From: "Hongzhi, Song" <hongzhi.song@windriver.com>
To: Khem Raj <raj.khem@gmail.com>
Cc: openembeded-devel <openembedded-devel@lists.openembedded.org>
Subject: Re: [meta-oe][PATCH 1/1] kernel-selftest: Add a recipe on kernel selftest
Date: Mon, 30 Jul 2018 20:14:21 +0800 [thread overview]
Message-ID: <72cde4ee-b3b7-ca7d-bb2b-edc567fd1837@windriver.com> (raw)
In-Reply-To: <CAMKF1so=Z61q3ZqGmmv6pzht5b9CK1_CFofNA9yKdXE-orX_3A@mail.gmail.com>
Hi Raj,
I can continue to fix the error with rpi that you build. But I found
that rpi gets kernel
from git://github.com/raspberrypi/linux.git;branch=rpi-4.14.y, which is
not belong
to openembeded. I think that the rpi recipe must make some special
change which
cause the error.
And the recipe in meta-oe is OK both with glibc and musl. So could you
build a project
with oe-core's kernel? If there is no problem, would you merge the recipe?
--Hongzhi
On 2018年07月30日 12:02, Khem Raj wrote:
> I can not get it compiling for musl and rpi with poky lsb equivalent
> distro until there are clean builds it won’t be possible for me to get
> this in
>
> On Sun, Jul 29, 2018 at 8:34 PM Hongzhi, Song
> <hongzhi.song@windriver.com <mailto:hongzhi.song@windriver.com>> wrote:
>
> ping
>
> --Hongzhi
>
>
> On 2018年07月27日 17:00, Hongzhi, Song wrote:
> > Hi Raj,
> >
> > Could you help to confirm that if the libelf exists in your
> > recipe-sysroot?
> >
> > Because I can't reproduce your error.
> >
> >
> > --Hongzhi
> >
> >
> > On 2018年07月27日 16:40, Khem Raj wrote:
> >> this fails on rpi
> >>
> >> ERROR: Logfile of failure stored in:
> >>
> /mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/temp/log.do_compile.22198
>
> >>
> >> Log data follows:
> >> | DEBUG: Executing shell function do_compile
> >> | NOTE: make -j 16 CROSS_COMPILE=arm-bec-linux-gnueabi- ARCH=arm
> >> CC=arm-bec-linux-gnueabi-gcc -march=armv7ve -mthumb
> -mfpu=neon-vfpv4
> >> -mfloat-abi=hard -mcpu=cortex-a7 -fstack-protector-strong
> >> -D_FORTIFY_SOURCE=2 -Wformat -Wformat-security
> -Werror=format-security
> >> -fstack-protector-strong -D_FORTIFY_SOURCE=2 -Wformat
> >> -Wformat-security -Werror=format-security
> >>
> --sysroot=/mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/recipe-sysroot
>
> >>
> >> AR=arm-bec-linux-gnueabi-ar LD=arm-bec-linux-gnueabi-ld
> >>
> --sysroot=/mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/recipe-sysroot
>
> >>
> >>
> DESTDIR=/mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/image
> >> -C
> >>
> /mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/kernel-selftest-1.0/tools/testing/selftests/bpf
> >> | make: Entering directory
> >>
> '/mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/kernel-selftest-1.0/tools/testing/selftests/bpf'
>
> >>
> >> | make -C ../../../lib/bpf
> >>
> OUTPUT=/mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/kernel-selftest-1.0/tools/testing/selftests/bpf/
>
> >>
> >> | make[1]: Entering directory
> >>
> '/mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/kernel-selftest-1.0/tools/lib/bpf'
>
> >>
> >> |
> >> | Auto-detecting system features:
> >> | ... libelf: [ ^[[31mOFF^[[m ]
> >> | ... bpf: [ ^[[31mOFF^[[m ]
> >> |
> >> | BPF API too old
> >> | make[1]: *** [Makefile:219: bpfdep] Error 255
> >> | make[1]: *** Waiting for unfinished jobs....
