Make thrust::sort use radix sort with more comparators #1884

bernhardmgruber · 2024-06-19T18:42:05Z

The implementation of thrust::sort checks whether the supplied comparator is thrust::less<Key> or thrust::greater<Key> to choose whether to use radix sort or merge sort. However, we have more comparison operators available: ::cuda::std::less/greater and all transparent comparators (Key is void). This PR lets the additional comparators also use radix sort.

The PR to make the thrust comparators aliases to the libcudacxx one's will not fix this immediately, since thrust will stay with its own types for a deprecation iteration of CCCL: #1872

Newly included: ::cuda::std::less/greater and all transparent comparators.

bernhardmgruber · 2024-06-19T18:43:02Z

thrust/testing/cuda/sort.cu

+// TODO(bgruber): use a single test case with a concatenated key list and a cartesion product with the comparators
+SimpleUnitTest<TestRadixSortDispatch, IntegralTypes> TestRadixSortDispatchIntegralInstance;
+SimpleUnitTest<TestRadixSortDispatch, FloatingPointTypes> TestRadixSortDispatchFPInstance;


In the CUB unit tests I can use the metal library for type manipulation. Is that also available for Thrust tests?

github-actions · 2024-06-20T01:02:19Z

🟨 CI finished in 6h 18m: Pass: 99%/249 | Total: 4d 21h | Avg: 28m 20s | Max: 55m 53s | Hits: 61%/246735

🟨 cub: Pass: 98%/131 | Total: 2d 19h | Avg: 30m 43s | Max: 51m 46s | Hits: 53%/107469

