mersenneforum.org - gpuOwL: an OpenCL program for Mersenne primality testing

mersenneforum.org (https://www.mersenneforum.org/index.php)

- GpuOwl (https://www.mersenneforum.org/forumdisplay.php?f=171)

- - gpuOwL: an OpenCL program for Mersenne primality testing (https://www.mersenneforum.org/showthread.php?t=22204)

v5.0 failed with repeatable error on a prime exponent

Please change code to issue message and log entry re such a problem, leave the entry in worktodo, and continue to the next assignment, not halt.[CODE]C:\msys64\home\ken\gpuowl-compile\v5.0>openowl.exe -user kriesel -cpu condorella-rx480 -device 0
2018-11-01 08:33:15 gpuowl 5.0-f604bb1
2018-11-01 08:33:15 condorella-rx480 -user kriesel -cpu condorella-rx480 -device 0
2018-11-01 08:33:15 condorella-rx480 756839 FFT 512K: Width 64x8, Height 64x8; 1.44 bits/word
2018-11-01 08:33:15 condorella-rx480 using long carry kernels
2018-11-01 08:33:16 condorella-rx480 Ellesmere-36x1266-@28:0.0 Radeon (TM) RX 480 Graphics
2018-11-01 08:33:19 condorella-rx480 OpenCL compilation in 3338 ms, with "-DEXP=756839u -DWIDTH=512u -DSMALL_HEIGHT=512u -DMIDDLE=1u -I. -cl-fast-relaxed-math
-cl-std=CL2.0 "
2018-11-01 08:33:19 condorella-rx480 756839.owl not found, starting from the beginnig.
2018-11-01 08:33:19 condorella-rx480 powerSmooth(756839, 3000) has 4370 bits
2018-11-01 08:33:22 condorella-rx480 756839.owl loaded: k 0, B1 3000, block 400, res64 808e9fd922e8a0af, stage 1, baseBits 0
2018-11-01 08:33:22 condorella-rx480 756839 B1=3000 B2=70000 (effective B2=70000) selected 4230 P-1 points in 0.01s
2018-11-01 08:33:23 condorella-rx480 756839 OK 800/757200 [ 0.11%], 0.55 ms/it; 1 muls, -16.00 ms/mul; ETA 0d 00:07; 661c814a3224d4dc (check 0.23s)
2018-11-01 08:33:23 condorella-rx480 756839 GCD no factor (0.22s)
2018-11-01 08:33:28 condorella-rx480 756839 10000/757200 [ 1.32%], 0.54 ms/it; 611 muls, 0.37 ms/mul; ETA 0d 00:07; b14737e6b0df37e1
2018-11-01 08:33:34 condorella-rx480 756839 20000/757200 [ 2.64%], 0.56 ms/it; 1084 muls, 0.23 ms/mul; ETA 0d 00:07; d4f722ecbcaf72c7
2018-11-01 08:33:40 condorella-rx480 756839 30000/757200 [ 3.96%], 0.56 ms/it; 917 muls, 0.22 ms/mul; ETA 0d 00:07; becefb17ea7bc907
2018-11-01 08:33:45 condorella-rx480 756839 40000/757200 [ 5.28%], 0.52 ms/it; 598 muls, 0.70 ms/mul; ETA 0d 00:06; 4027f55e5abb465c
2018-11-01 08:33:51 condorella-rx480 756839 50000/757200 [ 6.60%], 0.58 ms/it; 353 muls, -1.06 ms/mul; ETA 0d 00:07; fe729c99f45fc2de
2018-11-01 08:33:56 condorella-rx480 756839 60000/757200 [ 7.92%], 0.53 ms/it; 340 muls, 0.21 ms/mul; ETA 0d 00:06; e70f4a41ec67b2a6
2018-11-01 08:34:01 condorella-rx480 756839 70000/757200 [ 9.24%], 0.49 ms/it; 326 muls, 1.29 ms/mul; ETA 0d 00:06; 4c814bfdf2dec664
2018-11-01 08:34:07 condorella-rx480 756839 80000/757200 [10.57%], 0.52 ms/it; ETA 0d 00:06; 96f668f515cdccc7
2018-11-01 08:34:12 condorella-rx480 756839 90000/757200 [11.89%], 0.52 ms/it; ETA 0d 00:06; 8900ac33e8fba278
2018-11-01 08:34:17 condorella-rx480 756839 100000/757200 [13.21%], 0.52 ms/it; ETA 0d 00:06; 32b7785790aed404
2018-11-01 08:34:22 condorella-rx480 756839 110000/757200 [14.53%], 0.52 ms/it; ETA 0d 00:06; 7b488dc9aca7799b
2018-11-01 08:34:28 condorella-rx480 756839 120000/757200 [15.85%], 0.52 ms/it; ETA 0d 00:06; f854ef85204f30d0
2018-11-01 08:34:33 condorella-rx480 756839 130000/757200 [17.17%], 0.53 ms/it; ETA 0d 00:05; fe46348342c6122e
2018-11-01 08:34:38 condorella-rx480 756839 140000/757200 [18.49%], 0.52 ms/it; ETA 0d 00:05; 3b70d7db6ab3bec3
2018-11-01 08:34:43 condorella-rx480 756839 150000/757200 [19.81%], 0.52 ms/it; ETA 0d 00:05; 486178b2041b6c08
2018-11-01 08:34:49 condorella-rx480 756839 EE 160000/757200 [21.13%], 0.52 ms/it; ETA 0d 00:05; 0b79956769558c34 (check 0.25s)
2018-11-01 08:34:49 condorella-rx480 756839.owl loaded: k 800, B1 3000, block 400, res64 661c814a3224d4dc, stage 1, baseBits 0
2018-11-01 08:34:54 condorella-rx480 756839 10000/757200 [ 1.32%], 0.63 ms/it; 611 muls, -0.60 ms/mul; ETA 0d 00:08; b14737e6b0df37e1
