Display doesn't turn on for incoming phone call when proximity sensor triggered

[Matthew Exon]

I just received two phone calls. In each case the phone started ringing but the screen stayed black. In the first case I let it ring for a while hoping that it would magically wake up, since some bugs do seem to work that way. But nothing happened. I was unwilling to press the power button since in iOS that immediately hangs up the call. Eventually however I did press it, and the screen woke up. But I was too late to answer the call. When the person phoned back the screen was black again, so I pressed the power button and was then able to slide the slider to answer the call.

I don't think this is the same as Bug #1493574, since that bug seems to happen after the phone has been answered.

Syslog of about the right time attached. Phone reports the times of the calls as 15:41 and 15:47.

[Pat McGowan (pat-mcgowan)]

@Mathew can you add more info about the os version and model, have not seen this reported before

[Matthew Exon (ubuntubugs-mexon)]

OTA-8 and MX4 respectively. I haven't put the phone in r/w mode or installed anything apart from the app store, but I have got various services running in a chroot - MySQL, Apache, and Mnemosyne.

I've had other problems that might be related. For the past day or two incoming texts have made the phone vibrate but not play a sound, even though the volume was turned right up. I went to the setting where the sound plays and changed it, which proved that the phone was making noise, but it didn't fix the problem. Rebooting, however, did fix it. If I have the same incoming call problem again I'll add more info.

[Pat McGowan (pat-mcgowan)]

Wondering if its a memory issue, also check for crash files

[Matthew Exon (ubuntubugs-mexon)]

Presumably not this one?

phablet@ubuntu-phablet:~$ ls -lh /var/crash
total 40K
-rw-rw-r-- 1 phablet whoopsie 37K Dec 7 10:27 _usr_bin_content-hub-service.32011.crash

Not sure where the crash logs are kept...

[Matthew Exon (ubuntubugs-mexon)]

OK I get what the problem is: it's the proximity sensor. I have a flip-case for my phone that covers the proximity sensor. If the flip-case is closed the phone thinks it's next to my ear and switches off the screen.

So there are two things wrong here. First (IMO) it doesn't make sense to use the proximity sensor to control the screen when the phone call hasn't been answered yet. Second, the phone should at least switch the screen on when I open the flip case, but it doesn't.

I tested it with four phone calls just now:

1) Flip-case closed. Didn't switch on screen.
2) Flip-case open. Screen turned on fine
3) Flip-case closed, then opened once it started ringing. Still didn't switch on screen.
4) Flip-case open. Screen turned on fine.

syslog attached.

[Pat McGowan (pat-mcgowan)]

Not turning the screen on when proximity is active is an explicit decision to avoid previous pocket answering issues. The idea is what good is turning the screen on if you cannot see it.

I cannot reproduce the second issue of not lighting the screen after uncovering the sensor in either stable or proposed. Will ask someone else to test.

[Michał Sawicz (saviq)]

@Matthew, we've enabled the proximity sensor in this case to avoid accidental interactions with the incoming call or text notifications. It should, however turn on when you open your case, that's would be a bug.

[Jean-Baptiste Lallement (jibel)]

I didn't reproduce the second issue either with rc-proposed.

[Allan LeSage (allanlesage)]

Same, not able to reproduce this on arale. I've also tried with settings on "Lock when idle: Never" and having ignored a previous incoming SMS, still not able to reproduce this.

[Lorn Potter (lorn-potter)]

I can reproduce this.

Start with screen blanked.
Cover top of phone with something., so it covers the proximity sensor.
Call that phone from another one.
While it is ringing, lift the cover off the proximity sensor.

See that screen stays blanked and only way to unblank it is to push button.

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[Pat McGowan (pat-mcgowan)]

This sounds like bug #1543948 so suping it there, change back if you disagree.