[lfs-support] gcc-8.2.0 (8.3-systemd chapter 6.21 on x86_64)

[Hans Malissa]

Hi all,

Once again, I'm confused by the output from a test suite, in this case it's gcc-8.2.0 in chapter 6.21 of 8.3-systemd on x86_64.
When I run the tests as described in the book, su nobody -s /bin/bash -c "PATH=$PATH make -k check" exits with:

[...]
# of expected passes 54
make[4]: Leaving directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic/testsuite'
make[3]: Leaving directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic/testsuite'
make[3]: Entering directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic'
true DO=all multi-do # make
make[3]: Leaving directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic'
make[2]: Leaving directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic'
make[1]: Leaving directory '/usr/src/gcc-8.2.0/build'
make: *** [Makefile:2235: do-check] Error 2
make: Target 'check' not remade because of errors.

But when I run ../contrib/test_summary I don't see any 'unexpected failures'. I don't see any failures at all, and no errors either. So I'm not quite sure which error make -k check is referring to. What went wrong here? I'm pasting the output of ../contrib/test_summary below.
Thanks a lot,

Hans

# ../contrib/test_summary

cat <<'EOF' |
LAST_UPDATED: Obtained from SVN: tags/gcc_8_2_0_release revision 262993

Native configuration is x86_64-pc-linux-gnu

=== g++ tests ===


Running target unix

=== g++ Summary ===

# of expected passes 125327
# of expected failures 504
# of unsupported tests 4931
/usr/src/gcc-8.2.0/build/gcc/xg++ version 8.2.0 (GCC)

=== gcc tests ===


Running target unix

=== gcc Summary ===

# of expected passes 130917
# of expected failures 393
# of unsupported tests 2107
/usr/src/gcc-8.2.0/build/gcc/xgcc version 8.2.0 (GCC)

=== libatomic tests ===


Running target unix

=== libatomic Summary ===

# of expected passes 54
=== libgomp tests ===


Running target unix

=== libgomp Summary ===

# of expected passes 1837
# of unsupported tests 192
=== libitm tests ===


Running target unix

=== libitm Summary ===

# of expected passes 42
# of expected failures 3
# of unsupported tests 1
=== libstdc++ tests ===


Running target unix

=== libstdc++ Summary ===

# of expected passes 12199
# of expected failures 71
# of unsupported tests 306

Compiler version: 8.2.0 (GCC)
Platform: x86_64-pc-linux-gnu
configure flags: --prefix=/usr --enable-languages=c,c++ --disable-multilib --disable-bootstrap --disable-libmpx --with-system-zlib
EOF
Mail -s "Results for 8.2.0 (GCC) testsuite on x86_64-pc-linux-gnu" gcc-***@gcc.gnu.org &&
mv /usr/src/gcc-8.2.0/build/./gcc/testsuite/g++/g++.sum /usr/src/gcc-8.2.0/build/./gcc/testsuite/g++/g++.sum.sent &&
mv /usr/src/gcc-8.2.0/build/./gcc/testsuite/gcc/gcc.sum /usr/src/gcc-8.2.0/build/./gcc/testsuite/gcc/gcc.sum.sent &&
mv /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libatomic/testsuite/libatomic.sum /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libatomic/testsuite/libatomic.sum.sent &&
mv /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libgomp/testsuite/libgomp.sum /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libgomp/testsuite/libgomp.sum.sent &&
mv /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libitm/testsuite/libitm.sum /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libitm/testsuite/libitm.sum.sent &&
mv /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libstdc++-v3/testsuite/libstdc++.sum /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libstdc++-v3/testsuite/libstdc++.sum.sent &&
mv /usr/src/gcc-8.2.0/build/./gcc/testsuite/g++/g++.log /usr/src/gcc-8.2.0/build/./gcc/testsuite/g++/g++.log.sent &&
mv /usr/src/gcc-8.2.0/build/./gcc/testsuite/gcc/gcc.log /usr/src/gcc-8.2.0/build/./gcc/testsuite/gcc/gcc.log.sent &&
mv /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libatomic/testsuite/libatomic.log /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libatomic/testsuite/libatomic.log.sent &&
mv /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libgomp/testsuite/libgomp.log /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libgomp/testsuite/libgomp.log.sent &&
mv /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libitm/testsuite/libitm.log /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libitm/testsuite/libitm.log.sent &&
mv /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libstdc++-v3/testsuite/libstdc++.log /usr/src/gcc-8.2.0/build/./x86_64-pc-linux-gnu/libstdc++-v3/testsuite/libstdc++.log.sent &&
true

[Ken Moffat]