🔍 cpu: amd64 🔍
  🔍 amd64              Pass:  98%/123 | Total:  2d 14h | Avg: 30m 29s | Max: 51m 46s | Hits:  54%/100653
  🟩 arm64              Pass: 100%/8   | Total:  4h 34m | Avg: 34m 22s | Max: 39m 11s | Hits:  41%/6816  
🔍 ctk: 12.4 🔍
  🟩 11.1               Pass: 100%/15  | Total:  7h 02m | Avg: 28m 08s | Max: 49m 21s | Hits:  32%/11568 
  🟩 11.8               Pass: 100%/3   | Total:  2h 18m | Avg: 46m 17s | Max: 49m 03s | Hits:  41%/2556  
  🔍 12.4               Pass:  98%/113 | Total:  2d 09h | Avg: 30m 39s | Max: 51m 46s | Hits:  56%/93345 
🔍 cudacxx: nvcc12.4 🔍
  🟩 ClangCUDA17        Pass: 100%/2   | Total: 43m 08s | Avg: 21m 34s | Max: 21m 58s | Hits:  43%/1408  
  🟩 nvcc11.1           Pass: 100%/15  | Total:  7h 02m | Avg: 28m 08s | Max: 49m 21s | Hits:  32%/11568 
  🟩 nvcc11.8           Pass: 100%/3   | Total:  2h 18m | Avg: 46m 17s | Max: 49m 03s | Hits:  41%/2556  
  🔍 nvcc12.4           Pass:  98%/111 | Total:  2d 09h | Avg: 30m 49s | Max: 51m 46s | Hits:  56%/91937 
🔍 cudacxx_family: nvcc 🔍
  🟩 ClangCUDA          Pass: 100%/2   | Total: 43m 08s | Avg: 21m 34s | Max: 21m 58s | Hits:  43%/1408  
  🔍 nvcc               Pass:  98%/129 | Total:  2d 18h | Avg: 30m 52s | Max: 51m 46s | Hits:  53%/106061
🟨 cxx
  🟩 Clang9             Pass: 100%/6   | Total:  3h 06m | Avg: 31m 02s | Max: 38m 25s | Hits:  35%/4890  
  🟩 Clang10            Pass: 100%/3   | Total:  1h 43m | Avg: 34m 27s | Max: 36m 27s | Hits:  40%/2562  
  🟩 Clang11            Pass: 100%/4   | Total:  2h 17m | Avg: 34m 22s | Max: 35m 35s | Hits:  40%/3416  
  🟩 Clang12            Pass: 100%/4   | Total:  2h 15m | Avg: 33m 51s | Max: 34m 22s | Hits:  40%/3416  
  🟩 Clang13            Pass: 100%/4   | Total:  2h 16m | Avg: 34m 14s | Max: 35m 28s | Hits:  40%/3416  
  🟩 Clang14            Pass: 100%/4   | Total:  2h 24m | Avg: 36m 07s | Max: 37m 37s | Hits:  41%/3416  
  🟩 Clang15            Pass: 100%/4   | Total:  2h 17m | Avg: 34m 15s | Max: 35m 37s | Hits:  41%/3408  
  🟩 Clang16            Pass: 100%/4   | Total:  2h 14m | Avg: 33m 40s | Max: 34m 02s | Hits:  41%/3408  
  🟨 Clang17            Pass:  96%/26  | Total: 10h 46m | Avg: 24m 52s | Max: 37m 20s | Hits:  77%/21004 
  🟩 GCC6               Pass: 100%/2   | Total: 51m 35s | Avg: 25m 47s | Max: 26m 35s | Hits:  39%/1552  
  🟩 GCC7               Pass: 100%/6   | Total:  3h 01m | Avg: 30m 19s | Max: 34m 53s | Hits:  36%/4893  
  🟩 GCC8               Pass: 100%/6   | Total:  3h 01m | Avg: 30m 10s | Max: 34m 50s | Hits:  36%/4893  
  🟩 GCC9               Pass: 100%/6   | Total:  3h 11m | Avg: 31m 50s | Max: 36m 57s | Hits:  36%/4893  
