Closed
Bug 1134493
Opened 9 years ago
Closed 9 years ago
Keyboard closes after typing one character when lockscreen is disabled
Categories
(Core :: Panning and Zooming, defect)
Tracking
()
VERIFIED
FIXED
mozilla39
People
(Reporter: botond, Assigned: kats)
References
Details
(Keywords: dogfood, regression)
Attachments
(2 files)
2.61 KB,
patch
|
botond
:
review+
bajaj
:
approval-mozilla-b2g37+
|
Details | Diff | Splinter Review |
2.58 MB,
video/mp4
|
Details |
STR: 1. Open the Browser app. 2. Tap the address bar. The keyboard opens. 3. Tap a key. Expected results: A character is entered into the address bar. The keyboard remains open. Actual results: A character is entered into the address bar. The keyboard closes. The behaviour becomes even worse after this. When you re-open the keyboard and tap a key again, a character isn't even entered into the address bar - the keyboard just closes. I'm running with Gecko and Gaia pulled today. In particular, I verified that I have the patch from 1129344 applied, but I still see the problem. It basically makes the phone unusable...
Reporter | ||
Comment 1•9 years ago
|
||
This issue has largely gone away after a full rebuild + reflash with today's trunk.
Status: NEW → RESOLVED
Closed: 9 years ago
Resolution: --- → WORKSFORME
Reporter | ||
Comment 2•9 years ago
|
||
(In reply to Botond Ballo [:botond] from comment #1) > This issue has largely gone away after a full rebuild + reflash with today's > trunk. *Sigh*. No it hasn't - it's just intermittent. Seems to be particularly bad when the key pressed in <backspace>.
Status: RESOLVED → REOPENED
Resolution: WORKSFORME → ---
Comment 4•9 years ago
|
||
From bug 1135938 comment 2 Etienne, Kats - I've confirmed that backing out the gaia patch in bug 1130011 makes this go away. It seems that we're not *actually* catching any events in those listeners though, so I assume we're going through some different APZ patch and triggering touch/focus events when we don't want them.
Comment 5•9 years ago
|
||
It looks like bug 1130011 is not yet in 2.2, so this is only happening in 3.0.
status-b2g-v2.2:
--- → unaffected
status-b2g-master:
--- → affected
Assignee | ||
Comment 6•9 years ago
|
||
Hm, interesting. If you want to back out bug 1130011 for now I'm ok with that. I can take a look to see what's going on although it probably won't be until later this week. I'll have to repro this problem first since I haven't been seeing it on my Flame lately. Leaving ni on me for now.
Flags: needinfo?(bugmail.mozilla)
Assignee | ||
Updated•9 years ago
|
Flags: needinfo?(bugmail.mozilla)
Assignee | ||
Comment 7•9 years ago
|
||
I can repro this if I disable the lockscreen, which effectively removes the touch listeners on the system window. (If bug 1130016 gets fixed then this issue will be reproducible even without disabling the lockscreen). Looking into it.
Assignee | ||
Comment 8•9 years ago
|
||
I think this is happening because the code at http://mxr.mozilla.org/mozilla-central/source/layout/ipc/RenderFrameParent.cpp?rev=34a3bc7200c1#629 forces the hit region of the entire keyboard process to empty. So if there's stuff in the rocketbar search results behind the keyboard that will get hit instead, and dismiss the keyboard. When you first bring up the rocketbar the results are generally empty so the hit goes to the system process which sends it to the keyboard process via main-thread dispatch and so it works that time. Having listeners on the root window in the system process also helps because it forces APZ to wait for the response... I think.
Assignee: nobody → bugmail.mozilla
Flags: needinfo?(bugmail.mozilla)
Assignee | ||
Updated•9 years ago
|
Component: Gaia::Keyboard → Panning and Zooming
Product: Firefox OS → Core
Assignee | ||
Comment 9•9 years ago
|
||
Technically this is a regression from bug 1131840, and it was bug 1130011 that exposed it. Marking 2.2 as affected since bug 1131840 was uplifted there.
Blocks: 1131840
Assignee | ||
Comment 10•9 years ago
|
||
Attachment #8568738 -
Flags: review?(botond)
Assignee | ||
Comment 11•9 years ago
|
||
try |
https://treeherder.mozilla.org/#/jobs?repo=try&revision=0cd728a477b3
Updated•9 years ago
|
Flags: needinfo?(etienne)
Reporter | ||
Updated•9 years ago
|
Attachment #8568738 -
Flags: review?(botond) → review+
Reporter | ||
Comment 13•9 years ago
|
||
(In reply to Kartikaya Gupta (email:kats@mozilla.com) from comment #7) > I can repro this if I disable the lockscreen, which effectively removes the > touch listeners on the system window. By the way, this explains why the problem went away for me after a full flash (comment 1), which resets settings to defaults (which for the lockscreen is enabled), and then came back shortly after (comment 2) when I would have disabled the lockscreen again.
Assignee | ||
Comment 14•9 years ago
|
||
https://hg.mozilla.org/integration/mozilla-inbound/rev/64f22e3076ea
https://hg.mozilla.org/mozilla-central/rev/64f22e3076ea
Status: REOPENED → RESOLVED
Closed: 9 years ago → 9 years ago
status-firefox39:
--- → fixed
Resolution: --- → FIXED
Target Milestone: --- → mozilla39
Comment 16•9 years ago
|
||
Could you please also uplift to v2.2? Thanks!
Flags: needinfo?(bugmail.mozilla)
Assignee | ||
Comment 17•9 years ago
|
||
Comment on attachment 8568738 [details] [diff] [review] Account for mozpasspointerevents when setting force-ehr flag NOTE: Please see https://wiki.mozilla.org/Release_Management/B2G_Landing to better understand the B2G approval process and landings. [Approval Request Comment] Bug caused by (feature/regressing bug #): Caused by 1131840 but exposed by 1130011. User impact if declined: If the lockscreen is disabled, then it becomes very hard to type in the keyboard as it will dismiss itself frequently (almost always) Testing completed: locally Risk to taking this patch (and alternatives if risky): low risk patch; the bug is an oversight in my patches for bug 1131840 where I forgot to check the mozpasspointerevents attribute in one place. String or UUID changes made by this patch: none
Flags: needinfo?(bugmail.mozilla)
Attachment #8568738 -
Flags: approval-mozilla-b2g37?
Assignee | ||
Updated•9 years ago
|
Summary: Keyboard closes after typing one character → Keyboard closes after typing one character when lockscreen is disabled
Assignee | ||
Updated•9 years ago
|
Updated•9 years ago
|
Attachment #8568738 -
Flags: approval-mozilla-b2g37? → approval-mozilla-b2g37+
Comment 19•9 years ago
|
||
verify for 2.2. Add 'verifyme' for master branch. *v2.2 env info: Build ID 20150302162504 Gaia Revision 062d395273e70b8688a91b5a2b34c0cc7afc679f Gaia Date 2015-03-02 22:12:55 Gecko Revision https://hg.mozilla.org/releases/mozilla-b2g37_v2_2/rev/5a7cb3b9f78e Gecko Version 37.0 Device Name flame Firmware(Release) 4.4.2 Firmware(Incremental) eng.cltbld.20150302.195105 Firmware Date Mon Mar 2 19:51:14 EST 2015 Bootloader L1TC000118D0
Keywords: verifyme
Updated•9 years ago
|
Comment 20•9 years ago
|
||
This bug has been successfully verified on latest Flame v3.0. See attachment: verified_v3.0.mp4 Reproduce rate: 0/5 STR: 1.Disable screenlock. 2.Open Browser app. 3.Tap the address bar. **The keyboard opens. 4.Tap a key on keyboard. **The character is entered into the address bar.And the keyboard remains open as expected. Flame 3.0 build: Build ID 20150303010233 Gaia Revision c8ed1085a67490a1ecd7f275e5de9487e1b93b1d Gaia Date 2015-03-02 20:14:21 Gecko Revision https://hg.mozilla.org/mozilla-central/rev/0b3c520002ad Gecko Version 39.0a1 Device Name flame Firmware(Release) 4.4.2 Firmware(Incremental) eng.cltbld.20150303.044015 Firmware Date Tue Mar 3 04:40:27 EST 2015 Bootloader L1TC000118D0 ------------------------------------------------------------------------------- Leaving "verifyme" for firefox39.
Comment 21•9 years ago
|
||
Assignee | ||
Comment 22•9 years ago
|
||
This bug doesn't affect desktop, so changing status to verified.
Status: RESOLVED → VERIFIED
Keywords: verifyme
You need to log in
before you can comment on or make changes to this bug.
Description
•