Date: Thu, 20 Dec 2018 12:28:51 -0800 From: Mark Millard <marklmi@yahoo.com> To: freebsd-arm <freebsd-arm@freebsd.org>, FreeBSD Current <freebsd-current@freebsd.org>, ports-list freebsd <freebsd-ports@freebsd.org> Subject: Re: FreeBSD head -r341836 amd64->aarch64 cross-build of -r484783 ports via poudriere: devel/qt5-testlib hung-up during "Checking for POSIX monotonic clock" Message-ID: <AA9E699B-6C79-4FCF-8A74-BDFCAEB50835@yahoo.com> In-Reply-To: <8A63BA63-FBCB-40DA-B031-4C7F4B1F5118@yahoo.com> References: <190D54A2-D29A-4B62-B8FF-99514DBE92A0@yahoo.com> <FB89FE8A-7C64-4D16-926B-2C033BBB9D1C@yahoo.com> <8A63BA63-FBCB-40DA-B031-4C7F4B1F5118@yahoo.com>
next in thread | previous in thread | raw e-mail | index | archive | help
[A amd64->armv7 cross build shows interesting hang-up behavior as
well, apparently highly repeatable for my current context.]
On 2018-Dec-19, at 16:21, Mark Millard <marklmi at yahoo.com> wrote:
> [I attached to the hung-up process with gdb and looked around a =
little.]
>=20
> On 2018-Dec-19, at 13:58, Mark Millard <marklmi at yahoo.com> wrote:
>=20
>> [Looks like a race or some such for devel/qt5-testlib: retry of =
poudreire-devel
>> did not hang. The other hang-up seems to be repeating and I give some =
details.]
>>=20
>> On 2018-Dec-19, at 12:20, Mark Millard <marklmi at yahoo.com> wrote:
>>=20
>>> FYI: Based on FreeBSD head -r341836 (host and target) and ports =
-r484783 . This
>>> was a rebuild based on going from perl5.26 to perl5.28 without =
updating the ports
>>> tree and from system clang 6 for the prior FreeBSD-head context used =
to clang 7
>>> this time. (I'm not attributing causes here.) poudriere was using =
amd64-native
>>> tools for speeding up the cross-build.
>>>=20
>>> # grep -r =3Dperl5=3D /etc/ ~/src.configs/ /usr/local/etc/
>>> /etc/make.conf:DEFAULT_VERSIONS+=3Dperl5=3D5.28 gcc=3D8
>>> /usr/local/etc/poudriere.d/make.conf:DEFAULT_VERSIONS+=3Dperl5=3D5.28 =
gcc=3D8
>>>=20
>>> There was also a "print/texinfo:configure/runaway" but I've not =
looked into
>>> it at all yet and it may be a while before I do. The other ports =
attempted
>>> built fine as far as I can tell so far.
>>>=20
>>>=20
>>> The devel/qt5-testlib failure looks like:
>>>=20
>>> [00:00:13] Building 123 packages using 28 builders
>>> . . .
>>> [00:49:30] [10] [00:00:00] Building devel/qt5-testlib | =
qt5-testlib-5.11.2
>>> . . .