  🟩 GCC10              Pass: 100%/4   | Total:  2h 20m | Avg: 35m 14s | Max: 36m 59s | Hits:  41%/3416  
  🟩 GCC11              Pass: 100%/7   | Total:  4h 45m | Avg: 40m 51s | Max: 49m 03s | Hits:  41%/5964  
  🟩 GCC12              Pass: 100%/4   | Total:  2h 25m | Avg: 36m 24s | Max: 37m 48s | Hits:  41%/3408  
  🟨 GCC13              Pass:  96%/28  | Total: 11h 26m | Avg: 24m 30s | Max: 42m 38s | Hits:  72%/23004 
  🟩 Intel2023.2.0      Pass: 100%/3   | Total:  1h 55m | Avg: 38m 23s | Max: 43m 40s | Hits:  37%/2340  
  🟩 MSVC14.16          Pass: 100%/1   | Total: 49m 21s | Avg: 49m 21s | Max: 49m 21s | Hits:  38%/695   
  🟩 MSVC14.29          Pass: 100%/2   | Total:  1h 28m | Avg: 44m 21s | Max: 44m 44s | Hits:  38%/1390  
  🟩 MSVC14.39          Pass: 100%/3   | Total:  2h 25m | Avg: 48m 28s | Max: 51m 46s | Hits:  38%/2085  
🟨 cxx_family
  🟨 Clang              Pass:  98%/59  | Total:  1d 05h | Avg: 29m 52s | Max: 38m 25s | Hits:  55%/48936 
  🟨 GCC                Pass:  98%/63  | Total:  1d 07h | Avg: 29m 35s | Max: 49m 03s | Hits:  53%/52023 
  🟩 Intel              Pass: 100%/3   | Total:  1h 55m | Avg: 38m 23s | Max: 43m 40s | Hits:  37%/2340  
  🟩 MSVC               Pass: 100%/6   | Total:  4h 43m | Avg: 47m 14s | Max: 51m 46s | Hits:  38%/4170  
🟨 jobs
  🟩 Build              Pass: 100%/99  | Total:  2d 08h | Avg: 34m 02s | Max: 51m 46s | Hits:  39%/81909 
  🟨 DeviceLaunch       Pass:  87%/8   | Total:  2h 42m | Avg: 20m 15s | Max: 42m 38s | Hits:  94%/5964  
  🟩 GraphCapture       Pass: 100%/8   | Total:  2h 09m | Avg: 16m 12s | Max: 24m 00s | Hits:  99%/6816  
  🟨 HostLaunch         Pass:  87%/8   | Total:  2h 32m | Avg: 19m 00s | Max: 27m 52s | Hits:  99%/5964  
  🟩 TestGPU            Pass: 100%/8   | Total:  3h 30m | Avg: 26m 20s | Max: 31m 51s | Hits:  99%/6816  
🟨 std
  🟨 11                 Pass:  97%/34  | Total: 16h 42m | Avg: 29m 29s | Max: 44m 36s | Hits:  53%/27687 
  🟩 14                 Pass: 100%/37  | Total: 19h 27m | Avg: 31m 33s | Max: 49m 21s | Hits:  53%/30624 
  🟨 17                 Pass:  97%/36  | Total: 19h 09m | Avg: 31m 55s | Max: 49m 03s | Hits:  48%/29005 
  🟩 20                 Pass: 100%/24  | Total: 11h 45m | Avg: 29m 23s | Max: 51m 46s | Hits:  61%/20153 
🟨 gpu
  🟨 v100               Pass:  98%/131 | Total:  2d 19h | Avg: 30m 43s | Max: 51m 46s | Hits:  53%/107469
🟩 sm
  🟩 60;70;80;90        Pass: 100%/3   | Total:  2h 18m | Avg: 46m 17s | Max: 49m 03s | Hits:  41%/2556  
  🟩 90a                Pass: 100%/4   | Total:  1h 19m | Avg: 19m 53s | Max: 21m 51s | Hits:  41%/3408

🟩 thrust: Pass: 100%/118 | Total: 2d 02h | Avg: 25m 42s | Max: 55m 53s | Hits: 67%/139266

🟩 cpu
  🟩 amd64              Pass: 100%/110 | Total:  1d 23h | Avg: 25m 40s | Max: 55m 53s | Hits:  68%/129822
  🟩 arm64              Pass: 100%/8   | Total:  3h 29m | Avg: 26m 14s | Max: 28m 58s | Hits:  63%/9444  
🟩 ctk
  🟩 11.1               Pass: 100%/15  | Total:  6h 19m | Avg: 25m 19s | Max: 47m 12s | Hits:  63%/17705 
  🟩 11.8               Pass: 100%/3   | Total:  1h 49m | Avg: 36m 27s | Max: 41m 32s | Hits:  63%/3543  
  🟩 12.4               Pass: 100%/100 | Total:  1d 18h | Avg: 25m 26s | Max: 55m 53s | Hits:  68%/118018
🟩 cudacxx
  🟩 ClangCUDA17        Pass: 100%/2   | Total: 50m 08s | Avg: 25m 04s | Max: 25m 55s | Hits:  62%/2360  
  🟩 nvcc11.1           Pass: 100%/15  | Total:  6h 19m | Avg: 25m 19s | Max: 47m 12s | Hits:  63%/17705 
  🟩 nvcc11.8           Pass: 100%/3   | Total:  1h 49m | Avg: 36m 27s | Max: 41m 32s | Hits:  63%/3543  
  🟩 nvcc12.4           Pass: 100%/98  | Total:  1d 17h | Avg: 25m 27s | Max: 55m 53s | Hits:  68%/115658
🟩 cudacxx_family
  🟩 ClangCUDA          Pass: 100%/2   | Total: 50m 08s | Avg: 25m 04s | Max: 25m 55s | Hits:  62%/2360  
  🟩 nvcc               Pass: 100%/116 | Total:  2d 01h | Avg: 25m 43s | Max: 55m 53s | Hits:  68%/136906
🟩 cxx
  🟩 Clang9             Pass: 100%/6   | Total:  2h 29m | Avg: 24m 57s | Max: 28m 01s | Hits:  63%/7080  
  🟩 Clang10            Pass: 100%/3   | Total:  1h 19m | Avg: 26m 34s | Max: 30m 01s | Hits:  63%/3540  
  🟩 Clang11            Pass: 100%/4   | Total:  1h 53m | Avg: 28m 22s | Max: 31m 08s | Hits:  63%/4720  
  🟩 Clang12            Pass: 100%/4   | Total:  1h 44m | Avg: 26m 09s | Max: 30m 03s | Hits:  63%/4720  
  🟩 Clang13            Pass: 100%/4   | Total:  1h 44m | Avg: 26m 03s | Max: 27m 40s | Hits:  63%/4720  
  🟩 Clang14            Pass: 100%/4   | Total:  1h 49m | Avg: 27m 19s | Max: 30m 08s | Hits:  63%/4720  
  🟩 Clang15            Pass: 100%/4   | Total:  1h 47m | Avg: 26m 49s | Max: 29m 22s | Hits:  63%/4720  
  🟩 Clang16            Pass: 100%/4   | Total:  1h 50m | Avg: 27m 38s | Max: 30m 21s | Hits:  63%/4720  
  🟩 Clang17            Pass: 100%/18  | Total:  5h 40m | Avg: 18m 56s | Max: 30m 41s | Hits:  79%/21240 
  🟩 GCC6               Pass: 100%/2   | Total: 45m 38s | Avg: 22m 49s | Max: 25m 01s | Hits:  63%/2360  
  🟩 GCC7               Pass: 100%/6   | Total:  2h 30m | Avg: 25m 04s | Max: 29m 14s | Hits:  63%/7086  
  🟩 GCC8               Pass: 100%/6   | Total:  2h 31m | Avg: 25m 13s | Max: 27m 48s | Hits:  63%/7086  
  🟩 GCC9               Pass: 100%/6   | Total:  2h 37m | Avg: 26m 14s | Max: 30m 23s | Hits:  63%/7086  
  🟩 GCC10              Pass: 100%/4   | Total:  1h 54m | Avg: 28m 31s | Max: 31m 51s | Hits:  63%/4724  
  🟩 GCC11              Pass: 100%/7   | Total:  3h 40m | Avg: 31m 33s | Max: 41m 32s | Hits:  56%/8267  
  🟩 GCC12              Pass: 100%/4   | Total:  1h 57m | Avg: 29m 22s | Max: 34m 47s | Hits:  63%/4724  
  🟩 GCC13              Pass: 100%/20  | Total:  6h 35m | Avg: 19m 45s | Max: 37m 07s | Hits:  73%/23620 
  🟩 Intel2023.2.0      Pass: 100%/3   | Total:  1h 35m | Avg: 31m 56s | Max: 34m 55s | Hits:  63%/3549  
  🟩 MSVC14.16          Pass: 100%/1   | Total: 47m 12s | Avg: 47m 12s | Max: 47m 12s | Hits:  61%/1176  
  🟩 MSVC14.29          Pass: 100%/2   | Total:  1h 42m | Avg: 51m 03s | Max: 52m 00s | Hits:  61%/2352  
  🟩 MSVC14.39          Pass: 100%/6   | Total:  3h 36m | Avg: 36m 04s | Max: 55m 53s | Hits:  80%/7056  
🟩 cxx_family
  🟩 Clang              Pass: 100%/51  | Total: 20h 19m | Avg: 23m 54s | Max: 31m 08s | Hits:  69%/60180 
  🟩 GCC                Pass: 100%/55  | Total: 22h 32m | Avg: 24m 35s | Max: 41m 32s | Hits:  66%/64953 
  🟩 Intel              Pass: 100%/3   | Total:  1h 35m | Avg: 31m 56s | Max: 34m 55s | Hits:  63%/3549  
  🟩 MSVC               Pass: 100%/9   | Total:  6h 05m | Avg: 40m 38s | Max: 55m 53s | Hits:  73%/10584 
🟩 gpu
  🟩 v100               Pass: 100%/118 | Total:  2d 02h | Avg: 25m 42s | Max: 55m 53s | Hits:  67%/139266
🟩 jobs
  🟩 Build              Pass: 100%/99  | Total:  1d 22h | Avg: 28m 08s | Max: 55m 53s | Hits:  62%/116850
  🟩 TestCPU            Pass: 100%/11  | Total:  1h 49m | Avg:  9m 58s | Max: 21m 22s | Hits:  99%/12972 
  🟩 TestGPU            Pass: 100%/8   | Total:  2h 18m | Avg: 17m 20s | Max: 37m 07s | Hits:  89%/9444  
🟩 sm
  🟩 60;70;80;90        Pass: 100%/3   | Total:  1h 49m | Avg: 36m 27s | Max: 41m 32s | Hits:  63%/3543  
  🟩 90a                Pass: 100%/4   | Total: 58m 57s | Avg: 14m 44s | Max: 16m 00s | Hits:  63%/4724  
🟩 std
  🟩 11                 Pass: 100%/30  | Total: 10h 38m | Avg: 21m 16s | Max: 29m 33s | Hits:  69%/35418 
  🟩 14                 Pass: 100%/34  | Total: 15h 30m | Avg: 27m 21s | Max: 50m 07s | Hits:  67%/40122 
  🟩 17                 Pass: 100%/33  | Total: 15h 23m | Avg: 27m 59s | Max: 52m 00s | Hits:  65%/38946 
  🟩 20                 Pass: 100%/21  | Total:  9h 01m | Avg: 25m 48s | Max: 55m 53s | Hits:  69%/24780

