On Fri, Apr 17, 2020 at 11:03:26AM +, Matthias Klose wrote:
> Package: src:juman
> Version: 7.0-3.4
> Severity: normal
> Tags: sid bullseye
> User: debian-...@lists.debian.org
> Usertags: ftbfs-gcc-10
>
> Please keep this issue open in the bug tracker for the package it
> was filed for. If a fix in another package is required, please
> file a bug for the other package (or clone), and add a block in this
> package. Please keep the issue open until the package can be built in
> a follow-up test rebuild.
>
> The package fails to build in a test rebuild on at least amd64 with
> gcc-10/g++-10, but succeeds to build with gcc-9/g++-9. The
> severity of this report will be raised before the bullseye release,
> so nothing has to be done for the buster release.
> Making all in dic
> make[4]: Entering directory '/<>/dic'
> gcc -E -P JUMAN.connect.c | LANG=C sed "s/\(\#pragma\)/\;\1/" > JUMAN.connect
> ../makemat/makemat
> JUMAN.kankei parsing... done.
>
> JUMAN.connect parsing...
> /<>/makemat/.libs/makemat: U5f62U5bb9U8a5e is
> undefined in JUMAN.grammar .
> exit(11)
> make[4]: *** [Makefile:532: jumandic.tab] Error 11
Somewhat surprisingly, this package now builds successfully on current
sid without any changes. It still fails on bullseye.
It looks like this is because configure.ac is missing AC_PROG_CPP, which was
tolerated by earlier versions of the toolchain but apparently not anymore.
This causes an error that gets ignored but the build continues after that.
Making all in dic
make[4]: Entering directory '/build/1st/juman-7.0/dic'
Makefile:524: warning: ignoring prerequisites on suffix rule definition
/bin/bash: line 1: CPP@: command not found
../makemat/makemat
JUMAN.kankei parsing... done.
JUMAN.connect parsing... (table size 1254) done.
I've tested that adding AC_PROG_CPP makes the build fail the same way as
earlier.
No idea if the current build result is broken or not so leaving the severity
as-is.
--
Niko Tyni nt...@debian.org
