On Wed, 2022-08-03 at 09:25 +0200, Martin Jansa wrote:

You can use VIRTUAL-RUNTIME_sed variable instead of 'sed'
_everywhere_ and then change the preferred runtime provider in your
DISTRO config.

Similar case for "stat" from busybox in webOS OSE:
https://github.com/webosose/meta-webosose/search?q=VIRTUAL-RUNTIME_stat

and much worse case (because too many places add runtime dependency
on bash) to replace bash:
https://github.com/webosose/meta-webosose/search?q=VIRTUAL-RUNTIME_bash
see
https://bugzilla.yoctoproject.org/show_bug.cgi?id=9217#c5

This is where we should perhaps think about the filter variable option
Joshua and I have talked about periodically as that could handle this
in a much neater and generic way, at the unfortunate cost of
complexity.

The multilib code already tries to do this kind of remapping.

Cheers,

Richard

		Craig McQueen Embedded Systems Engineer
		t e w	: : :	+61 3 9780 4378 craig.mcqueen@... innerrange.com
Inner Range • 1 Millennium Court • Knoxfield • Victoria • 3180 • Australia

 

 

My local daily builds also run on master. I triggered a kirkstone build today, it builds successfully for me. Is it failing for specific machine ?

		Craig McQueen Embedded Systems Engineer
		t e w	: : :	+61 3 9780 4378 craig.mcqueen@... innerrange.com
Inner Range • 1 Millennium Court • Knoxfield • Victoria • 3180 • Australia

 

 

On 2022-08-01 11:45, Quentin Schulz wrote:

Hi Maik,

On 8/1/22 17:34, Maik Vermeulen wrote:

Hi Quentin,

Thank you for your response!

I added kernel-modules to the IMAGE_INSTALL_append, but it seems that the
modules are still not being loaded.
Is that the correct way?

Also I see that CONFIG_NF_TABLES is not set (with ~# zcat /proc/config.gz |
grep CONFIG_NF_ | grep TABLE)
Is that expected?

No, you would indeed need to enable those in your kernel config.

0.7 - that's pre-dunfell!

For master or kirkstone, see:
https://lists.yoctoproject.org/g/linux-yocto/message/11523?p=%2C%2C%2C20%2C0%2C0%2C0%3A%3Acreated%2C0%2Cmacleod%2C20%2C1%2C0%2C92659340

and the follow-up that I'll send soon to get the pass-rate to 100%.

You won't be able to use the WR Linux features directly but take a look at the image.inc
and template.conf files in:

https://github.com/WindRiver-Labs/wrlinux/tree/WRLINUX_10_21_BASE/templates/feature/nftables

for the syntax to use the kernel scc files for nftables. It may not be there if you're using an older kernel.

../Randy

Cheers,
Quentin

--
# Randy MacLeod
# Wind River Linux

Greetings

5 videos from Yocto Project talks at the Embedded Linux Conference North America are now available on YouTube:

BoF: The Yocto Project and OpenEmbedded Organization - Philip Balister, OpenSDR: https://youtu.be/Upvyb6VlICk

Configuring and Building a Heterogenous System Using the Yocto Project - Mark Hatle, AMD: https://youtu.be/SSFAp4LS6hs

Lessons Learned: Migrating a Production Platform to Yocto - Mitch Gaines, Farmblox: https://youtu.be/2mXxhlNM0_Q

Software Bill of Materials and Supply Chain with the Yocto Project - Joshua Watt, Garmin: https://youtu.be/6zms_qGmVqg

Yocto Project Autobuilders and the SWAT Team - Alexandre Belloni, Bootlin: https://youtu.be/SLHV9HbxKcM

For quick sharing, these videos are available on a special playlist: https://www.youtube.com/playlist?list=PL9E8231JvrhKff6RuVXJ8qlluqS4xG_9c

Please spread the news through social media. Feel free to reuse my own posts:
https://fosstodon.org/@MichaelOpdenacker/108754686397298056
https://www.linkedin.com/posts/michaelopdenacker_bof-the-yocto-project-and-openembedded-organization-activity-6960297794007904258-7eaB
https://twitter.com/mopdenacker/status/1554533737643159552

Cheers
Michael.

--
Michael Opdenacker, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

Hi Joshua,

It's 100% my fault, I'm mixed 2 variables:

ROOTFS_POSTPROCESS_COMMAND
IMAGE_POSTPROCESS_COMMAND

My "ln " command have to be placed into IMAGE_POSTPROCESS_COMMAND, not ROOTFS_POSTPROCESS_COMMAND.

