BBC: Algorithm-MinPerfHashTwoLevel-0.16 fails with perl 5.31.5

[eserte]

Description
The test suite of Algorithm-MinPerfHashTwoLevel-0.16 mostly passes with perls <= 5.31.4, but has only failures with 5.31.5 (@demerphq: FYI). See http://fast-matrix.cpantesters.org/?dist=Algorithm-MinPerfHashTwoLevel%200.16

[dur-randir]

Bisect points to

commit e8864db
Author: Matt Turner mattst88@gmail.com
Date: Wed Sep 4 21:48:56 2019 -0700

Clean up U8TO*_LE macro implementations

The code guarded by #ifndef U32_ALIGNMENT_REQUIRED attempts to optimize
byte-swapping by doing unaligned loads, but accessing data through
unaligned pointers is undefined behavior in C. Moreover, compilers are
more than capable of recognizing these open-coded byte-swap patterns and
emitting a bswap instruction, or an unaligned load instruction, or a
combined load, etc. There's no need for multiple paths to attain the
desired result.

See https://rt.perl.org/Ticket/Display.html?id=133495

[mattst88]

I'll take a look.

[mattst88]

Oh, I think it's because it has the same broken code copied from Perl as I removed in my couple of commits.

I'll try to submit a patch tomorrow.

[mattst88]

Does this fix it?

https://github.com/mattst88/Algorithm-MinPerfHashTwoLevel/tree/clean-up-endianness

[jkeenan]

This ticket was closed yesterday after commit 2e38307. However, just now I built blead at HEAD (a6d5b82) and tried to install Algorithm::MinPerfHashTwoLevel using cpanm. The installation failed.

The relevant part of the cpanm build.log is enormous. So I'll only cite the first 65 lines.

Searching Algorithm::MinPerfHashTwoLevel () on cpanmetadb ...
--> Working on Algorithm::MinPerfHashTwoLevel
Fetching http://www.cpan.org/authors/id/Y/YV/YVES/Algorithm-MinPerfHashTwoLevel-0.16.tar.gz
-> OK
Unpacking Algorithm-MinPerfHashTwoLevel-0.16.tar.gz
Entering Algorithm-MinPerfHashTwoLevel-0.16
Checking configure dependencies from META.json
Checking if you have ExtUtils::MakeMaker 6.58 ... Yes (7.38)
Configuring Algorithm-MinPerfHashTwoLevel-0.16
Running Makefile.PL
Checking if your kit is complete...
Looks good
Generating a Unix-style Makefile
Writing Makefile for Algorithm::MinPerfHashTwoLevel
Writing MYMETA.yml and MYMETA.json
-> OK
Checking dependencies from MYMETA.json ...
Checking if you have ExtUtils::MakeMaker 7.36 ... Yes (7.38)
Checking if you have Test::More 1.302164 ... Yes (1.302168)
Building and testing Algorithm-MinPerfHashTwoLevel-0.16
cp lib/Algorithm/MinPerfHashTwoLevel.pm blib/lib/Algorithm/MinPerfHashTwoLevel.pm
cp lib/Tie/Hash/MinPerfHashTwoLevel/OnDisk.pm blib/lib/Tie/Hash/MinPerfHashTwoLevel/OnDisk.pm
Running Mkbootstrap for MinPerfHashTwoLevel ()
chmod 644 "MinPerfHashTwoLevel.bs"
"/home/jkeenan/testing/blead/bin/perl" -MExtUtils::Command::MM -e 'cp_nonempty' -- MinPerfHashTwoLevel.bs blib/arch/auto/Algorithm/MinPerfHashTwoLevel/MinPerfHashTwoLevel.bs 644
"/home/jkeenan/testing/blead/bin/perl" "/home/jkeenan/testing/blead/lib/perl5/5.31.6/ExtUtils/xsubpp"  -typemap '/home/jkeenan/testing/blead/lib/perl5/5.31.6/ExtUtils/typemap'  MinPerfHashTwoLevel.xs > MinPerfHashTwoLevel.xsc
mv MinPerfHashTwoLevel.xsc MinPerfHashTwoLevel.c
cc -c  -I. -fwrapv -fno-strict-aliasing -pipe -fstack-protector-strong -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -O3   -DVERSION=\"0.16\" -DXS_VERSION=\"0.16\" -fPIC "-I/home/jkeenan/testing/blead/lib/perl5/5.31.6/x86_64-linux/CORE"   MinPerfHashTwoLevel.c
rm -f blib/arch/auto/Algorithm/MinPerfHashTwoLevel/MinPerfHashTwoLevel.so
cc  -shared -O2 -L/usr/local/lib -fstack-protector-strong  MinPerfHashTwoLevel.o  -o blib/arch/auto/Algorithm/MinPerfHashTwoLevel/MinPerfHashTwoLevel.so  \
      \
  
chmod 755 blib/arch/auto/Algorithm/MinPerfHashTwoLevel/MinPerfHashTwoLevel.so
"/home/jkeenan/testing/blead/bin/perl" -MExtUtils::Command::MM -e 'cp_nonempty' -- MinPerfHashTwoLevel.bs blib/arch/auto/Algorithm/MinPerfHashTwoLevel/MinPerfHashTwoLevel.bs 644
PERL_DL_NONLAZY=1 "/home/jkeenan/testing/blead/bin/perl" "-MExtUtils::Command::MM" "-MTest::Harness" "-e" "undef *Test::Harness::Switches; test_harness(0, 'blib/lib', 'blib/arch')" t/*.t
t/Algo_fail.t ... ok

#   Failed test 'state for seed is as expected'
#   at t/Algo_v5.t line 18.
#          got: 'd425545415e5d524558525b55514d555955435b51545b4d5b42524c465757554'
#     expected: '4475044405b585b4c5d575a5454485c5050465a50515e445247574d4752525c4'

#   Failed test 'test hash function 'fnorble''
#   at t/Algo_v5.t line 26.
#          got: '16125894952443580120'
#     expected: '17300635747579776751'

#   Failed test 'test hash function 'blah blah blah''
#   at t/Algo_v5.t line 26.
#          got: '14183480757785019034'
#     expected: '7551733764733872358'

#   Failed test 'test hash function 'blah blaH blah''
#   at t/Algo_v5.t line 26.
#          got: '6718570657053589147'
#     expected: '15370689535882563083'

#   Failed test 'simple hash A-Z'
#   at t/Algo_v5.t line 31.
#     Structures begin differing at:
#          $got->[0]{val_normalized} = 'J'
#     $expected->[0]{val_normalized} = 'N'
# $VAR1 = [
#           {
#             "h0" => '15684897285982551205',

Re-opening ticket.

Thank you very much.
Jim Keenan