✏ B2G beta (formerly Firefox OS)

Sorry for the late answer, I have built the new images and after flashing I recognized a strange behavior: The system starts as usual and one could use the phone but after a few seconds the screen freezes and gets black afterwards, in this mode the system discharges the battery very fast and one could only restart by removing the battery.
This image is able to handle the sim but wifi still doesn’t work and adb doesn’t work either. In the directory of the compiled images I found

but

was not in this directory.

Too bad that there is no adb. Does the phone boot/work better without the sim card? Some sim cards can cause problems, sometimes it helps testing the phone without them first.

I think thats about the same as i had a few days befor. Im now unshure if it is because i just flashed it again or because i flashed with the new build. So maybe try to rebuild today. For me that helped.

@fp1_wo_sw_updates i don’t find the first file in my FPOSOS build and in the FFOS build. I made a seach and picked a screenshoot:
FFOS:


FPOSOS:

Are you shure you have this file in the persist folder?
Regards, Novski

@fp1_wo_sw_updates I tried it without my SIM inserted and nothing changed.
@novski sure

Okay so I will try another build.

Just an idea, thanks for trying it out!

It’s not so important where the file is, but that it gets found. The logfile provided showed that there is a problem reading/opening the file, if I understood all that correctly.

More logfiles (logcat, dmesg) are needed. Just look for the filename inside the logs, so you don’t have to publish too much info. I think the software checks in both places.

Maybe not what you want to hear here, but there is a sailfish release out there since yesterday, if you want to test another alternative in the meantime. I’m not against mozilla, just want to mention it. It’s good to have fallbacks.

wow, its beautifull. But phonecalls don’t work on my side…
the display goes black after pressing the number and connect…
I’l gona test it a bit further…
Thanks for the note!
Regards, Novski

For a few days B2G always failed in the building process. I found out that they were caused by this bugfix: 832165 - [Build] GAIA_PATH not working when you set it in .userconfig file as env of |./build.sh|
@fp1_wo_sw_updates As you mentioned the file WCNSS_qcom_wlan_nv.bin should be found on two locations and according to the Makefiles in device-fairphone2 this should be realized by copying it to /persist and linking to that file:
FP2.mk line 76

#wlan driver
PRODUCT_COPY_FILES += \
    device/fairphone_devices/FP2/WCNSS_qcom_cfg.ini:system/etc/wifi/WCNSS_qcom_cfg.ini \
    device/fairphone_devices/FP2/WCNSS_qcom_wlan_nv.bin:persist/WCNSS_qcom_wlan_nv.bin

AndroidBoard.mk line 148

#Create symbolic links
$(shell mkdir -p $(TARGET_OUT_ETC)/firmware/wlan/prima;
ln -sf /persist/WCNSS_qcom_wlan_nv.bin
$(TARGET_OUT_ETC)/firmware/wlan/prima/WCNSS_qcom_wlan_nv.bin;
ln -sf /data/misc/wifi/WCNSS_qcom_cfg.ini
$(TARGET_OUT_ETC)/firmware/wlan/prima/WCNSS_qcom_cfg.ini)

These Makefiles confuse me a lot.

Hi. I only checked where in the the log file the system was searching for the wlan drivers and having issues. If they aren’t there in your build, just copy them there.

The Makefile as shown just creates two symbolic links from the “persist” folder to the newly created (mkdir) target etc folder. For me all this looks good. If you look at the system image, you should be able to find the file under:

etc/firmware/wlan/prima/WCNSS_qcom_wlan_nv.bin

as a symbolic link like this:
./system/etc/firmware/wlan/prima/WCNSS_qcom_wlan_nv.bin -> /persist/WCNSS_qcom_wlan_nv.bin

Maybe you have a broken symbolic link?

If /persist/WCNSS_qcom_wlan_nv.bin is not around or does not get copied to the image, the file will not be around. So you need the file that provides the persist “folder” as well.

Normally there should be a “persist.img” as well that supplies this folder.

Note: I hope this will not kill your system, but maybe you need to do something like fastboot flash persist persist.img. Have you found something like this for Firefox OS? Please check if something similar is done during the flash process.

@fp1_wo_sw_updates I don’t think that we have to flash the persist image because the same file also exists in the stock partition as you can see with the FPOSOS.
I’m not that familiar with repo do you know a way to revert all commits that have done after a certain day? Every time I try to build the image it stops on some really silly places. I’m gonna despair on that.

Update regarding the status of Firefox OS Beta.
Mozilla is currently under pressure in porting the project to the community and doesn’t have enough spare time to debug our problem.
So we have to wait until that is done in about 3-4 Weeks. Anyway maybe that is better as i see more new blobs coming up in the Fairphone repo. and soon in a update for FPOS.
The good thing is that one Developer from Mozilla has a FP2 Device and was today able to reproduce the Wi-Fi / Data fetching problems i reported. So he has the possibility to debug it him self as soon as he has a spare minute…
He will inform us via the bugtracker over at bugzilla. Link
And i will post it here as soon as i get informed.
So whe have to stick to the creepy system called android on our beloved Hardware for just a few weeks more.
Regards, Novski

2 Likes

@novski Thank you for your engagement onto this topic. I’m sorry that I didn’t post a bug, but I was not able for a long time to build a current version. Have you found out what could cause this problem? For me it seems that selinux is denying access to the driver file. Did Juan have a guess?
sebastku

As I said above, I also think it’s an access issue. And don’t worry about sibon-2.0.0 vs sibon-2.0.1 it’s the same stuff. I haven’t checked sibon-2.0.3 yet.

Interesting news for previous Firefox OS, now B2G. There is a Manifesto currently under development.

2 Likes

Hi there, I let you know Juan updated the repo.
So you can try to get it running now.

I had problems by only updateing the repo. FP2 got stuck in the grey “Loading…” page in startup cycle.
It would be interesting if that hapens to others as well.
Im currently downloading the full repo from scratch to go sure its fresh.

On Juan’s FP2 it works fine so, have fun! :grinning:
Regards, Novski

2 Likes

Hi,
there is a poll regarding the supported device by B2G, the new Firefox OS. Please vote for fairphone 2!!!

2 Likes

There ist a poll for Hardwaresupport of B2G…

What devices do you want to be supported next?

Regards, Novski

1 Like

Yes, it is the same that I reported just the last post. We can vote in order to have support for fairphone 2.

1 Like

Hi,
there is a Telegram group of B2G.
The wiki of the transition project.
IRC channel and Mozilla discussions.

2 Likes

My side is able to build the whole firefox code, and get .img files. but after i flashing system.img, boot.img, and recovery.img file, my phone is not able to enter home screen, and stop at a state where screen always display “loading…”. But adb command is working. and i check with logcat log, and see below logs, Does anyone has the same issue?
E/QC-QMI ( 340): qmi_ctl_tx_msg: qmi_qmux_tx_msg failed
E/QC-QMI ( 340): qmi_ctl_handle_request: qmi_ctl_tx_msg call failed
E/QC-QMI ( 340): qmi_qmux_open_connection: connection is disabled for conn_id=48
E/QC-QMI ( 340): qmi_qmux_tx_msg: failed to open inactive connd_id=48
E/QC-QMI ( 340): qmi_qmux: TX failed, connection inactive or in reset, conn_id=48, status_flags=4
E/QC-QMI ( 340): qmi_ctl_tx_msg: qmi_qmux_tx_msg failed
E/QC-QMI ( 340): qmi_ctl_handle_request: qmi_ctl_tx_msg call failed
E/QC-QMI ( 340): qmi_qmux_open_connection: connection is disabled for conn_id=57
E/QC-QMI ( 340): qmi_qmux_tx_msg: failed to open inactive connd_id=57
.
.
.
.
.
.
.
E/QC-QMI ( 340): qmi_qmux: TX failed, connection inactive or in reset, conn_id=48, status_flags=4
.
.
.
.
.
.
.
E/sdcard ( 322): missing packages.list; retrying
W/Atfwd_Sendcmd( 321): AtCmdFwd service not published, waiting… retryCnt : 4
E/sdcard ( 322): missing packages.list; retrying