This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

classification
Title: Python 2.5 CVS broken for HP-UX platform?
Type: Stage:
Components: Build Versions:
process
Status: closed Resolution:
Dependencies: Superseder:
Assigned To: Nosy List: anthonybaxter, nnorwitz, ranma, sf-robot
Priority: normal Keywords:

Created on 2005-06-01 13:05 by ranma, last changed 2022-04-11 14:56 by admin. This issue is now closed.

Messages (8)
msg25464 - (view) Author: Vincent Jamart (ranma) Date: 2005-06-01 13:05
While trying to compile Python 2.5 from the nighlty CVS
image, it raises the following errors on HP-UX. (OS
version= HP-UX 11.22 on ia64, compiler= aCC: HP
aC++/ANSI C B3910B A.06.00 [Aug 25 2004])

Compilation flags:
export CC=aCC
export CFLAGS="-Ae +DD64"
export LDFLAGS="+DD64"
make clean
./configure --prefix=/usr/local/python_cvs --with-cxx=aCC


(...)
creating Makefile
        aCC -c  -DNDEBUG -O  -I. -I./Include  
-DPy_BUILD_CORE -o
Modules/ccpython.o ./Modules/ccpython.cc
"/usr/include/sys/unistd.h", line 594: warning #2837-D:
omission of
explicit
          type is nonstandard ("int" assumed)
  inline truncate(const char *a, off_t b)       { return
__truncate64(a,b); }
         ^
 
"/usr/include/sys/unistd.h", line 595: warning #2837-D:
omission of
explicit
          type is nonstandard ("int" assumed)
  inline prealloc(int a, off_t b)               { return
__prealloc64(a,b); }
         ^
 
"/usr/include/sys/unistd.h", line 596: warning #2837-D:
omission of
explicit
          type is nonstandard ("int" assumed)
  inline lockf(int a, int b, off_t c)           { return
__lockf64(a,b,c); }
         ^
 
"/usr/include/sys/unistd.h", line 597: warning #2837-D:
omission of
explicit
          type is nonstandard ("int" assumed)
  inline ftruncate(int a, off_t b)              { return
__ftruncate64(a,b); }
         ^
 
"/usr/include/sys/stat.h", line 173: warning #2837-D:
omission of
explicit
          type is nonstandard ("int" assumed)
       inline lstat __((const char *, struct stat *));
              ^
 
"./Include/pyport.h", line 612: error #2035: #error
directive: "LONG_BIT
          definition appears wrong for platform (bad
gcc/glibc
config?)."
  #error "LONG_BIT definition appears wrong for
platform (bad gcc/glibc
config?)."
   ^
 
1 error detected in the compilation of
"./Modules/ccpython.cc".
*** Error exit code 2
 
Stop.
        aCC -c  -DNDEBUG -O  -I. -I./Include  
-DPy_BUILD_CORE -o
Modules/ccpython.o ./Modules/ccpython.cc
"/usr/include/sys/unistd.h", line 594: warning #2837-D:
omission of
explicit
          type is nonstandard ("int" assumed)
  inline truncate(const char *a, off_t b)       { return
__truncate64(a,b); }
         ^
 
"/usr/include/sys/unistd.h", line 595: warning #2837-D:
omission of
explicit
          type is nonstandard ("int" assumed)
  inline prealloc(int a, off_t b)               { return
__prealloc64(a,b); }
         ^
 
"/usr/include/sys/unistd.h", line 596: warning #2837-D:
omission of
explicit
          type is nonstandard ("int" assumed)
  inline lockf(int a, int b, off_t c)           { return
__lockf64(a,b,c); }
         ^
 
"/usr/include/sys/unistd.h", line 597: warning #2837-D:
omission of
explicit
          type is nonstandard ("int" assumed)
  inline ftruncate(int a, off_t b)              { return
__ftruncate64(a,b); }
         ^
 
"/usr/include/sys/stat.h", line 173: warning #2837-D:
omission of
explicit
          type is nonstandard ("int" assumed)
       inline lstat __((const char *, struct stat *));
              ^
 
"./Include/pyport.h", line 612: error #2035: #error
directive: "LONG_BIT
          definition appears wrong for platform (bad
gcc/glibc
config?)."
  #error "LONG_BIT definition appears wrong for
platform (bad gcc/glibc
config?)."
   ^
 
