This is the mail archive of the gdb-prs@sources.redhat.com mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

build/1481: build fails, native hppa2.0w-hp-hpux11.11, nfs, identical source timestamps


>Number:         1481
>Category:       build
>Synopsis:       build fails, native hppa2.0w-hp-hpux11.11, nfs,  identical source timestamps
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Dec 11 09:38:00 UTC 2003
>Closed-Date:
>Last-Modified:
>Originator:     mec.gnu@mindspring.com
>Release:        gdb 6.0
>Organization:
>Environment:
target = native, host = hppa2.0w-hp-hpux11.11
make is hpux make
c compiler is /opt/ansic/bin or /opt/aCC/bin
source tree is on an NFS drive
>Description:
My build blows up:

  Configuring in utils
  loading cache .././config.cache
  checking for a BSD compatible install.. /house/chastain/gdb/s0/utils/../install -sh -c
  checking host system type... hppa2.0w-hp-hpux11.11
  checking target system type... hppa2.0w-hp-hpux11.11
  checking build system type... hppa2.0w-hp-hpux11.11
  creating ./config.status
  creating Makefile
          cd /house/chastain/gdb/s0/utils && autoconf
  sh: autoconf:  not found
  *** Error exit code 127

The problem is that these files have identical timestamps:

  2002-11-23 04:01:30 UTC  utils/configure
  2002-11-23 04:01:30 UTC  utils/configure.in

This is a nasty edge case.  Single Unix Specification Version 3 for make(1) says:

  A target is considered out-of-date if it is older than any of its prerequisites or if it does not exist. 

And HPUX 11.11 "man make" says:

  A target is updated only if it depends on files that are newer than the target.
>How-To-Repeat:
Build gdb as described above (on a system with no autoconf).  I don't think the hppa part is significant except for the version of "make" used.  The NFS part is probably not important either -- it's not a clock skew problem between filesystems on different computers; it's a problem between two filesystems in the same directory.
>Fix:
(1) Require a correct make, such as Gnu Make.

(2) Check the source code for files that have identical time stamps and modify the time stamps by doing another cvs checkin.

(3) ??? Anything else?

I favor (2) but I'm open to discussion.  I like (2) because I think that gdb should build on as many systems as possible, even those with buggy "make" commands.  It's very convenient for us to generate files with identical time stamps in the source archive, but it's fragile to rely on "make" doing the right thing with them.

>Release-Note:
>Audit-Trail:
>Unformatted:


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]