Thank you to have answer

On Tue, Aug 2, 2022 at 6:30 AM Marta Rybczynska <rybczynska@...> wrote:

On Tue, Aug 2, 2022 at 4:49 PM Neal Caidin <ncaidin@...> wrote:

Current Dev Position: YP 4.1 M3

Next Deadline: 22nd August 2022 YP 4.1 M3 Build


Next Team Meetings:

Bug Triage meeting Thursday August 4th 7:30 am PDT (https://zoom.us/j/454367603?pwd=ZGxoa2ZXL3FkM3Y0bFd5aVpHVVZ6dz09)

Weekly Project Engineering Sync Tuesday August 2nd at 8 am PDT (https://zoom.us/j/990892712?pwd=cHU1MjhoM2x6ck81bkcrYjRrcmJsUT09)

Twitch - See https://www.twitch.tv/theyoctojester


Key Status/Updates:

YP 4.1 M2 was released

YP 3.1.18 passed QA and is due to be released

The debug file mapping issues have moved slightly further forward thanks to some help from Ross but are stalling due to insufficient developer time available to resolve the issues.

Some rust toolchain improvements did merge, including an automated rust toolchain test. The bulk of the rework is stalled on a handful of remaining issues, particularly a couple of reproducibility problems but also a mips n32 toolchain issue.

We have one open CVE on master against qemu, help to backport that fix would be very welcome to keep the numbers down. CVEs in kirkstone (LTS) significantly reduced.

Do you mean CVE-2022-35414?

Yes! We have a patch under review for kirkstone:
https://lists.openembedded.org/g/openembedded-core/message/168636

I think Richard is hoping someone will see if that also will work on
master and if so send a patch.

Steve

The autotest change allows for the list of targets to be built to be
larger. Add more targets to increase the build coverage of CI.
Unfortunately, there are some platforms that do not build everything.
So, clip those down to those that work for all. Also, there are a few
targets unique to certain platforms. So, expand those to have coverage.

Signed-off-by: Jon Mason <jon.mason@...>
Signed-off-by: Naveen Saini <naveen.kumar.saini@...>
---
ci/96b-nitrogen.yml | 14 ++++++++++++++
ci/arduino-nano-33-ble.yml | 15 +++++++++++++++
ci/base.yml | 2 ++
ci/nrf52840dk-nrf52840.yml | 14 ++++++++++++++
ci/stm32mp157c-dk2.yml | 7 +++++++
5 files changed, 52 insertions(+)

diff --git a/ci/96b-nitrogen.yml b/ci/96b-nitrogen.yml
index d033d02..9b685fe 100644
--- a/ci/96b-nitrogen.yml
+++ b/ci/96b-nitrogen.yml
@@ -4,3 +4,17 @@ header:
- ci/base.yml

machine: 96b-nitrogen
+
+target:
+ - zephyr-blinky
+ - zephyr-coap-client
+ - zephyr-coap-server
+ - zephyr-echo-client
+ - zephyr-hci-uart
+ - zephyr-helloworld
+ - zephyr-http-client
+ - zephyr-kernel-test-all
+ - zephyr-mqtt-publisher
+ - zephyr-peripheral-esp
+ - zephyr-peripheral-hr
+ - zephyr-philosophers
diff --git a/ci/arduino-nano-33-ble.yml b/ci/arduino-nano-33-ble.yml
index eb878b9..1035118 100644
--- a/ci/arduino-nano-33-ble.yml
+++ b/ci/arduino-nano-33-ble.yml
@@ -4,3 +4,18 @@ header:
- ci/base.yml

machine: arduino-nano-33-ble
+
+target:
+ - zephyr-blinky
+ - zephyr-coap-client
+ - zephyr-coap-server
+ - zephyr-echo-client
+ - zephyr-helloworld
+ - zephyr-http-client
+ - zephyr-kernel-test-all
+ - zephyr-mqtt-publisher
+ - zephyr-openthread-echo-client
+ - zephyr-openthread-rcp
+ - zephyr-peripheral-esp
+ - zephyr-peripheral-hr
+ - zephyr-philosophers
diff --git a/ci/base.yml b/ci/base.yml
index 678b45d..bab03d3 100644
--- a/ci/base.yml
+++ b/ci/base.yml
@@ -33,4 +33,6 @@ local_conf_header:
machine: unset

target:
+ - zephyr-helloworld
- zephyr-kernel-test-all
+ - zephyr-philosophers
diff --git a/ci/nrf52840dk-nrf52840.yml b/ci/nrf52840dk-nrf52840.yml
index 53ab8af..a0c1587 100644
--- a/ci/nrf52840dk-nrf52840.yml
+++ b/ci/nrf52840dk-nrf52840.yml
@@ -4,3 +4,17 @@ header:
- ci/base.yml

machine: nrf52840dk-nrf52840
+
+target:
+ - zephyr-blinky
+ - zephyr-coap-client
+ - zephyr-coap-server
+ - zephyr-echo-client
+ - zephyr-helloworld
+ - zephyr-http-client
+ - zephyr-kernel-test-all
+ - zephyr-lvgl
+ - zephyr-mqtt-publisher
+ - zephyr-peripheral-esp
+ - zephyr-peripheral-hr
+ - zephyr-philosophers
diff --git a/ci/stm32mp157c-dk2.yml b/ci/stm32mp157c-dk2.yml
index 8929b00..3dc04a5 100644
--- a/ci/stm32mp157c-dk2.yml
+++ b/ci/stm32mp157c-dk2.yml
@@ -4,3 +4,10 @@ header:
- ci/base.yml

machine: stm32mp157c-dk2
+
+target:
+ - zephyr-blinky
+ - zephyr-helloworld
+ - zephyr-kernel-test-all
+ - zephyr-philosophers
+ - zephyr-openamp-rsc-table
--
2.17.1

Trivial change to move stm32mp157c-dk2 to its alphabetical location.
This is relevant because it aligns the file to match the order on the
gilab pipeline entry.

