bug-autoconf
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: autoconf 2.52 testsuite fails.


From: Akim Demaille
Subject: Re: autoconf 2.52 testsuite fails.
Date: 10 Aug 2001 08:55:53 +0200
User-agent: Gnus/5.0808 (Gnus v5.8.8) XEmacs/21.4 (Academic Rigor)

Please, keep it public.

| On Thu, Aug 09, 2001 at 09:56:11AM +0200, Akim Demaille wrote:
| > | I'm stumped. 
|  
| > Same here :)
| 
| Could it be related to HP-UX having a mktemp that isn't really
| compatable with the OpenBSD mktemp derivative?

No idea.  The code to create the tmp dir seemed to have been working
up to now.  But you might be right, we'll see that afterward.

| ( incidently, I have the OpenBSD mktemp derivative installed 
| in /usr/local/bin, it's just after /usr/bin/mktemp in the search
| path ) 
|  
| > | Any ideas?
| 
| > Nope, but could you give a try to:
| 
| > cd tests
| > ./debug-5.sh
| > M4='m4 -dV' ../autoconf -A .. >log 2>&1
|  
| > and send the log?
| 
| Sure, it's attached. 
| 
| --Jack 
| 

> /tmp/acAc3224
> m4 debug: NONE: 0: input read from configure.ac
> m4trace:configure.ac:1: -1- id 1: AC_PLAIN_SCRIPT ...
> m4trace:configure.ac:1: -1- id 1: AC_PLAIN_SCRIPT([]) -> ???
> m4trace:configure.ac:1: -1- id 1: AC_PLAIN_SCRIPT(...) -> [AS_INIT
> 
> # Forbidden tokens and exceptions.
> m4_pattern_forbid([^_?A[CHUM]_])
> m4_pattern_forbid([_AC_])
> # Actually reserved by M4sh.
> m4_pattern_allow([^AS_FLAGS$])
> 
> m4_divert_push([BODY])dnl
> m4_wrap([m4_divert_pop([BODY])[]])dnl
> ]
> m4trace:configure.ac:1: -1- id 2: AS_INIT ...
> m4trace:configure.ac:1: -1- id 2: AS_INIT -> ???
> m4trace:configure.ac:1: -1- id 2: AS_INIT -> [m4_init
> 
> # Forbidden tokens and exceptions.
> m4_pattern_forbid([^_?AS_])
> ]
> m4trace:configure.ac:1: -1- id 3: m4_init ...
> m4trace:configure.ac:1: -1- id 3: m4_init -> ???
> m4trace:configure.ac:1: -1- id 3: m4_init -> [# We need a tmp directory.
> m4_ifndef([m4_tmpdir],
>           [m4_define([m4_tmpdir], [/tmp])])
> 
> # M4sugar reserves `m4_[A-Za-z0-9_]*'.  We'd need \b and +,
> # but they are not portable.
> m4_pattern_forbid([^m4_])
> m4_pattern_forbid([^dnl$])
> 
> # Check the divert push/pop perfect balance.
> m4_wrap([m4_ifdef([_m4_divert_diversion],
>          [m4_fatal([m4_init: unbalanced m4_divert_push:]
> m4_defn([m4_divert_stack]))])[]])
> 
> m4_divert_push([KILL])
> m4_wrap([m4_divert_pop([KILL])[]])
> ]
> m4trace:configure.ac:1: -1- id 4: m4_ifndef ...
> m4trace:configure.ac:1: -1- id 4: m4_ifndef([m4_tmpdir], 
> [m4_define([m4_tmpdir], [/tmp])]) -> ???
> m4trace:configure.ac:1: -1- id 4: m4_ifndef(...) -> [m4_ifdef([m4_tmpdir], 
> [], [m4_define([m4_tmpdir], [/tmp])])]
> m4trace:configure.ac:1: -1- id 5: m4_ifdef ...
> m4trace:configure.ac:1: -1- id 5: m4_ifdef([m4_tmpdir], [], 
> [m4_define([m4_tmpdir], [/tmp])]) -> ???
> m4trace:configure.ac:1: -1- id 5: m4_ifdef(...)
> m4trace:configure.ac:1: -1- id 6: m4_pattern_forbid ...
> m4trace:configure.ac:1: -1- id 6: m4_pattern_forbid([^m4_]) -> ???
> m4trace:configure.ac:1: -1- id 6: m4_pattern_forbid(...) -> 
> [m4_file_append(m4_defn([m4_tmpdir])/forbidden.rx, [^m4_])]
> m4trace:configure.ac:1: -1- id 7: m4_file_append ...

