Re: Subject: [GNU Autoconf 2.59c] testsuite: 21 failed

[Paul Eggert]

Ralf Menzel <address@hidden> writes:

> this is again a bug report for my (mostly) native Sparc/Solaris 9
> system. Seems like the native awk of Solaris 9 doesn't know about the
> builtin function `sub':

Thanks for mentioning this.  (That's what I get for porting to Solaris
from memory.)  I installed the following patch.

2005-06-17  Paul Eggert  <address@hidden>

        * lib/m4sugar/m4sh.m4 (as_awk_strverscmp): Port to Solaris /bin/awk.
        * doc/autoconf.texi:
        Don't mention Solaris versions so much, if a
        problem is common to all extant versions of Solaris.  Say "SunOS
        4" instead of "SunOS" for SunOS 4.
        (awk): Mention more of the limitations of traditional Awk.
        (cat): Don't talk about cat -v.

Index: lib/m4sugar/m4sh.m4
===================================================================
RCS file: /cvsroot/autoconf/autoconf/lib/m4sugar/m4sh.m4,v
retrieving revision 1.147
diff -p -u -r1.147 m4sh.m4
--- lib/m4sugar/m4sh.m4 16 Jun 2005 21:03:47 -0000      1.147
+++ lib/m4sugar/m4sh.m4 17 Jun 2005 20:36:27 -0000
@@ -1027,24 +1027,27 @@ _AS_PATH_WALK([$PATH], [echo "PATH: $as_
 # Output variables for comparing version numbers.
 m4_defun([_AS_VERSION_COMPARE_PREPARE],
 [[as_awk_strverscmp='
+  # Use only awk features that work with 7th edition Unix awk (1978).
+  # My, what an old awk you have, Mr. Solaris!
   END {
     while (length(v1) || length(v2)) {
       # Set d1 to be the next thing to compare from v1, and likewise for d2.
       # Normally this is a single character, but if v1 and v2 contain digits,
       # compare them as integers and fractions as strverscmp does.
-      d1 = v1; sub(/[^0-9].*/, "", d1); len1 = length(d1)
-      d2 = v2; sub(/[^0-9].*/, "", d2); len2 = length(d2)
-      if (len1 && len2) {
-       # v1 and v2 both have leading digits.
-       v1 = substr(v1, len1 + 1)
-       v2 = substr(v2, len1 + 1)
+      if (v1 ~ /^[0-9]/ && v2 ~ /^[0-9]/) {
+       # Split v1 and v2 into their leading digit string components d1 and d2,
+       # and advance v1 and v2 past the leading digit strings.
+       for (len1 = 1; substr(v1, len1 + 1) ~ /^[0-9]/; len1++) continue
+       for (len2 = 1; substr(v2, len2 + 1) ~ /^[0-9]/; len2++) continue
+       d1 = substr(v1, 1, len1); v1 = substr(v1, len1 + 1)
+       d2 = substr(v2, 1, len2); v2 = substr(v2, len2 + 1)
        if (d1 ~ /^0/) {
          if (d2 ~ /^0/) {
            # Compare two fractions.
-           do {
+           while (d1 ~ /^0/ && d2 ~ /^0/) {
              d1 = substr(d1, 2); len1--
              d2 = substr(d2, 2); len2--
-           } while (d1 ~ /^0/ && d2 ~ /^0/);
+           }
            if (len1 != len2 && ! (len1 && len2 && substr(d1, 1, 1) == 
substr(d2, 1, 1))) {
              # The two components differ in length, and the common prefix
              # contains only leading zeros.  Consider the longer to be less.
Index: doc/autoconf.texi
===================================================================
RCS file: /cvsroot/autoconf/autoconf/doc/autoconf.texi,v
retrieving revision 1.914
diff -p -u -r1.914 autoconf.texi
--- doc/autoconf.texi   16 Jun 2005 07:24:43 -0000      1.914
+++ doc/autoconf.texi   17 Jun 2005 20:36:27 -0000
@@ -3332,7 +3332,7 @@ best implementation.
 @acindex{PROG_GREP}
 @ovindex GREP
 On AIX the default @code{grep} silently truncates long lines on the
-input before matching.  On Solaris 9 @code{/usr/bin/grep} does not
+input before matching.  On Solaris, @code{/usr/bin/grep} does not
 understand the @option{-e} option.  On NeXT, @code{grep} understands only a
 single @option{-e} option.  This macro looks for @sc{gnu} Grep or
 else the best available @code{grep} or @code{ggrep} in the user's
@@ -3881,7 +3881,7 @@ Solaris 7).
 @c @fuindex vsprintf
 @prindex @code{vsprintf}
 The ISO C standard says @code{sprintf} and @code{vsprintf} return the
-number of bytes written, but on some old systems (SunOS 4 for
+number of bytes written, but on some ancient systems (SunOS 4 for
 instance) they return the buffer pointer instead.
 
 @item @code{sscanf}
@@ -4681,10 +4681,9 @@ additions.  Please help us keeping it as
 Paul Eggert notes that: ISO C 1999 says that @file{inttypes.h} includes
 @file{stdint.h}, so there's no need to include @file{stdint.h}
 separately in a standard environment.  Many implementations have
address@hidden but not @file{stdint.h} (e.g., Solaris 7), but I don't
address@hidden but not @file{stdint.h} (e.g., Solaris 7), but we don't
 know of any implementation that has @file{stdint.h} but not
address@hidden  Nor do I know of any free software that includes
address@hidden; @file{stdint.h} seems to be a creation of the committee.
address@hidden
 
 @item @file{linux/irda.h}
 @hdrindex{linux/irda.h}
@@ -5677,9 +5676,9 @@ b.o} solves the issue.
 
 @item Don't rely on correct @code{#line} support
 On Solaris 8, @command{c89} (Sun WorkShop 6 update 2 C 5.3 Patch
-111679-08 2002/05/09)) rejects @code{#line} directives whose line
-numbers are greater than 32767.  In addition, nothing in @sc{posix}
-makes this invalid.  That is the reason why Autoconf stopped issuing
+111679-08 2002/05/09)) diagnoses @code{#line} directives whose line
+numbers are greater than 32767.  In addition, nothing in Posix
+makes this invalid.  That is why Autoconf stopped issuing
 @code{#line} directives.
 @end table
 
@@ -9909,11 +9908,11 @@ bash-2.04$ @kbd{cat <<EOF
 
 Many older shells (including the Bourne shell) implement here-documents
 inefficiently.  And some shells mishandle large here-documents: for
-example, Solaris 8 @command{dtksh}, which is derived from
-Korn shell version M-12/28/93d, mishandles variable expansion that occurs
-on 1024-byte buffer boundaries within a here-document.  Users can
-generally fix these problems by using a faster or more reliable
-shell, e.g., by using the command @samp{bash ./configure} rather than
+example, Solaris @command{dtksh}, which is derived from Korn shell
+version M-12/28/93d, mishandles variable expansion that occurs on
+1024-byte buffer boundaries within a here-document.  Users can generally
+fix these problems by using a faster or more reliable shell, e.g., by
+using the command @samp{bash ./configure CONFIG_SHELL=bash} rather than
 plain @samp{./configure}.
 
 Some shells can be extremely inefficient when there are a lot of
@@ -10277,9 +10276,9 @@ of a ``bad substitution''.
 
 @sp 1
 
-The Solaris @command{sh} has a frightening bug in its interpretation
+Solaris @command{/bin/sh} has a frightening bug in its interpretation
 of this.  Imagine you need set a variable to a string containing
address@hidden@}}.  This @address@hidden character confuses the Solaris 
@command{sh}
address@hidden@}}.  This @address@hidden character confuses Solaris 
@command{/bin/sh}
 when the affected variable was already set.  This bug can be exercised
 by running:
 
@@ -10404,7 +10403,7 @@ releases of Solaris don't support it:
 
 @example
 $ @kbd{showrev -c /bin/sh | grep version}
-Command version: SunOS 5.9 Generic May 2002
+Command version: SunOS 5.10 Generic January 2005
 $ @kbd{echo $(echo blah)}
 syntax error: `(' unexpected
 @end example
@@ -10433,7 +10432,8 @@ confusion, insert a space between the tw
 
 When setting several variables in a row, be aware that the order of the
 evaluation is undefined.  For instance @samp{foo=1 foo=2; echo $foo}
-gives @samp{1} with sh on Solaris, but @samp{2} with Bash.  You must use
+gives @samp{1} with Solaris @command{/bin/sh}, but @samp{2} with Bash.
+You must use
 @samp{;} to enforce the order: @samp{foo=1; foo=2; echo $foo}.
 
 Don't rely on the following to find @file{subdir/program}:
@@ -10876,7 +10876,7 @@ directory, you must use @samp{. ./foo}.
 @prindex @command{!}
 The Unix version 7 shell did not support
 negating the exit status of commands with @command{!}, and this feature
-is still absent from more modern shells (e.g., Solaris 9 @command{sh}).
+is still absent from more modern shells (e.g., Solaris @command{/bin/sh}).
 Shell code like this:
 
 @example
@@ -10948,7 +10948,8 @@ $ @kbd{scanset='[<]'; case '\' in $scans
 $
 @end example
 
-Even with this, SunOS 5.7 ksh matches a backslash if the set contains any
+Even with this, Solaris @command{ksh} matches a backslash if the set
+contains any
 of the characters @samp{|}, @samp{&}, @samp{(}, or @samp{)}.
 
 Conversely, Tru64 @command{ksh} (circa 2003) erroneously always matches
@@ -10971,7 +10972,7 @@ ash-0.3.8 $ @kbd{case foo in esac;}
 
 Many shells still do not support parenthesized cases, which is a pity
 for those of us using tools that rely on balanced parentheses.  For
-instance, Solaris 8's Bourne shell:
+instance, Solaris @command{/bin/sh}:
 
 @example
 $ @kbd{case foo in (foo) echo foo;; esac}
@@ -11016,7 +11017,7 @@ etc.@: for a means to simulate @option{-
 Do not use backslashes in the arguments, as there is no consensus on
 their handling.  On @samp{echo '\n' | wc -l}, the @command{sh} of
 Digital Unix 4.0 and @acronym{MIPS RISC/OS} 4.52, answer 2, but the Solaris
address@hidden, Bash, and Zsh (in @command{sh} emulation mode) report 1.
address@hidden/bin/sh}, Bash, and Zsh (in @command{sh} emulation mode) report 1.
 Please note that the problem is truly @command{echo}: all the shells
 understand @samp{'\n'} as the string composed of a backslash and an
 @samp{n}.
@@ -11057,7 +11058,7 @@ trap to clean up before exiting.  If the
 nonzero status, the trap also exits with nonzero status so that the
 invoker can tell that an error occurred.
 
-Unfortunately, in some shells, such as Solaris 8 @command{sh}, an exit
+Unfortunately, in some shells, such as Solaris @command{/bin/sh}, an exit
 trap ignores the @code{exit} command's argument.  In these shells, a trap
 cannot determine whether it was invoked by plain @code{exit} or by
 @code{exit 1}.  Instead of calling @code{exit} directly, use the
@@ -11073,7 +11074,8 @@ of the environment variables.  Conversel
 received by the shell when it is launched should be imported as a shell
 variable marked as exported.
 
-Alas, many shells, such as Solaris 2.5, @sc{irix} 6.3, @sc{irix} 5.2,
+Alas, many shells, such as Solaris @command{/bin/sh},
address@hidden 6.3, @sc{irix} 5.2,
 @acronym{AIX} 4.1.5, and Digital Unix 4.0, forget to
 @command{export} the environment variables they receive.  As a result,
 two variables coexist: the environment variable and the shell
@@ -11099,8 +11101,8 @@ that you update.
 @item @command{false}
 @c ------------------
 @prindex @command{false}
-Don't expect @command{false} to exit with status 1: in the native Bourne
-shell of Solaris 8 it exits with status 255.
+Don't expect @command{false} to exit with status 1: in native
+Solaris it exits with status 255.
 
 
 @item @command{for}
@@ -11199,7 +11201,7 @@ printf %s -foo
 @item @command{read}
 @c ------------------
 @prindex @command{read}
-Not all shells support @option{-r} (Solaris 9 @command{sh} for example).
+Not all shells support @option{-r} (Solaris @command{/bin/sh} for example).
 
 
 @item @command{pwd}
@@ -11315,11 +11317,11 @@ shouldn't do anything that tests feature
 the host system.  But occasionally you may find it necessary to check
 whether some arbitrary file exists.  To do so, use @samp{test -f} or
 @samp{test -r}.  Do not use @samp{test -x}, because address@hidden does not
-have it.  Do not use @samp{test -e} either, because Solaris 2.5 does not
-have it.  To test for symbolic links on systems that have them, use
+have it.  Do not use @samp{test -e} either, because Solaris @command{/bin/sh}
+lacks it.  To test for symbolic links on systems that have them, use
 @samp{test -h} rather than @samp{test -L}; either form conforms to
 Posix 1003.1-2001, but older shells like Solaris 8
address@hidden support only @option{-h}.
address@hidden/bin/sh} support only @option{-h}.
 
 @item @command{test} (strings)
 @c ---------------------------
@@ -11329,7 +11331,7 @@ option (e.g., @address@hidden = "-n"}
 
 Contrary to a common belief, @samp{test -n @var{string}} and
 @samp{test -z @var{string}} @strong{are} portable.  Nevertheless many
-shells (such as Solaris 2.5, @acronym{AIX} 3.2, @sc{unicos} 10.0.0.6,
+shells (such as Solaris, @acronym{AIX} 3.2, @sc{unicos} 10.0.0.6,
 Digital Unix 4, etc.)@: have bizarre precedence and may be confused if
 @var{string} looks like an operator:
 
@@ -11409,7 +11411,7 @@ the @emph{last} command run: that before
 @command{exit} itself?''
 
 Bash considers @command{exit} to be the last command, while Zsh and
-Solaris 8 @command{sh} consider that when the trap is run it is
+Solaris @command{/bin/sh} consider that when the trap is run it is
 @emph{still} in the @command{exit}, hence it is the previous exit status
 that the trap receives:
 
@@ -11491,11 +11493,11 @@ The small set of tools you can expect to
 include some limitations you should be aware of.
 
 @table @asis
address@hidden @command{awk}
address@hidden Awk
 @c ----------------
address@hidden @command{awk}
-Don't leave white space before the opening parenthesis in a user function call;
address@hidden awk will reject it:
address@hidden Awk
+Don't leave white space before the opening parenthesis in a user function call.
+Posix does not allow this and @acronym{GNU} Awk rejects it:
 
 @example
 $ @kbd{gawk 'function die () @{ print "Aaaaarg!"  @}
@@ -11526,8 +11528,7 @@ bar
 foo
 @end example
 
-Some AWK, such as HPUX 11.0's native one, have regex engines fragile to
-inner anchors:
+Some Awk implementations, such as HPUX 11.0's native one, mishandle anchors:
 
 @example
 $ @kbd{echo xfoo | $AWK '/foo|^bar/ @{ print @}'}
@@ -11541,20 +11542,52 @@ bar
 
 @noindent
 Either do not depend on such patterns (i.e., use @samp{/^(.*foo|bar)/},
-or use a simple test to reject such address@hidden
+or use a simple test to reject such implementations.
 
address@hidden version 5.2 limits on the maximal length of regular
-expressions and literal strings in the program to 399.
address@hidden version 5.2 has an arbitrary limit of 399 on the the
+length of regular expressions and literal strings in an Awk program.
 
-Some @command{awk}, such as Solaris 9's native one, have a limit of 99
+Traditional Awk implementations derived from Unix version 7, such as
+Solaris @command{/bin/awk}, have many limitations and do not
+conform to Posix.  Nowadays @code{AC_PROG_AWK} (@pxref{Particular
+Programs}) will find you an Awk that doesn't have these problems, but if
+for some reason you prefer not to use @code{AC_PROG_AWK} you may need to
+address them.
+
+Traditional Awk does not support multidimensional arrays or user-defined
+functions.
+
+Traditional Awk does not support the @option{-v} option.  You can use
+assignments after the program instead, e.g., @command{$AWK '@{print v
address@hidden' v=x}; however, don't forget that such assignments are not
+evaluated until they are encountered (e.g., after any @code{BEGIN}
+action).
+
+Traditional Awk does not support the keywords @code{delete} or @code{do}.
+
+Traditional Awk does not support the expressions
address@hidden@address@hidden:@var{c}}, @address@hidden, 
@address@hidden@var{b}},
+or @address@hidden@var{b}}.
+
+Traditional Awk does not support the predefined @code{CONVFMT} variable.
+
+Traditional Awk supports only the predefined functions @code{exp},
address@hidden, @code{length}, @code{log}, @code{split}, @code{sprintf},
address@hidden, and @code{substr}.
+
+Traditional Awk @code{getline} is not at all compatible with Posix;
+avoid it.
+
+Traditional Awk @code{split} supports only two arguments.
+
+Traditional Awk has a limit of 99
 fields in a record.  You may be able to circumvent this problem by using
-the builtin function @code{split}.
address@hidden
 