Signed-off-by: Jon Mason <jon.mason@...>
Signed-off-by: Naveen Saini <naveen.kumar.saini@...>
---
.gitlab-ci.yml | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml
index 014123f..8d22654 100644
--- a/.gitlab-ci.yml
+++ b/.gitlab-ci.yml
@@ -77,9 +77,6 @@ intel-x86-64:
nrf52840dk-nrf52840:
extends: .build

-stm32mp157c-dk2:
- extends: .build
-
qemu-cortex-m3/testimage:
extends: .build

@@ -88,3 +85,6 @@ qemu-nios2:

qemu-x86/testimage:
extends: .build
+
+stm32mp157c-dk2:
+ extends: .build
--
2.17.1

Warnings are being seen in gitlab of
WARNING - Falling back to file-relative addressing of local include "base.yml"
WARNING - Update your layer to repo-relative addressing to avoid this warning
WARNING - Falling back to file-relative addressing of local include "meta-openembedded.yml"
WARNING - Update your layer to repo-relative addressing to avoid this warning

Make the relevant changes to resolve this issue

Signed-off-by: Jon Mason <jon.mason@...>
Signed-off-by: Naveen Saini <naveen.kumar.saini@...>
---
ci/96b-avenger96.yml | 2 +-
ci/96b-nitrogen.yml | 2 +-
ci/arduino-nano-33-ble.yml | 2 +-
ci/base.yml | 2 +-
ci/intel-x86-64.yml | 2 +-
ci/nrf52840dk-nrf52840.yml | 2 +-
ci/qemu-nios2.yml | 2 +-
ci/qemu-x86.yml | 2 +-
ci/stm32mp157c-dk2.yml | 2 +-
9 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/ci/96b-avenger96.yml b/ci/96b-avenger96.yml
index 6d632f1..bcdccf3 100644
--- a/ci/96b-avenger96.yml
+++ b/ci/96b-avenger96.yml
@@ -1,6 +1,6 @@
header:
version: 9
includes:
- - base.yml
+ - ci/base.yml

machine: 96b-avenger96
diff --git a/ci/96b-nitrogen.yml b/ci/96b-nitrogen.yml
index ecd96fb..d033d02 100644
--- a/ci/96b-nitrogen.yml
+++ b/ci/96b-nitrogen.yml
@@ -1,6 +1,6 @@
header:
version: 9
includes:
- - base.yml
+ - ci/base.yml

machine: 96b-nitrogen
diff --git a/ci/arduino-nano-33-ble.yml b/ci/arduino-nano-33-ble.yml
index ca332da..eb878b9 100644
--- a/ci/arduino-nano-33-ble.yml
+++ b/ci/arduino-nano-33-ble.yml
@@ -1,6 +1,6 @@
header:
version: 9
includes:
- - base.yml
+ - ci/base.yml

machine: arduino-nano-33-ble
diff --git a/ci/base.yml b/ci/base.yml
index fb1699c..678b45d 100644
--- a/ci/base.yml
+++ b/ci/base.yml
@@ -1,7 +1,7 @@
header:
version: 11
includes:
- - meta-openembedded.yml
+ - ci/meta-openembedded.yml

distro: zephyr

diff --git a/ci/intel-x86-64.yml b/ci/intel-x86-64.yml
index 525a38f..23e711c 100644
--- a/ci/intel-x86-64.yml
+++ b/ci/intel-x86-64.yml
@@ -1,6 +1,6 @@
header:
version: 9
includes:
- - base.yml
+ - ci/base.yml

machine: intel-x86-64
diff --git a/ci/nrf52840dk-nrf52840.yml b/ci/nrf52840dk-nrf52840.yml
index cbbf434..53ab8af 100644
--- a/ci/nrf52840dk-nrf52840.yml
+++ b/ci/nrf52840dk-nrf52840.yml
@@ -1,6 +1,6 @@
header:
version: 9
includes:
- - base.yml
+ - ci/base.yml

machine: nrf52840dk-nrf52840
diff --git a/ci/qemu-nios2.yml b/ci/qemu-nios2.yml
index c382582..b818371 100644
--- a/ci/qemu-nios2.yml
+++ b/ci/qemu-nios2.yml
@@ -1,7 +1,7 @@
header:
version: 9
includes:
- - base.yml
+ - ci/base.yml

