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.

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

openblas: 0.3.10 -> 0.3.12 #101780

Merged
merged 1 commit into from Oct 27, 2020
Merged

openblas: 0.3.10 -> 0.3.12 #101780

merged 1 commit into from Oct 27, 2020

Conversation

@glittershark
Copy link
Member

@glittershark glittershark commented Oct 26, 2020

Motivation for this change

Fixes #92458

Things done
  • Tested using sandboxing (nix.useSandbox on NixOS, or option sandbox in nix.conf on non-NixOS linux)
  • Built on platform(s)
    • NixOS
    • macOS
    • other Linux distributions
  • Tested via one or more NixOS test(s) if existing and applicable for the change (look inside nixos/tests)
  • Tested compilation of all pkgs that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review wip"
  • Tested execution of all binary files (usually in ./result/bin/)
  • Determined the impact on package closure size (by running nix path-info -S before and after)
  • Ensured that relevant documentation is up to date
  • Fits CONTRIBUTING.md.
@@ -144,12 +144,6 @@ stdenv.mkDerivation rec {
buildPackages.stdenv.cc
];

# Disable an optimisation which seems to cause issues, pending an
# upstream fix: https://github.com/xianyi/OpenBLAS/issues/2496
Copy link
Member Author

@glittershark glittershark Oct 26, 2020

this appears to have been closed upstream

@glittershark
Copy link
Member Author

@glittershark glittershark commented Oct 26, 2020

cc @FRidh

@FRidh
Copy link
Member

@FRidh FRidh commented Oct 26, 2020

Please test/build numpy, scipy, and octave.

@glittershark
Copy link
Member Author

@glittershark glittershark commented Oct 26, 2020

  • numpy
  • scipy
  • octave

@flokli
Copy link
Contributor

@flokli flokli commented Oct 26, 2020

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

@glittershark glittershark changed the base branch from master to staging Oct 26, 2020
@glittershark
Copy link
Member Author

@glittershark glittershark commented Oct 26, 2020

😢 that wasn't what I wanted

@glittershark
Copy link
Member Author

@glittershark glittershark commented Oct 26, 2020

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

done.

@flokli
Copy link
Contributor

@flokli flokli commented Oct 26, 2020

This needs a rebase.

@glittershark
Copy link
Member Author

@glittershark glittershark commented Oct 26, 2020

This needs a rebase.

done

@flokli
Copy link
Contributor

@flokli flokli commented Oct 26, 2020

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

@glittershark
Copy link
Member Author

@glittershark glittershark commented Oct 26, 2020

for real this time

@flokli flokli merged commit 8b98955 into NixOS:staging Oct 27, 2020
19 checks passed
@glittershark glittershark deleted the bump-openblas branch Oct 27, 2020
@jonringer
Copy link
Contributor

@jonringer jonringer commented Oct 27, 2020

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 <root@gws.fyi>
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
Copy link
Member Author

@glittershark glittershark commented Oct 27, 2020

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
Copy link
Contributor

@jonringer jonringer commented Oct 27, 2020

interesting, not sure why mine just segfaults

@glittershark
Copy link
Member Author

@glittershark glittershark commented Oct 28, 2020

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

@jonringer
Copy link
Contributor

@jonringer jonringer commented Oct 28, 2020

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

@andir
Copy link
Member

@andir andir commented Oct 28, 2020

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.

FRidh added a commit that referenced this issue Oct 28, 2020
Hash is updated but not the version.

This reverts commit 8b98955, reversing
changes made to 5e2fde0.
@FRidh
Copy link
Member

@FRidh FRidh commented Oct 28, 2020

Reverted in 4e29151.

@xaverdh
Copy link
Contributor

@xaverdh xaverdh commented Oct 28, 2020

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
Copy link
Member

@FRidh FRidh commented Oct 28, 2020

Version was there in 858e5dc before the force push.

dpausp added a commit to flyingcircusio/nixpkgs that referenced this issue Oct 28, 2020
Version number got lost in a force push upstream:

NixOS/nixpkgs#101780
@glittershark
Copy link
Member Author

@glittershark glittershark commented Oct 28, 2020

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

@TredwellGit TredwellGit mentioned this pull request Oct 30, 2020
3 tasks
nomeata added a commit to nomeata/nixpkgs that referenced this issue Jan 30, 2021
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

The motivation is to unbreak building `python36Packages.scipy` (see
issue 92458)
nomeata added a commit to nomeata/nixpkgs that referenced this issue Feb 2, 2021
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)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment