Age | Commit message (Collapse) | Author | Files | Lines |
|
* Allow script execution at the end of the boot.
* Yocto first boot service is install at for "sysinit.target" level.
AGL need to have a custom service, exec when target service system,
(SYSTEMD_SERVICE_AFTER), is up.
* All script inside directory "/etc/agl-postinsts" will be exec at
the first boot.
* After the first boot the service "run-agl-postinsts.service"
is disable.
Change-Id: Ib12b7e11b451b4a41c4a31f5cda3686679fe9c08
Signed-off-by: Ronan Le Martret <ronan.lemartret@iot.bzh>
(cherry picked from commit 67f1d7f631f20dfb4be3765ec32dfcd237d649c5)
|
|
Automotive packagegroup recipe kept, but empty.
Future specific automotive packages may be added later.
Change-Id: Iebfea2a0f0671c3fa3e03a5d96dd1eeb0453eaf0
Signed-off-by: Romain Forlot <romain.forlot@iot.bzh>
(cherry picked from commit 36086bd8afb8442829460a9e60b8314d45bf9bdf)
|
|
Since the Weston DRM backend does not work with the video card implementations
in VirtualBox and VMWare player, and has some outstanding issues on QEMU that
have resulted in GenIVI not enabling it[1], disable it for now. A comment has
been added to the QEMU weston.ini indicating that the DRM backend can be
re-enabled to potentially gain higher performance.
Additionally, the common weston.service systemd unit has been modified to
add the --tty option to weston-launch to start weston on /dev/tty1. This is
now required to avoid console messages writing on top of the HomeScreen in
QEMU, and is harmless to operation on the other platforms.
[1]: https://github.com/GENIVI/meta-genivi-dev/pull/34
Bug-AGL: SPEC-405
Change-Id: Ic0c0adeb15baf392e3ffa0bbd26c62d060df1ef5
Signed-off-by: Scott Murray <scott.murray@konsulko.com>
|
|
Fix the issue that radio app using r820t tuner driver some times crash.
Bug-AGL: SPEC-418
Change-Id: I1529ecba91c3988eea6b271d5f8ce6d2d8f1bb11
Signed-off-by: Harunobu Kurokawa <harunobu.kurokawa.dn@renesas.com>
|
|
System freezes tracing from In bcm2836_arm_irqchip_handle_irq
Bug-AGL: SPEC-366
Change-Id: Ifd7e587495950eab22e6b6bb29c093a923fa4161
Signed-off-by: Philippe Coval <philippe.coval@osg.samsung.com>
|
|
Change-Id: I9b512a907398de9c96dc46111c74bed4059f7b8e
Signed-off-by: Romain Forlot <romain.forlot@iot.bzh>
(cherry picked from commit b0cb0935aef4ced9c82d7972ed61f636e9036055)
|
|
Prevent pkg-config not found issue
This patch was in iotivity 1.1-rel branch,
and (wrongly) removed in 1.2.0
So it needed to be reintroduced in upstream 1.2-rel
and as a patch for 1.2.0, track upstream bug at:
https://jira.iotivity.org/browse/IOT-1219
Bug-AGL: SPEC-158
Thanks-to: Alison Chaiken
Change-Id: I6af59a590f8be489398623d29cf85c2f00000002
Signed-off-by: Philippe Coval <philippe.coval@osg.samsung.com>
(cherry picked from commit 7dd5a4100ffb71337c59046120c3c41547cb7f2b)
|
|
This changeset removes tar-native from ASSUME_PROVIDED in an attempt
to mitigate the tar-native extraction issue that we see on the jenkins slaves.
Is I understand the race-window, tar-native is just being extracted and already
in a preferred PATH while being extracted. The right fix would likely be
to extract first, then change PATH after the file is in place. A way to
do this is being researched.
Change-Id: I0b138a1bd27406e719ab0d711cf542a812fb8460
Signed-off-by: Jan-Simon Möller <jsmoeller@linuxfoundation.org>
|
|
This changeset patches the rpi with the needed changes to
allow bluetooth to work together with smack. The patches were
developed for porter and are a backport of patches submitted
for 4.10+ .