local_conf_header:
nonbuilding_tests: |
diff --git a/ci/qemu-x86.yml b/ci/qemu-x86.yml
index ba80dd3..df744a1 100644
--- a/ci/qemu-x86.yml
+++ b/ci/qemu-x86.yml
@@ -1,7 +1,7 @@
header:
version: 9
includes:
- - base.yml
+ - ci/base.yml

local_conf_header:
failing_tests: |
diff --git a/ci/stm32mp157c-dk2.yml b/ci/stm32mp157c-dk2.yml
index c833794..8929b00 100644
--- a/ci/stm32mp157c-dk2.yml
+++ b/ci/stm32mp157c-dk2.yml
@@ -1,6 +1,6 @@
header:
version: 9
includes:
- - base.yml
+ - ci/base.yml

machine: stm32mp157c-dk2
--
2.17.1

From: Peter Hoyes <Peter.Hoyes@...>

When TESTIMAGE_AUTO is enabled, the do_testimage task is inserted after
do_image_complete and before do_build so that the test suites
automatically run as part of the image build.

However, do_testdata_write is currently constrained to run only before
do_build, so it likely won't execute prior to do_testimage. Change the
"before" constraint to do_testimage do that the testdata is always
generated prior to running the testimage task.

Signed-off-by: Peter Hoyes <Peter.Hoyes@...>
Signed-off-by: Naveen Saini <naveen.kumar.saini@...>
---
meta-zephyr-core/classes/zephyrtest.bbclass | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/meta-zephyr-core/classes/zephyrtest.bbclass b/meta-zephyr-core/classes/zephyrtest.bbclass
index 248fd15..aa48e6c 100644
--- a/meta-zephyr-core/classes/zephyrtest.bbclass
+++ b/meta-zephyr-core/classes/zephyrtest.bbclass
@@ -50,4 +50,4 @@ python testdata_clean() {
os.remove(fname)
}

-addtask do_testdata_write before do_build after do_deploy
+addtask do_testdata_write before do_testimage after do_deploy
--
2.17.1

newlib fails to compile for nios2 architecture. Work around this by
using glibc instead.

Signed-off-by: Jon Mason <jon.mason@...>
Signed-off-by: Naveen Saini <naveen.kumar.saini@...>
---
meta-zephyr-bsp/conf/machine/qemu-nios2.conf | 2 ++
meta-zephyr-core/conf/distro/zephyr.conf | 2 +-
2 files changed, 3 insertions(+), 1 deletion(-)

diff --git a/meta-zephyr-bsp/conf/machine/qemu-nios2.conf b/meta-zephyr-bsp/conf/machine/qemu-nios2.conf
index de20320..c41f505 100644
--- a/meta-zephyr-bsp/conf/machine/qemu-nios2.conf
+++ b/meta-zephyr-bsp/conf/machine/qemu-nios2.conf
@@ -14,3 +14,5 @@ QB_OPT_APPEND = "-nographic"
QB_CPU = "-cpu nios2"

ARCH:qemu-nios2 = "nios2"
+
+TCLIBC = "glibc"
diff --git a/meta-zephyr-core/conf/distro/zephyr.conf b/meta-zephyr-core/conf/distro/zephyr.conf
index 6ecd421..bdf1821 100644
--- a/meta-zephyr-core/conf/distro/zephyr.conf
+++ b/meta-zephyr-core/conf/distro/zephyr.conf
@@ -4,7 +4,7 @@ DISTRO_VERSION = "1.0"

TARGET_VENDOR = "-yocto"

-TCLIBC = "newlib"
+TCLIBC ?= "newlib"

TEST_TARGET = "QemuTargetZephyr"
TEST_SUITES = "zephyr"
--
2.17.1

From: Peter Hoyes <Peter.Hoyes@...>

Now that TESTIMAGE_AUTO is available for Zephyr builds, enable it in
ci/testimage.yml and remove the redundant build_and_test base
configuration.

Remove testimage from Nios2 build as it is currently failing.

Signed-off-by: Peter Hoyes <Peter.Hoyes@...>
Signed-off-by: Naveen Saini <naveen.kumar.saini@...>
---
.gitlab-ci.yml | 17 +++--------------
ci/testimage.yml | 1 +
2 files changed, 4 insertions(+), 14 deletions(-)

diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml
index 68abd32..014123f 100644
--- a/.gitlab-ci.yml
+++ b/.gitlab-ci.yml
@@ -35,17 +35,6 @@ stages:
paths:
- $CI_PROJECT_DIR/work/build/tmp/work*/**/temp/log.do_*.*

