[sr #111381] autoconf-2.72.90 tests fail with traditional yacc, lex

--8323329-1714636915-1770178144=:1163927
Content-Type: TEXT/plain; CHARSET=utf-8
Content-Transfer-Encoding: QUOTED-PRINTABLE
Content-Disposition: inline

URL:=0A  <https://savannah.gnu.org/support/?111381>=0A=0A                 S=
ummary: autoconf-2.72.90 tests fail with traditional yacc,=0Alex=0A        =
           Group: Autoconf=0A               Submitter: eggert=0A           =
    Submitted: Tue 03 Feb 2026 08:08:58 PM PST=0A                Priority: =
5 - Unprioritized=0A                Severity: 2 - Minor=0A                 =
 Status: None=0A                 Privacy: Public=0A             Assigned to=
: None=0A        Originator Email:=0A             Open/Closed: Open=0A     =
    Discussion Lock: Unlocked=0A        Operating System: Solaris=0A=0A=0A =
   _______________________________________________________=0A=0AFollow-up C=
omments:=0A=0A=0A-------------------------------------------------------=0A=
Date: Tue 03 Feb 2026 08:08:58 PM PST By: Paul Eggert <eggert>=0AWhen I ran=
 "make check" with autoconf-2.72.90 on Solaris 10 sparc (where the=0Aonly G=
NU tool is GNU m4 1.4.19 which I built myself), the following tests=0Afail:=
=0A=0A453: AC_PROG_LEX with noyywrap                       FAILED=0A(semant=
ics.at:1306)=0A454: AC_PROG_LEX with yywrap                         FAILED=
=0A(semantics.at:1308)=0A517: AC_DECL_YYTEXT                               =
   FAILED=0A(acprograms.at:22)=0A=0APresumably this is because the tests ar=
e using /usr/ccs/bin/yacc and=0A/usr/ccs/bin/lex instead of GNU Bison and f=
lex. "lex -V" and "yacc -Qy" both=0Asay they are "Software Generation Utili=
ties (SGU) Solaris-ELF (4.0)"; the=0Aexecutables are both dated 2005-01-22.=
=0A=0AI'll attach a compressed tarball of the failed test output files.=0A=
=0AA search through the archives reports that=0A[https://lists.nongnu.org/a=
rchive/html/autoconf/2023-12/msg00000.html Zack was=0Aworking on fixing thi=
s for Autoconf in 2023], but I don't know what happened=0Awith that. A brie=
f look at the Autoconf bug tracker found sr #110266, which=0Amay be the sam=
e bug as this one.=0A=0ATo avoid further noise about this perhaps we should=
 remove the test cases? Or=0Aat least remove them if they use traditional y=
acc and lex?=0A=0A=0A=0A=0A=0A=0A=0A    ___________________________________=
____________________=0A=0AReply to this item at:=0A=0A  <https://savannah.g=
nu.org/support/?111381>=0A=0A______________________________________________=
_=0AMessage sent via Savannah=0Ahttps://savannah.gnu.org/=0A
--8323329-1714636915-1770178144=:1163927
Content-Type: APPLICATION/pgp-signature; name=signature.asc

-----BEGIN PGP SIGNATURE-----

iHUEABYIAB0WIQQk97aszIMMAvLLwm6qLAuaBUf3TgUCaYLGYAAKCRCqLAuaBUf3
Tq/SAQDW94BTdLpwGgYMfEbxm6kb48NCQ9IEoTMoPyQh2PSr2AEA7OZb9XtzOkYj
W80KisfpviMGGaiXYj4vNpX6aMgq7A4=
=H2PM
-----END PGP SIGNATURE-----

--8323329-1714636915-1770178144=:1163927--