maint-5.38: porting test failure; must run make regen

[jkeenan]

When smoke-testing the maint-5.38 branch on FreeBSD-14, I got failures in each run of t/porting/regen.t. See: https://perl.develop-help.com/db/5514634.

The individual test failure looks like this:

$ cd t;./perl harness -v porting/regen.t; cd -

ok 1 - generated keywords.c is up to date
ok 2 - generated keywords.h is up to date
ok 3 - generated perly.act is up to date
ok 4 - generated perly.h is up to date
ok 5 - generated perly.tab is up to date
ok 6 - generated charclass_invlists.h is up to date
ok 7 - generated uni_keywords.h is up to date
not ok 8 - generated uconfig.h is up to date
# Failed test 8 - generated uconfig.h is up to date at porting/regen.t line 122
#      got "config_h.SH"
# expected ""
ok 9 - generated regcharclass.h is up to date

I confirmed this failure on Ubuntu Linux.

Running make regen updated uconfig.h and appears to have fixed the test failures. (This problem appears to be just on maint-5.38, not maint-5.40.) Please review.

[steve-m-hay]

LGTM. Apologies for not doing that myself. I should have realized since I had exactly the same issue yesterday!

[jkeenan]

This pull request has failed some of its jobs in our GH CI; I re-started those jobs. The job called "dist-modules (latest)" has failed a second time. See: https://github.com/Perl/perl5/actions/runs/15375891354/job/43267333140?pr=23347.

Testing IO
  *** Testing IO ***
  $ cp -a dist/IO/. /tmp/R8oxHFb_Aq/.
  $ cp /tmp/yl8je4taFX/ppport.h .
  $ /usr/local/bin/perl Makefile.PL
  Generating a Unix-style Makefile
  Writing Makefile for IO
  Writing MYMETA.yml and MYMETA.json
  $ make test TEST_VERBOSE=0
  Running Mkbootstrap for IO ()
  chmod 644 "IO.bs"
  "/usr/local/bin/perl" -MExtUtils::Command::MM -e 'cp_nonempty' -- IO.bs blib/arch/auto/IO/IO.bs 644
  "/usr/local/bin/perl" "/usr/local/lib/perl5/site_perl/5.40.0/ExtUtils/xsubpp"  -typemap '/usr/local/lib/perl5/5.40.0/ExtUtils/typemap'  IO.xs > IO.xsc
  mv IO.xsc IO.c
  cc -c   -fwrapv -fno-strict-aliasing -pipe -fstack-protector-strong -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_FORTIFY_SOURCE=2 -O2   -DVERSION=\"1.52\" -DXS_VERSION=\"1.52\" -fPIC "-I/usr/local/lib/perl5/5.40.0/x86_64-linux-gnu/CORE"   IO.c
  cc -c   -fwrapv -fno-strict-aliasing -pipe -fstack-protector-strong -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_FORTIFY_SOURCE=2 -O2   -DVERSION=\"1.52\" -DXS_VERSION=\"1.52\" -fPIC "-I/usr/local/lib/perl5/5.40.0/x86_64-linux-gnu/CORE"   poll.c
  rm -f blib/arch/auto/IO/IO.so
  cc  -shared -O2 -L/usr/local/lib -fstack-protector-strong  IO.o poll.o  -o blib/arch/auto/IO/IO.so  \
        \
    
  chmod 755 blib/arch/auto/IO/IO.so
  cp IO.pm blib/lib/IO.pm
  cp lib/IO/File.pm blib/lib/IO/File.pm
  cp lib/IO/Socket/UNIX.pm blib/lib/IO/Socket/UNIX.pm
  cp lib/IO/Pipe.pm blib/lib/IO/Pipe.pm
  cp lib/IO/Socket.pm blib/lib/IO/Socket.pm
  cp lib/IO/Poll.pm blib/lib/IO/Poll.pm
  cp lib/IO/Socket/INET.pm blib/lib/IO/Socket/INET.pm
  cp lib/IO/Handle.pm blib/lib/IO/Handle.pm
  cp lib/IO/Select.pm blib/lib/IO/Select.pm
  cp lib/IO/Dir.pm blib/lib/IO/Dir.pm
  cp lib/IO/Seekable.pm blib/lib/IO/Seekable.pm
  PERL_DL_NONLAZY=1 "/usr/local/bin/perl" "-MExtUtils::Command::MM" "-MTest::Harness" "-e" "undef *Test::Harness::Switches; test_harness(0, 'blib/lib', 'blib/arch')" t/*.t
  t/cachepropagate-tcp.t ... ok
  t/cachepropagate-udp.t ... ok
  t/cachepropagate-unix.t .. ok
  t/gh17447.t .............. ok
  t/IO.t ................... ok
  t/io_const.t ............. ok
  t/io_dir.t ............... ok
  t/io_dup.t ............... ok
  t/io_file.t .............. ok
  t/io_file_export.t ....... ok
  
  #   Failed test 'tied handle read works in list context 2'
  #   at t/io_getline.t line 109.
  #          got: undef
  #     expected: 'Tick'
  
  #   Failed test 'tied handle read works in list context 3'
  #   at t/io_getline.t line 110.
  #          got: 'Tick'
  #     expected: 'tick'
  
  #   Failed test '$line unchanged'
  #   at t/io_getline.t line 117.
  #          got: undef
  #     expected: 'Tick'
  # Looks like you failed 3 tests of 37.
  t/io_getline.t ........... 
  Dubious, test returned 3 (wstat 768, 0x300)
  Failed 3/37 subtests 
  t/io_leak.t .............. skipped: No XS::APItest::sv_count() available
  t/io_linenum.t ........... ok
  t/io_multihomed.t ........ ok
  t/io_pipe.t .............. ok
  t/io_poll.t .............. ok
  t/io_sel.t ............... ok
  t/io_sock.t .............. ok
  t/io_sock_errstr.t ....... ok
  t/io_taint.t ............. ok
  t/io_tell.t .............. ok
  t/io_udp.t ............... ok
  t/io_unix.t .............. ok
  t/io_utf8.t .............. ok
  t/io_utf8argv.t .......... ok
  t/io_xs.t ................ ok
  
  Test Summary Report
  -------------------
  t/io_getline.t         (Wstat: 768 (exited 3) Tests: 37 Failed: 3)
    Failed tests:  33-34, 37
    Non-zero exit status: 3
  Files=26, Tests=16681, 26 wallclock secs ( 0.37 usr  0.03 sys +  0.71 cusr  0.18 csys =  1.29 CPU)
  Failed 1/26 test programs. 3/16681 subtests failed.
  Result: FAIL
  make: *** [Makefile:1061: test_dynamic] Error 255
  Failed: 512 (0x200)
Error: IO failed at make test

However, I wonder if this failure is relevant when what we're testing is a maintenance release for perl-5.38.

[steve-m-hay]

This pull request has failed some of its jobs in our GH CI; I re-started those jobs. The job called "dist-modules (latest)" has failed a second time. See: https://github.com/Perl/perl5/actions/runs/15375891354/job/43267333140?pr=23347.

[...]

However, I wonder if this failure is relevant when what we're testing is a maintenance release for perl-5.38.

Not sure what the problem there is, but the regen is definitely needed, so I've merged this PR.
Thanks for spotting this.