Amanda-Users

RE: amlabel not detecting a tape in device? BSDI5 and Tape.h

2003-08-21 16:04:16
Subject: RE: amlabel not detecting a tape in device? BSDI5 and Tape.h
From: "Chris Fairman" <cfairman AT succeed DOT net>
To: <amanda-users AT amanda DOT org>
Date: Thu, 21 Aug 2003 13:02:13 -0700
Thanks for your responses!

However it seems that the trouble is getting Amanda to build properly on
BSDi 5. There were some warnings that I originally ignored during the
configure script phase that I believe are impacting the way the binaries
are behaving. 

The configure script doesn't fail, and the build process looks clean.

Building 2.4.4p1 on BSDi configure gives the following warnings:

configure: WARNING: netinet/ip.h: present but cannot be compiled
configure: WARNING: netinet/ip.h: check for missing prerequisite
headers?
configure: WARNING: netinet/ip.h: proceeding with the preprocessor's
result

Config.log detail:
/usr/include/netinet/ip.h:70: field `ip_src' has incomplete type
/usr/include/netinet/ip.h:70: field `ip_dst' has incomplete type
/usr/include/netinet/ip.h:152: syntax error before `n_int32_t'
/usr/include/netinet/ip.h:154: field `ipt_addr' has incomplete type
/usr/include/netinet/ip.h:155: syntax error before `n_int32_t'

configure: WARNING: sys/mount.h: present but cannot be compiled
configure: WARNING: sys/mount.h: check for missing prerequisite headers?
configure: WARNING: sys/mount.h: proceeding with the preprocessor's
result

Config.log detail:
/usr/include/sys/ucred.h:48: `NGROUPS' undeclared here (not in a
function)
/usr/include/sys/ucred.h:68: `MAXLOGNAME' undeclared here (not in a
function)


configure: WARNING: sys/tape.h: present but cannot be compiled
configure: WARNING: sys/tape.h: check for missing prerequisite headers?
configure: WARNING: sys/tape.h: proceeding with the preprocessor's
result

config.log Detail:
        /usr/include/sys/tape.h:18: field `tp_dev' has incomplete type
        /usr/include/sys/tape.h:23: syntax error before `tpr_t'
        /usr/include/sys/tape.h:27: warning: `struct buf' declared
inside  parameter list/usr/include/sys/tape.h:27: warning: its scope is
only    this definition or declaration, which is probably not what you
want.


When building Amanda-2.5.0-20030716, two of these warnings go away.
However, I'm am still left with the tape.h warning.

checking sys/tape.h usability... no
checking sys/tape.h presence... yes
configure: WARNING: sys/tape.h: present but cannot be compiled
configure: WARNING: sys/tape.h: check for missing prerequisite headers?
configure: WARNING: sys/tape.h: proceeding with the preprocessor's
result

config.log detail:
/usr/include/sys/tape.h:18: field `tp_dev' has incomplete type
/usr/include/sys/tape.h:23: syntax error before `tpr_t'
/usr/include/sys/tape.h:27: warning: `struct buf' declared inside
parameter list/usr/include/sys/tape.h:27: warning: its scope is only
this definition or declaration, which is probably not what you want.

Reviewing the configure script in either release I can see that there is
a case statement that checks OS type. This case statement does not
include bsdi5*. I have edited the configure script to reflect the new
version but I am not sure that the options for Bsdi4 are appropriate for
Bsdi5:


bsdi4* | bsdi5*)
  lt_cv_deplibs_check_method='file_magic ELF [0-9][0-9]*-bit [ML]SB
(shared object|dynamic lib)'
  lt_cv_file_magic_cmd='/usr/bin/file -L'
  lt_cv_file_magic_test_file=/shlib/libc.so
  ;;

and:

bsdi4* | bsdi5*)
  version_type=linux
  need_version=no
  library_names_spec='${libname}${release}.so$versuffix
${libname}${release}.so$major $libname.so'
  soname_spec='${libname}${release}.so$major'
  finish_cmds='PATH="\$PATH:/sbin" ldconfig $libdir'
  shlibpath_var=LD_LIBRARY_PATH
  sys_lib_search_path_spec="/shlib /usr/lib /usr/X11/lib
/usr/contrib/lib /lib /usr/local/lib"
  sys_lib_dlsearch_path_spec="/shlib /usr/lib /usr/local/lib"
  export_dynamic_flag_spec=-rdynamic
  # the default ld.so.conf also contains /usr/contrib/lib and
  # /usr/X11R6/lib (/usr/X11 is a link to /usr/X11R6), but let us allow
  # libtool to hard-code these into programs
  ;;

This does not fix the tape.h warning.

I assume that this is the cause for Amanda's inability to correctly
access the device (and not detect a tape).

Any insight would be greatly appreciated...

Thanks in Advance,
-Chris


<Prev in Thread] Current Thread [Next in Thread>
  • RE: amlabel not detecting a tape in device? BSDI5 and Tape.h, Chris Fairman <=