>>> [07:31:31] [10] [06:42:01] Saved devel/qt5-testlib | =
qt5-testlib-5.11.2 wrkdir to: =
/usr/local/poudriere/data/wrkdirs/FBSDFSSDjailCortexA57-default/default/qt=
5-testlib-5.11.2.tar
>>> [07:31:32] [10] [06:42:02] Finished devel/qt5-testlib | =
qt5-testlib-5.11.2: Failed: configure/runaway
>>>=20
>>> With logs/errors/qt5-testlib-5.11.2.log showing:
>>>=20
>>> Checking for POSIX monotonic clock...=20
>>> + cd =
/wrkdirs/usr/ports/devel/qt5-testlib/work/qtbase-everywhere-src-5.11.2/con=
fig.tests/clock-monotonic && =
/wrkdirs/usr/ports/devel/qt5-testlib/work/qtbase-everywhere-src-5.11.2/bin=
/qmake "CONFIG -=3D qt debug_and_release app_bundle lib_bundle" "CONFIG =
+=3D shared warn_off console single_arch" =
/wrkdirs/usr/ports/devel/qt5-testlib/work/qtbase-everywhere-src-5.11.2/con=
fig.tests/clock-monotonic
>>> + cd =
/wrkdirs/usr/ports/devel/qt5-testlib/work/qtbase-everywhere-src-5.11.2/con=
fig.tests/clock-monotonic && MAKEFLAGS=3D make
>>> =3D>> Killing runaway build after 21600 seconds with no output
>>> =3D>> Cleaning up wrkdir
>>> =3D=3D=3D> Cleaning for qt5-testlib-5.11.2
>>> Killed
>>> build of devel/qt5-testlib | qt5-testlib-5.11.2 ended at Wed Dec 19 =
06:45:42 PST 2018
>>> build time: 06:41:46
>>> !!! build failure encountered !!!
>>>=20
>>>=20
>>> # less =
/wrkdirs/usr/ports/devel/qt5-testlib/work/qtbase-everywhere-src-5.11.2/con=
fig.log
>>> . . .
>>> test config.qtbase_corelib.libraries.librt succeeded
>>> executing config test clock-monotonic
>>> + cd =
/wrkdirs/usr/ports/devel/qt5-testlib/work/qtbase-everywhere-src-5.11.2/con=
fig.tests/clock-monotonic && =
/wrkdirs/usr/ports/devel/qt5-testlib/work/qtbase-everywhere-src-5.11.2/bin=
/qmake "CONFIG -=3D qt debug_and_release app_bundle lib_bundle" "CONFIG =
+=3D shared warn_off console single_arch" =
/wrkdirs/usr/ports/devel/qt5-testlib/work/qtbase-everywhere-src-5.11.2/con=
fig.tests/clock-monotonic
>>> + cd =
/wrkdirs/usr/ports/devel/qt5-testlib/work/qtbase-everywhere-src-5.11.2/con=
fig.tests/clock-monotonic && MAKEFLAGS=3D make
>>>=20
>>>=20
>>> Some supporting details of context:
>>>=20
>>> # uname -apKU
>>> FreeBSD FBSDFSSD 13.0-CURRENT FreeBSD 13.0-CURRENT #5 r341836M: Tue =
Dec 11 16:37:42 PST 2018 =
markmi@FBSDFSSD:/usr/obj/amd64_clang/amd64.amd64/usr/src/amd64.amd64/sys/G=
ENERIC-NODBG amd64 amd64 1300005 1300005
>>>=20
>>> # svnlite info /usr/ports/ | grep "Re[plv]"
>>> Relative URL: ^/head
>>> Repository Root: svn://svn.freebsd.org/ports
>>> Repository UUID: 35697150-7ecd-e111-bb59-0022644237b5
>>> Revision: 484783
>>> Last Changed Rev: 484783
>>>=20
>>=20
>> I started poudriere up again with just the 2 needing to be rebuilt =
(plus
>> what depends on the 2). devel/qt5-testlib quickly completed just =
fine:
>>=20
>> [00:02:16] [02] [00:00:00] Building devel/qt5-testlib | =
qt5-testlib-5.11.2
>> [00:04:54] [02] [00:02:38] Finished devel/qt5-testlib | =
qt5-testlib-5.11.2: Success
>>=20
>>=20
>> In the prior build that had the hang-ups I looked and dor =
print/texinfo :
>>=20
>> /wrkdirs/usr/ports/print/texinfo/work/texinfo-6.5/config.log shows =
for its
>> hang-up:
>>=20
>> . . .
>> configure:6639: checking for alloca
>> configure:6676: /nxb-bin/usr/bin/cc -o conftest -O2 -pipe =
-mcpu=3Dcortex-a57 -DLIBICONV_PLUG -g -fno-strict-aliasing =
-mcpu=3Dcortex-a57 -DLIBICONV_PLUG -D_THREAD_SAFE conftest.c >&5
>> configure:6676: $? =3D 0
>> configure:6684: result: yes
>> configure:6794: checking for C/C++ restrict keyword
>> configure:6821: /nxb-bin/usr/bin/cc -c -O2 -pipe -mcpu=3Dcortex-a57 =
-DLIBICONV_PLUG -g -fno-strict-aliasing -mcpu=3Dcortex-a57 =
-DLIBICONV_PLUG -D_THREAD_SAFE conftest.c >&5
>> configure:6821: $? =3D 0
>> configure:6829: result: __restrict
>> configure:6844: checking whether // is distinct from /
>>=20
>>=20
>> In the poudriere re-run print/texinfo seems to be not progressing:
>>=20
>> root 87913 0.0 0.0 12920 3668 0 I 13:29 =
0:00.06 | | `-- sh: =
poudriere[FBSDFSSDjailCortexA57-default][01]: build_pkg =
(texinfo-6.5_1,1) (sh)
>> root 88869 0.0 0.0 12920 3660 0 I 13:29 =
0:00.00 | | `-- sh: =
poudriere[FBSDFSSDjailCortexA57-default][01]: build_pkg =
(texinfo-6.5_1,1) (sh)
>> root 88870 0.0 0.0 10412 1848 0 IJ 13:29 =
0:00.01 | | `-- /usr/bin/make -C /usr/ports/print/texinfo =
configure
>> root 88974 0.0 0.0 10272 1812 0 IJ 13:30 =
0:00.00 | | `-- /bin/sh -e -c (cd =
/wrkdirs/usr/ports/print/texinfo/work/texinfo-6.5 && =
_LATE_CONFIGURE_ARGS=3D"" ; if [=20
>> root 89283 0.0 0.0 11160 2108 0 IJ 13:30 =
0:00.10 | | `-- /bin/sh ./configure --enable-nls =
--prefix=3D/usr/local --localstatedir=3D/var --mandir=3D/usr/local/man =
--di
>> root 89692 0.0 0.0 227368 14504 0 IJ 13:30 =
0:00.03 | | `-- /usr/local/bin/qemu-aarch64-static =
wc //dev/null
>> root 89694 0.0 0.0 227424 14596 0 IJ 13:30 =
0:00.01 | | `-- /usr/local/bin/qemu-aarch64-static =
wc //dev/null
>> root 89695 0.0 0.0 227584 14720 0 IJ 13:30 =
0:00.01 | | `-- wc: zygote (qemu-aarch64-static)
>>=20
>>=20
>> So it appears that:
>>=20
>> /usr/local/bin/qemu-aarch64-static wc //dev/null
>>=20
>> is hanging-up (again).
>>=20
>>=20
>> Given that these are hangups I'll note that this is a Ryzen
>> Threadripper 1950X context and is running under Hyper-V from
>> Windows 10's 1809 update. I gave it 28 logical processors and
>> have it to have the virtual NUMA topology match the topology of
>> the physical hardware: "Use Hardware Topology". (Processors
>> 28, NUMA nodes 2, Sockets 1, Hardware threads per core 2.)
