openblas: 0.3.10 -> 0.3.12 by glittershark · Pull Request #101780 · NixOS/nixpkgs

glittershark · 2020-10-26T17:30:23Z

Motivation for this change

Fixes #92458

Things done

glittershark · 2020-10-26T17:41:21Z

pkgs/development/libraries/science/math/openblas/default.nix

this appears to have been closed upstream

glittershark · 2020-10-26T17:55:45Z

cc @FRidh

FRidh · 2020-10-26T18:01:43Z

Please test/build numpy, scipy, and octave.

glittershark · 2020-10-26T18:10:44Z

numpy
scipy
octave

flokli · 2020-10-26T22:38:57Z

octave seems to be fine too - this should however target staging due to the amount of rebuilds.

glittershark · 2020-10-26T22:41:12Z

😢 that wasn't what I wanted

glittershark · 2020-10-26T22:45:00Z

octave seems to be fine too - this should however target staging due to the amount of rebuilds.

done.

flokli · 2020-10-26T22:47:26Z

This needs a rebase.

glittershark · 2020-10-26T22:47:52Z

This needs a rebase.

done

flokli · 2020-10-26T23:08:33Z

Apparently this wasn't pushed, or at least GitHub still shows a conflict.

glittershark · 2020-10-26T23:25:11Z

for real this time

jonringer · 2020-10-27T18:26:49Z

this breaks numpy for me, or at least now, it segfaults when running tests with 32 cores

store/gvkhgh813chlsvdsnxgwhv72xns08xcp-python3-3.8.6/bin/python3.8 -c 'import numpy; numpy.test("fast", verbose=10)'
builder for '/nix/store/s7azf4dald37gw9sa9shpvvpgldnjwgz-python3.8-numpy-1.19.1.drv' failed with exit code 139
error: build of '/nix/store/s7azf4dald37gw9sa9shpvvpgldnjwgz-python3.8-numpy-1.19.1.drv' failed
3b4cd4f4dacd7db4011e88d819a1ae5acc3f3aff is the first bad commit
commit 3b4cd4f4dacd7db4011e88d819a1ae5acc3f3aff
Author: Griffin Smith <[email protected]>
Date:   Mon Oct 26 13:08:56 2020 -0400

    openblas: 0.3.10 -> 0.3.12

 ...-optimised-aarch64-dgemm_beta-pending-fix.patch | 26 ----------------------
 .../libraries/science/math/openblas/default.nix    |  8 +------
 2 files changed, 1 insertion(+), 33 deletions(-)
 delete mode 100644 pkgs/development/libraries/science/math/openblas/0001-Disable-optimised-aarch64-dgemm_beta-pending-fix.patch
bisect run success
[11:22:55] jon@nixos ~/projects/nixpkgs ((3b4cd4f4dac...)|BISECTING)
$ git bisect run nix-build -A python3Packages.numpy

testing now if turning down the cores helps

glittershark · 2020-10-27T19:40:11Z

can confirm that python3Packages.numpy builds fine on my machine:

❯ lscpu
Architecture:                    x86_64
CPU op-mode(s):                  32-bit, 64-bit
Byte Order:                      Little Endian
Address sizes:                   39 bits physical, 48 bits virtual
CPU(s):                          12
On-line CPU(s) list:             0-11
Thread(s) per core:              2
Core(s) per socket:              6
Socket(s):                       1
NUMA node(s):                    1
Vendor ID:                       GenuineIntel
CPU family:                      6
Model:                           158
Model name:                      Intel(R) Core(TM) i9-8950HK CPU @ 2.90GHz
Stepping:                        10
CPU MHz:                         3459.563
CPU max MHz:                     4800.0000
CPU min MHz:                     800.0000
BogoMIPS:                        5799.77
Virtualization:                  VT-x
L1d cache:                       192 KiB
L1i cache:                       192 KiB
L2 cache:                        1.5 MiB
L3 cache:                        12 MiB
NUMA node0 CPU(s):               0-11
Vulnerability Itlb multihit:     KVM: Mitigation: Split huge pages
Vulnerability L1tf:              Mitigation; PTE Inversion; VMX conditional cache flushes, SMT vulnerable
Vulnerability Mds:               Mitigation; Clear CPU buffers; SMT vulnerable
Vulnerability Meltdown:          Mitigation; PTI
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1:        Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:        Mitigation; Full generic retpoline, IBPB conditional, IBRS_FW, STIBP conditional, RSB filling
Vulnerability Srbds:             Mitigation; Microcode
Vulnerability Tsx async abort:   Mitigation; Clear CPU buffers; SMT vulnerable
Flags:                           fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtop
                                 ology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm a
                                 bm 3dnowprefetch cpuid_fault epb invpcid_single pti ssbd ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm mpx rdseed adx smap clflush
                                 opt intel_pt xsaveopt xsavec xgetbv1 xsaves dtherm ida arat pln pts hwp hwp_notify hwp_act_window hwp_epp md_clear flush_l1d

jonringer · 2020-10-27T20:46:38Z

interesting, not sure why mine just segfaults

glittershark · 2020-10-28T02:01:10Z

I'm not sure how this all works, but can we ask Hydra to build it for us?

jonringer · 2020-10-28T02:58:41Z

I think ofborg was able to build this for us. Still, this shouldn't be happening :(

andir · 2020-10-28T08:44:34Z

This PR did not bump the version attribute. I am a bit confused how all of you did run tests but not notice the hash mismatch.