-# Workaround for Zephyr not currectly handling TESTIMAGE_AUTO
-.build_and_test:
- extends: .setup
- script:
- - KASFILES=$(./ci/jobs-to-kas "$CI_JOB_NAME")
- - kas shell --update --force-checkout $KASFILES -c 'cat conf/*.conf'
- - kas build $KASFILES
- - kas build $KASFILES -c testimage
- - ./ci/check-warnings $KAS_WORK_DIR/build/warnings.log
-
-
#
# Prep stage, update repositories once
#
@@ -92,10 +81,10 @@ stm32mp157c-dk2:
extends: .build

qemu-cortex-m3/testimage:
- extends: .build_and_test
+ extends: .build

-qemu-nios2/testimage:
+qemu-nios2:
extends: .build

qemu-x86/testimage:
- extends: .build_and_test
+ extends: .build
diff --git a/ci/testimage.yml b/ci/testimage.yml
index 7ef051b..83e17a7 100644
--- a/ci/testimage.yml
+++ b/ci/testimage.yml
@@ -6,3 +6,4 @@ local_conf_header:
IMAGE_CLASSES += "testimage"
TEST_TARGET = "QemuTargetZephyr"
TEST_SUITES = "zephyr"
+ TESTIMAGE_AUTO = "1"
--
2.17.1

Mostly stolen from meta-arm

NOTE: this differs from upstream in that the kirkstone branch is being
used instead of the master branch in ci/base.yml

Signed-off-by: Jon Mason <jon.mason@...>
Signed-off-by: Naveen Saini <naveen.kumar.saini@...>
---
.gitlab-ci.yml | 101 +++++++++++++++++++++++++++++++++++++
ci/96b-avenger96.yml | 6 +++
ci/96b-nitrogen.yml | 6 +++
ci/arduino-nano-33-ble.yml | 6 +++
ci/base.yml | 36 +++++++++++++
ci/check-machine-coverage | 26 ++++++++++
ci/check-warnings | 19 +++++++
ci/intel-x86-64.yml | 6 +++
ci/jobs-to-kas | 19 +++++++
ci/logging.yml | 13 +++++
ci/meta-openembedded.yml | 11 ++++
ci/nrf52840dk-nrf52840.yml | 6 +++
ci/qemu-cortex-m3.yml | 12 +++++
ci/qemu-nios2.yml | 10 ++++
ci/qemu-x86.yml | 10 ++++
ci/stm32mp157c-dk2.yml | 6 +++
ci/testimage.yml | 8 +++
ci/update-repos | 40 +++++++++++++++
18 files changed, 341 insertions(+)
create mode 100644 .gitlab-ci.yml
create mode 100644 ci/96b-avenger96.yml
create mode 100644 ci/96b-nitrogen.yml
create mode 100644 ci/arduino-nano-33-ble.yml
create mode 100644 ci/base.yml
create mode 100755 ci/check-machine-coverage
create mode 100755 ci/check-warnings
create mode 100644 ci/intel-x86-64.yml
create mode 100755 ci/jobs-to-kas
create mode 100644 ci/logging.yml
create mode 100644 ci/meta-openembedded.yml
create mode 100644 ci/nrf52840dk-nrf52840.yml
create mode 100644 ci/qemu-cortex-m3.yml
create mode 100644 ci/qemu-nios2.yml
create mode 100644 ci/qemu-x86.yml
create mode 100644 ci/stm32mp157c-dk2.yml
create mode 100644 ci/testimage.yml
create mode 100755 ci/update-repos

diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml
new file mode 100644
index 0000000..68abd32
--- /dev/null
+++ b/.gitlab-ci.yml
@@ -0,0 +1,101 @@
+image: ghcr.io/siemens/kas/kas:latest-release
+
+stages:
+ - prep
+ - build
+
+# Common job fragment to get a worker ready
+.setup:
+ stage: build
+ interruptible: true
+ variables:
+ KAS_WORK_DIR: $CI_PROJECT_DIR/work
+ KAS_REPO_REF_DIR: $CI_BUILDS_DIR/persist/repos
+ SSTATE_DIR: $CI_BUILDS_DIR/persist/sstate
+ DL_DIR: $CI_BUILDS_DIR/persist/downloads
+ BB_LOGCONFIG: $CI_PROJECT_DIR/ci/logging.yml
+ before_script:
+ - echo KAS_WORK_DIR = $KAS_WORK_DIR
+ - echo SSTATE_DIR = $SSTATE_DIR
+ - echo DL_DIR = $DL_DIR
+ - rm -rf $KAS_WORK_DIR
+ - mkdir --verbose --parents $KAS_WORK_DIR $KAS_REPO_REF_DIR $SSTATE_DIR $DL_DIR
+
+# Generalised fragment to do a Kas build
+.build:
+ extends: .setup
+ script:
+ - KASFILES=$(./ci/jobs-to-kas "$CI_JOB_NAME")
+ - kas shell --update --force-checkout $KASFILES -c 'cat conf/*.conf'
+ - kas build $KASFILES
+ - ./ci/check-warnings $KAS_WORK_DIR/build/warnings.log
+ artifacts:
+ name: "logs"
+ when: on_failure
+ paths:
+ - $CI_PROJECT_DIR/work/build/tmp/work*/**/temp/log.do_*.*
+
+# Workaround for Zephyr not currectly handling TESTIMAGE_AUTO
+.build_and_test:
+ extends: .setup
+ script:
+ - KASFILES=$(./ci/jobs-to-kas "$CI_JOB_NAME")
+ - kas shell --update --force-checkout $KASFILES -c 'cat conf/*.conf'
+ - kas build $KASFILES
+ - kas build $KASFILES -c testimage
+ - ./ci/check-warnings $KAS_WORK_DIR/build/warnings.log
+
+
+#
+# Prep stage, update repositories once
+#
+update-repos:
+ extends: .setup
+ stage: prep
+ script:
+ - flock --verbose --timeout 60 $KAS_REPO_REF_DIR ./ci/update-repos
+
+
+#
+# Bootstrap stage, machine coverage
+#
+
+# What percentage of machines in the layer do we build
+machine-coverage:
+ stage: prep
+ interruptible: true
+ script:
+ - ./ci/check-machine-coverage
+ coverage: '/Coverage: \d+/'
+
+
+#
+# Build stage, the actual build jobs
+#
+
+96b-avenger96:
+ extends: .build
+
+96b-nitrogen:
+ extends: .build
+
+arduino-nano-33-ble:
+ extends: .build
+
+intel-x86-64:
+ extends: .build
+
+nrf52840dk-nrf52840:
+ extends: .build
+
+stm32mp157c-dk2:
+ extends: .build
+
+qemu-cortex-m3/testimage:
+ extends: .build_and_test
+
+qemu-nios2/testimage:
+ extends: .build
+
+qemu-x86/testimage:
+ extends: .build_and_test
diff --git a/ci/96b-avenger96.yml b/ci/96b-avenger96.yml
new file mode 100644
index 0000000..6d632f1
--- /dev/null
+++ b/ci/96b-avenger96.yml
@@ -0,0 +1,6 @@
+header:
+ version: 9
+ includes:
+ - base.yml
+
+machine: 96b-avenger96
diff --git a/ci/96b-nitrogen.yml b/ci/96b-nitrogen.yml
new file mode 100644
index 0000000..ecd96fb
--- /dev/null
+++ b/ci/96b-nitrogen.yml
@@ -0,0 +1,6 @@
+header:
+ version: 9
+ includes:
+ - base.yml
+
+machine: 96b-nitrogen
diff --git a/ci/arduino-nano-33-ble.yml b/ci/arduino-nano-33-ble.yml
new file mode 100644
index 0000000..ca332da
--- /dev/null
+++ b/ci/arduino-nano-33-ble.yml
@@ -0,0 +1,6 @@
+header:
+ version: 9
+ includes:
+ - base.yml
+
+machine: arduino-nano-33-ble
diff --git a/ci/base.yml b/ci/base.yml
new file mode 100644
index 0000000..fb1699c
--- /dev/null
+++ b/ci/base.yml
@@ -0,0 +1,36 @@
+header:
+ version: 11
+ includes:
+ - meta-openembedded.yml
+
+distro: zephyr
+
+defaults:
+ repos:
+ refspec: kirkstone
+
+repos:
+ meta-zephyr:
+ layers:
+ meta-zephyr-core:
+ meta-zephyr-bsp:
+
+ poky:
+ url: https://git.yoctoproject.org/git/poky
+ layers:
+ meta:
+ meta-poky:
+
+env:
+ BB_LOGCONFIG: ""
+
+local_conf_header:
+ base: |
+ BB_SERVER_TIMEOUT = "60"
+ CONF_VERSION = "2"
+ INHERIT += "rm_work"
+
+machine: unset
+
+target:
+ - zephyr-kernel-test-all
diff --git a/ci/check-machine-coverage b/ci/check-machine-coverage
new file mode 100755
index 0000000..19f9571
--- /dev/null
+++ b/ci/check-machine-coverage
@@ -0,0 +1,26 @@
+#! /usr/bin/env python3
+
+from pathlib import Path
+import sys
+
+metazephyr = Path.cwd()
+
+if metazephyr.name != "meta-zephyr":
+ print("Not running inside meta-zephyr")
+ sys.exit(1)
+
+# All machine configurations
+machines = metazephyr.glob("meta-zephyr-bsp/conf/machine/*.conf")
+machines = set(p.stem for p in machines)
+
+# All kas files
+kas = metazephyr.glob("ci/*.yml")
+kas = set(p.stem for p in kas)
+
+missing = machines - kas
+print(f"The following machines are missing: {', '.join(sorted(missing))}.")
+
+covered = len(machines) - len(missing)
+total = len(machines)
+percent = int(covered / total * 100)
+print(f"Coverage: {percent}%")
diff --git a/ci/check-warnings b/ci/check-warnings
new file mode 100755
index 0000000..9d08010
--- /dev/null
+++ b/ci/check-warnings
@@ -0,0 +1,19 @@
+#! /bin/bash
+
+# Expects the path to a log file as $1, and if this file has any content
+# then display the contents and exit with an error code.
+
+set -e -u
+
+LOGFILE=$1
+
+LINES=$(grep --invert-match "relocations in \.text" $LOGFILE | wc -l)
+if test "$LINES" -ne 0; then
+ echo ==============================
+ echo The build had warnings/errors:
+ echo ==============================
+ cat $LOGFILE
+ exit 1
+fi
+
+exit 0
diff --git a/ci/intel-x86-64.yml b/ci/intel-x86-64.yml
new file mode 100644
index 0000000..525a38f
--- /dev/null
+++ b/ci/intel-x86-64.yml
@@ -0,0 +1,6 @@
+header:
+ version: 9
+ includes:
+ - base.yml
+
+machine: intel-x86-64
diff --git a/ci/jobs-to-kas b/ci/jobs-to-kas
new file mode 100755
index 0000000..7057970
--- /dev/null
+++ b/ci/jobs-to-kas
@@ -0,0 +1,19 @@
+#! /bin/bash
+
+# Read a GitLab CI job name on $1 and transform it to a
+# list of Kas yaml files
+
+set -e -u
+
+# Read Job namne from $1 and split on /
+IFS=/ read -r -a PARTS<<<$1
+
+# Prefix each part with ci/
+PARTS=("${PARTS[@]/#/ci/}")
+
+# Suffix each part with .yml
+PARTS=("${PARTS[@]/%/.yml}")
+
+# Print colon-separated
+IFS=":"
+echo "${PARTS[*]}"
diff --git a/ci/logging.yml b/ci/logging.yml
new file mode 100644
index 0000000..3af1029
--- /dev/null
+++ b/ci/logging.yml
@@ -0,0 +1,13 @@
+# Python logging configuration to write all warnings to a separate file
+version: 1
+
+handlers:
+ warnings:
+ class: logging.FileHandler
+ level: WARNING
+ filename: warnings.log
+ formatter: BitBake.logfileFormatter
+
+loggers:
+ BitBake:
+ handlers: [warnings]
diff --git a/ci/meta-openembedded.yml b/ci/meta-openembedded.yml
new file mode 100644
index 0000000..bed338d
--- /dev/null
+++ b/ci/meta-openembedded.yml
@@ -0,0 +1,11 @@
+header:
+ version: 11
+
+repos:
+ meta-openembedded:
+ url: https://git.openembedded.org/meta-openembedded
+ layers:
+ meta-filesystems:
+ meta-networking:
+ meta-oe:
+ meta-python:
diff --git a/ci/nrf52840dk-nrf52840.yml b/ci/nrf52840dk-nrf52840.yml
new file mode 100644
index 0000000..cbbf434
--- /dev/null
+++ b/ci/nrf52840dk-nrf52840.yml
@@ -0,0 +1,6 @@
+header:
+ version: 9
+ includes:
+ - base.yml
+
+machine: nrf52840dk-nrf52840
diff --git a/ci/qemu-cortex-m3.yml b/ci/qemu-cortex-m3.yml
new file mode 100644
index 0000000..b01480c
--- /dev/null
+++ b/ci/qemu-cortex-m3.yml
@@ -0,0 +1,12 @@
+header:
+ version: 11
+ includes:
+ - ci/base.yml
+
+local_conf_header:
+ nonbuilding_tests: |
+ ZEPHYRTESTS:remove = "common context pending poll sleep"
+ qemu_opts: |
+ QB_OPT_APPEND = "-icount shift=3,align=off,sleep=on -rtc clock=vm"
+
+machine: qemu-cortex-m3
diff --git a/ci/qemu-nios2.yml b/ci/qemu-nios2.yml
new file mode 100644
index 0000000..c382582
--- /dev/null
+++ b/ci/qemu-nios2.yml
@@ -0,0 +1,10 @@
+header:
+ version: 9
+ includes:
+ - base.yml
+
+local_conf_header:
+ nonbuilding_tests: |
+ ZEPHYRTESTS:remove = "interrupt"
+
+machine: qemu-nios2
diff --git a/ci/qemu-x86.yml b/ci/qemu-x86.yml
new file mode 100644
index 0000000..ba80dd3
--- /dev/null
+++ b/ci/qemu-x86.yml
@@ -0,0 +1,10 @@
+header:
+ version: 9
+ includes:
+ - base.yml
+
+local_conf_header:
+ failing_tests: |
+ ZEPHYRTESTS:remove = "pending"
+
+machine: qemu-x86
diff --git a/ci/stm32mp157c-dk2.yml b/ci/stm32mp157c-dk2.yml
new file mode 100644
index 0000000..c833794
--- /dev/null
+++ b/ci/stm32mp157c-dk2.yml
@@ -0,0 +1,6 @@
+header:
+ version: 9
+ includes:
+ - base.yml
+
+machine: stm32mp157c-dk2
diff --git a/ci/testimage.yml b/ci/testimage.yml
new file mode 100644
index 0000000..7ef051b
--- /dev/null
+++ b/ci/testimage.yml
@@ -0,0 +1,8 @@
+header:
+ version: 11
+
+local_conf_header:
+ testimage: |
+ IMAGE_CLASSES += "testimage"
+ TEST_TARGET = "QemuTargetZephyr"
+ TEST_SUITES = "zephyr"
diff --git a/ci/update-repos b/ci/update-repos
new file mode 100755
index 0000000..fa638aa
--- /dev/null
+++ b/ci/update-repos
@@ -0,0 +1,40 @@
+#! /usr/bin/env python3
+
+# Update clones of the repositories we need in KAS_REPO_REF_DIR to speed up fetches
+
+import sys
+import os
+import subprocess
+import pathlib
+
+def repo_shortname(url):
+ # Taken from Kas (Repo.__getattr__) to ensure the logic is right
+ from urllib.parse import urlparse
+ url = urlparse(url)
+ return ('{url.netloc}{url.path}'
+ .format(url=url)
+ .replace('@', '.')
+ .replace(':', '.')
+ .replace('/', '.')
+ .replace('*', '.'))
+
+repositories = (
+ "https://git.yoctoproject.org/git/poky",
+ "https://git.openembedded.org/meta-openembedded",
+)
+
+if __name__ == "__main__":
+ if "KAS_REPO_REF_DIR" not in os.environ:
+ print("KAS_REPO_REF_DIR needs to be set")
+ sys.exit(1)
+
+ base_repodir = pathlib.Path(os.environ["KAS_REPO_REF_DIR"])
+
+ for repo in repositories:
+ repodir = base_repodir / repo_shortname(repo)
+ if repodir.exists():
+ print("Updating %s..." % repo)
+ subprocess.run(["git", "-C", repodir, "fetch"], check=True)
+ else:
+ print("Cloning %s..." % repo)
+ subprocess.run(["git", "clone", "--bare", repo, repodir], check=True)
--
2.17.1