>=20
> Attaching to the stuck process via gdb and looking at the backtrace
> shows:
>=20
> (gdb) attach 89695
> Attaching to program: /usr/local/bin/qemu-aarch64-static, process =
89695
> [New LWP 101548 of process 89695]
> [Switching to LWP 100804 of process 89695]
> _pselect () at _pselect.S:3
> 3 PSEUDO(pselect)
>=20
> (gdb) bt
> #0 _pselect () at _pselect.S:3
> #1 0x00000000601da57f in __thr_pselect (count=3D12, =
rfds=3D0x7ffffffe3650, wfds=3D0x0, efds=3D0x0, timo=3D0x0, =
mask=3D0x7ffffffe3600) at /usr/src/lib/libthr/thread/thr_syscalls.c:378
> #2 0x000000006004928d in do_freebsd_select (env=3D0x860edfb18, =
n=3D<optimized out>, rfd_addr=3D140736934698744, wfd_addr=3D<optimized =
out>, efd_addr=3D0, target_tv_addr=3D0)
> at =
/wrkdirs/usr/ports/emulators/qemu-user-static/work/qemu-bsd-user-4ef7d07/b=
sd-user/freebsd/os-time.h:468
> #3 do_freebsd_syscall (cpu_env=3D0x860edfb18, num=3D93, arg1=3D12, =
arg2=3D140736934698744, arg3=3D0, arg4=3D0, arg5=3D0, arg6=3D274914043516,=
arg7=3D274913946564, arg8=3D6579811)
> at =
/wrkdirs/usr/ports/emulators/qemu-user-static/work/qemu-bsd-user-4ef7d07/b=
sd-user/syscall.c:1106
> #4 0x000000006003903c in target_cpu_loop (env=3D0x860edfb18) at =
/wrkdirs/usr/ports/emulators/qemu-user-static/work/qemu-bsd-user-4ef7d07/b=
sd-user/aarch64/target_arch_cpu.h:100
> #5 0x0000000060038e09 in cpu_loop (env=3D0xc) at =
/wrkdirs/usr/ports/emulators/qemu-user-static/work/qemu-bsd-user-4ef7d07/b=
sd-user/main.c:121
> #6 0x0000000060039ecb in main (argc=3D<optimized out>, =
argv=3D0x7fffffffd360) at =
/wrkdirs/usr/ports/emulators/qemu-user-static/work/qemu-bsd-user-4ef7d07/b=
sd-user/main.c:513
>=20
> (gdb) up
> #1 0x00000000601da57f in __thr_pselect (count=3D12, =
rfds=3D0x7ffffffe3650, wfds=3D0x0, efds=3D0x0, timo=3D0x0, =
mask=3D0x7ffffffe3600) at /usr/src/lib/libthr/thread/thr_syscalls.c:378
> 378 ret =3D __sys_pselect(count, rfds, wfds, efds, timo, =
mask);
> (gdb) print *rfds
> $1 =3D {__fds_bits =3D {2048, 0 <repeats 15 times>}}
>=20
> (gdb) info threads
> Id Target Id Frame=20
> * 1 LWP 100804 of process 89695 _pselect () at _pselect.S:3
> 2 LWP 101548 of process 89695 _umtx_op_err () at =
/usr/src/lib/libthr/arch/amd64/amd64/_umtx_op_err.S:37
>=20
> (gdb) thread 2=20
> [Switching to thread 2 (LWP 101548 of process 89695)]
> #0 _umtx_op_err () at =
/usr/src/lib/libthr/arch/amd64/amd64/_umtx_op_err.S:37
> 37 RSYSCALL_ERR(_umtx_op)
>=20
> (gdb) bt
> #0 _umtx_op_err () at =
/usr/src/lib/libthr/arch/amd64/amd64/_umtx_op_err.S:37
> #1 0x00000000601d2ec0 in _thr_umtx_timedwait_uint (mtx=3D0x861027008, =
id=3D<optimized out>, clockid=3D<optimized out>, abstime=3D<optimized =
out>, shared=3D<optimized out>)
> at /usr/src/lib/libthr/thread/thr_umtx.c:236
> #2 0x00000000601dc6f8 in cond_wait_user (cvp=3D<optimized out>, =
mp=3D0x860515b00, abstime=3D0x0, cancel=3D1) at =
/usr/src/lib/libthr/thread/thr_cond.c:307
> #3 cond_wait_common (cond=3D<optimized out>, mutex=3D<optimized out>, =
abstime=3D0x0, cancel=3D1) at /usr/src/lib/libthr/thread/thr_cond.c:367
> #4 0x00000000601438bc in qemu_futex_wait (ev=3D<optimized out>, =
val=3D4294967295) at util/qemu-thread-posix.c:350
> #5 qemu_event_wait (ev=3D0x62735d10 <rcu_call_ready_event>) at =
util/qemu-thread-posix.c:445
> #6 0x000000006014a92a in call_rcu_thread (opaque=3D<optimized out>) =
at util/rcu.c:255
> #7 0x00000000601dc376 in thread_start (curthread=3D0x860518e00) at =
/usr/src/lib/libthr/thread/thr_create.c:291
> #8 0x0000000000000000 in ?? ()
> Backtrace stopped: Cannot access memory at address 0x7fffdfdfc000
>=20
In a separate rebuild of 124 ports print/texinfo repeated its
problem.