Bug-AGL: SPEC-293
Change-Id: I96d04a709de6ad7843840d9ae154f041173948a2
Signed-off-by: Jan-Simon Möller <jsmoeller@linuxfoundation.org>
|
|
This patch is ported from meta-agl-extra
Commit details below:
commit c50805d25ba95473e8b4d1eb28d1203a328cd77a
Author: José Bollo <jose.bollo@iot.bzh>
Change-Id: I54a40935d6d2bbd9c267985c9e23a14bdffb7a3c
Signed-off-by: Karthik Ramanan <a0393906@ti.com>
(cherry picked from commit 554f8d394c626d3a151127e67caddc3196e3ffc1)
|
|
This change add support for the official touchscreen that is connected throught the DSI-1
port to the board.
Change-Id: If85490ca69310922b8bf314681d1d8459995c61e
Signed-off-by: Jan-Simon Möller <jsmoeller@linuxfoundation.org>
|
|
Change-Id: Id5a1a8e0cabf8181342103aa8aae8413ceb30d5c
Signed-off-by: Jan-Simon Möller <jsmoeller@linuxfoundation.org>
|
|
Apply patch from iotivity-1.2.1:
* [PATCH 081/341] yocto: Add aarch64 for DragonBoard-410c
Bug-AGL: SPEC-158
Change-Id: I06f56ae860760002a6376866488e1f55624f50cc
Signed-off-by: Philippe Coval <philippe.coval@osg.samsung.com>
|
|
Add ENABLE_CMA setting that enables required kernel configuration
to allow the dynamic CMA to function.
Change-Id: I0b0ab0fd7fd64454ea3cc9a89344a98897838e11
Signed-off-by: Matt Ranostay <matt.ranostay@konsulko.com>
(cherry picked from commit fefb358dca0ac89b970c77671b2efda2a8cf99c0)
|
|
Allow the raspberrypi3 GPU to use the CMA dynamic memory split
to boost performance.
Change-Id: Ia72e7b83f7cca2b6b8a622c632345c748547429b
Signed-off-by: Matt Ranostay <matt.ranostay@konsulko.com>
(cherry picked from commit db6bd4b3cb83a97ccde587286e16ebbaaff57d02)
|
|
For dynamically allocating of CMA blocks the low and high water marks
settings determine when the GPU requests or releases memory from the
CPU.
Change-Id: Iad494c101fec0884bb687c5ea51c9657b7c17bd9
Signed-off-by: Matt Ranostay <matt.ranostay@konsulko.com>
(cherry picked from commit 6bcd449f3b6f9705dfe940d2011fce6243bac7de)
|
|
Change-Id: I90abab207aa0b874c4043e7aec9a4bb3b85f713e
Signed-off-by: Jan-Simon Möller <jsmoeller@linuxfoundation.org>
|
|
Set CMDLINE_DEBUG in linux-raspberrypi_4.4.bbappend to disable the debug
logging by the kernel and systemd that was spamming the console. As a side
benefit, this noticeably speeds up booting.
Change-Id: Ic99b1626ea20a27d3d157bcb952ec30cd50baeb2
Signed-off-by: Scott Murray <scott.murray@konsulko.com>
|
|
ivi-controller
The meta-arago layer included a number of ivi shell patches that are
not needed in AGL. One multi screen patch conflicts with another
bbappend causing build issues as it's already included by a less
specific layer. Further, the added "simple TI layout controller",
autolaunch, and screenshooter features are all unneeded in AGL. These
are removed to reduce the amount of non-essential patches we carry within
AGL. As a result of this, all weston.ini directives which use the
"simple TI layout controller" directives are removed and weston.ini
is switched to use the standard ivi-controller so that WindowManager may
run and access the ILM API. The end result is that WindowManager and
HomeScreen may now be launched after boot successfully.
Change-Id: I8d777fbd4c02acbf9d3ccaeb215b651a8562d5bb
Signed-off-by: Matt Porter <mporter@konsulko.com>
|
|
Change-Id: I8661db791c60feac41814173b666710409480c7c
Signed-off-by: Toshiaki Isogai <tisogai@jp.adit-jv.com>
|
|
Removed linux-firmware-ibt since it is only necessary for the
Joule platform.
Change-Id: Ia4916fbc777299df4a094874ad1585a43a46d39b
Signed-off-by: Kien Dinh <kien.t.dinh@intel.com>
|
|
Bug-AGL: SPEC-366
Change-Id: Id15035817fcf110d52354fe4822af93d49b8c9ec
Signed-off-by: Anton Gerasimov <anton@advancedtelematic.com>
|
|
The psplash is oriented 180 degrees rotated compared with the
HomeScreen application, this change will ensure that they are in
sync
Change-Id: Ic2a588ff3a061ff500d4050a6915fe5c899a1e7e
Signed-off-by: Karthik Ramanan <a0393906@ti.com>
|
|
It was discovered while building iotivity-1.2.0 using Intel BSP
Bug-AGL: SPEC-158
Change-Id: I4c735a276d3a484a330ef1ea8611105e6a84e35a
Signed-off-by: Philippe Coval <philippe.coval@osg.samsung.com>
|
|
The latest dra7xx-evm purchased from ti.com comes with the OSD
display, this change will allow the AGL nightly images to work
on the latest dra7xx-evm by default.
