
🟨 cxx
  🟩 Clang14            Pass: 100%/1   | Total: 12m 44s | Avg: 12m 44s | Max: 12m 44s
  🟩 Clang15            Pass: 100%/1   | Total: 13m 43s | Avg: 13m 43s | Max: 13m 43s
  🟩 Clang16            Pass: 100%/1   | Total: 15m 20s | Avg: 15m 20s | Max: 15m 20s
  🟩 Clang17            Pass: 100%/1   | Total: 15m 37s | Avg: 15m 37s | Max: 15m 37s
  🟨 Clang18            Pass:  75%/4   | Total: 57m 22s | Avg: 14m 20s | Max: 17m 25s
  🟥 GCC10              Pass:   0%/1   | Total:  4m 18s | Avg:  4m 18s | Max:  4m 18s
  🟩 GCC11              Pass: 100%/1   | Total: 14m 26s | Avg: 14m 26s | Max: 14m 26s
  🟨 GCC12              Pass:  50%/2   | Total: 34m 39s | Avg: 17m 19s | Max: 17m 58s
  🟩 GCC13              Pass: 100%/4   | Total: 46m 27s | Avg: 11m 36s | Max: 13m 08s
  🟩 MSVC14.36          Pass: 100%/1   | Total:  9m 06s | Avg:  9m 06s | Max:  9m 06s | Hits: 388%/261   
  🟩 MSVC14.39          Pass: 100%/1   | Total: 12m 07s | Avg: 12m 07s | Max: 12m 07s | Hits: 388%/261   
  🟩 NVHPC24.7          Pass: 100%/2   | Total: 14m 30s | Avg:  7m 15s | Max:  7m 17s
🟨 cxx_family
  🟨 Clang              Pass:  87%/8   | Total:  1h 54m | Avg: 14m 20s | Max: 17m 25s
  🟨 GCC                Pass:  75%/8   | Total:  1h 39m | Avg: 12m 28s | Max: 17m 58s
  🟩 MSVC               Pass: 100%/2   | Total: 21m 13s | Avg: 10m 36s | Max: 12m 07s | Hits: 388%/522   
  🟩 NVHPC              Pass: 100%/2   | Total: 14m 30s | Avg:  7m 15s | Max:  7m 17s
🟨 cudacxx_family
  🟨 nvcc               Pass:  85%/20  | Total:  4h 10m | Avg: 12m 30s | Max: 17m 58s | Hits: 388%/522   
🟨 gpu
  🟨 v100               Pass:  85%/20  | Total:  4h 10m | Avg: 12m 30s | Max: 17m 58s | Hits: 388%/522   
🟨 jobs
  🟨 Build              Pass:  94%/18  | Total:  3h 34m | Avg: 11m 56s | Max: 16m 41s | Hits: 388%/522   
  🟥 Test               Pass:   0%/2   | Total: 35m 23s | Avg: 17m 41s | Max: 17m 58s
🟩 sm
  🟩 90                 Pass: 100%/1   | Total: 10m 09s | Avg: 10m 09s | Max: 10m 09s
  🟩 90a                Pass: 100%/1   | Total: 11m 14s | Avg: 11m 14s | Max: 11m 14s

👃 Inspect Changes

Modifications in project?

	Project
	CCCL Infrastructure
	libcu++
	CUB
	Thrust
+/-	CUDA Experimental
	python
	CCCL C Parallel Library
	Catch2Helper

Modifications in project or dependencies?