Post by Hans Malissa
Hi all,
Once again, I'm confused by the output from a test suite, in this case it's gcc-8.2.0 in chapter 6.21 of 8.3-systemd on x86_64.
[...]
# of expected passes 54
make[4]: Leaving directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic/testsuite'
make[3]: Leaving directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic/testsuite'
make[3]: Entering directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic'
true DO=all multi-do # make
make[3]: Leaving directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic'
make[2]: Leaving directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic'
make[1]: Leaving directory '/usr/src/gcc-8.2.0/build'
make: *** [Makefile:2235: do-check] Error 2
make: Target 'check' not remade because of errors.
But when I run ../contrib/test_summary I don't see any 'unexpected failures'. I don't see any failures at all, and no errors either. So I'm not quite sure which error make -k check is referring to. What went wrong here? I'm pasting the output of ../contrib/test_summary below.
Thanks a lot,
Hans

(snipped test_summary, it is similar to what I got on an i3 skylake
although I had different totals for expected passes and unsupported
tests in both g++ and libstdc++.)

As always with testsuites, the detail can happen a lot earlier. My
run of make check on that machine was logged, and the first 12
lines, up to the point where it started to compile test progs, were:

make[2]: Nothing to be done for 'check'.
true DO=all multi-do # make
make btest stest ztest edtest ttest ctestg ctesta
make[2]: Nothing to be done for 'check'.
make[2]: Entering directory '/building/gcc-8.2.0/build/fixincludes'
autogen -T ../../fixincludes/check.tpl ../../fixincludes/inclhack.def
make[2]: autogen: Command not found
make[2]: *** [Makefile:176: check] Error 127
make[2]: Leaving directory '/building/gcc-8.2.0/build/fixincludes'
make[1]: *** [Makefile:3674: check-fixincludes] Error 2
make[2]: Nothing to be done for 'check'.
make[3]: Entering directory '/building/gcc-8.2.0/build/libbacktrace'

The Error 127 (no such file) translates becomes Error 2 in the
check-fixincludes target, and that (plus any other errors, if there
were any - not in your case) carries through to the final status.

ĸen
--
Is it about a bicycle ?
--
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Do not top post on this list.

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

h

[Hans Malissa]

Post by Ken Moffat

Post by Hans Malissa
Hi all,
Once again, I'm confused by the output from a test suite, in this case it's gcc-8.2.0 in chapter 6.21 of 8.3-systemd on x86_64.
[...]
# of expected passes 54
make[4]: Leaving directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic/testsuite'
make[3]: Leaving directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic/testsuite'
make[3]: Entering directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic'
true DO=all multi-do # make
make[3]: Leaving directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic'
make[2]: Leaving directory '/usr/src/gcc-8.2.0/build/x86_64-pc-linux-gnu/libatomic'
make[1]: Leaving directory '/usr/src/gcc-8.2.0/build'
make: *** [Makefile:2235: do-check] Error 2
make: Target 'check' not remade because of errors.
But when I run ../contrib/test_summary I don't see any 'unexpected failures'. I don't see any failures at all, and no errors either. So I'm not quite sure which error make -k check is referring to. What went wrong here? I'm pasting the output of ../contrib/test_summary below.
Thanks a lot,
Hans

(snipped test_summary, it is similar to what I got on an i3 skylake
although I had different totals for expected passes and unsupported
tests in both g++ and libstdc++.)
As always with testsuites, the detail can happen a lot earlier. My
run of make check on that machine was logged, and the first 12
make[2]: Nothing to be done for 'check'.
true DO=all multi-do # make
make btest stest ztest edtest ttest ctestg ctesta
make[2]: Nothing to be done for 'check'.
make[2]: Entering directory '/building/gcc-8.2.0/build/fixincludes'
autogen -T ../../fixincludes/check.tpl ../../fixincludes/inclhack.def
make[2]: autogen: Command not found
make[2]: *** [Makefile:176: check] Error 127
make[2]: Leaving directory '/building/gcc-8.2.0/build/fixincludes'
make[1]: *** [Makefile:3674: check-fixincludes] Error 2
make[2]: Nothing to be done for 'check'.
make[3]: Entering directory '/building/gcc-8.2.0/build/libbacktrace'
The Error 127 (no such file) translates becomes Error 2 in the
check-fixincludes target, and that (plus any other errors, if there
were any - not in your case) carries through to the final status.
Äžen

