Spurious failures of some silent*.test tests when cc != gcc (was: Testsuite failures on Tru64 OSF 5.1)

[Stefano Lattarini]

On Sunday 14 November 2010, Stefano Lattarini wrote:
> 
> > FAIL: silentlex.test (exit: 1)
> > ==============================
> > 
> > /tmp/am/build-alphaev67-dec-osf5.1/tests:/tmp/local/alphaev67-dec-osf5.1/bin
> >   :/tmp/bin:/bin:/usr/bin:/sbin:/usr/sbin:/usr/opt/svr4/bin:/etc:/usr/etc
> >   :/usr/dt/bin:/usr/bin/X11:/usr/ucb
> > silentlex: running flex --version
> > flex version 2.5.3
> > + pwd
> > /tmp/am/build-alphaev67-dec-osf5.1/tests/silentlex.dir
> > + set -e
> > + mkdir sub
> > + cat
> > + cat
> > + cat
> > + cat
> > + cp foo.l sub/bar.l
> > + aclocal-1.11a -Werror
> > + automake-1.11a --foreign -Werror -Wall --add-missing
> > configure.in:5: installing `./compile'
> > configure.in: installing `./ylwrap'
> > + autoconf
> > + ./configure --enable-silent-rules
>
> [MEGA-CUT]
>
> > + rm -f foo.c sub/bar.c
> > + ./configure am_cv_CC_dependencies_compiler_type=gcc --enable-silent-rules
>
> OK, so here I force gcc depmode ...
>
> > checking for a BSD-compatible install... ./install-sh -c
> > checking whether build environment is sane... yes
> > checking for a thread-safe mkdir -p... ./install-sh -c -d
> > checking for gawk... no
> > checking for mawk... no
> > checking for nawk... nawk
> > checking whether make sets $(MAKE)... yes
> > checking for style of include used by make... GNU
> > checking for gcc... no
> > checking for cc... cc
> > checking for C compiler default output file name... a.out
> > checking whether the C compiler works... yes
> > checking whether we are cross compiling... no
> > checking for suffix of executables...
> > checking for suffix of object files... o
> > checking whether we are using the GNU C compiler... no
> > checking whether cc accepts -g... yes
> > checking for cc option to accept ISO C89... none needed
> > checking dependency style of cc... (cached) gcc
> > checking whether cc understands -c and -o together... yes
> > checking for flex... flex
> > checking lex output file root... lex.yy
> > checking lex library... -ll
> > checking whether yytext is a pointer... yes
> > checking that generated files are newer than configure... done
> > configure: creating ./config.status
> > config.status: creating Makefile
> > config.status: creating sub/Makefile
> > config.status: executing depfiles commands
> > + make
> > cc: Severe: No such file or directory
> > ... file is '.deps/bar.TPo'
>
> ... but the C compiler is clearly not gcc-compatible in this respect.
>
Oh, and I'm seeing a similar failure with the SunStudio C compiler!  So
my explanation is probably correct, and I can write and test a patch even
without access to a Tru64/OSF system.
 
> My bad I guess.  The solution might be splitting silentlex.test in two
> sister tests:
>  - silentlex-generic.test, trying only default depmode and disabled
>    depmode; and
>  - silentlex-gcc.test, trying trying gcc depmode and having `gcc' in
>    $required.
> 
> And maybe something similar should be done for silentyacc.test too.
>
*And* for silent5.test.

I'll try to cook up a patch once my patches "Fix spurious Lex/Yacc-related
failures in tests `silent*.test' w.r.t. Solaris XPG4 make.":
 <http://lists.gnu.org/archive/html/automake-patches/2010-11/msg00221.html>
have been applied (to spare myself unnecessary conflicts).

Regards,
   Stefano