>From here

> m4trace:configure.ac:1: -2- id 8: m4_defn ...
> m4trace:configure.ac:1: -2- id 8: m4_defn([m4_tmpdir]) -> ???
> m4trace:configure.ac:1: -2- id 8: m4_defn(...) -> [m4_ifndef([m4_tmpdir],
>            [m4_fatal([m4_defn: undefined macro: m4_tmpdir])])dnl
> m4_builtin([defn], [m4_tmpdir])]
> m4trace:configure.ac:1: -2- id 9: m4_ifndef ...
> m4trace:configure.ac:1: -2- id 9: m4_ifndef([m4_tmpdir], [m4_fatal([m4_defn: 
> undefined macro: m4_tmpdir])]) -> ???
> m4trace:configure.ac:1: -2- id 9: m4_ifndef(...) -> [m4_ifdef([m4_tmpdir], 
> [], [m4_fatal([m4_defn: undefined macro: m4_tmpdir])])]
> m4trace:configure.ac:1: -2- id 10: m4_ifdef ...
> m4trace:configure.ac:1: -2- id 10: m4_ifdef([m4_tmpdir], [], 
> [m4_fatal([m4_defn: undefined macro: m4_tmpdir])]) -> ???
> m4trace:configure.ac:1: -2- id 10: m4_ifdef(...)
> m4trace:configure.ac:1: -2- id 11: dnl ...
> m4trace:configure.ac:1: -2- id 11: dnl -> ???
> m4trace:configure.ac:1: -2- id 11: dnl
> m4trace:configure.ac:1: -2- id 12: m4_builtin ...
> m4trace:configure.ac:1: -2- id 12: m4_builtin([defn], [m4_tmpdir]) -> ???
> m4trace:configure.ac:1: -2- id 12: m4_builtin(...) -> [[/tmp/acAc3224]]

to here, it's looking for the tmp dir name, and found it.  Fine.

> m4trace:configure.ac:1: -1- id 7: 
> m4_file_append([/tmp/acAc3224/forbidden.rx], [^m4_]) -> ???
> m4trace:configure.ac:1: -1- id 7: m4_file_append(...) -> [m4_syscmd([cat 
> >>/tmp/acAc3224/forbidden.rx <<_m4eof
> ^m4_
> _m4eof
> ])
> m4_if(m4_sysval, [0], [],
>       [m4_fatal([m4_file_append: cannot write: /tmp/acAc3224/forbidden.rx])])]
> m4trace:configure.ac:1: -1- id 13: m4_syscmd ...
> m4trace:configure.ac:1: -1- id 13: m4_syscmd([cat 
> >>/tmp/acAc3224/forbidden.rx <<_m4eof
> ^m4_
> _m4eof
> ]) -> ???
> sh: cat:  not found.

This is the actual failure.  I don't understand the error msg at all.
I doubt that the tmp dir what not created, I suppose `autoconf' would
have failed, but who knows...

To check that it was created, please run

        ./debug-5.sh
        ../autoconf -A .. -d -v

the verbose messages should refer to the tmp dir it used, and with -d,
it should not be removed.  So you should be able to check if it was
created or not.

Other than that, 

> m4trace:configure.ac:1: -1- id 13: m4_syscmd(...)
> m4trace:configure.ac:1: -1- id 14: m4_if ...
> m4trace:configure.ac:1: -2- id 15: m4_sysval ...
> m4trace:configure.ac:1: -2- id 15: m4_sysval -> ???
> m4trace:configure.ac:1: -2- id 15: m4_sysval -> [127]
> m4trace:configure.ac:1: -1- id 14: m4_if([127], [0], [], 
> [m4_fatal([m4_file_append: cannot write: /tmp/acAc3224/forbidden.rx])]) -> ???
> m4trace:configure.ac:1: -1- id 14: m4_if(...) -> [m4_fatal([m4_file_append: 
> cannot write: /tmp/acAc3224/forbidden.rx])]
> m4trace:configure.ac:1: -1- id 16: m4_fatal ...
> m4trace:configure.ac:1: -1- id 16: m4_fatal([m4_file_append: cannot write: 
> /tmp/acAc3224/forbidden.rx]) -> ???
> m4trace:configure.ac:1: -1- id 16: m4_fatal(...) -> 
> [m4_errprintn(m4_location[: error: m4_file_append: cannot write: 
> /tmp/acAc3224/forbidden.rx])dnl

Good, the error was caught.  But was it clearly displayed?  I don't
remember...

Hm...  Just a stupid idea...  Could you try:

cat >/tmp/foo <<_m4eof
^m4_
_m4eof

and

cat >>/tmp/bar <<_m4eof
^m4_
_m4eof

?  Thanks!



reply via email to

[Prev in Thread] Current Thread [Next in Thread]