Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

OpenBLAS test suite failures on ARM neoverse_v1 #314

Open
casparvl opened this issue Aug 8, 2023 · 10 comments
Open

OpenBLAS test suite failures on ARM neoverse_v1 #314

casparvl opened this issue Aug 8, 2023 · 10 comments
Labels
2023.06-software.eessi.io 2023.06 version of software.eessi.io aarch64 related to Arm 64-bit targets (aarch64) pilot-2023.06 tests Related to software testing

Comments

@casparvl
Copy link
Collaborator

casparvl commented Aug 8, 2023 •
edited by boegel
Loading

We're seeing somewhat higher number of OpenBLAS test suite failures on ARM neoverse_v1 in:

For now, we've increased the number of accepted failures, but we'll need to investigate this issue in more detail. Are these test just out of numerical bounds? Or way off?

See also:

It did not seem to be a problem however in:

which build OpenBLAS 0.3.15 on ARM neoverse_v1 (see here)
@casparvl casparvl mentioned this issue Aug 8, 2023
Merged
@boegel
Copy link
Contributor

boegel commented Aug 8, 2023

@casparvl Can you include the high-level summary that shows how many tests fail, and maybe also dig up more details on the failing tests (should probably go in a gist), which we can use when reporting this upstream (or getting some input on this from @bartoldeman)?

@boegel boegel added pilot-2023.06 aarch64 related to Arm 64-bit targets (aarch64) tests Related to software testing labels Aug 8, 2023
@casparvl
Copy link
Collaborator Author

casparvl commented Aug 8, 2023 

SUMMARY                 nb test run     numerical error         other error
================        ===========     =================       ================
REAL                    1315683         107     (0.008%)        0       (0.000%)
DOUBLE PRECISION        1314777         66      (0.005%)        0       (0.000%)
COMPLEX                 773609          97      (0.013%)        0       (0.000%)
COMPLEX16               776246          74      (0.010%)        0       (0.000%)

--> ALL PRECISIONS      4180315         344     (0.008%)        0       (0.000%)


== 2023-07-19 12:16:15,506 openblas.py:113 INFO 4180315 LAPACK tests run - 344 failed due to numerical errors - 0 failed due to other errors

More detailed (though not complete) test log from bot/easybuild/build/OpenBLAS/0.3.21/GCC-12.2.0/OpenBLAS-0.3.21/lapack-netlib/TESTING/testing_results.txt:

