AIX 7.1 Smoke Tests

[p5pRT]

Migrated from rt.perl.org#131734 (status was 'open')

Searchable as RT131734$

[p5pRT]

From carlos@carlosguevara.com

Hi,

I saw Jim Keenan's TPC talk and have been submitting Perl 5 smoke test
results for various OSs.

However, I've run into a problem with AIX 7.1. I've never used Git, but
I've managed to figure out that commit
7ad910c on 2017-06-09 broke the make
process, and it's remained broken since. Since this commit is so old, and
because of my limited knowledge of Git and the smoke testing process, I'm
not sure how to proceed.

For reference, these are the before / after smoke test results​:
http​://perl5.test-smoke.org/report/56555
http​://perl5.test-smoke.org/report/56558

The outout of 'perl -V' from both runs is included at the end of this
message.

I'd be happy to help if you have any requests or suggestions.

Regards,
Carlos Guevara

#####

Summary of my perl5 (revision 5 version 26 subversion 0) configuration​:

  Platform​:
  osname=aix
  osvers=7.1.0.0
  archname=aix
  uname='aix sovma206 1 7 00c3c6f54c00 '
  config_args='-Dcc=gcc -Dprefix=/home/smoke/bin/perl-5.26.0 -des'
  hint=recommended
  useposix=true
  d_sigaction=define
  useithreads=undef
  usemultiplicity=undef
  use64bitint=undef
  use64bitall=undef
  uselongdouble=undef
  usemymalloc=n
  default_inc_excludes_dot=define
  bincompat5005=undef
  Compiler​:
  cc='gcc -maix32'
  ccflags ='-D_ALL_SOURCE -D_ANSI_C_SOURCE -D_POSIX_SOURCE
-DUSE_NATIVE_DLOPEN -fwrapv -fno-strict-aliasing -pipe -D_LARGE_FILES
-D_FORTIFY_SOURCE=2'
  optimize='-O'
  cppflags='-D_ALL_SOURCE -D_ANSI_C_SOURCE -D_POSIX_SOURCE
-DUSE_NATIVE_DLOPEN -fwrapv -fno-strict-aliasing -pipe'
  ccversion=''
  gccversion='4.8.5'
  gccosandvers=''
  intsize=4
  longsize=4
  ptrsize=4
  doublesize=8
  byteorder=4321
  doublekind=4
  d_longlong=define
  longlongsize=8
  d_longdbl=define
  longdblsize=8
  longdblkind=0
  ivtype='long'
  ivsize=4
  nvtype='double'
  nvsize=8
  Off_t='off_t'
  lseeksize=8
  alignbytes=8
  prototype=define
  Linker and Libraries​:
  ld='gcc -maix32'
  ldflags =' -Wl,-brtl -Wl,-bdynamic -Wl,-b32 -Wl,-bmaxdata​:0x80000000'
  libpth=/opt/freeware/lib
/opt/freeware/lib/gcc/powerpc-ibm-aix7.1.0.0/4.8.5/include-fixed /usr/lib
/lib /usr/ccs/lib /usr/lib64
  libs=-lpthread -lbind -lnsl -lgdbm -ldbm -ldb -ldl -lld -lm -lcrypt -lc
  perllibs=-lpthread -lbind -lnsl -ldl -lld -lm -lcrypt -lc
  libc=/lib/libc.a
  so=a
  useshrplib=false
  libperl=libperl.a
  gnulibc_version=''
  Dynamic Linking​:
  dlsrc=dl_aix.xs
  dlext=so
  d_dlsymun=undef
  ccdlflags='-Xlinker
-bE​:/home/smoke/bin/perl-5.26.0/lib/5.26.0/aix/CORE/perl.exp'
  cccdlflags=' '
  lddlflags=' -Wl,-bhalt​:4 -Wl,-G -Wl,-bI​:$(PERL_INC)/perl.exp
-Wl,-bE​:$(BASEEXT).exp -Wl,-bnoentry -lc -lm'

