Use double buffer for block_scan

[pfultz2]

This avoids incorrect concurrent writes. And fixes the nonzero test failure on navi.

[migraphx-bot]

Test	Batch	Rate new 557157	Rate old 225873	Diff	Compare
torchvision-resnet50	64	2,830.24	2,829.18	0.04%	✅
torchvision-resnet50_fp16	64	6,487.35	6,499.27	-0.18%	✅
torchvision-densenet121	32	2,091.20	2,098.86	-0.36%	✅
torchvision-densenet121_fp16	32	3,672.12	3,663.93	0.22%	✅
torchvision-inceptionv3	32	1,583.55	1,585.30	-0.11%	✅
torchvision-inceptionv3_fp16	32	2,569.10	2,570.22	-0.04%	✅
cadene-inceptionv4	16	702.61	703.00	-0.06%	✅
cadene-resnext64x4	16	691.69	692.19	-0.07%	✅
slim-mobilenet	64	8,335.78	8,333.06	0.03%	✅
slim-nasnetalarge	64	225.55	225.49	0.02%	✅
slim-resnet50v2	64	2,665.93	2,664.66	0.05%	✅
bert-mrpc-onnx	8	820.33	822.01	-0.20%	✅
bert-mrpc-tf	1	388.52	388.70	-0.05%	✅
pytorch-examples-wlang-gru	1	303.94	303.71	0.07%	✅
pytorch-examples-wlang-lstm	1	317.13	320.29	-0.99%	✅
torchvision-resnet50_1	1	600.58	605.77	-0.86%	✅
torchvision-inceptionv3_1	1	342.52	341.44	0.32%	✅
cadene-dpn92_1	1	399.34	399.88	-0.13%	✅
cadene-resnext101_1	1	328.84	328.75	0.02%	✅
slim-vgg16_1	1	458.43	460.49	-0.45%	✅
slim-mobilenet_1	1	2,130.59	2,127.55	0.14%	✅
slim-inceptionv4_1	1	219.15	220.20	-0.48%	✅
onnx-taau-downsample	1	302.57	303.83	-0.41%	✅
dlrm-criteoterabyte	1	21.59	21.61	-0.12%	✅
dlrm-criteoterabyte_fp16	1	40.66	40.62	0.09%	✅
agentmodel	1	nan	nan	nan%	❌
unet_fp16	2	55.71	54.72	1.81%	✅
resnet50v1_fp16	1	950.60	954.24	-0.38%	✅
bert_base_cased_fp16	64	967.53	903.07	7.14%	🔆
bert_large_uncased_fp16	32	303.96	285.54	6.45%	🔆
bert_large_fp16	1	166.74	166.50	0.15%	✅
distilgpt2_fp16	16	1,273.27	1,282.05	-0.69%	✅

This build is not recommended to merge 🔴

[migraphx-bot]

    :white_check_mark:bert-mrpc-onnx: PASSED: MIGraphX meets tolerance

    :white_check_mark:bert-mrpc-tf: PASSED: MIGraphX meets tolerance

    :white_check_mark:pytorch-examples-wlang-gru: PASSED: MIGraphX meets tolerance

    :white_check_mark:pytorch-examples-wlang-lstm: PASSED: MIGraphX meets tolerance

    :white_check_mark:torchvision-resnet50_1: PASSED: MIGraphX meets tolerance

    :white_check_mark:torchvision-inceptionv3_1: PASSED: MIGraphX meets tolerance

    :white_check_mark:cadene-dpn92_1: PASSED: MIGraphX meets tolerance

    :white_check_mark:cadene-resnext101_1: PASSED: MIGraphX meets tolerance

    :white_check_mark:slim-vgg16_1: PASSED: MIGraphX meets tolerance

    :white_check_mark:slim-mobilenet_1: PASSED: MIGraphX meets tolerance

    :white_check_mark:slim-inceptionv4_1: PASSED: MIGraphX meets tolerance

    :white_check_mark:dlrm-criteoterabyte: PASSED: MIGraphX meets tolerance

❌agentmodel: ERROR - check error output

Traceback (most recent call last):
File "/src/AMDMIGraphX/tools/accuracy/accuracy_checker.py", line 336, in
main()
File "/src/AMDMIGraphX/tools/accuracy/accuracy_checker.py", line 254, in main
pred_migx = np.array(model.run(params)[-1])
RuntimeError: /src/AMDMIGraphX/src/targets/gpu/device/include/migraphx/gpu/device/visit.hpp:140: hip_visit_views_impl: Ranks must be the same

    :white_check_mark:unet: PASSED: MIGraphX meets tolerance

    :white_check_mark:resnet50v1: PASSED: MIGraphX meets tolerance

🔴bert_base_cased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output

🔴bert_large_uncased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output

    :white_check_mark:bert_large: PASSED: MIGraphX meets tolerance

🔴distilgpt2_fp16: FAILED: MIGraphX is not within tolerance - check verbose output