If the user has an older EVM, then the uEnv.txt should be modified
to match the connected displays. The complete list of supported
LCDs can be found by checking the /boot folder of the the rootfs.
Change-Id: I7f7ac06a49670eb8b7446a6b27854bee4b76ace6
Signed-off-by: Karthik Ramanan <a0393906@ti.com>
|
|
Tested and verified on the Rev H EVM with OSD display
Should work on all other supported displays too.
Change-Id: Ib8f210ebf84a0f36fbc87c4553d13ff984874051
Signed-off-by: Karthik Ramanan <a0393906@ti.com>
|
|
The audio bug is fixed (SPEC-365), so that it should be
revert this commit to enable sound driver.
This reverts commit 1a70e5e8f48f17f604ab243dfb3ff01f7aa435f6.
Change-Id: I362a900e86cdf7888e5726a6a0a56daeb0d33145
Signed-off-by: Harunobu Kurokawa <harunobu.kurokawa.dn@renesas.com>
|
|
This enables wifi dongle of ralink, for example BUFFALO wifi dongle.
Change-Id: Iab408fd5f54feecfadb4f25adab978b0df89d66d
Signed-off-by: Kazumasa Mitsunari <kazumasa_mitsunari_za@mail.toyota.co.jp>
|
|
modules
Remove some launched applications that either don't exist or were used for basic
BSP testing. Add drm-backend and ivi-input-controller modules. Fixes weston config
so the HomeScreen can now be launched.
Change-Id: I2c6e507ccd244c26e8b5f5cf0ad0f68ae9a18495
Signed-off-by: Matt Porter <mporter@konsulko.com>
|
|
This adds neccessary firmware to support the Joule board, including wifi and
bluetooth.
Change-Id: Ia6973415113119aaf60fff0dbb494027fba96fd5
Signed-off-by: Kien Dinh <kien.t.dinh@intel.com>
|
|
The Porter board currently has a problem that pulseaudio can not be used due to the sound driver.
I enabled USB Audio as a workaround, but we had to manually change the default sound device.
As a further workaround, disable the SoC sound driver on the Porter board.
Change-Id: Ie53b75ea0bc252b82910e518da62d2748147fc90
Signed-off-by: Naoto Yamaguchi <i33399_YAMAGUCHI@aisin-aw.co.jp>
|
|
Change-Id: I11b245aea28b76f17cba3cdc8c964129bbaf906b
Signed-off-by: Dominig ar Foll (Intel Open Source) <dominig.arfoll@fridu.net>
|
|
This patch is to avoid the problem that pulseaudio can not be used in the Porter board.
This patch does the following two correspondences.
1. Enable USB Audio config
2. Corresponding to problem of self-destruction by outputting a lot of printk
Change-Id: I58682c0755e04e87ec198dd0494d6f9380ebe4ce
Signed-off-by: Naoto Yamaguchi <i33399_YAMAGUCHI@aisin-aw.co.jp>
|
|
If HDMI and LVDS were both enabled, parameters about screens
become inconsistent on weston. As a workaround until root cause
is fixed, disabling LVDS from weston.ini.
Change-Id: I150c0ba5e9655974f7dc365c2168fe7e2d7d93ab
Signed-off-by: Tadao Tanikawa <tanikawa.tadao@jp.panasonic.com>
|
|
Add the commented out mode line for using the GeChic 1502i in 1080p so
that it's easy to enable it for the CES demo setup.
Signed-off-by: Scott Murray <scott.murray@konsulko.com>
Change-Id: Idcdf25188d3391403ee29c5bda7568c652b4c730
|
|
This patch is required to use SubSurface in the AGL Distro application.
This patch is necessary for at least navigation application.
Change-Id: I6faaafed3db52f048bc329050de290cba67a546b
Signed-off-by: Naoto Yamaguchi <i33399_YAMAGUCHI@aisin-aw.co.jp>
|
|
Change-Id: Id899777abafd84bc2c7eb9d3ba0a22906a45c7cb
Signed-off-by: Dominig ar Foll (Intel Open Source) <dominig.arfoll@fridu.net>
|
|
AGL demo UI uses screen virtically. This splash screen could be the same
orientation.