Kirkstone was broken for meta-zephyr for weeks, but no one noticed because Gitlab CI is only running on master. Backport the relevant patches from master so this can be more easily detected, and I'll run a nightly kirkstone to my personal Gitlab CI setup (which can be monitored at https://gitlab.com/jonmason00/meta-zephyr/-/pipelines).

NOTE: I was conflicted between squashing the uniquely CI patches into a single one to hide some of the bug fixesin he originals, versus pulling them back to match what was in master. I did the latter, but the former might be more acceptable.

CI run for this series can be seen at https://gitlab.com/jonmason00/meta-zephyr/-/pipelines/602974431

Thanks,
Jon

--

Jon Mason (5):
qemu-nios2: use glibc
CI: add Gitlab CI support
CI: use path to avoid warning
CI: move stm32mp157c-dk2 to be alphabetical
CI: add more targets

Peter Hoyes (2):
zephyrtest: Enable use of TESTIMAGE_AUTO
CI: Use TESTIMAGE_AUTO

.gitlab-ci.yml | 90 ++++++++++++++++++++
ci/96b-avenger96.yml | 6 ++
ci/96b-nitrogen.yml | 20 +++++
ci/arduino-nano-33-ble.yml | 21 +++++
ci/base.yml | 38 +++++++++
ci/check-machine-coverage | 26 ++++++
ci/check-warnings | 19 +++++
ci/intel-x86-64.yml | 6 ++
ci/jobs-to-kas | 19 +++++
ci/logging.yml | 13 +++
ci/meta-openembedded.yml | 11 +++
ci/nrf52840dk-nrf52840.yml | 20 +++++
ci/qemu-cortex-m3.yml | 12 +++
ci/qemu-nios2.yml | 10 +++
ci/qemu-x86.yml | 10 +++
ci/stm32mp157c-dk2.yml | 13 +++
ci/testimage.yml | 9 ++
ci/update-repos | 40 +++++++++
meta-zephyr-bsp/conf/machine/qemu-nios2.conf | 2 +
meta-zephyr-core/classes/zephyrtest.bbclass | 2 +-
meta-zephyr-core/conf/distro/zephyr.conf | 2 +-
21 files changed, 387 insertions(+), 2 deletions(-)
create mode 100644 .gitlab-ci.yml
create mode 100644 ci/96b-avenger96.yml
create mode 100644 ci/96b-nitrogen.yml
create mode 100644 ci/arduino-nano-33-ble.yml
create mode 100644 ci/base.yml
create mode 100755 ci/check-machine-coverage
create mode 100755 ci/check-warnings
create mode 100644 ci/intel-x86-64.yml
create mode 100755 ci/jobs-to-kas
create mode 100644 ci/logging.yml
create mode 100644 ci/meta-openembedded.yml
create mode 100644 ci/nrf52840dk-nrf52840.yml
create mode 100644 ci/qemu-cortex-m3.yml
create mode 100644 ci/qemu-nios2.yml
create mode 100644 ci/qemu-x86.yml
create mode 100644 ci/stm32mp157c-dk2.yml
create mode 100644 ci/testimage.yml
create mode 100755 ci/update-repos

--
2.17.1

Hi Joshua,

Everything is around the /etc/version file. I know that this file is generated automatically and this is hardcoded in Yocto. I've to check how to found a workaround on this.

Sorry for this.

Have a great day