2018-11-01 08:35:00 condorella-rx480 756839 20000/757200 [ 2.64%], 0.54 ms/it; 1084 muls, 0.38 ms/mul; ETA 0d 00:07; d4f722ecbcaf72c7
2018-11-01 08:35:06 condorella-rx480 756839 30000/757200 [ 3.96%], 0.55 ms/it; 917 muls, 0.27 ms/mul; ETA 0d 00:07; becefb17ea7bc907
2018-11-01 08:35:11 condorella-rx480 756839 40000/757200 [ 5.28%], 0.53 ms/it; 598 muls, 0.53 ms/mul; ETA 0d 00:06; 4027f55e5abb465c
2018-11-01 08:35:17 condorella-rx480 756839 50000/757200 [ 6.60%], 0.55 ms/it; 353 muls, -0.12 ms/mul; ETA 0d 00:06; fe729c99f45fc2de
2018-11-01 08:35:22 condorella-rx480 756839 60000/757200 [ 7.92%], 0.48 ms/it; 340 muls, 1.72 ms/mul; ETA 0d 00:06; e70f4a41ec67b2a6
2018-11-01 08:35:28 condorella-rx480 756839 70000/757200 [ 9.24%], 0.52 ms/it; 326 muls, 0.55 ms/mul; ETA 0d 00:06; 4c814bfdf2dec664
2018-11-01 08:35:33 condorella-rx480 756839 80000/757200 [10.57%], 0.52 ms/it; ETA 0d 00:06; 96f668f515cdccc7
2018-11-01 08:35:38 condorella-rx480 756839 90000/757200 [11.89%], 0.52 ms/it; ETA 0d 00:06; 8900ac33e8fba278
2018-11-01 08:35:44 condorella-rx480 756839 100000/757200 [13.21%], 0.52 ms/it; ETA 0d 00:06; 32b7785790aed404
2018-11-01 08:35:49 condorella-rx480 756839 110000/757200 [14.53%], 0.52 ms/it; ETA 0d 00:06; 7b488dc9aca7799b
2018-11-01 08:35:54 condorella-rx480 756839 120000/757200 [15.85%], 0.52 ms/it; ETA 0d 00:06; f854ef85204f30d0
2018-11-01 08:35:59 condorella-rx480 756839 130000/757200 [17.17%], 0.52 ms/it; ETA 0d 00:05; fe46348342c6122e
2018-11-01 08:36:05 condorella-rx480 756839 140000/757200 [18.49%], 0.52 ms/it; ETA 0d 00:05; 3b70d7db6ab3bec3
2018-11-01 08:36:10 condorella-rx480 756839 150000/757200 [19.81%], 0.52 ms/it; ETA 0d 00:05; 486178b2041b6c08
2018-11-01 08:36:15 condorella-rx480 756839 EE 160000/757200 [21.13%], 0.52 ms/it; ETA 0d 00:05; 0b79956769558c34 (check 0.23s)
2018-11-01 08:36:15 condorella-rx480 756839.owl loaded: k 800, B1 3000, block 400, res64 661c814a3224d4dc, stage 1, baseBits 0
2018-11-01 08:36:21 condorella-rx480 756839 10000/757200 [ 1.32%], 0.64 ms/it; 611 muls, -0.77 ms/mul; ETA 0d 00:08; b14737e6b0df37e1
2018-11-01 08:36:27 condorella-rx480 756839 20000/757200 [ 2.64%], 0.55 ms/it; 1084 muls, 0.35 ms/mul; ETA 0d 00:07; d4f722ecbcaf72c7
2018-11-01 08:36:32 condorella-rx480 756839 30000/757200 [ 3.96%], 0.51 ms/it; 917 muls, 0.63 ms/mul; ETA 0d 00:06; becefb17ea7bc907
2018-11-01 08:36:38 condorella-rx480 756839 40000/757200 [ 5.28%], 0.52 ms/it; 598 muls, 0.64 ms/mul; ETA 0d 00:06; 4027f55e5abb465c
2018-11-01 08:36:43 condorella-rx480 756839 50000/757200 [ 6.60%], 0.57 ms/it; 353 muls, -0.64 ms/mul; ETA 0d 00:07; fe729c99f45fc2de
2018-11-01 08:36:49 condorella-rx480 756839 60000/757200 [ 7.92%], 0.55 ms/it; 340 muls, -0.21 ms/mul; ETA 0d 00:06; e70f4a41ec67b2a6
2018-11-01 08:36:54 condorella-rx480 756839 70000/757200 [ 9.24%], 0.54 ms/it; 326 muls, 0.01 ms/mul; ETA 0d 00:06; 4c814bfdf2dec664
2018-11-01 08:36:59 condorella-rx480 756839 80000/757200 [10.57%], 0.52 ms/it; ETA 0d 00:06; 96f668f515cdccc7
2018-11-01 08:37:05 condorella-rx480 756839 90000/757200 [11.89%], 0.52 ms/it; ETA 0d 00:06; 8900ac33e8fba278
2018-11-01 08:37:10 condorella-rx480 756839 100000/757200 [13.21%], 0.53 ms/it; ETA 0d 00:06; 32b7785790aed404
2018-11-01 08:37:15 condorella-rx480 756839 110000/757200 [14.53%], 0.53 ms/it; ETA 0d 00:06; 7b488dc9aca7799b
2018-11-01 08:37:21 condorella-rx480 756839 120000/757200 [15.85%], 0.52 ms/it; ETA 0d 00:06; f854ef85204f30d0
2018-11-01 08:37:26 condorella-rx480 756839 130000/757200 [17.17%], 0.53 ms/it; ETA 0d 00:05; fe46348342c6122e
2018-11-01 08:37:31 condorella-rx480 756839 140000/757200 [18.49%], 0.53 ms/it; ETA 0d 00:05; 3b70d7db6ab3bec3
2018-11-01 08:37:36 condorella-rx480 756839 150000/757200 [19.81%], 0.53 ms/it; ETA 0d 00:05; 486178b2041b6c08
2018-11-01 08:37:42 condorella-rx480 756839 EE 160000/757200 [21.13%], 0.52 ms/it; ETA 0d 00:05; 0b79956769558c34 (check 0.25s)
2018-11-01 08:37:42 condorella-rx480 3 sequential errors, will stop.
2018-11-01 08:37:42 condorella-rx480 Exiting because "too many errors"
2018-11-01 08:37:42 condorella-rx480 Bye[/CODE]

