next branch towards YP 2.6 (thud)
Description
Environment
Attachments
- 09 Apr 2019, 04:32 PM
relates to
Activity
Walt Miner June 4, 2019 at 10:22 PM
Close for HH RC1
Scott Murray April 26, 2019 at 4:39 PM
Upgrade merged into master, please open new issues for any problems per Walt's request.
Walt Miner April 26, 2019 at 12:50 PM
Now that the sandbox has been merged to master I think we should RESOLVE this issue and open new issues for anything that is found going forward.
jose bollo April 9, 2019 at 4:50 PM
I attached the diff between my upgrade of dbus and the one of scott because I found issues. Some of the given changes are just formatting and don't care. But some other are really different.. The weird affair is that confusing BusResult and dbus_bool_t might invert the logic as shows their definitions:
In dbus/dbus-types.h and dbus/dbus-macros.h, the standard expected definition:
typedef dbus_uint32_t dbus_bool_t;
#ifndef TRUE
# define TRUE 1
#endif
#ifndef FALSE
# define FALSE 0
#endifBut in bus/bus.h the definition of BusResult is surprizing:
typedef const struct BusResultStruct { int dummy; } *BusResult;
static const BusResult BUS_RESULT_TRUE = (BusResult)0x0;
static const BusResult BUS_RESULT_FALSE = (BusResult)0x1;
static const BusResult BUS_RESULT_LATER = (BusResult)0x2;I'm not the author and can't explain why it is defined that way. Anyway the confusion is to avoid!!!
Stephane Desneux April 5, 2019 at 1:26 PM(edited)
@Harunobu Kurokawa: meta-agl updated (sandbox/sdesneux/thud-upgrade branch)... Rebuilding..
Now I get HMI correctly 
Sounds good !
Details
Details
Assignee
Reporter
Fix versions
Labels
Contract ID
Priority
Major
Preparation / Uprev to Yocto Project 2.6 (thud) as base of AGL.
Latest status (20190404)
Main boards status
Machine
Build status
HMI
4A
Comment
qemux86-64
OK
OK
OK with tweak**
4A starts only if AVIRT is not used (workaround: rm /usr/libexec/agl/smixer/etc/smixer-4a-avirt.json && reboot -f). AVIRT failure is very probably due to missing kernel config fragment in qemu img => another issue
m3ulcb
OK
OK
OK with tweak
4A starts only if AVIRT is not used (workaround: rm /usr/libexec/agl/smixer/etc/smixer-4a-avirt.json && reboot -f)
intel-corei7-64
OK
OK
OK**
Observing issue SPEC-2211 on up2
dra7xx-evm
unknown
unknown
unknown
raspberrypi3
OK
unknown
unknown
Does not boot, stuck at "Starting kernel"
m3ulcb-nogfx
OK
OK (none)
unknown
h3ulcb
OK
OK
OK with tweak
same as m3ulcb
h3. Other boards status
Machine
Build status
Runtime status
Comment
beaglebone
unknown
unknown
cyclone5
unknown
unknown
dragonboard-410c
unknown
unknown
[sdesneux: Should it still be supported?]
dragonboard-820c
unknown
unknown
h3-salvator-x
unknown
unknown
imx6qsabreauto
unknown
unknown
imx6qsabrelite
unknown
unknown
joule
unknown
unknown
nitrogen6x
unknown
unknown
qemuarm
unknown
unknown
qemuarm64
unknown
unknown
raspberrypi2
unknown
unknown
[sdesneux: Should it still be supported?]
raspberrypi3-64
unknown
unknown
h2. Build instructions
IMPORTANT: it seems that the new handling of git submodules introduced recently on thud breaks the validity of "old" caches (in particular the download cache). So if you hit some issues with do_unpack tasks on agl-service-* packages, try to restart the build after a full wipe of the bitbake caches.
To build for thud, use the following commands:
repo init -b sandbox/sdesneux/thud-upgrade -m default.xml -u [https://gerrit.automotivelinux.org/gerrit/AGL/AGL-repo] repo sync [run aglsetup, bitbake ...]