1 error detected in the compilation of
"./Modules/ccpython.cc".
*** Error exit code 2
 
Stop.
msg25465 - (view) Author: Vincent Jamart (ranma) Date: 2005-06-16 10:09
Logged In: YES 
user_id=150220

We nailed down the problem. 
Apparantly the option "+DD64" is not used by 'make'. 
Nevertheless we added this to CPPFLAGS and CFLAGS. 
msg25466 - (view) Author: Anthony Baxter (anthonybaxter) (Python triager) Date: 2005-06-19 09:01
Logged In: YES 
user_id=29957

It's not clear from your message - did you find a fix for
this? It seems like this should be closed as notabug...

msg25467 - (view) Author: Vincent Jamart (ranma) Date: 2005-06-28 15:05
Logged In: YES 
user_id=150220

This is how we managed to solve the problem here, but it's
not a clean solution:

#install script for HP-UX 11.xx
CPPFLAGS="+DD64 -I/usr/local/bzip2-1.0.3/include"
export CPPFLAGS
CC="/opt/aCC/bin/aCC -Ae +DD64"
export CC
CFLAGS="-Ae +DD64"
export CFLAGS
CXX="/opt/aCC/bin/aCC -Ae"
export CXX
CXXFLAGS="-Ae +DD64"
export CXXFLAGS
LDFLAGS="+DD64 -L/usr/local/bzip2-1.0.3/lib"
export LDFLAGS
cd dist/src
#
# Do not use the --without-gcc flag because the CC env.var.
will be
# reset to 'cc' (which do not want)
# Additionally we're not using --with-cxx because we want
the system to use
# the CC as it is defined above. The reason is that
apparantly CFLAGS is not
# taken into account thus all options should go into the
definition of CC
#
./configure --prefix=/usr/local/python --without-threads
#
# Finally after the configure we need to edit the resulting
Makefile
# by hand because 'ld' will be used to link the extension
modules. However
# 'ld' does not know the '+DD64' flag and we should link
with aCC instead.
# However there is no way to ask 'configure' to link with
aCC instead of ld.
msg25468 - (view) Author: Vincent Jamart (ranma) Date: 2005-06-29 06:50
Logged In: YES 
user_id=150220

This is how we managed to solve the problem here, but it's
not a clean solution:

#install script for HP-UX 11.xx
CPPFLAGS="+DD64 -I/usr/local/bzip2-1.0.3/include"
export CPPFLAGS
CC="/opt/aCC/bin/aCC -Ae +DD64"
export CC
CFLAGS="-Ae +DD64"
export CFLAGS
CXX="/opt/aCC/bin/aCC -Ae"
export CXX
CXXFLAGS="-Ae +DD64"
export CXXFLAGS
LDFLAGS="+DD64 -L/usr/local/bzip2-1.0.3/lib"
export LDFLAGS
cd dist/src
#
# Do not use the --without-gcc flag because the CC env.var.
will be
# reset to 'cc' (which do not want)
# Additionally we're not using --with-cxx because we want
the system to use
# the CC as it is defined above. The reason is that
apparantly CFLAGS is not
# taken into account thus all options should go into the
definition of CC
#
./configure --prefix=/usr/local/python --without-threads
#
# Finally after the configure we need to edit the resulting
Makefile
# by hand because 'ld' will be used to link the extension
modules. However
# 'ld' does not know the '+DD64' flag and we should link
with aCC instead.
# However there is no way to ask 'configure' to link with
aCC instead of ld.
msg25469 - (view) Author: Neal Norwitz (nnorwitz) * (Python committer) Date: 2005-09-20 06:21
Logged In: YES 
user_id=33168

I think there was a fix checked in recently to address some
(hopefully all) of these issues.  Does the current CVS build
on your HPUX box?
msg25470 - (view) Author: Anthony Baxter (anthonybaxter) (Python triager) Date: 2005-09-20 06:32
Logged In: YES 
user_id=29957

As far as I know, the patches contributed by Elemental
Security make current CVS HEAD work fine on ia64 HP/UX.
Please do test them, though - I think Guido put some notes
in the toplevel README about how to get the build to work.
msg25471 - (view) Author: SourceForge Robot (sf-robot) Date: 2006-03-07 03:24
Logged In: YES 
user_id=1312539

This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
History
Date User Action Args
2022-04-11 14:56:11adminsetgithub: 42040
2005-06-01 13:05:50ranmacreate