Characteristics of this binary (from libperl)​:
  Compile-time options​:
  HAS_TIMES
  PERLIO_LAYERS
  PERL_COPY_ON_WRITE
  PERL_DONT_CREATE_GVSV
  PERL_HASH_FUNC_ONE_AT_A_TIME_HARD
  PERL_MALLOC_WRAP
  PERL_OP_PARENT
  PERL_PRESERVE_IVUV
  USE_LARGE_FILES
  USE_LOCALE
  USE_LOCALE_COLLATE
  USE_LOCALE_CTYPE
  USE_LOCALE_NUMERIC
  USE_LOCALE_TIME
  USE_PERLIO
  USE_PERL_ATOF
  Built under aix
  Compiled at Jul 8 2017 17​:39​:11
  @​INC​:
  /home/smoke/bin/perl-5.26.0/lib/site_perl/5.26.0/aix
  /home/smoke/bin/perl-5.26.0/lib/site_perl/5.26.0
  /home/smoke/bin/perl-5.26.0/lib/5.26.0/aix
  /home/smoke/bin/perl-5.26.0/lib/5.26.0

[p5pRT]

From @jkeenan

On Mon, 10 Jul 2017 19​:04​:09 GMT, carlos@​carlosguevara.com wrote​:

Hi,

I saw Jim Keenan's TPC talk and have been submitting Perl 5 smoke test
results for various OSs.

However, I've run into a problem with AIX 7.1. I've never used Git, but
I've managed to figure out that commit
7ad910c on 2017-06-09 broke the make
process, and it's remained broken since.

I suspect you are correct, given that (a) 7ad910c merged ExtUtils-MakeMaker 7.28 into blead; (b) there were significant changes to cpan/ExtUtils-MakeMaker/lib/ExtUtils/MM_AIX.pm in that commit; and (c) CPANtesters reports on AIX are completely lacking (see​: http​://matrix.cpantesters.org/?dist=ExtUtils-MakeMaker%207.28).

But, as a confirmation measure, would it be possible (if you have time) for you to do a 'git checkout' of the commit *immediately preceding* 7ad910c -- 7ad910c^ or 42752ac -- and configure and build perl at that point? If 'make' completes successfully, then you have indeed isolated the offending commit.

Chris Williams and any other EUMM experts -- can you take a look?

Thank you very much.

--
James E Keenan (jkeenan@​cpan.org)

[p5pRT]

The RT System itself - Status changed from 'new' to 'open'

[p5pRT]

From carlos@carlosguevara.com

Yes, 'make' completes successfully on
42752ac.
The complete smoke test report is at​:
http​://perl5.test-smoke.org/report/56555

Please let me know if I can be of any further assistance.

On Mon, Jul 10, 2017 at 3​:23 PM, James E Keenan via RT <
perlbug-followup@​perl.org> wrote​:

On Mon, 10 Jul 2017 19​:04​:09 GMT, carlos@​carlosguevara.com wrote​:

Hi,

I saw Jim Keenan's TPC talk and have been submitting Perl 5 smoke test
results for various OSs.

However, I've run into a problem with AIX 7.1. I've never used Git, but
I've managed to figure out that commit
7ad910c on 2017-06-09 broke the make
process, and it's remained broken since.

I suspect you are correct, given that (a) 7ad910c merged
ExtUtils-MakeMaker 7.28 into blead; (b) there were significant changes to
cpan/ExtUtils-MakeMaker/lib/ExtUtils/MM_AIX.pm in that commit; and (c)
CPANtesters reports on AIX are completely lacking (see​:
http​://matrix.cpantesters.org/?dist=ExtUtils-MakeMaker%207.28).

But, as a confirmation measure, would it be possible (if you have time)
for you to do a 'git checkout' of the commit *immediately preceding*
7ad910c -- 7ad910c^ or 42752ac -- and
configure and build perl at that point? If 'make' completes successfully,
then you have indeed isolated the offending commit.

Chris Williams and any other EUMM experts -- can you take a look?

Thank you very much.

--
James E Keenan (jkeenan@​cpan.org)