On Thu, Jun 02, 2016 at 02:26:47PM +0800, Fengwei Yin wrote:

On 2016/6/2 13:07, Leo Yan wrote:

...

Hi John,

[ + linaro-android ML ]

I have two questions for working on AOSP Hikey:

• When use "adb shell" to execute commands, then it will return back

the execution value and error code. Usually we will get back the new line character as "\r\n", but I found for AOSP releasing it will only return back "\n" for new line.

So is this decided by system's booting script or decided by shell's building? I want to change back to use "\r\n" for the format.

• Second question is related with adb's permission. On my pc I need

add "sudo" or change to root "sudo -s" for adb commands, otherwise it will report "insufficient permission".

This is okay when I manually send adb commands, but when use python script to execute adb commands it also need add "sudo" for add adb related commands.

So want to check if there have more convienent method so can let normal user to use adb command?

we need "sudo" for adb/fastboot because udev uses 0664 permission to create entry in /dev. You could change the udev rules to use 0666. Then "sudo" is not necessary to run adb/fastboot. Same thing applied to usb serial as well.

Thanks, Fengwei.

At beginning I reluctantly add udev rules by self. But now looks this is the most smooth method[1].

[1] http://wiki.cyanogenmod.org/w/UDEV

Thanks, Leo Yan

On 2 June 2016 at 12:41, Jorge Ramirez jorge.ramirez-ortiz@linaro.org wrote:

the LISA pm [1] tools also check for a property to determine if the Android system booted properly; to do that, it calls devlib [2] boot_completed = boolean(self.getprop('sys.boot_completed'))

We fail this test on the HiKey AOSP. Anyone knows if there is any reason for this is not being set? or should devlib just remove this check?

I have a fairly old (May 14) AOSP build but Hikey does set that property. I rely on that property too when I'm too lazy to switch on the display :)

This is what I see:

hikey:/ # getprop | grep boot [dev.bootcomplete]: [1] [init.svc.bootanim]: [stopped] [ro.boot.console]: [ttyFIQ0] [ro.boot.hardware]: [hikey] [ro.boot.selinux]: [permissive] [ro.bootimage.build.date]: [Sat May 14 15:28:06 IST 2016] [ro.bootimage.build.date.utc]: [1463219886] [ro.bootimage.build.fingerprint]: [generic/hikey/hikey:6.0.1/MASTER/pundir05141528:userdebug/test-keys] [ro.bootloader]: [unknown] [ro.bootmode]: [unknown] [ro.runtime.firstboot]: [39332] [service.bootanim.exit]: [1] [sys.boot_completed]: [1]

Regards, Amit Pundir

[1] https://github.com/ARM-software/lisa [2] https://github.com/ARM-software/devlib/blob/master/devlib/target.py#L744

---

linaro-android mailing list linaro-android@lists.linaro.org https://lists.linaro.org/mailman/listinfo/linaro-android

On 2 June 2016 at 15:10, Jorge Ramirez jorge.ramirez-ortiz@linaro.org wrote:

On 06/02/2016 11:20 AM, Amit Pundir wrote:

...

On 2 June 2016 at 12:41, Jorge Ramirez jorge.ramirez-ortiz@linaro.org wrote:

...

the LISA pm [1] tools also check for a property to determine if the Android system booted properly; to do that, it calls devlib [2] boot_completed = boolean(self.getprop('sys.boot_completed'))

We fail this test on the HiKey AOSP. Anyone knows if there is any reason for this is not being set? or should devlib just remove this check?

I have a fairly old (May 14) AOSP build but Hikey does set that property. I rely on that property too when I'm too lazy to switch on the display :)

This is what I see:

hikey:/ # getprop | grep boot [dev.bootcomplete]: [1] [init.svc.bootanim]: [stopped] [ro.boot.console]: [ttyFIQ0] [ro.boot.hardware]: [hikey] [ro.boot.selinux]: [permissive] [ro.bootimage.build.date]: [Sat May 14 15:28:06 IST 2016] [ro.bootimage.build.date.utc]: [1463219886] [ro.bootimage.build.fingerprint]: [generic/hikey/hikey:6.0.1/MASTER/pundir05141528:userdebug/test-keys] [ro.bootloader]: [unknown] [ro.bootmode]: [unknown] [ro.runtime.firstboot]: [39332] [service.bootanim.exit]: [1] [sys.boot_completed]: [1]

Regards, Amit Pundir

thanks Amit. I see the following on my end

hikey:/ $ getprop | grep boot [service.bootanim.exit]: [0] hikey:/ $ getprop | grep boot [init.svc.bootanim]: [running]

--> Boot Animation is still running. Are you sure you have booted to UI? I doiubt that. Can you please check or share your logcat for failures.

[ro.bootimage.build.date]: [Tue May 24 16:42:46 CEST 2016] [ro.bootimage.build.date.utc]: [1464100966] [ro.bootimage.build.fingerprint]: [generic/hikey/hikey:6.0.1/MASTER/jramir05241642:userdebug/test-keys] [ro.bootloader]: [unknown] [ro.bootmode]: [unknown] [service.bootanim.exit]: [0]

do I need to have a monitor plugged for boot to complete? I am working headless

Did you get this sorted out?

With my current build, I see:

[sys.boot_completed]: [1]

On Thu, Jun 2, 2016 at 2:57 AM, Jorge Ramirez jorge.ramirez-ortiz@linaro.org wrote:

On 06/02/2016 11:44 AM, Amit Pundir wrote:

...

On 2 June 2016 at 15:10, Jorge Ramirez jorge.ramirez-ortiz@linaro.org wrote:

...

On 06/02/2016 11:20 AM, Amit Pundir wrote:

...

On 2 June 2016 at 12:41, Jorge Ramirez jorge.ramirez-ortiz@linaro.org wrote:

...

the LISA pm [1] tools also check for a property to determine if the Android system booted properly; to do that, it calls devlib [2] boot_completed = boolean(self.getprop('sys.boot_completed'))

We fail this test on the HiKey AOSP. Anyone knows if there is any reason for this is not being set? or should devlib just remove this check?

I have a fairly old (May 14) AOSP build but Hikey does set that property. I rely on that property too when I'm too lazy to switch on the display :)

This is what I see:

hikey:/ # getprop | grep boot [dev.bootcomplete]: [1] [init.svc.bootanim]: [stopped] [ro.boot.console]: [ttyFIQ0] [ro.boot.hardware]: [hikey] [ro.boot.selinux]: [permissive] [ro.bootimage.build.date]: [Sat May 14 15:28:06 IST 2016] [ro.bootimage.build.date.utc]: [1463219886] [ro.bootimage.build.fingerprint]: [generic/hikey/hikey:6.0.1/MASTER/pundir05141528:userdebug/test-keys] [ro.bootloader]: [unknown] [ro.bootmode]: [unknown] [ro.runtime.firstboot]: [39332] [service.bootanim.exit]: [1] [sys.boot_completed]: [1]

Regards, Amit Pundir

thanks Amit. I see the following on my end

hikey:/ $ getprop | grep boot [service.bootanim.exit]: [0] hikey:/ $ getprop | grep boot [init.svc.bootanim]: [running]

--> Boot Animation is still running. Are you sure you have booted to UI? I doiubt that. Can you please check or share your logcat for failures.

ah ok I see errors... attached.