...
Matrix order=   15, type=10, seed= 798,1691,2423, 745, result  5 is 8.389E+06
 Matrix order=   15, type=11, seed= 931,1787, 557,2429, result  5 is 8.389E+06
 Matrix order=   15, type=17, seed= 529,3615,1764,1221, result  5 is 8.389E+06
 Matrix order=   15, type=18, seed=3991, 625,3539,2581, result  5 is 8.389E+06
 Matrix order=   15, type=19, seed=2400,1821, 218,1365, result  5 is 8.389E+06
 Matrix order=   15, type=20, seed=2155,2073,3686, 149, result  5 is 8.389E+06
 Matrix order=   15, type=21, seed=3823,2194,3510,3029, result  5 is 8.389E+06
 Matrix order=   15, type=22, seed=2951, 608,1256,3481, result  5 is 8.389E+06
 Matrix order=   15, type=23, seed=1243,2927, 263, 357, result  5 is 8.389E+06
 Matrix order=   15, type=24, seed=1487,3956,2976,3649, result  5 is 8.389E+06
 Matrix order=   15, type=25, seed=2605,2211,3982,2721, result  5 is 8.389E+06
 Matrix order=   15, type=26, seed=   8,1600,1726,2817, result  5 is 8.389E+06
 Matrix order=   20, type= 8, seed= 624, 693,1681,   9, result  5 is 8.389E+06
 Matrix order=   20, type=10, seed= 563,2705, 428, 633, result  5 is 8.389E+06
 Matrix order=   20, type=12, seed= 112,1059,2403,2121, result  5 is 8.389E+06
 Matrix order=   20, type=13, seed=3448,3524, 410,2697, result  5 is 8.389E+06
 Matrix order=   20, type=14, seed= 248,   6,3354,3273, result  5 is 8.389E+06
 Matrix order=   20, type=17, seed= 723,4024,2193,2265, result  5 is 8.389E+06
 Matrix order=   20, type=18, seed= 828, 842,1787,3733, result  5 is 8.389E+06
 Matrix order=   20, type=19, seed= 252,3327,3300,1389, result  5 is 8.389E+06
 Matrix order=   20, type=20, seed=1904,2346,3199,2437, result  5 is 8.389E+06
 Matrix order=   20, type=21, seed=4066,  26,1253, 221, result  5 is 8.389E+06
 Matrix order=   20, type=22, seed=3234,3664,3237,3473, result  5 is 8.389E+06
 Matrix order=   20, type=23, seed= 481, 957,3230,3689, result  5 is 8.389E+06
 Matrix order=   20, type=24, seed=2871,2618,1987,2949, result  5 is 8.389E+06
 Matrix order=   20, type=25, seed=2604,3233, 644,3529, result  5 is 8.389E+06
 Matrix order=   20, type=26, seed=1465,2266,3976,3833, result  5 is 8.389E+06
 SGV drivers:     64 out of   1092 tests failed to pass the threshold
...
 Matrix order=   20, type=10, seed= 563,2705, 428, 633, result  5 is 4.504D+15
 Matrix order=   20, type=12, seed= 112,1059,2403,2121, result  5 is 4.504D+15
 Matrix order=   20, type=13, seed=3448,3524, 410,2697, result  5 is 4.504D+15
 Matrix order=   20, type=17, seed= 723,4024,2193,2265, result  5 is 4.504D+15
 Matrix order=   20, type=18, seed= 828, 842,1787,3733, result  5 is 4.504D+15
 Matrix order=   20, type=19, seed= 252,3327,3300,1389, result  5 is 4.504D+15
 Matrix order=   20, type=20, seed=1904,2346,3199,2437, result  5 is 4.504D+15
 Matrix order=   20, type=21, seed=4066,  26,1253, 221, result  5 is 4.504D+15
 Matrix order=   20, type=22, seed=3234,3664,3237,3473, result  5 is 4.504D+15
 Matrix order=   20, type=23, seed= 481, 957,3230,3689, result  5 is 4.504D+15
 Matrix order=   20, type=24, seed=2871,2618,1987,2949, result  5 is 4.504D+15
 Matrix order=   20, type=25, seed=2604,3233, 644,3529, result  5 is 4.504D+15
 Matrix order=   20, type=26, seed=1465,2266,3976,3833, result  5 is 4.504D+15
 DGV drivers:     34 out of   1092 tests failed to pass the threshold
...
 Matrix order=    6, type=16, seed= 852,1170,4018,2529, result  5 is 4.504D+15
 Matrix order=    8, type=20, seed= 202,3818,2830,2373, result  5 is 4.504D+15
 Matrix order=    8, type=25, seed= 582,1680, 289,1745, result  5 is 4.504D+15
 Matrix order=   10, type= 8, seed= 573,2931,1057, 529, result  5 is 4.504D+15
 Matrix order=   10, type=10, seed= 308, 212,3333,  73, result  5 is 4.504D+15
 Matrix order=   10, type=13, seed=3036,2372,2403,2421, result  5 is 4.504D+15
 Matrix order=   15, type=12, seed=2288,1863,3505, 805, result  5 is 4.504D+15
 Matrix order=   15, type=13, seed=3415,2880,2965,2889, result  5 is 4.504D+15
 Matrix order=   15, type=16, seed= 681,2778,2914,2225, result  5 is 4.504D+15
 Matrix order=   15, type=17, seed= 304,3863, 350,1657, result  5 is 4.504D+15
 Matrix order=   15, type=21, seed=2362, 980, 534,3365, result  5 is 4.504D+15
 Matrix order=   15, type=25, seed=2888,1855, 900,2225, result  5 is 4.504D+15
 Matrix order=   20, type= 8, seed= 636, 119,  47,2269, result  5 is 4.504D+15
 Matrix order=   20, type= 9, seed=1007,2268,2913,  29, result  5 is 4.504D+15
