Summary: | dev-python/statsmodels-0.13.2: failed (test phase) (x86) | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Joonas Niilola <juippis> |
Component: | Current packages | Assignee: | Python Gentoo Team <python> |
Status: | CONFIRMED --- | ||
Severity: | normal | CC: | gentoo, matoro_gentoo, mgorny, proxy-maint, sci-mathematics |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | x86 | ||
OS: | Linux | ||
See Also: | https://github.com/statsmodels/statsmodels/issues/8250 | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Bug Depends on: | |||
Bug Blocks: | 908367 | ||
Attachments: | build.log |
Description
Joonas Niilola
![]() Please expand -march=native. # gcc -march=native -E -v - </dev/null 2>&1 | grep cc1 /usr/libexec/gcc/i686-pc-linux-gnu/11.2.1/cc1 -E -quiet -v - -march=znver2 -mmmx -mpopcnt -msse -msse2 -msse3 -mssse3 -msse4.1 -msse4.2 -mavx -mavx2 -msse4a -mno-fma4 -mno-xop -mfma -mno-avx512f -mbmi -mbmi2 -maes -mpclmul -mno-avx512vl -mno-avx512bw -mno-avx512dq -mno-avx512cd -mno-avx512er -mno-avx512pf -mno-avx512vbmi -mno-avx512ifma -mno-avx5124vnniw -mno-avx5124fmaps -mno-avx512vpopcntdq -mno-avx512vbmi2 -mno-gfni -mno-vpclmulqdq -mno-avx512vnni -mno-avx512bitalg -mno-avx512bf16 -mno-avx512vp2intersect -mno-3dnow -madx -mabm -mno-cldemote -mclflushopt -mclwb -mclzero -mcx16 -mno-enqcmd -mf16c -mfsgsbase -mfxsr -mno-hle -msahf -mno-lwp -mlzcnt -mmovbe -mno-movdir64b -mno-movdiri -mmwaitx -mno-pconfig -mno-pku -mno-prefetchwt1 -mprfchw -mno-ptwrite -mrdpid -mrdrnd -mrdseed -mno-rtm -mno-serialize -mno-sgx -msha -mno-shstk -mno-tbm -mno-tsxldtrk -mno-vaes -mno-waitpkg -mwbnoinvd -mxsave -mxsavec -mxsaveopt -mxsaves -mno-amx-tile -mno-amx-int8 -mno-amx-bf16 -mno-uintr -mno-hreset -mno-kl -mno-widekl -mno-avxvnni --param l1-cache-size=32 --param l1-cache-line-size=64 --param l2-cache-size=512 -mtune=znver2 -dumpbase - -march=i686 seems to fail with same error. This is a container if that helps (so no virtualization). (^ that being said I have a feint feeling after enabling virtualization in my UEFI my x86's has been acting differently.) And just to confirm, tests pass on my stable-amd64 env: ============== 16046 passed, 715 skipped, 127 xfailed, 360 warnings in 174.23s (0:02:54) ============== Well, obviously it's a precision problem, though I was hoping it happens with some older hardware. I'm not sure if upstream is really interested in test failures on 32-bit x86 though, especially given the number of issues open. Could you please try the current stable and see whether this is a regression? 0.13.1 has the same errors. ======== 42 failed, 15987 passed, 714 skipped, 127 xfailed, 4817 warnings in 233.44s (0:03:53) ======== * ERROR: dev-python/statsmodels-0.13.1::gentoo failed (test phase): * tests fail with python3.9 (now it's down to 42 from 43) And to complete: == 86 failed, 14079 passed, 688 skipped, 128 xfailed, 6752 warnings, 329 errors in 221.39s (0:03:41) == * ERROR: dev-python/statsmodels-0.12.2-r1::gentoo failed (test phase): * tests fail with python3.9 So I guess 0.13.2 is good to be stabilized? Could you also expand -march=native on amd64 on that machine? Yes, please stabilize and I'll try to figure out x86 later. # gcc -march=native -E -v - </dev/null 2>&1 | grep cc1 /usr/libexec/gcc/x86_64-pc-linux-gnu/11.2.1/cc1 -E -quiet -v - -march=znver2 -mmmx -mpopcnt -msse -msse2 -msse3 -mssse3 -msse4.1 -msse4.2 -mavx -mavx2 -msse4a -mno-fma4 -mno-xop -mfma -mno-avx512f -mbmi -mbmi2 -maes -mpclmul -mno-avx512vl -mno-avx512bw -mno-avx512dq -mno-avx512cd -mno-avx512er -mno-avx512pf -mno-avx512vbmi -mno-avx512ifma -mno-avx5124vnniw -mno-avx5124fmaps -mno-avx512vpopcntdq -mno-avx512vbmi2 -mno-gfni -mno-vpclmulqdq -mno-avx512vnni -mno-avx512bitalg -mno-avx512bf16 -mno-avx512vp2intersect -mno-3dnow -madx -mabm -mno-cldemote -mclflushopt -mclwb -mclzero -mcx16 -mno-enqcmd -mf16c -mfsgsbase -mfxsr -mno-hle -msahf -mno-lwp -mlzcnt -mmovbe -mno-movdir64b -mno-movdiri -mmwaitx -mno-pconfig -mno-pku -mno-prefetchwt1 -mprfchw -mno-ptwrite -mrdpid -mrdrnd -mrdseed -mno-rtm -mno-serialize -mno-sgx -msha -mno-shstk -mno-tbm -mno-tsxldtrk -mno-vaes -mno-waitpkg -mwbnoinvd -mxsave -mxsavec -mxsaveopt -mxsaves -mno-amx-tile -mno-amx-int8 -mno-amx-bf16 -mno-uintr -mno-hreset -mno-kl -mno-widekl -mno-avxvnni --param l1-cache-size=32 --param l1-cache-line-size=64 --param l2-cache-size=512 -mtune=znver2 -dumpbase - Could you try reporting this upstream? https://github.com/statsmodels/statsmodels/issues I suppose one option is to increase the tolerance, either unconditionally or conditionally to x86. One more idea: could you try if appending `-mfpmath=sse` to CFLAGS fixes the tests? Unfortunately it makes no difference. I've filed bug upstream. 0.14.0 continuing to get worse. 48 failed, 17561 passed, 296 skipped, 139 xfailed, 1021 warnings in 131.87s (0:02:11) The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=830cd7719231b4c34d98a000035154e5d53d5d02 commit 830cd7719231b4c34d98a000035154e5d53d5d02 Author: Sam James <sam@gentoo.org> AuthorDate: 2023-12-25 07:33:20 +0000 Commit: Sam James <sam@gentoo.org> CommitDate: 2023-12-25 07:33:20 +0000 dev-python/pandas: drop statsmodels dep on x86 It's a test-only optional dep and statsmodels is busted on x86. Bug: https://bugs.gentoo.org/840770 Bug: https://bugs.gentoo.org/905736 Bug: https://bugs.gentoo.org/908367 Signed-off-by: Sam James <sam@gentoo.org> dev-python/pandas/pandas-1.5.3.ebuild | 4 ++-- dev-python/pandas/pandas-2.1.1.ebuild | 4 ++-- dev-python/pandas/pandas-2.1.3.ebuild | 4 ++-- dev-python/pandas/pandas-2.1.4.ebuild | 4 ++-- 4 files changed, 8 insertions(+), 8 deletions(-) |