1191226 - [Shinano][Yukon] Define supported network types for devices

Status: RESOLVED FIXED

(Firefox OS Graveyard :: GonkIntegration, defect)

Description

[:gerard-majax]

After AOSP port cleanup there is no more ro.telephony.default_network value being defined on those platforms. We should add our own property on both so that user can select proper radio technology:
 - ro.moz.ril.0.network_types=gsm,wcdma,lte

Comment 1

[:gerard-majax]

Attachment: Shinano platform PR

Comment 2

[:gerard-majax]

Attachment: Yukon platform PR

Comment 3

[:gerard-majax]

Edgar confirmed me on IRC that without either ro.moz.ril.X.network_types nor ro.telephony.default_network, it is expected we fallback on a small set of radio technologies. He also confirmed me that the preferred way to handle this is via ro.moz.ril property.

Comment 4

[Michael Wu [:mwu]]

How does the AOSP port determine the default set of radio technologies? ro.telephony.default_network seems to be what AOSP uses, so if that's not set, the AOSP port shouldn't be working either. (unless they've added a new mechanism for this..)

Comment 5

[Michael Wu [:mwu]]

After reading the previous comments, I guess my question is really: why don't we have ro.telephony.default_network anymore?

Comment 6

[:gerard-majax]

(In reply to Michael Wu [:mwu] from comment #5)
> After reading the previous comments, I guess my question is really: why
> don't we have ro.telephony.default_network anymore?

Because this was needed upto AOSP KK. With L I have been told it's not useful anymore.

Comment 7

[Michael Wu [:mwu]]

Does this mean AOSP doesn't use it anymore? If that's the case, do we know what new thing AOSP uses?

Comment 8

[:gerard-majax]

(In reply to Michael Wu [:mwu] from comment #7)
> Does this mean AOSP doesn't use it anymore? If that's the case, do we know
> what new thing AOSP uses?

I don't know but I'm asking Alin :). Hsin-Yi or Edgar, do you know?

Comment 9

[:gerard-majax]

Right, change of plan, turns out this is still used in L: http://androidxref.com/5.1.1_r6/xref/frameworks/base/telephony/java/android/telephony/TelephonyManager.java#915

So we'll PR Sony's upstream repo to add it back (Shinano, Rhine and Yukon platform) and then we'll merge the Shinano and Yukon.

Comment 10

[:gerard-majax]

Right so there's something I am misunderstanding here. We have ro.telephony.default_network that looks to be interpreted by B2G as the maximum capacities of the device: setting it to 7 makes me unable to select above H+; setting it to 1 makes me stuck on GSM only. Yet the naming of this property suggests that this is used to select the default value, not the capacities.

Does this means our RIL expects absolutely to have ro.moz.ril.X.network_types to be defined, and only fallback on the ro.telephony.default_network as a safe fallback maximum value for the device's capacities?

Comment 11

[Edgar Chen [:edgar]]

(In reply to Alexandre LISSY :gerard-majax from comment #10)
> Right so there's something I am misunderstanding here. We have
> ro.telephony.default_network that looks to be interpreted by B2G as the
> maximum capacities of the device: setting it to 7 makes me unable to select
> above H+; setting it to 1 makes me stuck on GSM only. Yet the naming of this
> property suggests that this is used to select the default value, not the
> capacities.
> 
> Does this means our RIL expects absolutely to have
> ro.moz.ril.X.network_types to be defined, and only fallback on the
> ro.telephony.default_network as a safe fallback maximum value for the
> device's capacities?

Exactly, our RIL expects to have `ro.moz.ril.X.network_types` defined. If not defined, we will try to get the *possible* capability from `ro.telephony.default_network` (which is from AOSP). [1]

[1] https://dxr.mozilla.org/mozilla-central/source/dom/mobileconnection/gonk/MobileConnectionService.js#507-545

Comment 12

[:gerard-majax]

And you use this possible as the maximum value, right?

Comment 13

[Edgar Chen [:edgar]]

(In reply to Alexandre LISSY :gerard-majax from comment #12)
> And you use this possible as the maximum value, right?

That's right.

Comment 14

[:gerard-majax]

Attachment: Yukon platform PR

Comment 15

[:gerard-majax]

Attachment: Shinano platform PR

The Shinano merge also include a couple of Camera-related header commit that landed. I can do a separate PR if you prefer.

Comment 16

[:gerard-majax]

https://github.com/mozilla-b2g/device-sony-yukon/commit/885b19f0d2f16a0afd667ea013f56f66244efa76
https://github.com/lissyx/device-sony-shinano/tree/bug1191226