 @item @command{cat}
 @c ----------------
 @prindex @command{cat}
-Don't rely on any option.  The option @option{-v}, which displays
-non-printing characters, @emph{seems} portable, though.
+Don't rely on any option.
 
 
 @item @command{cc}
@@ -11621,7 +11654,7 @@ newline encoding.
 @prindex @command{cp}
 Avoid the @option{-r} option, since its behavior is not specified by
 Posix.  Use @option{-R} instead.  On @acronym{GNU} hosts the two options
-are equivalent, but on Solaris 9 hosts (for example) @command{cp -r}
+are equivalent, but on Solaris hosts (for example) @command{cp -r}
 reads from pipes instead of replicating them.
 
 Some @command{cp} implementations (e.g., BSD/OS 4.2) do not allow
@@ -11631,7 +11664,7 @@ avoid this problem, omit the trailing sl
 /tmp/newdir/} if @file{/tmp/newdir} does not exist.
 
 @c This is thanks to Ian.
-SunOS @command{cp} does not support @option{-f}, although its
+SunOS 4 @command{cp} does not support @option{-f}, although its
 @command{mv} does.  It's possible to deduce why @command{mv} and
 @command{cp} are different with respect to @option{-f}.  @command{mv}
 prompts by default before overwriting a read-only file.  @command{cp}
@@ -11657,7 +11690,7 @@ the system calls that @command{cp} uses;
 @command{cp} implementations use @code{utimes}, which has
 1-microsecond resolution.  These newer implementations include GNU
 coreutils 5.0.91 or later, and Solaris 8 (sparc) patch 109933-02 or
-later.  Unfortunately as of September 2003 there is still no system
+later.  Unfortunately as of June 2005 there is still no system
 call to set time stamps to the full nanosecond resolution.
 
 Bob Proulx notes that @samp{cp -p} always @emph{tries} to copy
@@ -11884,7 +11917,7 @@ not work on long input lines.  To work a
 @c -----------------
 @prindex @command{find}
 The option @option{-maxdepth} seems to be @acronym{GNU} specific.
-Tru64 v5.1, address@hidden 1.5 and Solaris 2.5 @command{find}
+Tru64 v5.1, address@hidden 1.5 and Solaris @command{find}
 commands do not understand it.
 
 The replacement of @address@hidden@}} is guaranteed only if the argument is
@@ -11905,7 +11938,7 @@ while @acronym{GNU} @command{find} repor
 @c -----------------
 @prindex @command{grep}
 Don't use @samp{grep -q} to suppress output, because many @command{grep}
-implementations (e.g., Solaris 10) do not support @option{-q}.
+implementations (e.g., Solaris) do not support @option{-q}.
 Don't use @samp{grep -s} to suppress output either, because Posix
 says @option{-s} does not suppress output, only some error messages;
 also, the @option{-s} option of traditional @command{grep} behaved
@@ -11916,8 +11949,8 @@ status of @code{grep} to determine wheth
 
 Some traditional @command{grep} implementations do not work on long
 input lines.  Also, many implementations do not support multiple regexps
-with @option{-e}: they either reject @samp{-e} entirely (e.g., Solaris
-10) or honor only the last pattern (e.g., @acronym{IRIX} 6.5).  To
+with @option{-e}: they either reject @samp{-e} entirely (e.g., Solaris)
+or honor only the last pattern (e.g., @acronym{IRIX} 6.5).  To
 work around these problems, invoke @code{AC_PROG_GREP} and then use
 @code{$GREP}.
 
@@ -12005,14 +12038,14 @@ None of @command{mkdir}'s options are po
 
 Posix does not clearly specify whether @samp{mkdir -p foo}
 should succeed when @file{foo} is a symbolic link to an already-existing
-directory.  GNU Coreutils 5.1.0 @command{mkdir} succeeds, but Solaris 9
+directory.  GNU Coreutils 5.1.0 @command{mkdir} succeeds, but Solaris
 @command{mkdir} fails.
 