	Project
	CCCL Infrastructure
	libcu++
	CUB
	Thrust
+/-	CUDA Experimental
	python
	CCCL C Parallel Library
	Catch2Helper

🏃‍ Runner counts (total jobs: 20)

#	Runner
12	`linux-amd64-cpu16`
4	`linux-arm64-cpu16`
2	`windows-amd64-cpu16`
2	`linux-amd64-gpu-v100-latest-1`

caugonnet · 2026-03-20T06:33:55Z

/ok to test 69d714b

cudax/include/cuda/experimental/__stf/stackable/DESIGN.md

Replay all five transformation steps on the CPU after finalize() and EXPECT each element matches, ensuring the graph_scope RAII test verifies correctness rather than just running without error. Made-with: Cursor

Both operations must abort when called inside a nested context (i.e. after push/graph_scope). The tests use the same SIGABRT handler pattern as the existing stackable error checks. Made-with: Cursor

Verifies that read-only data is auto-pushed as read in nested graph scopes and that the original host buffer is not modified after finalize, confirming write-back is skipped for read-only data. Made-with: Cursor

Computes sqrt of 1..1024 via Newton's Babylonian method, iterating until max |change| < 1e-12. Demonstrates while_graph_scope with reduce-based convergence checking in ~80 lines, as a simpler introduction than the existing Jacobi examples. Made-with: Cursor

caugonnet · 2026-03-20T07:02:03Z

/ok to test f7c3c19

wait() returns a value and requires a copyable scalar type; slice<int> causes an incomplete-type compilation error. Switch to scalar_view<int> which is the intended usage pattern for wait(). Made-with: Cursor

Made-with: Cursor

Move graph_scope_guard and while_graph_scope_guard out of the stackable_ctx class into standalone definitions in stackable_ctx.cuh, consistent with repeat_graph_scope_guard which was already standalone. The nested-name syntax (stackable_ctx::graph_scope_guard) is preserved via forward declarations inside the class. Factory methods are now defined out-of-line after the guard classes. Reduces stackable_ctx_impl.cuh from 1654 to 1441 lines. Made-with: Cursor

The previous commit placed guard definitions after the UNITTESTED_FILE section, causing incomplete-type errors in the inline unit tests. Move all guard definitions (graph_scope_guard, while_graph_scope_guard, repeat_graph_scope_guard) before the #ifdef UNITTESTED_FILE block. Made-with: Cursor

caugonnet · 2026-03-20T08:12:01Z

/ok to test 7f486e5

…of CI job

caugonnet · 2026-03-20T09:34:41Z

/ok to test 5c8f6f0

github-actions · 2026-03-20T09:52:28Z

🥳 CI Workflow Results

🟩 Finished in 16m 14s: Pass: 100%/48 | Total: 4h 39m | Max: 14m 37s | Hits: 99%/26011

See results here.

This STF-specific dot file cleanup utility doesn't belong in the benchmarks directory. It demangles and simplifies CUDA STF template names in dot graph output, so it belongs alongside the STF code. Made-with: Cursor

caugonnet force-pushed the stackable_ctx_data branch from 1a454b0 to dcb75d2 Compare October 31, 2024 09:02

miscco reviewed Nov 4, 2024

View reviewed changes

caugonnet added the stf Sequential Task Flow programming model label Nov 7, 2024

caugonnet force-pushed the stackable_ctx_data branch from ff0ba38 to fb3de98 Compare November 7, 2024 12:36

caugonnet commented Jan 14, 2025

View reviewed changes

cudax/include/cuda/experimental/stf.cuh Outdated Show resolved Hide resolved

caugonnet changed the title ~~stackable stf resources~~ [STF] stackable stf resources Jan 14, 2025

andralex reviewed Jan 16, 2025

View reviewed changes

cudax/examples/stf/binary_fhe_stackable.cu Outdated Show resolved Hide resolved

andralex reviewed Jan 16, 2025

View reviewed changes

cudax/examples/stf/binary_fhe_stackable.cu Outdated Show resolved Hide resolved

andralex reviewed Jan 16, 2025

View reviewed changes

cudax/include/cuda/experimental/__stf/utility/stackable_ctx.cuh Outdated Show resolved Hide resolved

caugonnet mentioned this pull request Jan 22, 2025

[FEA]: STF logical token should be compatible with the freeze API #3475

Closed

1 task

caugonnet self-assigned this Jan 22, 2025