[QUOTE=kriesel;499252]Please change code to issue message and log entry re such a problem, leave the entry in worktodo, and continue to the next assignment, not halt.[CODE]C:\msys64\home\ken\gpuowl-compile\v5.0>openowl.exe -user kriesel -cpu condorella-rx480 -device 0
2018-11-01 08:33:15 gpuowl 5.0-f604bb1
2018-11-01 08:33:15 condorella-rx480 -user kriesel -cpu condorella-rx480 -device 0
2018-11-01 08:33:15 condorella-rx480 756839 FFT 512K: Width 64x8, Height 64x8; 1.44 bits/word
2018-11-01 08:33:15 condorella-rx480 using long carry kernels
2018-11-01 08:33:16 condorella-rx480 Ellesmere-36x1266-@28:0.0 Radeon (TM) RX 480 Graphics
2018-11-01 08:33:19 condorella-rx480 OpenCL compilation in 3338 ms, with "-DEXP=756839u -DWIDTH=512u -DSMALL_HEIGHT=512u -DMIDDLE=1u -I. -cl-fast-relaxed-math
-cl-std=CL2.0 "
2018-11-01 08:33:19 condorella-rx480 756839.owl not found, starting from the beginnig.
2018-11-01 08:33:19 condorella-rx480 powerSmooth(756839, 3000) has 4370 bits
2018-11-01 08:33:22 condorella-rx480 756839.owl loaded: k 0, B1 3000, block 400, res64 808e9fd922e8a0af, stage 1, baseBits 0
2018-11-01 08:33:22 condorella-rx480 756839 B1=3000 B2=70000 (effective B2=70000) selected 4230 P-1 points in 0.01s
2018-11-01 08:33:23 condorella-rx480 756839 OK 800/757200 [ 0.11%], 0.55 ms/it; 1 muls, -16.00 ms/mul; ETA 0d 00:07; 661c814a3224d4dc (check 0.23s)
2018-11-01 08:33:23 condorella-rx480 756839 GCD no factor (0.22s)
2018-11-01 08:33:28 condorella-rx480 756839 10000/757200 [ 1.32%], 0.54 ms/it; 611 muls, 0.37 ms/mul; ETA 0d 00:07; b14737e6b0df37e1
2018-11-01 08:33:34 condorella-rx480 756839 20000/757200 [ 2.64%], 0.56 ms/it; 1084 muls, 0.23 ms/mul; ETA 0d 00:07; d4f722ecbcaf72c7
2018-11-01 08:33:40 condorella-rx480 756839 30000/757200 [ 3.96%], 0.56 ms/it; 917 muls, 0.22 ms/mul; ETA 0d 00:07; becefb17ea7bc907
2018-11-01 08:33:45 condorella-rx480 756839 40000/757200 [ 5.28%], 0.52 ms/it; 598 muls, 0.70 ms/mul; ETA 0d 00:06; 4027f55e5abb465c
2018-11-01 08:33:51 condorella-rx480 756839 50000/757200 [ 6.60%], 0.58 ms/it; 353 muls, -1.06 ms/mul; ETA 0d 00:07; fe729c99f45fc2de
2018-11-01 08:33:56 condorella-rx480 756839 60000/757200 [ 7.92%], 0.53 ms/it; 340 muls, 0.21 ms/mul; ETA 0d 00:06; e70f4a41ec67b2a6
2018-11-01 08:34:01 condorella-rx480 756839 70000/757200 [ 9.24%], 0.49 ms/it; 326 muls, 1.29 ms/mul; ETA 0d 00:06; 4c814bfdf2dec664
2018-11-01 08:34:07 condorella-rx480 756839 80000/757200 [10.57%], 0.52 ms/it; ETA 0d 00:06; 96f668f515cdccc7
2018-11-01 08:34:12 condorella-rx480 756839 90000/757200 [11.89%], 0.52 ms/it; ETA 0d 00:06; 8900ac33e8fba278
2018-11-01 08:34:17 condorella-rx480 756839 100000/757200 [13.21%], 0.52 ms/it; ETA 0d 00:06; 32b7785790aed404
2018-11-01 08:34:22 condorella-rx480 756839 110000/757200 [14.53%], 0.52 ms/it; ETA 0d 00:06; 7b488dc9aca7799b
2018-11-01 08:34:28 condorella-rx480 756839 120000/757200 [15.85%], 0.52 ms/it; ETA 0d 00:06; f854ef85204f30d0
2018-11-01 08:34:33 condorella-rx480 756839 130000/757200 [17.17%], 0.53 ms/it; ETA 0d 00:05; fe46348342c6122e
2018-11-01 08:34:38 condorella-rx480 756839 140000/757200 [18.49%], 0.52 ms/it; ETA 0d 00:05; 3b70d7db6ab3bec3
2018-11-01 08:34:43 condorella-rx480 756839 150000/757200 [19.81%], 0.52 ms/it; ETA 0d 00:05; 486178b2041b6c08
2018-11-01 08:34:49 condorella-rx480 756839 EE 160000/757200 [21.13%], 0.52 ms/it; ETA 0d 00:05; 0b79956769558c34 (check 0.25s)
2018-11-01 08:34:49 condorella-rx480 756839.owl loaded: k 800, B1 3000, block 400, res64 661c814a3224d4dc, stage 1, baseBits 0
2018-11-01 08:34:54 condorella-rx480 756839 10000/757200 [ 1.32%], 0.63 ms/it; 611 muls, -0.60 ms/mul; ETA 0d 00:08; b14737e6b0df37e1
2018-11-01 08:35:00 condorella-rx480 756839 20000/757200 [ 2.64%], 0.54 ms/it; 1084 muls, 0.38 ms/mul; ETA 0d 00:07; d4f722ecbcaf72c7
2018-11-01 08:35:06 condorella-rx480 756839 30000/757200 [ 3.96%], 0.55 ms/it; 917 muls, 0.27 ms/mul; ETA 0d 00:07; becefb17ea7bc907
2018-11-01 08:35:11 condorella-rx480 756839 40000/757200 [ 5.28%], 0.53 ms/it; 598 muls, 0.53 ms/mul; ETA 0d 00:06; 4027f55e5abb465c
2018-11-01 08:35:17 condorella-rx480 756839 50000/757200 [ 6.60%], 0.55 ms/it; 353 muls, -0.12 ms/mul; ETA 0d 00:06; fe729c99f45fc2de
2018-11-01 08:35:22 condorella-rx480 756839 60000/757200 [ 7.92%], 0.48 ms/it; 340 muls, 1.72 ms/mul; ETA 0d 00:06; e70f4a41ec67b2a6
2018-11-01 08:35:28 condorella-rx480 756839 70000/757200 [ 9.24%], 0.52 ms/it; 326 muls, 0.55 ms/mul; ETA 0d 00:06; 4c814bfdf2dec664
2018-11-01 08:35:33 condorella-rx480 756839 80000/757200 [10.57%], 0.52 ms/it; ETA 0d 00:06; 96f668f515cdccc7
2018-11-01 08:35:38 condorella-rx480 756839 90000/757200 [11.89%], 0.52 ms/it; ETA 0d 00:06; 8900ac33e8fba278
2018-11-01 08:35:44 condorella-rx480 756839 100000/757200 [13.21%], 0.52 ms/it; ETA 0d 00:06; 32b7785790aed404
2018-11-01 08:35:49 condorella-rx480 756839 110000/757200 [14.53%], 0.52 ms/it; ETA 0d 00:06; 7b488dc9aca7799b
2018-11-01 08:35:54 condorella-rx480 756839 120000/757200 [15.85%], 0.52 ms/it; ETA 0d 00:06; f854ef85204f30d0
2018-11-01 08:35:59 condorella-rx480 756839 130000/757200 [17.17%], 0.52 ms/it; ETA 0d 00:05; fe46348342c6122e
2018-11-01 08:36:05 condorella-rx480 756839 140000/757200 [18.49%], 0.52 ms/it; ETA 0d 00:05; 3b70d7db6ab3bec3
2018-11-01 08:36:10 condorella-rx480 756839 150000/757200 [19.81%], 0.52 ms/it; ETA 0d 00:05; 486178b2041b6c08
2018-11-01 08:36:15 condorella-rx480 756839 EE 160000/757200 [21.13%], 0.52 ms/it; ETA 0d 00:05; 0b79956769558c34 (check 0.23s)
2018-11-01 08:36:15 condorella-rx480 756839.owl loaded: k 800, B1 3000, block 400, res64 661c814a3224d4dc, stage 1, baseBits 0
2018-11-01 08:36:21 condorella-rx480 756839 10000/757200 [ 1.32%], 0.64 ms/it; 611 muls, -0.77 ms/mul; ETA 0d 00:08; b14737e6b0df37e1
2018-11-01 08:36:27 condorella-rx480 756839 20000/757200 [ 2.64%], 0.55 ms/it; 1084 muls, 0.35 ms/mul; ETA 0d 00:07; d4f722ecbcaf72c7
2018-11-01 08:36:32 condorella-rx480 756839 30000/757200 [ 3.96%], 0.51 ms/it; 917 muls, 0.63 ms/mul; ETA 0d 00:06; becefb17ea7bc907
2018-11-01 08:36:38 condorella-rx480 756839 40000/757200 [ 5.28%], 0.52 ms/it; 598 muls, 0.64 ms/mul; ETA 0d 00:06; 4027f55e5abb465c
2018-11-01 08:36:43 condorella-rx480 756839 50000/757200 [ 6.60%], 0.57 ms/it; 353 muls, -0.64 ms/mul; ETA 0d 00:07; fe729c99f45fc2de
2018-11-01 08:36:49 condorella-rx480 756839 60000/757200 [ 7.92%], 0.55 ms/it; 340 muls, -0.21 ms/mul; ETA 0d 00:06; e70f4a41ec67b2a6
2018-11-01 08:36:54 condorella-rx480 756839 70000/757200 [ 9.24%], 0.54 ms/it; 326 muls, 0.01 ms/mul; ETA 0d 00:06; 4c814bfdf2dec664
2018-11-01 08:36:59 condorella-rx480 756839 80000/757200 [10.57%], 0.52 ms/it; ETA 0d 00:06; 96f668f515cdccc7
2018-11-01 08:37:05 condorella-rx480 756839 90000/757200 [11.89%], 0.52 ms/it; ETA 0d 00:06; 8900ac33e8fba278
2018-11-01 08:37:10 condorella-rx480 756839 100000/757200 [13.21%], 0.53 ms/it; ETA 0d 00:06; 32b7785790aed404
2018-11-01 08:37:15 condorella-rx480 756839 110000/757200 [14.53%], 0.53 ms/it; ETA 0d 00:06; 7b488dc9aca7799b
2018-11-01 08:37:21 condorella-rx480 756839 120000/757200 [15.85%], 0.52 ms/it; ETA 0d 00:06; f854ef85204f30d0
2018-11-01 08:37:26 condorella-rx480 756839 130000/757200 [17.17%], 0.53 ms/it; ETA 0d 00:05; fe46348342c6122e
2018-11-01 08:37:31 condorella-rx480 756839 140000/757200 [18.49%], 0.53 ms/it; ETA 0d 00:05; 3b70d7db6ab3bec3
2018-11-01 08:37:36 condorella-rx480 756839 150000/757200 [19.81%], 0.53 ms/it; ETA 0d 00:05; 486178b2041b6c08
2018-11-01 08:37:42 condorella-rx480 756839 EE 160000/757200 [21.13%], 0.52 ms/it; ETA 0d 00:05; 0b79956769558c34 (check 0.25s)
2018-11-01 08:37:42 condorella-rx480 3 sequential errors, will stop.
2018-11-01 08:37:42 condorella-rx480 Exiting because "too many errors"
2018-11-01 08:37:42 condorella-rx480 Bye[/CODE][/QUOTE]

Same issue:
[CODE]2018-11-01 09:59:19 gpuowl 5.0--mod
2018-11-01 09:59:19 0 -user selroc -cpu 0 -device 0
2018-11-01 09:59:19 0 756839 FFT 512K: Width 64x8, Height 64x8; 1.44 bits/word
2018-11-01 09:59:19 0 using long carry kernels
2018-11-01 09:59:19 0 gfx803-36x1360-@4a:0.0 Ellesmere [Radeon RX 470/480]
2018-11-01 09:59:20 0 OpenCL compilation in 1083 ms, with "-DEXP=756839u -DWIDTH=512u -DSMALL_HEIGHT=512u -DMIDDLE=1u -I. -cl-fast-relaxed-math -cl-std=CL2.0 "
2018-11-01 09:59:20 0 756839.owl not found, starting from the beginnig.
2018-11-01 09:59:21 0 756839 OK 800/757200 [ 0.11%], 0.53 ms/it; ETA 0d 00:07; 24ac239d8eb8ffa2 (check 0.26s)
2018-11-01 09:59:26 0 756839 10000/757200 [ 1.32%], 0.53 ms/it; ETA 0d 00:07; e0f756a0e6b027cf
2018-11-01 09:59:31 0 756839 20000/757200 [ 2.64%], 0.53 ms/it; ETA 0d 00:07; c24d9712d700c29e
2018-11-01 09:59:37 0 756839 30000/757200 [ 3.96%], 0.53 ms/it; ETA 0d 00:06; ef92f116fa7b7853
2018-11-01 09:59:42 0 756839 40000/757200 [ 5.28%], 0.53 ms/it; ETA 0d 00:06; 56bee347346be732
2018-11-01 09:59:47 0 756839 50000/757200 [ 6.60%], 0.53 ms/it; ETA 0d 00:06; 88a1922073d97c57
2018-11-01 09:59:53 0 756839 60000/757200 [ 7.92%], 0.53 ms/it; ETA 0d 00:06; 2dd5ee5cdfe0c62a
2018-11-01 09:59:58 0 756839 70000/757200 [ 9.24%], 0.53 ms/it; ETA 0d 00:06; 70439075d84ca857
2018-11-01 10:00:03 0 756839 80000/757200 [10.57%], 0.53 ms/it; ETA 0d 00:06; 953b2f1c170a9def
2018-11-01 10:00:08 0 756839 90000/757200 [11.89%], 0.53 ms/it; ETA 0d 00:06; 0e275a89b9c39b27
2018-11-01 10:00:14 0 756839 100000/757200 [13.21%], 0.53 ms/it; ETA 0d 00:06; 046a3e1ad36681e9
2018-11-01 10:00:19 0 756839 110000/757200 [14.53%], 0.53 ms/it; ETA 0d 00:06; fad6fff7757f9a66
2018-11-01 10:00:24 0 756839 120000/757200 [15.85%], 0.53 ms/it; ETA 0d 00:06; b9e7f5cc6fc13dc0
2018-11-01 10:00:30 0 756839 130000/757200 [17.17%], 0.53 ms/it; ETA 0d 00:06; 79b53436131c503b
2018-11-01 10:00:35 0 756839 140000/757200 [18.49%], 0.53 ms/it; ETA 0d 00:05; 04d1642ce8add525
2018-11-01 10:00:40 0 756839 150000/757200 [19.81%], 0.53 ms/it; ETA 0d 00:05; dde1480d8d123ee9
2018-11-01 10:00:46 0 756839 EE 160000/757200 [21.13%], 0.53 ms/it; ETA 0d 00:05; 76f44754c8e05f8c (check 0.26s)
2018-11-01 10:00:46 0 756839.owl loaded: k 800, B1 0, block 400, res64 24ac239d8eb8ffa2, stage 1, baseBits 0
2018-11-01 10:00:50 0 Stopping, please wait..
2018-11-01 10:00:50 0 756839 OK 7600/757200 [ 1.00%], 0.57 ms/it; ETA 0d 00:07; a8f484a5db49912f (check 0.23s)
2018-11-01 10:00:50 0 Exiting because "stop requested"
2018-11-01 10:00:50 0 Bye[/CODE]