> >> | No libelf found
> >> | make[1]: *** [Makefile:216: elfdep] Error 255
> >> | HOSTCC
> >>
> /mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/kernel-selftest-1.0/tools/testing/selftests/bpf/fixdep.o
> >> | HOSTLD
> >>
> /mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/kernel-selftest-1.0/tools/testing/selftests/bpf/fixdep-in.o
> >> | LINK
> >>
> /mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/kernel-selftest-1.0/tools/testing/selftests/bpf/fixdep
> >> | make[1]: Leaving directory
> >>
> '/mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/kernel-selftest-1.0/tools/lib/bpf'
>
> >>
> >> | make: *** [Makefile:33:
> >>
> /mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/kernel-selftest-1.0/tools/testing/selftests/bpf/libbpf.a]
>
> >>
> >> Error 2
> >> | make: Leaving directory
> >>
> '/mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/kernel-selftest-1.0/tools/testing/selftests/bpf'
>
> >>
> >> | ERROR: oe_runmake failed
> >> | WARNING:
> >>
> /mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/temp/run.do_compile.22198:1
> >> exit 1 from 'exit 1'
> >> | ERROR: Function failed: do_compile (log file is located at
> >>
> /mnt/a/oe/build/tmp/work/raspberrypi3-bec-linux-gnueabi/kernel-selftest/1.0-r0/temp/log.do_compile.22198)
>
> >>
> >>
> >> On Mon, Jul 16, 2018 at 2:51 AM Hongzhi, Song
> >> <hongzhi.song@windriver.com
> <mailto:hongzhi.song@windriver.com>> wrote:
> >>>
> >>>
> >>> On 2018年07月14日 00:23, Burton, Ross wrote:
> >>>> On 13 July 2018 at 15:06, Hongzhi.Song
> <hongzhi.song@windriver.com <mailto:hongzhi.song@windriver.com>>
> >>>> wrote:
> >>>>> +LIC_FILES_CHKSUM =
> >>>>> "file://COPYING;md5=d7810fab7487fb0aad327b76f1be7cd7 \
> >>>> Why are you shipping your own copy of the kernel's COPYING
> file even
> >>>> though you copy another in that prefunc?
> >>>>
> >>>> Isn't it possible to just depend on kernel-source and build
> directly
> >>>> from the kernel source tree?
> >>>>
> >>>>> +# for bpf and vm
> >>>>> +DEPENDS = " \
> >>>>> + elfutils \
> >>>>> + libcap \
> >>>>> + libcap-ng \
> >>>>> + fuse \
> >>>>> + util-linux \
> >>>>> + rsync-native \
> >>>>> +"
> >>>> Really not convinced these dependencies are accurate.
> >>>>
> >>>>> +TEST_LIST = "bpf \
> >>>>> + vm \
> >>>>> +"
> >>>> You're not listing memfd in here, but that is the only place
> which
> >>>> uses fuse as far as I can tell.
> >>>>
> >>>> My suggestion is to trim the DEPENDS back to the core minimum
> and use
> >>>> PACKAGECONIG to select what directories get built. This means
> you can
> >>>> have optional test suites where the dependencies are not in
> oe-core
> >>>> (fuse for memfd, for example), and get the right RDEPENDS too.
> >>>>
> >>>>> + oe_runmake -C ${S}/tools/testing/selftests/${i}
> >>>>> INSTALL_PATH=${D}/opt/kselftest/${i} install
> >>>> FHS says /opt is for sysadmin-installed tools that are not
> package
> >>>> managed. You're building a package. You're also building a
> package
> >>>> which is essentially ptest. I'd say inherit ptest, write a
> runner,
> >>>> and put all the binaries in $PTESTDIR.
> >>> Hi Burton,
> >>>
> >>> kernel-selftest is designed to be shared by ptest and oe-self,
> and
> >>> thus is
> >>> not supposed to inherit ptest.
> >>>
> >>> The suggestions you mentioned above will be modified soon.
> >>>
> >>> --Hongzhi
> >>>
> >>>> Ross
> >>>>
> >
>
next prev parent reply other threads:[~2018-07-30 12:12 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-07-13 14:06 [meta-oe][PATCH v2 0/1] Add a recipe on kernel selftest Hongzhi.Song
2018-07-13 14:06 ` [meta-oe][PATCH 1/1] kernel-selftest: " Hongzhi.Song
2018-07-13 16:23 ` Burton, Ross
2018-07-16 9:53 ` Hongzhi, Song
2018-07-27 8:40 ` Khem Raj
2018-07-27 9:00 ` Hongzhi, Song
2018-07-30 3:36 ` Hongzhi, Song
2018-07-30 4:02 ` Khem Raj
2018-07-30 12:14 ` Hongzhi, Song [this message]
2018-07-30 18:38 ` Khem Raj
2018-07-31 11:14 ` Hongzhi, Song
2018-07-13 14:06 ` [OE-core][PATCH] x86: remove "qemu-usermode" from MACHINE_FEATURES_BACKFILL_CONSIDERED Hongzhi.Song
2018-07-13 14:13 ` Alexander Kanavin
2018-07-14 8:44 ` Hongzhi, Song
2018-07-13 14:59 ` akuster808
2018-07-13 14:10 ` [meta-oe][PATCH v2 0/1] Add a recipe on kernel selftest Hongzhi, Song
2018-07-13 15:02 ` akuster808
2018-07-16 10:04 ` Hongzhi, Song
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=72cde4ee-b3b7-ca7d-bb2b-edc567fd1837@windriver.com \
--to=hongzhi.song@windriver.com \
--cc=openembedded-devel@lists.openembedded.org \
--cc=raj.khem@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox