Date: Wed, 19 Dec 2018 16:21:42 -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: <8A63BA63-FBCB-40DA-B031-4C7F4B1F5118@yahoo.com> In-Reply-To: <FB89FE8A-7C64-4D16-926B-2C033BBB9D1C@yahoo.com> References: <190D54A2-D29A-4B62-B8FF-99514DBE92A0@yahoo.com> <FB89FE8A-7C64-4D16-926B-2C033BBB9D1C@yahoo.com>
next in thread | previous in thread | raw e-mail | index | archive | help
[I attached to the hung-up process with gdb and looked around a little.]
On 2018-Dec-19, at 13:58, Mark Millard <marklmi at yahoo.com> wrote:
> [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.)
Attaching to the stuck process via gdb and looking at the backtrace
shows:
(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)
(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
(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>}}
(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
(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)
(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
=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?8A63BA63-FBCB-40DA-B031-4C7F4B1F5118>