Hash is updated but not the version. This reverts commit 8b98955, reversing changes made to 5e2fde0.

FRidh · 2020-10-28T08:48:29Z

Reverted in 4e29151.

xaverdh · 2020-10-28T09:16:20Z

This PR did not bump the version attribute. I am a bit confused how all of you did run tests but not notice the hash mismatch.

Maybe it was lost in the rebase / force push? Does github record a history of pull requests somehow?

FRidh · 2020-10-28T09:33:27Z

Version was there in 858e5dc before the force push.

Version number got lost in a force push upstream: NixOS#101780

glittershark · 2020-10-28T14:25:29Z

🤦 sorry about that - that's what I get for trying to multitask as much as I was

this backports the effect of the following commits from `master` to 20.09: * f52263c treewide: Start to break up static overlay * d1d536c openblas: 0.3.10 -> 0.3.12 * f715602 Revert "openblas: 0.3.10 -> 0.3.12" * 840c201 Merge pull request NixOS#101715 from r-ryantm/auto-update/openblas * e1a59dd openblas: 0.3.10 -> 0.3.12 * 4e29151 Revert "Merge pull request NixOS#101780 from glittershark/bump-openblas" * 3b4cd4f openblas: 0.3.10 -> 0.3.12 * 692d219 Merge staging-next into staging * 7902256 openblas: enable multiple outputs * 92d7b38 openblas: enable on ppc64le * 0137860 openblas: 0.3.12 -> 0.3.13 The motivation is to unbreak building `python36Packages.scipy` (see issue 92458)

glittershark mentioned this pull request Oct 26, 2020

python36Packages.scipy: segmentation fault when running tests #92458

Closed

ofborg bot requested a review from ttuegel October 26, 2020 17:40

glittershark commented Oct 26, 2020

View reviewed changes

pkgs/development/libraries/science/math/openblas/default.nix Outdated

Copy link

Member Author

glittershark Oct 26, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

this appears to have been closed upstream

glittershark changed the base branch from master to staging October 26, 2020 22:40

glittershark requested review from FRidh, Profpatsch, edolstra, edwtjo, infinisil, jonringer, mmahut, nbp, peti and tfc as code owners October 26, 2020 22:40

glittershark force-pushed the bump-openblas branch from 858e5dc to f280a98 Compare October 26, 2020 22:44

ofborg bot added the 2.status: merge conflict This PR has merge conflicts with the target branch label Oct 26, 2020

openblas: 0.3.10 -> 0.3.12

3b4cd4f

glittershark force-pushed the bump-openblas branch from f280a98 to 3b4cd4f Compare October 26, 2020 23:24

ofborg bot removed the 2.status: merge conflict This PR has merge conflicts with the target branch label Oct 26, 2020

flokli merged commit 8b98955 into NixOS:staging Oct 27, 2020

glittershark deleted the bump-openblas branch October 27, 2020 14:00

FRidh added a commit that referenced this pull request Oct 28, 2020

Revert "Merge pull request #101780 from glittershark/bump-openblas"

4e29151

Hash is updated but not the version. This reverts commit 8b98955, reversing changes made to 5e2fde0.

This was referenced Oct 28, 2020

openblas: 0.3.10 -> 0.3.12 #101715

Merged

systemd: introduce more options for a more minimal build #101886

Merged

dpausp added a commit to flyingcircusio/nixpkgs that referenced this pull request Oct 28, 2020

Fix OpenBLAS version

b37976e

Version number got lost in a force push upstream: NixOS#101780

TredwellGit mentioned this pull request Oct 30, 2020

openblas: 0.3.10 -> 0.3.12 #102126

Merged

3 tasks

nomeata mentioned this pull request Jan 7, 2021

[staging-20.09] openblas: 0.3.10 -> 0.3.12 #108709

Merged

FRidh mentioned this pull request Jan 8, 2021

Python3x vs Python3xFull # pkgs broken and overlay without effect #108773

Closed

Uh oh!

Conversation

glittershark commented Oct 26, 2020

Motivation for this change

Things done

Uh oh!

glittershark Oct 26, 2020

Choose a reason for hiding this comment

Uh oh!

glittershark commented Oct 26, 2020

Uh oh!

FRidh commented Oct 26, 2020

Uh oh!

glittershark commented Oct 26, 2020

Uh oh!

flokli commented Oct 26, 2020

Uh oh!

glittershark commented Oct 26, 2020

Uh oh!

glittershark commented Oct 26, 2020

Uh oh!

flokli commented Oct 26, 2020

Uh oh!

glittershark commented Oct 26, 2020

Uh oh!

flokli commented Oct 26, 2020

Uh oh!

glittershark commented Oct 26, 2020

Uh oh!

jonringer commented Oct 27, 2020

Uh oh!

glittershark commented Oct 27, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jonringer commented Oct 27, 2020

Uh oh!

glittershark commented Oct 28, 2020

Uh oh!

jonringer commented Oct 28, 2020

Uh oh!

andir commented Oct 28, 2020

Uh oh!

FRidh commented Oct 28, 2020

Uh oh!

xaverdh commented Oct 28, 2020

Uh oh!

FRidh commented Oct 28, 2020

Uh oh!

glittershark commented Oct 28, 2020

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

glittershark commented Oct 27, 2020 •

edited

Loading