Re: Configuring GCC 10.3 on PPC Mac OS X 10.4.11/Tiger for build reveals flaws in configure scripts

[Jonathan Wakely <jwakely.gcc@gmail.com>]

On Fri, 25 Nov 2022 at 10:20, Peter Dyballa via Gcc-help
<gcc-help@gcc.gnu.org> wrote:
>
> Hello!
>
> I am just showing the faults and the code causing them, no cure, when building with MacPorts:
>
> ----------------------
> Configuring stage 1 in ./isl
>
> ..../gcc10-bootstrap/work/gcc-10.3.0/isl/configure: line 1: cd: /opt/local/libexec/gcc10-bootstrap: No such file or directory
> checking for powerpc-apple-darwin8.11.0-gcc...
>
> checking whether we are using the GNU C compiler... (cached) yes
> ..../gcc10-bootstrap/work/gcc-10.3.0/isl/configure: line 5363: test: =: unary operator expected

isl is not part of GCC, and I don't see this error in the version of
isl that gcc-10.3.0/contrib/download_prerequisites uses.

Using isl is optional anyway, you can just leave it out if you want.
Either way, this seems to be a problem with the version of isl that
MacPorts uses, not a GCC problem.

> Configuring stage 1 in ./libbacktrace
>
> checking for objcopy... no
> checking for readelf... readelf
> checking whether objcopy supports debuglink... ..../gcc10-bootstrap/work/gcc-10.3.0/libbacktrace/configure: line 13557: --add-gnu-debuglink=x: command not found

This line number doesn't match what's shown below, as it's coming from
line 13560 below (which is line 13496 in my copy of the gcc-10.30
sources, so maybe MacPorts patched it?)

> no
> checking whether tests can run... yes
>
> 13553   { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether objcopy supports debuglink" >&5
> 13554   $as_echo_n "checking whether objcopy supports debuglink... " >&6; }
> 13555   if ${libbacktrace_cv_objcopy_debuglink+:} false; then :
> 13556     $as_echo_n "(cached) " >&6
> 13557   else
> 13558     if test -n "${with_target_subdir}"; then
> 13559     libbacktrace_cv_objcopy_debuglink=no
> 13560   elif ${OBJCOPY} --add-gnu-debuglink=x /bin/ls /tmp/ls$$; then
> 13561     rm -f /tmp/ls$$
> 13562     libbacktrace_cv_objcopy_debuglink=yes
> 13563   else
> 13564     libbacktrace_cv_objcopy_debuglink=no
> 13565   fi
> 13566   fi
> 13567   { $as_echo "$as_me:${as_lineno-$LINENO}: result: $libbacktrace_cv_objcopy_debuglink" >&5
> 13568   $as_echo "$libbacktrace_cv_objcopy_debuglink" >&6; }
> 13569    if test "$libbacktrace_cv_objcopy_debuglink" = yes; then
> 13570     HAVE_OBJCOPY_DEBUGLINK_TRUE=
> 13571     HAVE_OBJCOPY_DEBUGLINK_FALSE='#'
> 13572   else
> 13573     HAVE_OBJCOPY_DEBUGLINK_TRUE='#'
> 13574     HAVE_OBJCOPY_DEBUGLINK_FALSE=
> 13575   fi
> ----------------------
>
> Configuring stage 2 in ./libbacktrace
>
> checking for readelf... readelf
> checking whether objcopy supports debuglink... ..../gcc10-bootstrap/work/gcc-10.3.0/libbacktrace/configure: line 13557: --add-gnu-debuglink=x: command not found
> no
> checking whether tests can run... yes
> ==================================
>
>
> Configuring stage 2 in powerpc-apple-darwin8.11.0/libstdc++-v3
>
> checking for uchar.h... no
> no

This is harmless, but I see the problem and will fix it.

> checking for int64_t... yes

> Configuring stage 2 in ./gcc
>
> checking linker --sysroot support... no
> checking __stack_chk_fail in target C library... checking for __stack_chk_fail... no
> no __stack_chk_fail on this target
> no
> checking sys/sdt.h in the target C library... no


I'll see if I can find this one too. Also harmless though.