👃 Inspect Changes

Modifications in project?

	Project
	CCCL Infrastructure
	libcu++
	CUB
+/-	Thrust
	CUDA Experimental

Modifications in project or dependencies?

	Project
	CCCL Infrastructure
	libcu++
+/-	CUB
+/-	Thrust
	CUDA Experimental

🏃‍ Runner counts (total jobs: 249)

#	Runner
178	`linux-amd64-cpu16`
40	`linux-amd64-gpu-v100-latest-1`
16	`linux-arm64-cpu16`
15	`windows-amd64-cpu16`

github-actions · 2024-06-23T22:39:19Z

🟩 CI finished in 4d 03h: Pass: 100%/249 | Total: 4d 22h | Avg: 28m 26s | Max: 55m 53s | Hits: 62%/248439

🟩 cub: Pass: 100%/131 | Total: 2d 19h | Avg: 30m 54s | Max: 51m 46s | Hits: 54%/109173

🟩 cpu
  🟩 amd64              Pass: 100%/123 | Total:  2d 14h | Avg: 30m 40s | Max: 51m 46s | Hits:  55%/102357
  🟩 arm64              Pass: 100%/8   | Total:  4h 34m | Avg: 34m 22s | Max: 39m 11s | Hits:  41%/6816  
🟩 ctk
  🟩 11.1               Pass: 100%/15  | Total:  7h 02m | Avg: 28m 08s | Max: 49m 21s | Hits:  32%/11568 
  🟩 11.8               Pass: 100%/3   | Total:  2h 18m | Avg: 46m 17s | Max: 49m 03s | Hits:  41%/2556  
  🟩 12.4               Pass: 100%/113 | Total:  2d 10h | Avg: 30m 51s | Max: 51m 46s | Hits:  57%/95049 
🟩 cudacxx
  🟩 ClangCUDA17        Pass: 100%/2   | Total: 43m 08s | Avg: 21m 34s | Max: 21m 58s | Hits:  43%/1408  
  🟩 nvcc11.1           Pass: 100%/15  | Total:  7h 02m | Avg: 28m 08s | Max: 49m 21s | Hits:  32%/11568 
  🟩 nvcc11.8           Pass: 100%/3   | Total:  2h 18m | Avg: 46m 17s | Max: 49m 03s | Hits:  41%/2556  
  🟩 nvcc12.4           Pass: 100%/111 | Total:  2d 09h | Avg: 31m 01s | Max: 51m 46s | Hits:  57%/93641 
🟩 cudacxx_family
  🟩 ClangCUDA          Pass: 100%/2   | Total: 43m 08s | Avg: 21m 34s | Max: 21m 58s | Hits:  43%/1408  
  🟩 nvcc               Pass: 100%/129 | Total:  2d 18h | Avg: 31m 02s | Max: 51m 46s | Hits:  54%/107765
🟩 cxx
  🟩 Clang9             Pass: 100%/6   | Total:  3h 06m | Avg: 31m 02s | Max: 38m 25s | Hits:  35%/4890  
  🟩 Clang10            Pass: 100%/3   | Total:  1h 43m | Avg: 34m 27s | Max: 36m 27s | Hits:  40%/2562  
  🟩 Clang11            Pass: 100%/4   | Total:  2h 17m | Avg: 34m 22s | Max: 35m 35s | Hits:  40%/3416  
  🟩 Clang12            Pass: 100%/4   | Total:  2h 15m | Avg: 33m 51s | Max: 34m 22s | Hits:  40%/3416  
  🟩 Clang13            Pass: 100%/4   | Total:  2h 16m | Avg: 34m 14s | Max: 35m 28s | Hits:  40%/3416  
  🟩 Clang14            Pass: 100%/4   | Total:  2h 24m | Avg: 36m 07s | Max: 37m 37s | Hits:  41%/3416  
  🟩 Clang15            Pass: 100%/4   | Total:  2h 17m | Avg: 34m 15s | Max: 35m 37s | Hits:  41%/3408  
  🟩 Clang16            Pass: 100%/4   | Total:  2h 14m | Avg: 33m 40s | Max: 34m 02s | Hits:  41%/3408  
  🟩 Clang17            Pass: 100%/26  | Total: 10h 59m | Avg: 25m 21s | Max: 37m 20s | Hits:  78%/21856 
  🟩 GCC6               Pass: 100%/2   | Total: 51m 35s | Avg: 25m 47s | Max: 26m 35s | Hits:  39%/1552  
  🟩 GCC7               Pass: 100%/6   | Total:  3h 01m | Avg: 30m 19s | Max: 34m 53s | Hits:  36%/4893  
  🟩 GCC8               Pass: 100%/6   | Total:  3h 01m | Avg: 30m 10s | Max: 34m 50s | Hits:  36%/4893  
  🟩 GCC9               Pass: 100%/6   | Total:  3h 11m | Avg: 31m 50s | Max: 36m 57s | Hits:  36%/4893  
  🟩 GCC10              Pass: 100%/4   | Total:  2h 20m | Avg: 35m 14s | Max: 36m 59s | Hits:  41%/3416  
  🟩 GCC11              Pass: 100%/7   | Total:  4h 45m | Avg: 40m 51s | Max: 49m 03s | Hits:  41%/5964  
  🟩 GCC12              Pass: 100%/4   | Total:  2h 25m | Avg: 36m 24s | Max: 37m 48s | Hits:  41%/3408  
  🟩 GCC13              Pass: 100%/28  | Total: 11h 36m | Avg: 24m 52s | Max: 42m 38s | Hits:  73%/23856 
  🟩 Intel2023.2.0      Pass: 100%/3   | Total:  1h 55m | Avg: 38m 23s | Max: 43m 40s | Hits:  37%/2340  
  🟩 MSVC14.16          Pass: 100%/1   | Total: 49m 21s | Avg: 49m 21s | Max: 49m 21s | Hits:  38%/695   
  🟩 MSVC14.29          Pass: 100%/2   | Total:  1h 28m | Avg: 44m 21s | Max: 44m 44s | Hits:  38%/1390  
  🟩 MSVC14.39          Pass: 100%/3   | Total:  2h 25m | Avg: 48m 28s | Max: 51m 46s | Hits:  38%/2085  
🟩 cxx_family
  🟩 Clang              Pass: 100%/59  | Total:  1d 05h | Avg: 30m 04s | Max: 38m 25s | Hits:  56%/49788 
  🟩 GCC                Pass: 100%/63  | Total:  1d 07h | Avg: 29m 45s | Max: 49m 03s | Hits:  54%/52875 
  🟩 Intel              Pass: 100%/3   | Total:  1h 55m | Avg: 38m 23s | Max: 43m 40s | Hits:  37%/2340  
  🟩 MSVC               Pass: 100%/6   | Total:  4h 43m | Avg: 47m 14s | Max: 51m 46s | Hits:  38%/4170  
🟩 gpu
  🟩 v100               Pass: 100%/131 | Total:  2d 19h | Avg: 30m 54s | Max: 51m 46s | Hits:  54%/109173
🟩 jobs
  🟩 Build              Pass: 100%/99  | Total:  2d 08h | Avg: 34m 02s | Max: 51m 46s | Hits:  39%/81909 
  🟩 DeviceLaunch       Pass: 100%/8   | Total:  2h 54m | Avg: 21m 49s | Max: 42m 38s | Hits:  94%/6816  
  🟩 GraphCapture       Pass: 100%/8   | Total:  2h 09m | Avg: 16m 12s | Max: 24m 00s | Hits:  99%/6816  
  🟩 HostLaunch         Pass: 100%/8   | Total:  2h 42m | Avg: 20m 19s | Max: 27m 52s | Hits:  99%/6816  
  🟩 TestGPU            Pass: 100%/8   | Total:  3h 30m | Avg: 26m 20s | Max: 31m 51s | Hits:  99%/6816  
🟩 sm
  🟩 60;70;80;90        Pass: 100%/3   | Total:  2h 18m | Avg: 46m 17s | Max: 49m 03s | Hits:  41%/2556  
  🟩 90a                Pass: 100%/4   | Total:  1h 19m | Avg: 19m 53s | Max: 21m 51s | Hits:  41%/3408  
🟩 std
  🟩 11                 Pass: 100%/34  | Total: 16h 53m | Avg: 29m 48s | Max: 44m 36s | Hits:  54%/28539 
  🟩 14                 Pass: 100%/37  | Total: 19h 27m | Avg: 31m 33s | Max: 49m 21s | Hits:  53%/30624 
  🟩 17                 Pass: 100%/36  | Total: 19h 21m | Avg: 32m 16s | Max: 49m 03s | Hits:  50%/29857 
  🟩 20                 Pass: 100%/24  | Total: 11h 45m | Avg: 29m 23s | Max: 51m 46s | Hits:  61%/20153

