I was reading
https://bugzilla.yoctoproject.org/show_bug.cgi?id=14096
https://bugzilla.yoctoproject.org/show_bug.cgi?id=13244
and your e-mail about the make issue in perl, but in my case it looks a bit
different I think.
It was with the make-native dependency, but happens only in
selftest test_ccache_tool:
2022-06-09 11:30:21,855 - oe-selftest - INFO - test_ccache_tool
(buildoptions.ImageOptionsTests.test_ccache_tool)
2022-06-09 12:53:41,855 - oe-selftest - INFO - Keepalive message
2022-06-09 14:17:01,856 - oe-selftest - INFO - Keepalive message
2022-06-09 15:40:21,856 - oe-selftest - INFO - Keepalive message
^^ was what triggered me to look where it got stuck.
I can build perl-native fine in my other oe-core nodistro build directory
and now I've built it ok also in regular poky build directory as in:
5f249a407f6 vim: put xxd in its own package
While the selftest was stuck I was checking where exactly it got stuck in
the WORKDIR and
oe_runmake -j 1 installman
'DESTDIR=/OE/build/poky/build-st/tmp/work/x86_64-linux/perl-native/5.36.0-r0/image'
install 2>&1 | tee installman.log
finished fine, while:
oe_runmake -j 2 installman
'DESTDIR=/OE/build/poky/build-st/tmp/work/x86_64-linux/perl-native/5.36.0-r0/image'
install 2>&1 | tee installman.log
was getting stuck:
....
Generating a Unix-style Makefile
Writing Makefile for Sys::Hostname
cd ext/XS-Typemap/ && ../../miniperl_top -I../../lib Makefile.PL \
INSTALLDIRS=perl INSTALLMAN1DIR=none INSTALLMAN3DIR=none \
PERL_CORE=1 LIBPERL_A=libperl.so.5.36.0 PERL="../../miniperl_top"
Generating a Unix-style Makefile
Writing Makefile for XS::APItest
Generating a Unix-style Makefile
Writing Makefile for XS::Typemap
With -n, it looks like it got stuck just before building buildtoc:
....
cd ext/XS-APItest/ && ../../miniperl_top -I../../lib Makefile.PL \
INSTALLDIRS=perl INSTALLMAN1DIR=none INSTALLMAN3DIR=none \
PERL_CORE=1 LIBPERL_A=libperl.so.5.36.0 PERL="../../miniperl_top"
cd ext/XS-Typemap/ && ../../miniperl_top -I../../lib Makefile.PL \
INSTALLDIRS=perl INSTALLMAN1DIR=none INSTALLMAN3DIR=none \
PERL_CORE=1 LIBPERL_A=libperl.so.5.36.0 PERL="../../miniperl_top"
./miniperl_top -f pod/buildtoc -q
./miniperl_top installman
--destdir=/OE/build/poky/build-st/tmp/work/x86_64-linux/perl-native/5.36.0-r0/image
make[1]: Leaving directory
'/OE/build/poky/build-st/tmp/work/x86_64-linux/perl-native/5.36.0-r0/perl-5.36.0'
It didn't start miniperl_top (which itself works fine):
martin@jama
/OE/build/poky/build-st/tmp/work/x86_64-linux/perl-native/5.36.0-r0/perl-5.36.0
$ ./miniperl_top -f pod/buildtoc -q
martin@jama
/OE/build/poky/build-st/tmp/work/x86_64-linux/perl-native/5.36.0-r0/perl-5.36.0
$ echo $?
0
martin 3016692 0.0 0.0 99860 2788 ? S 17:47 0:00 make -j
40
DESTDIR=/OE/build/poky/build-st/tmp/work/x86_64-linux/perl-native/5.36.0-r0/image
install
martin 3016738 99.9 0.0 101440 4184 ? R 17:47 56:21 make
install.man
Attaching to process 3016738
Reading symbols from
/OE/build/poky/build-st/tmp/work/x86_64-linux/perl-native/5.36.0-r0/recipe-sysroot-native/usr/bin/make...
(No debugging symbols found in
/OE/build/poky/build-st/tmp/work/x86_64-linux/perl-native/5.36.0-r0/recipe-sysroot-native/usr/bin/make)
Reading symbols from
/OE/build/poky/build-st/tmp/sysroots-uninative/x86_64-linux/lib/libc.so.6...
