[Clfs-support] Locale errors during SVN pure64 glibc check

[Mark McCorkell]

I'm in the process of doing a CLFS pure64 build as something to
experiment with. During the glibc 'make check' inside a chroot jail in
section 10.7 I'm getting 14 test fails in the localedata section. The
only other fail is the expected annexc fail.

Google returned a suggestion of running

cp ../glibc-x.xx/iconvdata/gconv-modules iconvdata

before doing 'make check' but there's no mention of this being needed in
the current book. I haven't tried this yet.

Architecture          : x86_64
Host Operating System : debian sid-amd64
CLFS Book             : Version SVN-20090409-x86_64-Pure64
Section               : 10.7. Glibc-2.9
Command               : make -k check >glibc-check-log 2>&1 ; grep Error glibc-check-log
Error Message         : 
make[2]: *** [/sources/glibc-build/localedata/tst_mblen.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/tst_mbrlen.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/tst_mbrtowc.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/tst_mbsrtowcs.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/tst_mbstowcs.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/tst_mbtowc.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/tst_swscanf.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/tst_wcrtomb.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/tst_wcsrtombs.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/tst_wcstombs.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/tst_wctob.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/tst_wctomb.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/bug-iconv-trans.out] Error 1
make[2]: *** [/sources/glibc-build/localedata/tst-wctype.out] Error 1
make[1]: *** [localedata/tests] Error 2
make[2]: [/sources/glibc-build/posix/annexc.out] Error 1 (ignored)
make: *** [check] Error 2

Example of one of the localedata test failures: 

localedata/tst_mblen.out

mblen:de_DE.ISO-8859-1:1:1:2:F:the return value is different from an
expected value
mblen:de_DE.ISO-8859-1:2:1:2:S:PASSED
mblen:de_DE.ISO-8859-1:3:1:2:S:PASSED
mblen:de_DE.ISO-8859-1:4:1:2:S:PASSED
mblen:de_DE.ISO-8859-1:5:1:2:S:PASSED
mblen:de_DE.ISO-8859-1:6:1:3:S:PASSED
mblen:en_US.ANSI_X3.4-1968:1:1:2:S:PASSED
mblen:en_US.ANSI_X3.4-1968:2:1:2:S:PASSED
mblen:en_US.ANSI_X3.4-1968:3:1:2:S:PASSED
mblen:en_US.ANSI_X3.4-1968:4:1:2:S:PASSED
mblen:en_US.ANSI_X3.4-1968:5:1:2:S:PASSED
mblen:en_US.ANSI_X3.4-1968:6:1:3:S:PASSED
mblen:ja_JP.EUC-JP:1:1:2:F:the return value is different from an
expected value
mblen:ja_JP.EUC-JP:2:1:2:F:the return value is different from an
expected value
mblen:ja_JP.EUC-JP:3:1:2:F:the return value is different from an
expected value
mblen:ja_JP.EUC-JP:4:1:1:S:PASSED
mblen:ja_JP.EUC-JP:4:1:2:S:PASSED
mblen:ja_JP.EUC-JP:5:1:2:S:PASSED
mblen:ja_JP.EUC-JP:6:1:3:S:PASSED

There are 13 more failures, all with the same errors for the
de_DE.ISO-8859-1 and ja_JP.EUC-JP locales followed by a failure in
localedata/bug-iconv-trans.out stating 'iconv_open failed'.

Have I missed something earlier in the build that could have prevented
these, possibly something on my host (see below)? I think my host
binutils/glibc are slightly newer than is recommended.

I don't plan on using either of these locales in the final system, so
could I carry on regardless or is this an indicator of something bigger
that might cause issues if left unresolved? Any advice/suggestions would
be appreciated.

Host configuration:

sh version-check.sh

Binutils: (GNU Binutils for Debian) 2.19.1
bison (GNU Bison) 2.4.1
bzip2,  Version 1.0.5, 10-Dec-2007.
Coreutils:  7.2
diff (GNU diffutils) 2.8.1
find (GNU findutils) 4.4.0
GNU Awk 3.1.5
gcc (Debian 4.3.3-5) 4.3.3
GNU C Library stable release version 2.9,
GNU grep 2.5.4
gzip 1.3.12
Linux 2.6.29-1-amd64
GNU Make 3.81
ncurses 5.7.20090404
patch 2.5.9
GNU sed version 4.1.5
tar (GNU tar) 1.22
makeinfo (GNU texinfo) 4.13

-- 
Mark McCorkell <markmccorkell at tiscali.co.uk>