I think it is again an FFT selection problem.

[QUOTE=SELROC;499235]I just did a couple of quick tests on v5.0, I am a bit confused.

859433 is a prime.
[URL]https://www.mersenne.org/report_exponent/?exp_lo=859433[/URL]

*Test1 on amdgpu-pro:
[URL]https://www.mersenneforum.org/attachment.php?attachmentid=19201&stc=1&d=1541063294[/URL]

*Test2 on rocm:
[URL]https://www.mersenneforum.org/attachment.php?attachmentid=19202&stc=1&d=1541063326[/URL]

the result is "C" in both cases.

BTW, in this case rocm is slower than amdgpu-pro. Rocm is faster on large exponents.[/QUOTE]
I'm a little surprised that worked as well as it did, since it's under 2 bits/word at 1.64.

V5.0 above 2 bits/word PRP-1

Nonzero (primenet goal) B2 values are being reported as zero.
Final residues for primes look odd to me, not a PRP3 of a prime. Is this a side effect of the simultaneous P-1 or what?[CODE]{"exponent":"1257787", "worktype":"PRP,P-1", "status":"P", "program":{"name":"gpuowl", "version":"5.0-f604bb1"}, "timestamp":"2018-11-01 14:08:03 UTC", "user":"kriesel", "computer":"condorella-rx480", "aid":"0", "fft-length":524288, "res64":"a65a1c89ee43505d", "b2":"0", "base":{"b1":"20000", "bias":{"2":19}, "res64":"a65a1c89ee43505d"}}
{"exponent":"1398269", "worktype":"PRP,P-1", "status":"P", "program":{"name":"gpuowl", "version":"5.0-f604bb1"}, "timestamp":"2018-11-01 14:20:36 UTC", "user":"kriesel", "computer":"condorella-rx480", "aid":"0", "fft-length":524288, "res64":"f47e9bd4350b65c2", "b2":"0", "base":{"b1":"20000", "bias":{"2":19}, "res64":"f47e9bd4350b65c2"}}
{"exponent":"2976221", "worktype":"PRP,P-1", "status":"P", "program":{"name":"gpuowl", "version":"5.0-f604bb1"}, "timestamp":"2018-11-01 14:49:56 UTC", "user":"kriesel", "computer":"condorella-rx480", "aid":"0", "fft-length":524288, "res64":"9154d3e4ec8180c1", "b2":"0", "base":{"b1":"40000", "bias":{"2":19}, "res64":"9154d3e4ec8180c1"}}
{"exponent":"3021377", "worktype":"PRP,P-1", "status":"P", "program":{"name":"gpuowl", "version":"5.0-f604bb1"}, "timestamp":"2018-11-01 15:17:54 UTC", "user":"kriesel", "computer":"condorella-rx480", "aid":"0", "fft-length":524288, "res64":"508313999383015c", "b2":"0", "base":{"b1":"40000", "bias":{"2":19}, "res64":"508313999383015c"}}[/CODE]0.51 ms/it is good.
Both B1 and B2 were specified in worktodo entries.

[QUOTE=SELROC;499253]
I think it is again an FFT selection problem.[/QUOTE]
Yes, in the sense there's no implemented fft length small enough in gpuowl for such small exponents. Even long carry has its limits, and I recall Preda stating some time ago its limit was around 2 bits/word. Maybe it's ~1.5-1.6 bits. p=756839 is 1.44bits/word, too low apparently; 859433 seemed to work at 1.64 bits/word. (At some point, doesn't the signal get lost in the noise?)