Reading symbols from
/OE/build/poky/build-st/tmp/sysroots-uninative/x86_64-linux/lib/.debug/libc.so.6...
Reading symbols from
/OE/build/poky/build-st/tmp/sysroots-uninative/x86_64-linux/lib/ld-linux-x86-64.so.2...
Reading symbols from
/OE/build/poky/build-st/tmp/sysroots-uninative/x86_64-linux/lib/.debug/ld-linux-x86-64.so.2...
(No debugging symbols found in
/OE/build/poky/build-st/tmp/sysroots-uninative/x86_64-linux/lib/.debug/ld-linux-x86-64.so.2)
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/usr/lib64/libthread_db.so.1".
0x000055c946453a1f in ?? ()
(gdb) bt
#0 0x000055c946453a1f in ?? ()
#1 0x000055c946452bf3 in ?? ()
#2 0x000055c946453b98 in ?? ()
#3 0x000055c946452bf3 in ?? ()
#4 0x000055c946453b98 in ?? ()
#5 0x000055c946452bf3 in ?? ()
#6 0x000055c946453b98 in ?? ()
#7 0x000055c946452bf3 in ?? ()
#8 0x000055c946453b98 in ?? ()
#9 0x000055c946452bf3 in ?? ()
#10 0x000055c946453fd6 in update_goal_chain ()
#11 0x000055c946438df3 in main ()
(gdb)
I've reproduced this twice today, so I was assuming it's just new separate
issue from new perl version - that's why I was asking if someone else is
seeing this behavior now.
I'll try to reproduce this with ccache enabled in regular build (I normally
don't use poky nor selftest). If it fails again, I'll enable debug for
make-native (or switch back to make-4.3 from gentoo).
Regards,
On Thu, Jun 9, 2022 at 6:21 PM Richard Purdie <
richard.pur...@linuxfoundation.org> wrote:
> On Thu, 2022-06-09 at 14:44 +0200, Martin Jansa wrote:
> > On Mon, Jun 6, 2022 at 2:02 PM Alexander Kanavin <alex.kana...@gmail.com>
> wrote:
> > > Signed-off-by: Alexander Kanavin <a...@linutronix.de>
> > > ---
> > > .../{perlcross_1.3.7.bb => perlcross_1.4.bb} | 2 +-
> > > .../perl/files/perl-configpm-switch.patch | 2 +-
> > > .../perl/files/perl-dynloader.patch | 2 +-
> > > .../perl/files/perl-rdepends.txt | 49 ++++++++++++-------
> > > .../perl/{perl_5.34.1.bb => perl_5.36.0.bb} | 2 +-
> > > 5 files changed, 35 insertions(+), 22 deletions(-)
> > > rename meta/recipes-devtools/perl-cross/{perlcross_1.3.7.bb =>
> perlcross_1.4.bb} (92%)
> > > rename meta/recipes-devtools/perl/{perl_5.34.1.bb => perl_5.36.0.bb}
> (99%)
> > >
> >
> >
> > Have anyone else seen perl.do_install getting stuck with this new shiny
> version as well?
> >
> > 690770 martin 20 0 10020 3700 3196 S 0.0 0.0 0:00.00 │
> └─ /bin/sh
> /OE/build/poky/build-st/tmp/work/x86_64-linux/perl-native/5.36.0-r0/temp/run.do_install.690768
> > 690771 martin 20 0 99860 1948 1364 S 0.0 0.0 0:00.00 │
> └─ make -j 40
> DESTDIR=/OE/build/poky/build-st/tmp/work/x86_64-linux/perl-native/5.36.0-r0/image
> install
> > 690816 martin 20 0 99M 3340 1428 R 102.1 0.0 3h06:44 │
> └─ make install.man
>
> I've seen perl and the kernel hang in make for some versions, I did try
> and put a warning about the known Fedora issue. Was this with the
> recent make-native dependency added? If so that would indicate it is
> potentially a bug in the current make version :(
>
> Cheers,
>
> Richard
>
>
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#166786):
https://lists.openembedded.org/g/openembedded-core/message/166786
Mute This Topic: https://lists.openembedded.org/mt/91575167/21656
Group Owner: openembedded-core+ow...@lists.openembedded.org
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-