In a rebuild targeting armv7, multimedia/gstreamer1-qt hung-up and timed
out. Another poudriere run also hung-up:
root 33719 0.0 0.0 12920 3528 0 I 11:40 0:00.03 | =
| `-- sh: poudriere[FBSDFSSDjailArmV7-default][02]: build_pkg =
(gstreamer1-qt5-1.2.0_14) (sh)
root 41551 0.0 0.0 12920 3520 0 I 11:43 0:00.00 | =
| `-- sh: poudriere[FBSDFSSDjailArmV7-default][02]: =
build_pkg (gstreamer1-qt5-1.2.0_14) (sh)
root 41552 0.0 0.0 10340 1744 0 IJ 11:43 0:00.01 | =
| `-- /usr/bin/make -C /usr/ports/multimedia/gstreamer1-qt =
FLAVOR=3Dqt5 build
root 41566 0.0 0.0 10236 1796 0 IJ 11:43 0:00.00 | =
| `-- /bin/sh -e -c (cd =
/wrkdirs/usr/ports/multimedia/gstreamer1-qt/work-qt5/.build; if ! =
/usr/bin/env QT_SELE
root 41567 0.0 0.0 89976 12896 0 IJ 11:43 0:00.07 | =
| `-- /usr/local/bin/qemu-arm-static ninja -j28 -v all
root 41585 0.0 0.0 102848 25056 0 IJ 11:43 0:00.10 | =
| |-- /usr/local/bin/qemu-arm-static =
/usr/local/bin/cmake -E cmake_autogen /wrkdirs/usr/ports/multimedia/g
root 41586 0.0 0.0 102852 25072 0 IJ 11:43 0:00.11 | =
| `-- /usr/local/bin/qemu-arm-static =
/usr/local/bin/cmake -E cmake_autogen /wrkdirs/usr/ports/multimedia/g
or as top showed it:
41552 root 1 52 0 10M 1744K 0 wait 15 0:00 =
0.00% /usr/bin/make -C /usr/ports/multimedia/gstreamer1-qt FLAVOR=3Dqt5 =
build
41566 root 1 52 0 10M 1796K 0 wait 1 0:00 =
0.00% /bin/sh -e -c (cd =
/wrkdirs/usr/ports/multimedia/gstreamer1-qt/work-qt5/.build; if ! =
/usr/bin/env QT_SELECT=3Dqt5 QMAKEMODULES
41567 root 2 52 0 88M 13M 0 select 4 0:00 =
0.00% /usr/local/bin/qemu-arm-static ninja -j28 -v all
41585 root 2 52 0 100M 24M 0 kqread 8 0:00 =
0.00% /usr/local/bin/qemu-arm-static /usr/local/bin/cmake -E =
cmake_autogen /wrkdirs/usr/ports/multimedia/gstreamer1-qt/work-qt5/.
41586 root 2 52 0 100M 24M 0 kqread 22 0:00 =
0.00% /usr/local/bin/qemu-arm-static /usr/local/bin/cmake -E =
cmake_autogen /wrkdirs/usr/ports/multimedia/gstreamer1-qt/work-qt5/.
So: waiting in kqread. Repeated tries have gotten the same result
so far.
If this keeps up, later I may be able to try a native FreeBSD boot
instead of Hyper-V use on the same machine.
=3D=3D=3D
Mark Millard
marklmi at yahoo.com
( dsl-only.net went
away in early 2018-Mar)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?AA9E699B-6C79-4FCF-8A74-BDFCAEB50835>