Perl build on VMS: test phase fails on READ.T only.

[p5pRT]

Migrated from rt.perl.org#949 (status was 'resolved')

Searchable as RT949$

[p5pRT]

From [Unknown Contact. See original ticket]

stephen bauer wrote​:

 I am trying to install Perl 5\.005\_03 on a VAX
 running VMS 6\.2
 
 I am using DECC compiler V5\.6\, MMS version 3\.2\,
 Multinet V4\.1B and Decnet phase V\, V6\.3 ECO 10\.

For what it is worth I did not have any trouble with
a machine equipped as similar to yours as I can scrounge
up​: DEC C V5.3-006 on OpenVMS VAX V6.2, MMS V3.1-03,
V4.0 Rev C, Decnet phase IV.

That is "mms test" returned for me​:

All tests successful.
u=18.52 s=0 cu=0 cs=0 files=165 tests=6163

 In the Test phase of the build\, I get 1 error\.
 It is on the module \[\.T\.OP\]READ\.T
 
 Below is documentation of the error\, etc\.
 
 In the compile/link phase\, there is one
 informational message\.  I will include it\, in
 case it is relevant\.
 
 I did the "MMS" twice with "MMS realclean" 
 in between\.
 
 I got the source from the internet \(latest\_tar\.tar\)
 We un\-g\-zipped it and untar'ed  it on a PC\.
 We then FTP'd the modules from the PC to the VAX\.
 Somehow the FTP did not copy VMSISH\.H\, so I went back
 and FTP'ed that one by itself\.  I have re\-FTP'ed
 the READ\.T but it stil fails the test\.

This was the source of some of the trouble you had with missing files
when you ran configure.com. Try using gunzip.exe and vmstar.exe on your
VAX. These URLs are from README.vms​:

  2) GUNZIP/GZIP.EXE for VMS available from a number of web/ftp sites.
  http​://www.fsf.org/order/ftp.html
  ftp​://ftp.uu.net/archive/systems/gnu/diffutils*.tar.gz
  ftp​://gatekeeper.dec.com/pub/GNU/diffutils*.tar.gz
  ftp​://ftp.gnu.org/pub/gnu/diffutils*.tar.gz
  http​://www.openvms.digital.com/cd/GZIP/
  ftp​://ftp.digital.com/pub/VMS/

  3) VMS TAR also available from a number of web/ftp sites.
  ftp​://ftp.lp.se/vms/
  http​://www.openvms.digital.com/cd/VMSTAR/
  ftp​://ftp.digital.com/pub/VMS/

[snip]

 MCR $29$dua53​:\[perl\_5\_005\_03\]miniperl\.exe "\-I\[\-\-\-\.lib\]" "\-I\[\-\-\-\.lib\]" 
 \[\-\-\-\.lib\.E
 xtUtils\]xsubpp  \-typemap \[\-\-\-\.lib\.ExtUtils\]typemap STDIO\.xs >STDIO\.C
 CC/DECC 
 /Include=\[\]/Standard=Relaxed\_ANSI/Prefix=All/Obj=\.obj/NoList/Define=\("
 VE
 RSION=""2\.1"""\,"XS\_VERSION=""2\.1"""\)/Include=\(\[\-\-\-\]\)/Optimize  STDIO\.c
                     if \(\(retsts = sys$setddir\(&dirdsc\,0\,0\)\) & 1\) ST\(0\) 
 = &PL\_sv\_
 yes;
         \.\.\.\.\.\.\.\.\.\.\.\.\.\.\.\.\.\.\.\.\.\.\.\.\.\.^
 %CC\-I\-IMPLICITFUNC\, In this statement\, the identifier "sys$setddir" is
  implicitly declared as a function\.
                 At line number 208 in stdio\.xs\.
 
 %VCG\-I\-SUMMARY\, Completed with 0 error\(s\)\, 0 warning\(s\)\, and
                 1 informational messages\.
                 At line number 513 in stdio\.c\.
 \.
 \.
 \.
 $ mms test
 \.
 \.
 \.
 \[\.op\]read\.\.\.\.\.\.\.\.\.\.\.\.\.\.\.FAILED on test 1
 \.
 \.
 \.
 Failed 1 test\, 98\.19% okay\.\<CR>
 u=39\.79  s=0  cu=0  cs=0  files=164  tests=5994\<CR>

For what it is worth the informational message from the compiler and
the failure of read.t are unrelated. (The informational message was
the DECC compiler building the VMS​::Stdio module which is not
tested by read.t). I too saw the informational message with DECC 5.3-006.

Do you by chance have any logicals such as "T" or "OP" or "READ",
that is, does​:

  show logical t/table=*

turn up anything? That might have been part of the trouble.
The other thing that looks suspicious is that you defined
PERL_ROOT and did a SET DEF PERL_ROOT​:[000000]. Although
the build procedure was originally created to avoid any trouble
if you did that, I am not sure how well tested it is.
That is if you did​:

  set def $29$DUA53​:[PERL_5_005_03]
  mms test

You might get that test to pass.

At any rate, if I were you I would definitely start with a fresh
kit, transferred using binary ftp directly to the VAX and
gunzip then tar -xvf it.

Peter Prymmer

[p5pRT]

From [Unknown Contact. See original ticket]

Abandoning pre-5.6.0 install bugs

[p5pRT]

From The RT System itself

Abandoning pre-5.6.0 install bugs