Change-Id: I62333f009dca94da79c8257ee7da3069dd86bc2e
Signed-off-by: Tasuku Suzuki <tasuku.suzuki@qt.io>
|
|
This version of bluez, a little forward of 5.43,s
includes working features for using BLE GATT services
and attributes.
This is needed for a demo that uses DIGIPASS SecureClick
of Vasco.
To be removed when yocto upgrades to bluez 5.44 or higher.
Signed-off-by: José Bollo <jose.bollo@iot.bzh>
|
|
Change-Id: Ic60a3d8271a14729032f0757df8aa2f8108b7798
Signed-off-by: Dominig ar Foll (Intel Open Source) <dominig.arfoll@fridu.net>
|
|
Change-Id: Ie854777d28f455082bafdee07f36ab21c025dc68
Signed-off-by: Dominig ar Foll <dominig.arfoll@fridu.net>
|
|
runtime dependency with x11
Change-Id: I175b0d9900d61546cb0f116a1d698ee1f60b84b7
Signed-off-by: Dominig ar Foll (Intel Open Source) <dominig.arfoll@fridu.net>
Correct libva incorrect runtime dependency with x11
Change-Id: Ic4231c217c8c63575a98d370c898b8cb04e3a980
Signed-off-by: Dominig ar Foll (Intel Open Source) <dominig.arfoll@fridu.net>
|
|
* missing namespace into defconf
Change-Id: I9f292f8186fb771f8eb8c82ad8e7e985c474e9ea
Signed-off-by: Ronan <ronan.lemartret@iot.bzh>
|
|
* weston-init provides weston.service as weston.
We only want one recipe which provides weston.service
Change-Id: I2b9db570d466cd8f4007e2c28e0231f77d0a36ca
Signed-off-by: Ronan <ronan.lemartret@iot.bzh>
|
|
meta-agl-security only contains one recipe for xmlsec so it's not necessary
to maintain a full layer for this
Change-Id: I66b3eeaa0e777fcc0c94bdf562b7188b01641d86
Signed-off-by: Stephane Desneux <stephane.desneux@iot.bzh>
|
|
Change-Id: I9c00765c3c1eece18a80c7b9f16a79a2bb9d2e5a
Signed-off-by: Ronan <ronan.lemartret@iot.bzh>
|
|
Upstream YP introduced
http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?h=krogoth&id=529bbe2cc2cebeec0ee6cef27aba2d1aafe79d38
Thus this adds a follow-up fix for openssl.
Change-Id: Ifb55c3022596a6105662618f4cd08bd0165f5a6b
Signed-off-by: Romain Forlot <romain.forlot@iot.bzh>
|
|
* optee-os use python script to build so,
We need to add:
inherit pythonnative
If not we use host python during the build.
* During the build, optee-os use script:
- scripts/pem_to_pub_c.py
- scripts/sign.py
Those script depend on python-pycrypto
"from Crypto.PublicKey import RSA"
So we need to add python-pycrypto native to recipes
* The python-pycrypto recipes come from meta-security (out of AGL):
https://layers.openembedded.org/layerindex/recipe/39101/
Change-Id: I64340544c9a23a8e9ef62aca348fc08722d3dcc2
Signed-off-by: Ronan <ronan.lemartret@iot.bzh>
|
|
Due to:
Can't install libva-dev-1.7.0-r0@corei7_64: Can't install libva-egl1-1.7.0-r0@corei7_64: no package provides libva-x11
See
https://jenkins-new.automotivelinux.org/job/sandbox-snapshot-master/MACHINE=intel-corei7-64,label=agl-test-slave/146/console
We need to fix libva through a bbappend.
This reverts commit c678e2d040c6071f3b6bec593d3c6271504a92e7.
Change-Id: I486aa82484f8b62c65b7ae1a7607f75ade26ce91
Signed-off-by: Jan-Simon Möller <jsmoeller@linuxfoundation.org>
|
|
ROOTFS_SIZE is recomputed for each image type based on total size of
rootfs directory. If rootfs directory gets smaller between creation of
rootfs image and SD image, it fill result in corrupted SD image.
It should not happen in theory, but it happens occasionally for a reason
I could not identify.
Change-Id: I1c888c179fdcec32c3cacbbce7b2d34ec70bbb17
Signed-off-by: Anton Gerasimov <anton@advancedtelematic.com>
|