 Not all @code{mkdir -p} implementations are thread-safe.  When it is not
 and you call @code{mkdir -p a/b} and @code{mkdir -p a/c} at the same
 time, both will detect that @file{a/} is missing, one will create
 @file{a/}, then the other will try to create @file{a/} and die with a
address@hidden exists} error.  At least Solaris 8, NetBSD 1.6, and OpenBSD
address@hidden exists} error.  At least Solaris 10, NetBSD 1.6, and OpenBSD
 3.4 have an unsafe @code{mkdir -p}.  GNU Coreutils (since Fileutils
 version 4.0c), FreeBSD 5.0, and NetBSD-current are known to have a
 race-free @code{mkdir -p}.  This possible race is harmful in parallel
@@ -12132,7 +12165,7 @@ Input should have reasonably long lines,
 an input buffer limited to 4000 bytes.
 
 Alternation, @samp{\|}, is common but Posix does not require its
-support, so it should be avoided in portable scripts.  Solaris 8
+support, so it should be avoided in portable scripts.  Solaris
 @command{sed} does not support alternation; e.g., @samp{sed '/a\|b/d'}
 deletes only lines that contain the literal string @samp{a|b}.
 
@@ -12142,7 +12175,7 @@ Nested parenthesization in patterns (e.g
 quite portable to modern hosts, but is not supported by some older
 @command{sed} implementations like SVR3.
 
-Some @command{sed} implementations, e.g., Solaris 10,
+Some @command{sed} implementations, e.g., Solaris,
 restrict the special role of the asterisk to one-character regular expressions.
 This may lead to unexpected behavior:
 
@@ -12193,7 +12226,7 @@ supported it.
 
 Posix requires that you must not have any white space between
 @samp{!} and the following command.  It is OK to have blanks between
-the address and the @samp{!}.  For instance, on Solaris 8:
+the address and the @samp{!}.  For instance, on Solaris:
 
 @example
 $ @kbd{echo "foo" | sed -n '/bar/ ! p'}
@@ -12312,7 +12345,7 @@ are executed by the shell, all its weakn
 @item @code{$<}
 Posix says that the @samp{$<} construct in makefiles can be
 used only in inference rules and in the @samp{.DEFAULT} rule; its
-meaning in ordinary rules is unspecified.  Solaris 8's @command{make}
+meaning in ordinary rules is unspecified.  Solaris @command{make}
 for instance will replace it with the empty string.  OpenBSD (3.0 and
 later) @command{make} will diagnose these uses and error out.
 
@@ -12644,7 +12677,7 @@ f.c: if.c
 @cindex @code{VPATH} and automatic rule rewriting
 @cindex automatic rule rewriting and @code{VPATH}
 
-Some @command{make} implementations, such as SunOS @command{make} or
+Some @command{make} implementations, such as SunOS 4 @command{make} or
 OSF1/Tru64 @command{make}, will search prerequisites in @code{VPATH} and
 rewrite all their occurrences in the rule appropriately.
 
@@ -12671,7 +12704,7 @@ f.c: if.c
 
 @noindent
 However the "prerequisite rewriting" still applies here.  So if
address@hidden is in @file{../pkg/src}, SunOS @command{make} and OSF1/Tru64
address@hidden is in @file{../pkg/src}, SunOS 4 @command{make} and OSF1/Tru64
 @command{make} will execute
 
 @smallexample
@@ -12714,7 +12747,7 @@ install-HEADERS: $(HEADERS)
         done
 @end example
 
-The above @code{install-HEADERS} rule is not SunOS-proof because @code{for
+The above @code{install-HEADERS} rule is not SunOS-4-proof because @code{for
 i in $(HEADERS);} will be expanded as @code{for i in f.h g.h h.h;}
 where @code{f.h} and @code{g.h} are plain words and are hence
 subject to @code{VPATH} adjustments.
@@ -12819,7 +12852,7 @@ Dependency @file{../foo} was found in @f
 directory does not even exist, this just means that the simplification
 occurred before the file was checked for.)
 
-For the record here is how SunOS @command{make} behaves on this
+For the record here is how SunOS 4 @command{make} behaves on this
 very same example.
 @smallexample
 % make