[CODE]#draft qa suite for checking false negatives in PRP-1
#bounds listed are greater of primenet or gpu72 indicated at mersenne.ca
#tf are gpu72 target indicated at mersenne.ca

#following block that are commented out are less than 2 bits/word at gpuowl 512k fft length minimum
#B1=1,B2=25;PRP=0,1,2,107,-1,44,2
#B1=1,B2=25;PRP=0,1,2,127,-1,44,2
#B1=6,B2=150;PRP=0,1,2,521,-1,44,2
#B1=7,B2=175;PRP=0,1,2,607,-1,44,2
#B1=14,B2=350;PRP=0,1,2,1279,-1,44,2
#B1=25,B2=625;PRP=0,1,2,2203,-1,44,2
#B1=26,B2=650;PRP=0,1,2,2281,-1,44,2
#B1=26,B2=650;PRP=0,1,2,3217,-1,44,2
#B1=48,B2=1200;PRP=0,1,2,4253,-1,44,2
#B1=49,B2=1225;PRP=0,1,2,4423,-1,44,2
#B1=108,B2=2700;PRP=0,1,2,9689,-1,44,2
#B1=111,B2=2775;PRP=0,1,2,9941,-1,44,2
#B1=125,B2=3125;PRP=0,1,2,11213,-1,44,2
#B1=223,B2=5575;PRP=0,1,2,19937,-1,44,2
#B1=243,B2=6075;PRP=0,1,2,21701,-1,44,2
#B1=260,B2=6500;PRP=0,1,2,23209,-1,44,2
#B1=498,B2=12450;PRP=0,1,2,44497,-1,44,2
#B1=964,B2=24100;PRP=0,1,2,86243,-1,44,2
#B1=10000,B2=60000;PRP=0,1,2,110503,-1,44,2
#B1=10000,B2=70000;PRP=0,1,2,132049,-1,44,2
#B1=10000,B2=130000;PRP=0,1,2,216091,-1,44,2
#B1=20000,B2=500000;PRP=0,1,2,756839,-1,44,2
#B1=20000,B2=580000;PRP=0,1,2,859433,-1,44,2

#following seem suitable for gpuowl v3.x-5.0 supported fft lengths
B1=20000,B2=240000;PRP=0,1,2,1257787,-1,60,2
B1=20000,B2=300000;PRP=0,1,2,1398269,-1,60,2
B1=40000,B2=560000;PRP=0,1,2,2976221,-1,64,2
B1=40000,B2=560000;PRP=0,1,2,3021377,-1,64,2
B1=80000,B2=1440000;PRP=0,1,2,6972593,-1,67,2
B1=150000,B2=3150000;PRP=0,1,2,13466917,-1,69,2
B1=260000,B2=6500000;PRP=0,1,2,20996011,-1,69,2
B1=280000,B2=7000000;PRP=0,1,2,24036583,-1,70,2
B1=310000,B2=7750000;PRP=0,1,2,25964951,-1,70,2
B1=350000,B2=8750000;PRP=0,1,2,30402457,-1,71,2
B1=380000,B2=9500000;PRP=0,1,2,32582657,-1,71,2
B1=430000,B2=11610000;PRP=0,1,2,37156667,-1,71,2
B1=480000,B2=12960000;PRP=0,1,2,42643801,-1,72,2
B1=490000,B2=13230000;PRP=0,1,2,43112609,-1,72,2
B1=650000,B2=17550000;PRP=0,1,2,57885161,-1,73,2
B1=810000,B2=13860000;PRP=0,1,2,74207281,-1,74,2
B1=840000,B2=23520000;PRP=0,1,2,77232917,-1,74,2

#B1=,B2=;PRP=0,1,2,,-1,,2
[/CODE]Adding a minimum exponent column would be good here, or just a line about the minimum supported exponent:[CODE]2018-11-01 10:31:24 FFT maxExp W H M
2018-11-01 10:31:24 0.5M 10.2M 512 512 1
2018-11-01 10:31:24 1.0M 20.0M 1024 512 1
2018-11-01 10:31:24 1.0M 20.0M 512 1024 1
2018-11-01 10:31:24 2.0M 39.3M 1024 1024 1
2018-11-01 10:31:24 2.0M 39.3M 512 2048 1
2018-11-01 10:31:24 2.0M 39.3M 2048 512 1
2018-11-01 10:31:24 2.5M 48.9M 512 512 5
2018-11-01 10:31:24 4.0M 77.3M 1024 2048 1
2018-11-01 10:31:24 4.0M 77.3M 2048 1024 1
2018-11-01 10:31:24 4.0M 77.3M 4096 512 1
2018-11-01 10:31:24 4.5M 86.7M 512 512 9
2018-11-01 10:31:24 5.0M 96.1M 1024 512 5
2018-11-01 10:31:24 5.0M 96.1M 512 1024 5
2018-11-01 10:31:24 8.0M 151.8M 2048 2048 1
2018-11-01 10:31:24 8.0M 151.8M 4096 1024 1
2018-11-01 10:31:24 9.0M 170.3M 1024 512 9
2018-11-01 10:31:24 9.0M 170.3M 512 1024 9
2018-11-01 10:31:24 10.0M 188.7M 1024 1024 5
2018-11-01 10:31:24 10.0M 188.7M 512 2048 5
2018-11-01 10:31:24 10.0M 188.7M 2048 512 5
2018-11-01 10:31:24 16.0M 298.1M 4096 2048 1
2018-11-01 10:31:24 18.0M 334.3M 1024 1024 9
2018-11-01 10:31:24 18.0M 334.3M 512 2048 9
2018-11-01 10:31:24 18.0M 334.3M 2048 512 9
2018-11-01 10:31:24 20.0M 370.4M 1024 2048 5
2018-11-01 10:31:24 20.0M 370.4M 2048 1024 5
2018-11-01 10:31:24 20.0M 370.4M 4096 512 5
2018-11-01 10:31:24 36.0M 656.2M 1024 2048 9
2018-11-01 10:31:24 36.0M 656.2M 2048 1024 9
2018-11-01 10:31:24 36.0M 656.2M 4096 512 9
2018-11-01 10:31:24 40.0M 727.0M 2048 2048 5
2018-11-01 10:31:24 40.0M 727.0M 4096 1024 5
2018-11-01 10:31:24 72.0M 1287.5M 2048 2048 9
2018-11-01 10:31:24 72.0M 1287.5M 4096 1024 9
2018-11-01 10:31:24 80.0M 1426.4M 4096 2048 5
2018-11-01 10:31:24 144.0M 2525.2M 4096 2048 9[/CODE]