 Matrix order=   20, type=10, seed= 104,1378,2159,1485, result  5 is 4.504D+15
 Matrix order=   20, type=11, seed= 927, 874,3720,2349, result  5 is 4.504D+15
 Matrix order=   20, type=14, seed=2659,2147, 552,1661, result  5 is 4.504D+15
 Matrix order=   20, type=21, seed=2012,3064,3728,2789, result  5 is 4.504D+15
 DGV drivers:     18 out of   1092 tests failed to pass the threshold
...
 N=  30 M=   0, P=   5, type  8, test  2, ratio=  555331.
 N=   3 M=   0, P=  20, type  8, test  2, ratio= 0.889025E+07
 N=  30 M=   0, P=  20, type  8, test  2, ratio= 0.219820E+07
 M=   3 P=   0, N=  30, type  7, test  1, ratio= 0.800042E+07
 N=  30 M=   3, P=   0, type  7, test  1, ratio= 0.772279E+07
 M=   3 P=   5, N=  30, type  7, test  1, ratio= 0.852669E+07
 N=  30 M=   3, P=   5, type  8, test  2, ratio= 0.119398E+07
 N=   3 M=   3, P=  20, type  8, test  2, ratio= 0.998240E+07
 N=  30 M=   3, P=  20, type  8, test  2, ratio=  976876.
 N=   3 M=  10, P=   5, type  8, test  2, ratio= 0.121475E+08
 N=  30 M=  10, P=   5, type  7, test  1, ratio= 0.127652E+08
 N=  30 M=  10, P=   5, type  8, test  2, ratio=  714258.
 M=  10 P=  20, N=  30, type  7, test  1, ratio= 0.510405E+07
 N=  30 M=  10, P=  20, type  8, test  2, ratio= 0.208493E+07
 GQR:     14 out of   1728 tests failed to pass the threshold
...
 Matrix order=   20, type=16, seed= 647,2328,1944, 557, result  5 is 8.389E+06
 CGV drivers:      1 out of   1092 tests failed to pass the threshold
...
 CHESV , UPLO='U', N =    3, type  2, test  1, ratio = 0.10874E+07
 CHESVX, FACT='N', UPLO='U', N =    3, type  2, test  1, ratio = 0.10874E+07
 CHESV , UPLO='U', N =    5, type  9, test  1, ratio = 0.32042E+07
 CHESVX, FACT='N', UPLO='U', N =    5, type  9, test  1, ratio = 0.32042E+07
 CHESV , UPLO='U', N =    5, type 10, test  1, ratio = 0.35394E+07
 CHESVX, FACT='N', UPLO='U', N =    5, type 10, test  1, ratio = 0.35394E+07
 CHESV , UPLO='U', N =   10, type  2, test  1, ratio = 0.16458E+07
 CHESVX, FACT='N', UPLO='U', N =   10, type  2, test  1, ratio = 0.16458E+07
 CHESV , UPLO='U', N =   10, type  7, test  1, ratio = 0.14266E+06
 CHESVX, FACT='N', UPLO='U', N =   10, type  7, test  1, ratio = 0.14266E+06
 CHESV , UPLO='U', N =   10, type  9, test  1, ratio = 0.31372E+07
 CHESVX, FACT='N', UPLO='U', N =   10, type  9, test  1, ratio = 0.31372E+07
 CHESV , UPLO='U', N =   10, type 10, test  1, ratio = 0.16814E+07
 CHESVX, FACT='N', UPLO='U', N =   10, type 10, test  1, ratio = 0.16814E+07
 CHESV , UPLO='U', N =   50, type  2, test  1, ratio = 0.86168E+07
 CHESVX, FACT='N', UPLO='U', N =   50, type  2, test  1, ratio = 0.86168E+07
 CHESV , UPLO='U', N =   50, type  7, test  1, ratio = 0.80869E+06
 CHESVX, FACT='N', UPLO='U', N =   50, type  7, test  1, ratio = 0.80869E+06
 CHESV , UPLO='U', N =   50, type  8, test  1, ratio = 0.52818E+06
 CHESVX, FACT='N', UPLO='U', N =   50, type  8, test  1, ratio = 0.52818E+06
 CHESV , UPLO='U', N =   50, type  9, test  1, ratio = 0.85655E+07
 CHESVX, FACT='N', UPLO='U', N =   50, type  9, test  1, ratio = 0.85655E+07
 CHESV , UPLO='U', N =   50, type 10, test  1, ratio = 0.32095E+07
 CHESVX, FACT='N', UPLO='U', N =   50, type 10, test  1, ratio = 0.32095E+07
 CHE drivers:     24 out of   1072 tests failed to pass the threshold
...
 Matrix order=   12, type=24, seed=2693,2404,3046,2957, result  5 is 4.504D+15
 Matrix order=   12, type=25, seed= 578, 723, 929,1637, result  5 is 4.504D+15
 Matrix order=   12, type=26, seed=2061,1512,1968, 125, result  5 is 4.504D+15
 Matrix order=   20, type=16, seed= 647,2328,1944, 557, result  5 is 4.504D+15
 Matrix order=   20, type=17, seed=1585,3902,3906,1293, result  5 is 4.504D+15
 Matrix order=   20, type=18, seed=1063,2113,3640,2685, result  5 is 4.504D+15
 Matrix order=   20, type=19, seed=2305,3879, 305, 381, result  5 is 4.504D+15
 Matrix order=   20, type=20, seed= 897,3616, 121,2173, result  5 is 4.504D+15
 Matrix order=   20, type=21, seed= 676,3851,3089,3965, result  5 is 4.504D+15
 Matrix order=   20, type=22, seed=1763,2855,1954,1469, result  5 is 4.504D+15
 Matrix order=   20, type=23, seed=1338,3437,3180,3285, result  5 is 4.504D+15
 Matrix order=   20, type=24, seed=3522,1685,3785, 813, result  5 is 4.504D+15
 Matrix order=   20, type=25, seed=2931,3978,3195,3781, result  5 is 4.504D+15
 Matrix order=   20, type=26, seed= 449,1038,1670,1437, result  5 is 4.504D+15
 ZGV drivers:     53 out of   1092 tests failed to pass the threshold

@casparvl
Copy link
Collaborator Author

casparvl commented Aug 8, 2023

Just for my future self: that file is also available from the host on

/mnt/shared/home/bot/eessi-bot-software-layer/jobs/2023.07/pr_309/event_8d3f0620-2624-11ee-870f-e48080cd3472/run_005/linux_aarch64_neoverse_v1/eessi-2023.06-software/previous_tmp/build_step/bot/easybuild/build/OpenBLAS/0.3.21/GCC-12.2.0/OpenBLAS-0.3.21/lapack-netlib/TESTING/testing_results.txt

No need to go into the container first...

@casparvl
Copy link
Collaborator Author

casparvl commented Aug 8, 2023

Gist with the full log is here

@casparvl
Copy link
Collaborator Author

casparvl commented Aug 8, 2023

Created upstream issue OpenMathLib/OpenBLAS#4187

@casparvl
Copy link
Collaborator Author

casparvl commented Aug 9, 2023

Docs about how to interpret LAPACK test results: https://netlib.org/lapack/faq.html#_how_do_i_interpret_lapack_testing_failures

@casparvl
Copy link
Collaborator Author

casparvl commented Aug 10, 2023 •
edited
Loading

For Reference, on Neoverse_n1 we get some failuers as well, but way less. A manual build with

source /cvmfs/pilot.eessi-hpc.org/versions/2023.06/init/bash
module load EasyBuild
eb OpenBLAS-0.3.21-GCC-12.2.0.eb --disable-cleanup-builddir --disable-cleanup-tmpdir -r --rebuild

showed:


                        -->   LAPACK TESTING SUMMARY  <--
SUMMARY                 nb test run     numerical error         other error
================        ===========     =================       ================
REAL                    1315683         21      (0.002%)        0       (0.000%)
DOUBLE PRECISION        1314777         30      (0.002%)        0       (0.000%)
COMPLEX                 776527          3       (0.000%)        0       (0.000%)
COMPLEX16               777338          1       (0.000%)        0       (0.000%)

--> ALL PRECISIONS      4184325         55      (0.001%)        0       (0.000%)

Detailed logs (lapack_tests.txt):

