956255 - B2G RIL: no need to deactivate data call before radio off

Status: RESOLVED WONTFIX

(Firefox OS Graveyard :: RIL, defect)

Description

[Szu-Yu Chen [:aknow] (inactive)]

We don't have to do it. Remove the code.

Comment 1

[Szu-Yu Chen [:aknow] (inactive)]

Attachment: Part 1: No need to deactivate data call before radio off

Comment 2

[Szu-Yu Chen [:aknow] (inactive)]

Attachment: Part 2: Modify test

Comment 3

[Vicamo Yang [:vicamo][:vyang]]

(In reply to Szu-Yu Chen [:aknow] from comment #0)
> We don't have to do it. Remove the code.

Why? We had them merged in bug 908603.  So what's the effect after reverting them here?

Comment 4

[Jessica Jong [:jessica] (inactive)]

(In reply to Vicamo Yang [:vicamo][:vyang] (PTO Dec. 21 ~ Jan. 5) from comment #3)
> (In reply to Szu-Yu Chen [:aknow] from comment #0)
> > We don't have to do it. Remove the code.
> 
> Why? We had them merged in bug 908603.  So what's the effect after reverting
> them here?

In Bug 909688 we have handled data connectivity lost cases properly, so there should be no problem in gecko side if we remove 'powerOffRadioSafely()' now, but I think we still need modem's confirmation if they are okay with this. Please see Bug 856553 comment 98.

Comment 5

[Szu-Yu Chen [:aknow] (inactive)]

Bug 883065, looks like modem had already resolved the problem, so we don't have to deactivate the data call by ourself. In addition, I will also check the patch on unagi and fugu device to make sure there is no problem on modem.

Comment 6

[Szu-Yu Chen [:aknow] (inactive)]

(In reply to Szu-Yu Chen [:aknow] from comment #5)
> Bug 883065, looks like modem had already resolved the problem, so we don't
> have to deactivate the data call by ourself. In addition, I will also check
> the patch on unagi and fugu device to make sure there is no problem on modem.

Verified on unagi and fugu.
Data connection works well after turn off and then turn on the radio.

Comment 7

[Hsin-Yi Tsai (she/her) [:hsinyi]]

(In reply to Szu-Yu Chen [:aknow] from comment #6)
> (In reply to Szu-Yu Chen [:aknow] from comment #5)
> > Bug 883065, looks like modem had already resolved the problem, so we don't
> > have to deactivate the data call by ourself. In addition, I will also check
> > the patch on unagi and fugu device to make sure there is no problem on modem.
> 
> Verified on unagi and fugu.
> Data connection works well after turn off and then turn on the radio.

I also tested with buri, and it worked as expected.

Comment 8

[Hsin-Yi Tsai (she/her) [:hsinyi]]

Comment on attachment 8355467 [details] [diff] [review]
Part 1: No need to deactivate data call before radio off

Review of attachment 8355467 [details] [diff] [review]:
-----------------------------------------------------------------

Don't forget that we'd need to remove "Cu.import("resource://gre/modules/Promise.jsm");"

Comment 9

[Szu-Yu Chen [:aknow] (inactive)]

Attachment: Part 1#2: No need to deactivate data call before radio off

address comment

Comment 10

[Hsin-Yi Tsai (she/her) [:hsinyi]]

Comment on attachment 8357005 [details] [diff] [review]
Part 1#2: No need to deactivate data call before radio off

Review of attachment 8357005 [details] [diff] [review]:
-----------------------------------------------------------------

Cool, thank you for the work.

Comment 11

[Hsin-Yi Tsai (she/her) [:hsinyi]]

(In reply to Szu-Yu Chen [:aknow] from comment #6)
> (In reply to Szu-Yu Chen [:aknow] from comment #5)
> > Bug 883065, looks like modem had already resolved the problem, so we don't
> > have to deactivate the data call by ourself. In addition, I will also check
> > the patch on unagi and fugu device to make sure there is no problem on modem.
> 
> Verified on unagi and fugu.
> Data connection works well after turn off and then turn on the radio.

Did you test fugu and fugu-DSDS both? If no, please help make sure they both are covered. And in fugu-DSDS, we'd carefully check the cases that data call is on sim1 and on sim2. Thank you!

Comment 12

[Szu-Yu Chen [:aknow] (inactive)]

Attachment: [final] Part 1: No need to deactivate data call before radio off. r=hsinyi

Comment 13

[Szu-Yu Chen [:aknow] (inactive)]

Attachment: [final] Part 2: Modify test. r=hsinyi

Comment 14

[Szu-Yu Chen [:aknow] (inactive)]

https://tbpl.mozilla.org/?tree=Try&rev=daf651fef6c4

Comment 15

[Hsin-Yi Tsai (she/her) [:hsinyi]]

Finally got reply with the perspective of modem: confirmation

"The reason for PowerOffSafely is not related to local behavior, rather it's about being nice to the network.
Locally the modem will clean up everything on power off and, as you noticed, we are told that the data call is down via DATA_CALL_LIST_CHANGED message.
However on the network side the network isn't notified that the device is going down and doesn't get the chance to clean up resources associated with that data call (at least until some timeout happen on their side, assuming they are some contingency in place to cleanup stale connections). Explicitly deactivating the data call on power down is the device being a good citizen and allows the network to free up the resources more readily."

Though we don't see re-connection problem with this patch, to make our FFOS a good citizen to the network world, let's keep what we have now. Close as WONTFIX. Sorry for that, and thanks again!