Okay, I redirected the output of make -k check to a file, and I do indeed get something similar, the same ‘Error 127’, right in the beginning:

make[1]: Entering directory '/usr/src/gcc-8.2.0/build'
make[2]: Entering directory '/usr/src/gcc-8.2.0/build/fixincludes'
autogen -T ../../fixincludes/check.tpl ../../fixincludes/inclhack.def
make[2]: autogen: Command not found
make[2]: *** [Makefile:176: check] Error 127
make[2]: Leaving directory '/usr/src/gcc-8.2.0/build/fixincludes'
make[1]: *** [Makefile:3674: check-fixincludes] Error 2
make[2]: Entering directory '/usr/src/gcc-8.2.0/build/gcc'
Making a new config file...
[
]

So that seems to be comparable to what you got.
Does that mean that I can go ahead and install gcc-8.2.0 or is something wrong?
Thanks a lot,

Hans

[Ken Moffat]

Post by Hans Malissa

Post by Ken Moffat
As always with testsuites, the detail can happen a lot earlier. My
run of make check on that machine was logged, and the first 12
make[2]: Nothing to be done for 'check'.
true DO=all multi-do # make
make btest stest ztest edtest ttest ctestg ctesta
make[2]: Nothing to be done for 'check'.
make[2]: Entering directory '/building/gcc-8.2.0/build/fixincludes'
autogen -T ../../fixincludes/check.tpl ../../fixincludes/inclhack.def
make[2]: autogen: Command not found
make[2]: *** [Makefile:176: check] Error 127
make[2]: Leaving directory '/building/gcc-8.2.0/build/fixincludes'
make[1]: *** [Makefile:3674: check-fixincludes] Error 2
make[2]: Nothing to be done for 'check'.
make[3]: Entering directory '/building/gcc-8.2.0/build/libbacktrace'
The Error 127 (no such file) translates becomes Error 2 in the
check-fixincludes target, and that (plus any other errors, if there
were any - not in your case) carries through to the final status.
ĸen

make[1]: Entering directory '/usr/src/gcc-8.2.0/build'
make[2]: Entering directory '/usr/src/gcc-8.2.0/build/fixincludes'
autogen -T ../../fixincludes/check.tpl ../../fixincludes/inclhack.def
make[2]: autogen: Command not found
make[2]: *** [Makefile:176: check] Error 127
make[2]: Leaving directory '/usr/src/gcc-8.2.0/build/fixincludes'
make[1]: *** [Makefile:3674: check-fixincludes] Error 2
So that seems to be comparable to what you got.
Does that mean that I can go ahead and install gcc-8.2.0 or is something wrong?

Yes, install it, nothing is wrong.

For more on autogen, see
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=29700
http://savannah.gnu.org/projects/autogen

There is no way that can be built at this early stage (it appears
to need libguile, libxml2, libregex).

ĸen
--
Is it about a bicycle ?
--
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Do not top post on this list.

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail

[Hans Malissa]

As always with testsuites, the detail can happen a lot earlier. My
run of make check on that machine was logged, and the first 12
make[2]: Nothing to be done for 'check'.
true DO=all multi-do # make
make btest stest ztest edtest ttest ctestg ctesta
make[2]: Nothing to be done for 'check'.
make[2]: Entering directory '/building/gcc-8.2.0/build/fixincludes'
autogen -T ../../fixincludes/check.tpl ../../fixincludes/inclhack.def
make[2]: autogen: Command not found
make[2]: *** [Makefile:176: check] Error 127
make[2]: Leaving directory '/building/gcc-8.2.0/build/fixincludes'
make[1]: *** [Makefile:3674: check-fixincludes] Error 2
make[2]: Nothing to be done for 'check'.
make[3]: Entering directory '/building/gcc-8.2.0/build/libbacktrace'
The Error 127 (no such file) translates becomes Error 2 in the
check-fixincludes target, and that (plus any other errors, if there
were any - not in your case) carries through to the final status.
Äžen

Okay, I redirected the output of make -k check to a file, and I do indeed get something similar, the same ‘Error 127’, right in the beginning:

make[1]: Entering directory '/usr/src/gcc-8.2.0/build'
make[2]: Entering directory '/usr/src/gcc-8.2.0/build/fixincludes'
autogen -T ../../fixincludes/check.tpl ../../fixincludes/inclhack.def
make[2]: autogen: Command not found
make[2]: *** [Makefile:176: check] Error 127
make[2]: Leaving directory '/usr/src/gcc-8.2.0/build/fixincludes'
make[1]: *** [Makefile:3674: check-fixincludes] Error 2


So that seems to be comparable to what you got.
Does that mean that I can go ahead and install gcc-8.2.0 or is something wrong?

Yes, install it, nothing is wrong.

For more on autogen, see
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=29700
http://savannah.gnu.org/projects/autogen

There is no way that can be built at this early stage (it appears
to need libguile, libxml2, libregex).

Äžen

Great, thanks for the explanation. This sounds to me as if everyone should get the same error, since nobody would have autogen at this stage.

Hans