Matrix order=    8, type=25, seed= 889,2561,1319, 841, result  5 is 8.389E+06
 Matrix order=   10, type=17, seed=1420,1879,1628,  85, result  5 is 8.389E+06
 Matrix order=   10, type=24, seed=1210,3887,1474,2237, result  5 is 8.389E+06
 Matrix order=   10, type=26, seed=3592,3268,3086,1181, result  5 is 8.389E+06
 Matrix order=   15, type=18, seed=3991, 625,3539,2581, result  5 is 8.389E+06
 Matrix order=   15, type=20, seed=2155,2073,3686, 149, result  5 is 8.389E+06
 Matrix order=   15, type=21, seed=3823,2194,3510,3029, result  5 is 8.389E+06
 Matrix order=   15, type=24, seed=1487,3956,2976,3649, result  5 is 8.389E+06
 Matrix order=   15, type=25, seed=2605,2211,3982,2721, result  5 is 8.389E+06
 Matrix order=   15, type=26, seed=   8,1600,1726,2817, result  5 is 8.389E+06
 Matrix order=   20, type=16, seed= 904,3323, 291,1585, result  5 is 8.389E+06
 Matrix order=   20, type=17, seed= 723,4024,2193,2265, result  5 is 8.389E+06
 Matrix order=   20, type=18, seed= 828, 842,1787,3733, result  5 is 8.389E+06
 Matrix order=   20, type=20, seed=1904,2346,3199,2437, result  5 is 8.389E+06
 Matrix order=   20, type=21, seed=4066,  26,1253, 221, result  5 is 8.389E+06
 Matrix order=   20, type=22, seed=3234,3664,3237,3473, result  5 is 8.389E+06
 Matrix order=   20, type=23, seed= 481, 957,3230,3689, result  5 is 8.389E+06
 Matrix order=   20, type=24, seed=2871,2618,1987,2949, result  5 is 8.389E+06
 Matrix order=   20, type=25, seed=2604,3233, 644,3529, result  5 is 8.389E+06
 SGV drivers:     19 out of   1092 tests failed to pass the threshold
...
Matrix order=    8, type=25, seed= 582,1680, 289,1745, result  5 is 8.389E+06
 Matrix order=   15, type=17, seed= 304,3863, 350,1657, result  5 is 8.389E+06
 SGV drivers:      2 out of   1092 tests failed to pass the threshold
...
 Matrix order=   10, type=24, seed=1210,3887,1474,2237, result  5 is 4.504D+15
 Matrix order=   15, type=20, seed=2155,2073,3686, 149, result  5 is 4.504D+15
 Matrix order=   15, type=21, seed=3823,2194,3510,3029, result  5 is 4.504D+15
 Matrix order=   15, type=24, seed=1487,3956,2976,3649, result  5 is 4.504D+15
 Matrix order=   15, type=25, seed=2605,2211,3982,2721, result  5 is 4.504D+15
 Matrix order=   20, type=12, seed= 112,1059,2403,2121, result  5 is 4.504D+15
 Matrix order=   20, type=17, seed= 723,4024,2193,2265, result  5 is 4.504D+15
 Matrix order=   20, type=18, seed= 828, 842,1787,3733, result  5 is 4.504D+15
 Matrix order=   20, type=20, seed=1904,2346,3199,2437, result  5 is 4.504D+15
 Matrix order=   20, type=24, seed=2871,2618,1987,2949, result  5 is 4.504D+15
 DGV drivers:     10 out of   1092 tests failed to pass the threshold
...
 Matrix order=   15, type=17, seed= 304,3863, 350,1657, result  5 is 4.504D+15
 Matrix order=   15, type=21, seed=2362, 980, 534,3365, result  5 is 4.504D+15
 Matrix order=   15, type=25, seed=2888,1855, 900,2225, result  5 is 4.504D+15
 Matrix order=   20, type=16, seed=3366,1285,3695, 789, result  5 is 4.504D+15
 Matrix order=   20, type=19, seed=1811,3312,1916,3573, result  5 is 4.504D+15
 Matrix order=   20, type=21, seed=2012,3064,3728,2789, result  5 is 4.504D+15
 DGV drivers:      6 out of   1092 tests failed to pass the threshold
...
N=  30 M=   0, P=   5, type  8, test  2, ratio=  555331.
 N=   3 M=   0, P=  20, type  8, test  2, ratio= 0.889025E+07
 N=  30 M=   0, P=  20, type  8, test  2, ratio= 0.219820E+07
 M=   3 P=   0, N=  30, type  7, test  1, ratio= 0.800042E+07
 N=  30 M=   3, P=   0, type  7, test  1, ratio= 0.772279E+07
 M=   3 P=   5, N=  30, type  7, test  1, ratio= 0.852669E+07
 N=  30 M=   3, P=   5, type  8, test  2, ratio= 0.119398E+07
 N=   3 M=   3, P=  20, type  8, test  2, ratio= 0.998240E+07
 N=  30 M=   3, P=  20, type  8, test  2, ratio=  976876.
 N=   3 M=  10, P=   5, type  8, test  2, ratio= 0.121475E+08
 N=  30 M=  10, P=   5, type  7, test  1, ratio= 0.127652E+08
 N=  30 M=  10, P=   5, type  8, test  2, ratio=  714258.
 M=  10 P=  20, N=  30, type  7, test  1, ratio= 0.510405E+07
 N=  30 M=  10, P=  20, type  8, test  2, ratio= 0.208493E+07
 GQR:     14 out of   1728 tests failed to pass the threshold
...
Matrix order=   10, type=23, seed=2605, 629,2321,1285, result  5 is 8.389E+06
 Matrix order=   12, type=19, seed=1401, 178,4093, 253, result  5 is 8.389E+06
 Matrix order=   20, type=19, seed=2305,3879, 305, 381, result  5 is 8.389E+06
 CGV drivers:      3 out of   1092 tests failed to pass the threshold
...
 Matrix order=   20, type=16, seed= 647,2328,1944, 557, result  5 is 4.504D+15
 ZGV drivers:      1 out of   1092 tests failed to pass the threshold

 All tests for ZGV drivers  passed the threshold (   1092 tests run)
...