🟩 thrust: Pass: 100%/118 | Total: 2d 02h | Avg: 25m 42s | Max: 55m 53s | Hits: 67%/139266

🟩 cpu
  🟩 amd64              Pass: 100%/110 | Total:  1d 23h | Avg: 25m 40s | Max: 55m 53s | Hits:  68%/129822
  🟩 arm64              Pass: 100%/8   | Total:  3h 29m | Avg: 26m 14s | Max: 28m 58s | Hits:  63%/9444  
🟩 ctk
  🟩 11.1               Pass: 100%/15  | Total:  6h 19m | Avg: 25m 19s | Max: 47m 12s | Hits:  63%/17705 
  🟩 11.8               Pass: 100%/3   | Total:  1h 49m | Avg: 36m 27s | Max: 41m 32s | Hits:  63%/3543  
  🟩 12.4               Pass: 100%/100 | Total:  1d 18h | Avg: 25m 26s | Max: 55m 53s | Hits:  68%/118018
🟩 cudacxx
  🟩 ClangCUDA17        Pass: 100%/2   | Total: 50m 08s | Avg: 25m 04s | Max: 25m 55s | Hits:  62%/2360  
  🟩 nvcc11.1           Pass: 100%/15  | Total:  6h 19m | Avg: 25m 19s | Max: 47m 12s | Hits:  63%/17705 
  🟩 nvcc11.8           Pass: 100%/3   | Total:  1h 49m | Avg: 36m 27s | Max: 41m 32s | Hits:  63%/3543  
  🟩 nvcc12.4           Pass: 100%/98  | Total:  1d 17h | Avg: 25m 27s | Max: 55m 53s | Hits:  68%/115658
🟩 cudacxx_family
  🟩 ClangCUDA          Pass: 100%/2   | Total: 50m 08s | Avg: 25m 04s | Max: 25m 55s | Hits:  62%/2360  
  🟩 nvcc               Pass: 100%/116 | Total:  2d 01h | Avg: 25m 43s | Max: 55m 53s | Hits:  68%/136906
🟩 cxx
  🟩 Clang9             Pass: 100%/6   | Total:  2h 29m | Avg: 24m 57s | Max: 28m 01s | Hits:  63%/7080  
  🟩 Clang10            Pass: 100%/3   | Total:  1h 19m | Avg: 26m 34s | Max: 30m 01s | Hits:  63%/3540  
  🟩 Clang11            Pass: 100%/4   | Total:  1h 53m | Avg: 28m 22s | Max: 31m 08s | Hits:  63%/4720  
  🟩 Clang12            Pass: 100%/4   | Total:  1h 44m | Avg: 26m 09s | Max: 30m 03s | Hits:  63%/4720  
  🟩 Clang13            Pass: 100%/4   | Total:  1h 44m | Avg: 26m 03s | Max: 27m 40s | Hits:  63%/4720  
  🟩 Clang14            Pass: 100%/4   | Total:  1h 49m | Avg: 27m 19s | Max: 30m 08s | Hits:  63%/4720  
  🟩 Clang15            Pass: 100%/4   | Total:  1h 47m | Avg: 26m 49s | Max: 29m 22s | Hits:  63%/4720  
  🟩 Clang16            Pass: 100%/4   | Total:  1h 50m | Avg: 27m 38s | Max: 30m 21s | Hits:  63%/4720  
  🟩 Clang17            Pass: 100%/18  | Total:  5h 40m | Avg: 18m 56s | Max: 30m 41s | Hits:  79%/21240 
  🟩 GCC6               Pass: 100%/2   | Total: 45m 38s | Avg: 22m 49s | Max: 25m 01s | Hits:  63%/2360  
  🟩 GCC7               Pass: 100%/6   | Total:  2h 30m | Avg: 25m 04s | Max: 29m 14s | Hits:  63%/7086  
  🟩 GCC8               Pass: 100%/6   | Total:  2h 31m | Avg: 25m 13s | Max: 27m 48s | Hits:  63%/7086  
  🟩 GCC9               Pass: 100%/6   | Total:  2h 37m | Avg: 26m 14s | Max: 30m 23s | Hits:  63%/7086  
  🟩 GCC10              Pass: 100%/4   | Total:  1h 54m | Avg: 28m 31s | Max: 31m 51s | Hits:  63%/4724  
  🟩 GCC11              Pass: 100%/7   | Total:  3h 40m | Avg: 31m 33s | Max: 41m 32s | Hits:  56%/8267  
  🟩 GCC12              Pass: 100%/4   | Total:  1h 57m | Avg: 29m 22s | Max: 34m 47s | Hits:  63%/4724  
  🟩 GCC13              Pass: 100%/20  | Total:  6h 35m | Avg: 19m 45s | Max: 37m 07s | Hits:  73%/23620 
  🟩 Intel2023.2.0      Pass: 100%/3   | Total:  1h 35m | Avg: 31m 56s | Max: 34m 55s | Hits:  63%/3549  
  🟩 MSVC14.16          Pass: 100%/1   | Total: 47m 12s | Avg: 47m 12s | Max: 47m 12s | Hits:  61%/1176  
  🟩 MSVC14.29          Pass: 100%/2   | Total:  1h 42m | Avg: 51m 03s | Max: 52m 00s | Hits:  61%/2352  
  🟩 MSVC14.39          Pass: 100%/6   | Total:  3h 36m | Avg: 36m 04s | Max: 55m 53s | Hits:  80%/7056  
🟩 cxx_family
  🟩 Clang              Pass: 100%/51  | Total: 20h 19m | Avg: 23m 54s | Max: 31m 08s | Hits:  69%/60180 
  🟩 GCC                Pass: 100%/55  | Total: 22h 32m | Avg: 24m 35s | Max: 41m 32s | Hits:  66%/64953 
  🟩 Intel              Pass: 100%/3   | Total:  1h 35m | Avg: 31m 56s | Max: 34m 55s | Hits:  63%/3549  
  🟩 MSVC               Pass: 100%/9   | Total:  6h 05m | Avg: 40m 38s | Max: 55m 53s | Hits:  73%/10584 
🟩 gpu
  🟩 v100               Pass: 100%/118 | Total:  2d 02h | Avg: 25m 42s | Max: 55m 53s | Hits:  67%/139266
🟩 jobs
  🟩 Build              Pass: 100%/99  | Total:  1d 22h | Avg: 28m 08s | Max: 55m 53s | Hits:  62%/116850
  🟩 TestCPU            Pass: 100%/11  | Total:  1h 49m | Avg:  9m 58s | Max: 21m 22s | Hits:  99%/12972 
  🟩 TestGPU            Pass: 100%/8   | Total:  2h 18m | Avg: 17m 20s | Max: 37m 07s | Hits:  89%/9444  
🟩 sm
  🟩 60;70;80;90        Pass: 100%/3   | Total:  1h 49m | Avg: 36m 27s | Max: 41m 32s | Hits:  63%/3543  
  🟩 90a                Pass: 100%/4   | Total: 58m 57s | Avg: 14m 44s | Max: 16m 00s | Hits:  63%/4724  
🟩 std
  🟩 11                 Pass: 100%/30  | Total: 10h 38m | Avg: 21m 16s | Max: 29m 33s | Hits:  69%/35418 
  🟩 14                 Pass: 100%/34  | Total: 15h 30m | Avg: 27m 21s | Max: 50m 07s | Hits:  67%/40122 
  🟩 17                 Pass: 100%/33  | Total: 15h 23m | Avg: 27m 59s | Max: 52m 00s | Hits:  65%/38946 
  🟩 20                 Pass: 100%/21  | Total:  9h 01m | Avg: 25m 48s | Max: 55m 53s | Hits:  69%/24780