The upper limit of exponent is checked for. I'd prefer it continued on with the next worktodo entry though.
[CODE]C:\msys64\home\ken\gpuowl-compile\v5.0>openowl.exe -user kriesel -cpu condorella-rx480 -device 0
2018-11-01 11:26:09 gpuowl 5.0-f604bb1
2018-11-01 11:26:09 condorella-rx480 -user kriesel -cpu condorella-rx480 -device 0
2018-11-01 11:26:09 condorella-rx480 3321928171 FFT 147456K: Width 512x8, Height 256x8, Middle 9; 22.00 bits/word
2018-11-01 11:26:09 condorella-rx480 FFT size too small for exponent (22.00 bits/word).
2018-11-01 11:26:09 condorella-rx480 Exiting because "FFT size too small"
2018-11-01 11:26:09 condorella-rx480 Bye[/CODE]

V4.7 appears to be correctly skipping P-1, crashes anyway

Worktodo line is

PRP=0,1,2,48500017,-1,73,2

[CODE]C:\msys64\home\ken\gpuowl-compile\v4.7>openowl-v4.7-5b01b65-w64
2018-11-01 11:39:25 gpuowl 4.7-5b01b65
2018-11-01 11:39:25 FFT 2560K: Width 512 (64x8), Height 512 (64x8), Middle 5; 18.50 bits/word
2018-11-01 11:39:25 Note: using short carry kernels
2018-11-01 11:39:26 Ellesmere-36x1266-@28:0.0 Radeon (TM) RX 480 Graphics
2018-11-01 11:39:29 OpenCL compilation in 3343 ms, with "-DEXP=48500017u -DWIDTH=512u -DSMALL_HEIGHT=512u -DMIDDLE=5u -I. -cl-fast-relaxed-math -cl-std=CL2.0 "

2018-11-01 11:39:30 PRP M(48500017), FFT 2560K, 18.50 bits/word, B1 0, B2 0
2018-11-01 11:39:31 OK loaded: 0/48500017, B1 0, blockSize 400, 0000000000000003 (expected 0000000000000003)
2018-11-01 11:39:31 Selected 0 P-1 trial points
Assertion failed!

Program: C:\msys64\home\ken\gpuowl-compile\v4.7\openowl-v4.7-5b01b65-W64.exe
File: Signal.cpp, Line 14

Expression: oldHandler

This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.[/CODE]

A big "Thank You" to Preda for taking out some of those asserts as that looks like it was
the final fix that made GpuOwl ( OpenOwl ? ) working on Windoze again.

And also, another Thanks for writing this fabulous program in the first place !

A fix is in, please re-try.

A longer explanation of what happened:
1. due to PRP-1 not allowing "division-by-base" (where base != 3) anymore, the final iteration was changed to exponent - 1 (from 'exponent' as it was before), and the final condition for is-prime became:
residue == base (when base == 3^(even-power), i.e. B1 != 0), or
residue == -3 (when base == 3, i.e. B1 == 0).

2. this last test, comparing residue == -3, was botched.

3. I only tested with B1!=0 (the interesting new case that was risky). I disregarded testing prime with B1==0.

Alleviating the problem:
- the final res64 is correct. In the (very unlikely) case that somebody got a res64 0xfffffffffffffffc when testing with B1==0, he should re-do the test because most likely he found a prime.

[QUOTE=preda;499241]Thanks, that's a genuine error. I'll fix ASAP (24h).

The final residue shows the computation is fine, it's just the logic for deciding prime/not-prime at the end that's broken. Will fix.[/QUOTE]

[QUOTE=tServo;499282]A big "Thank You" to Preda for taking out some of those asserts as that looks like it was
the final fix that made GpuOwl ( OpenOwl ? ) working on Windoze again.

And also, another Thanks for writing this fabulous program in the first place ![/QUOTE]

Thanks for the nice words, and for your help with figuring out the Windows build!

[QUOTE=kriesel;499249]I think so. As I understand it, the computation needs to be carried past iteration p, to the next multiple of the block size, so the final error check can be done against it. Yes it looks odd because in earlier versions it was displayed differently.[/QUOTE]
Yes, the PRP is continued past the end (Exponent - 1) to reach a multiple of block-size where an error-check can be done. (the default block-size is 400).

[QUOTE]
Good catch on the C on a prime. We should check a few known primes, probably at every major release if not minor release. It's also an argument for double checking with different software or at least different offset, every exponent.

A list of tests to be done on every major release (which would necessarily get updated as program features change) would be a good thing. I wonder what Preda's testing consists of.[/QUOTE]
I appreciate the help I get with testing! I don't always have the time to do the right thing myself.

The testing I usually do is manual, spot testing a few situations. Recently I was very concerned with the new PRP-1 and its ability to correctly detect factors; I disregarded the old B1==0.

Not to forget: the Gerbicz Error Check (GEC) is an important ingredient in my overall evaluation of correctness; thus I'm particularly uncomfortable about the elements not covered by GEC: P-1 first-stage, and the GCD accumulation.