@laraPPr laraPPr mentioned this issue Aug 23, 2023
Open
@boegel boegel mentioned this issue Sep 26, 2023
Merged
@boegel boegel mentioned this issue Oct 13, 2023
Merged
@juntangc
Copy link

juntangc commented Nov 1, 2023

How much performance benefit does -ftree-vectorize offer? If you take it out for neoversev1, the number of errors will be 110 instead of 344.

			-->   LAPACK TESTING SUMMARY  <--
SUMMARY             	nb test run 	numerical error   	other error  
================   	===========	=================	================  
REAL             	1317867		0	(0.000%)	0	(0.000%)	
DOUBLE PRECISION	1316961		14	(0.001%)	0	(0.000%)	
COMPLEX          	774701		96	(0.012%)	0	(0.000%)	
COMPLEX16         	778430		0	(0.000%)	0	(0.000%)	

--> ALL PRECISIONS	4187959		110	(0.003%)	0	(0.000%)

@juntangc
Copy link

juntangc commented Nov 21, 2023 •
edited
Loading

@casparvl , I have created a PR (easybuilders/easybuild-easyconfigs#19280) to address most of the numerical errors by disabling tree-vectorize. This is also disabled in OpenBLAS main branch and OpenBLAS release 0.3.23. See
OpenMathLib/OpenBLAS@32566bf
OpenMathLib/OpenBLAS@5780952

For users who have a concern of potential performance degradation, I have validated that all these are small numerical differences. Details can be found at OpenMathLib/OpenBLAS#4335.

@boegel
Copy link
Contributor

boegel commented Jan 19, 2024

Indeed, for OpenBLAS 0.3.23 and newer we don't see an increased number of numerical test failures anymore on aarch64/neoverse_v1:

from /cvmfs/software.eessi.io/versions/2023.06/software/linux/aarch64/neoverse_v1/software/OpenBLAS/0.3.23-GCC-12.3.0//easybuild/easybuild-OpenBLAS-0.3.23-20231128.004334.log.bz2:

                        -->   LAPACK TESTING SUMMARY  <--
SUMMARY                 nb test run     numerical error         other error
================        ===========     =================       ================
REAL                    1328283         0       (0.000%)        0       (0.000%)
DOUBLE PRECISION        1327377         14      (0.001%)        0       (0.000%)
COMPLEX                 788035          0       (0.000%)        0       (0.000%)
COMPLEX16               789102          0       (0.000%)        0       (0.000%)

--> ALL PRECISIONS      4232797         14      (0.000%)        0       (0.000%)

from /cvmfs/software.eessi.io/versions/2023.06/software/linux/aarch64/neoverse_v1/software/OpenBLAS/0.3.24-GCC-13.2.0/easybuild/easybuild-OpenBLAS-0.3.24-20240102.205808.log.bz2:

                        -->   LAPACK TESTING SUMMARY  <--
SUMMARY                 nb test run     numerical error         other error
================        ===========     =================       ================
REAL                    1328283         0       (0.000%)        0       (0.000%)
DOUBLE PRECISION        1329105         0       (0.000%)        0       (0.000%)
COMPLEX                 772635          36      (0.005%)        0       (0.000%)
COMPLEX16               773702          36      (0.005%)        0       (0.000%)

--> ALL PRECISIONS      4203725         72      (0.002%)        0       (0.000%)

@boegel boegel added the 2023.06-software.eessi.io 2023.06 version of software.eessi.io label Jan 19, 2024
@boegel boegel mentioned this issue Jan 19, 2024
Merged
trz42 pushed a commit to trz42/software-layer that referenced this issue Apr 10, 2024
@poksumdo
Merge pull request EESSI#314 from trz42/nessi-2023.06-eb-4.9.1-system
cc6c6a6 
{2023.06}[system] EasyBuild v4.9.1
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
2023.06-software.eessi.io 2023.06 version of software.eessi.io aarch64 related to Arm 64-bit targets (aarch64) pilot-2023.06 tests Related to software testing
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@boegel @casparvl @juntangc