👃 Inspect Changes

Modifications in project?

	Project
	CCCL Infrastructure
	libcu++
	CUB
+/-	Thrust
	CUDA Experimental

Modifications in project or dependencies?

	Project
	CCCL Infrastructure
	libcu++
+/-	CUB
+/-	Thrust
	CUDA Experimental

🏃‍ Runner counts (total jobs: 249)

#	Runner
178	`linux-amd64-cpu16`
40	`linux-amd64-gpu-v100-latest-1`
16	`linux-arm64-cpu16`
15	`windows-amd64-cpu16`

Make thrust::sort use radix sort with more comparators

9298a76

Newly included: ::cuda::std::less/greater and all transparent comparators.

bernhardmgruber requested review from a team as code owners June 19, 2024 18:42

bernhardmgruber requested review from alliepiper and ericniebler June 19, 2024 18:42

bernhardmgruber added the thrust For all items related to Thrust. label Jun 19, 2024

bernhardmgruber commented Jun 19, 2024

View reviewed changes

miscco approved these changes Jun 20, 2024

View reviewed changes

bernhardmgruber enabled auto-merge (squash) June 23, 2024 22:23

bernhardmgruber merged commit bb44c7d into NVIDIA:main Jun 23, 2024
265 checks passed

bernhardmgruber deleted the radix_fix branch June 23, 2024 22:42

bernhardmgruber mentioned this pull request Jul 8, 2024

Port thrust::merge[_by_key] to CUB #1817

Merged

9 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Make thrust::sort use radix sort with more comparators #1884

Make thrust::sort use radix sort with more comparators #1884

bernhardmgruber commented Jun 19, 2024 •

edited

Loading

bernhardmgruber Jun 19, 2024

github-actions bot commented Jun 20, 2024

🟨 cub: Pass: 98%/131 | Total: 2d 19h | Avg: 30m 43s | Max: 51m 46s | Hits: 53%/107469

🟩 thrust: Pass: 100%/118 | Total: 2d 02h | Avg: 25m 42s | Max: 55m 53s | Hits: 67%/139266

👃 Inspect Changes

Modifications in project?

Modifications in project or dependencies?

🏃‍ Runner counts (total jobs: 249)

github-actions bot commented Jun 23, 2024

🟩 cub: Pass: 100%/131 | Total: 2d 19h | Avg: 30m 54s | Max: 51m 46s | Hits: 54%/109173

🟩 thrust: Pass: 100%/118 | Total: 2d 02h | Avg: 25m 42s | Max: 55m 53s | Hits: 67%/139266

👃 Inspect Changes

Modifications in project?

Modifications in project or dependencies?

🏃‍ Runner counts (total jobs: 249)

Make thrust::sort use radix sort with more comparators #1884

Make thrust::sort use radix sort with more comparators #1884

Conversation

bernhardmgruber commented Jun 19, 2024 • edited Loading

bernhardmgruber Jun 19, 2024

Choose a reason for hiding this comment

github-actions bot commented Jun 20, 2024

🟨 cub: Pass: 98%/131 | Total: 2d 19h | Avg: 30m 43s | Max: 51m 46s | Hits: 53%/107469

🟩 thrust: Pass: 100%/118 | Total: 2d 02h | Avg: 25m 42s | Max: 55m 53s | Hits: 67%/139266

👃 Inspect Changes

Modifications in project?

Modifications in project or dependencies?

🏃‍ Runner counts (total jobs: 249)

github-actions bot commented Jun 23, 2024

🟩 cub: Pass: 100%/131 | Total: 2d 19h | Avg: 30m 54s | Max: 51m 46s | Hits: 54%/109173

🟩 thrust: Pass: 100%/118 | Total: 2d 02h | Avg: 25m 42s | Max: 55m 53s | Hits: 67%/139266

👃 Inspect Changes

Modifications in project?

Modifications in project or dependencies?

🏃‍ Runner counts (total jobs: 249)

bernhardmgruber commented Jun 19, 2024 •

edited

Loading