863807 - [B2G][Email] User is unable to set up IMAP account

Status: VERIFIED FIXED

(Firefox OS Graveyard :: Gaia::E-Mail, defect)

Description

[Natalya Kot [:nkot]]

Attachment: logcat

Description:
The user is unable to setup Gmail account, "Please wait while your account is set up" message displays, keeps spinning and never gets loaded

Prerequisites:
WiFi is set to On
Signal strength is "Very good"

Repro Steps:
1) Updated to Inari Build ID: 20130419070202
2) Open Email app to set up a new email account(gmail)
3) Enter correct credentials into fields: name, email address, password
4) Tap "Next" to complete setup

Actual:
"Please wait while your account is set up" message with the loading spinner displays, unable to set up account

Expected:
After short loading period a dialog appears "Create another account" or "Continue to mail" and user can successfully launch mailbox by selecting "Continue to mail"

Environmental  Variables:
Kernel Date: Feb 21
Gecko: http://hg.mozilla.org/releases/mozilla-b2g18_v1_0_1/rev/c3ee6b445cf5
Gaia: 393a946b9d60294697838c5ef7e0548d47c3281a

Notes:
Repro frequency: 100%, 2/2 devices
See attached logcat for more info

Was able to browse websites same time during setting up Gmail accout
Was able to set up Hotmail account successfully

Comment 1

[Chris Kreinbring [:CKreinbring]]

Also seen on Unagi 1.1 version 20130419070205
Kernel Date: Dec 5
Gecko: http://hg.mozilla.org/releases/mozilla-b2g18/rev/82fdff82a2d0
Gaia: a24cf9f35f7414402edda3aef16f1fdcea21d785

Comment 2

[Naoki Hirata :nhirata (please use needinfo instead of cc)]

Tried to setup yahoo account and failed.  hotmail is successful.

Comment 3

[Andrew Sutherland [:asuth] (he/him)]

We need the fix on bug 863769 (and its logic fix), then things will work again.

Comment 4

[Dietrich Ayala (:dietrich)]

Functional regression blocking major feature, blocking+.

Andrew, can you add a test for account creation when this gets fixed?

Comment 5

[Andrew Overholt [:overholt]]

Andrew, feel free to re-/un-assign if you're not the correct person here.  tef+ bugs assigned to nobody@ look bad :)

Comment 8

[Andrew Sutherland [:asuth] (he/him)]

Bug 863769 and bug 863770 are effectively fixed now (fixes landed on all branches but waiting on landing on mozilla-central), but there is still a correctness problem in multi-process mode that was not surfaced in our back-end tests because those do not currently run multi-process, and the gecko mozTCPSocket fixes are not multi-process either.

This error manifests itself on the device as:
I/GeckoDump(  602): ERR: onerror reporting: NS_ERROR_FAILURE: Component returned failure code: 0x80004005 (NS_ERROR_FAILURE) [nsITCPSocketChild.send] @ jar:file:///system/b2g/omni.ja!/components/TCPSocket.js : 444

Comment 9

[Andrew Sutherland [:asuth] (he/him)]

(In reply to Andrew Sutherland (:asuth) from comment #8)
> and the
> gecko mozTCPSocket fixes are not multi-process either.

s/fixes/tests/

And I was incorrect about that; the existing TCPSocket tests *are* also run IPC, but they are still run with chrome privileges, so compartment/wrapper problems are not going to show up there either...



And this problem could conceivably just be in our JS code if we are somehow trying to pass in something that does not conform to:
  NS_ENSURE_TRUE(JS_IsArrayBufferObject(obj, aCx), NS_ERROR_FAILURE);
but would conform to:
  new_stream.setData(data, byteOffset, byteLength);

a gdb pass and/or some debug logging is required to figure out what's going on there.  I just got back from Madrid today and am thoroughly wrecked and will be crashing shortly.

Comment 10

[Andrew Sutherland [:asuth] (he/him)]

Bug 864775 is the final fix for this.  Fix plus build fix are on birch, landing on trunk and branches soon.

Comment 11

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Will the tef+ on *this* bug cause that bug to get uplifted to the branches, or does somebody need to get poked about those uplifts?

Comment 12

[Andrew Overholt [:overholt]]

(In reply to David Baron [:dbaron] (don't cc:, use needinfo? instead) from comment #11)
> Will the tef+ on *this* bug cause that bug to get uplifted to the branches,
> or does somebody need to get poked about those uplifts?

Everything was tef+'d except bug 864775 and I've just set the flag there.

Comment 13

[Andrew Sutherland [:asuth] (he/him)]

Uplifts occurred yesterday for v1-train and very early this morning for v1.0.1 on bug 864775, marking fixed.  The e-mail app should be as operational as it ever was before and slightly faster/more cpu efficient.

Comment 14

[Natalya Kot [:nkot]]

Andrew,

issue reoccurs today, on both v1.0.1 and v1-train builds
- NOT able to set up IMAP account, exactly same AR - keeps spinning, never gets loaded

Inari device
Gecko  http://hg.mozilla.org/releases/mozilla-b2g18_v1_0_1/rev/5d454864b668
Gaia   052253176f3ad495a1086e47a28fffbf07613388
Build  2013-04-25-070201

Unagi device
Gecko  http://hg.mozilla.org/releases/mozilla-b2g18/rev/6c2493de1441
Gaia   9c6e9decdd1333f3b6dbed22538a85361cf944dc
Build  2013-04-25-070204


!It was working fine on 4/24 builds, passed the smoketest for both v1.1.0 and v1.0.1

Comment 15

[Andrew Sutherland [:asuth] (he/him)]

logcat, please

Comment 16

[Natalya Kot [:nkot]]

hm.. apparently it was something wrong with my account, was able to sign in successfully with different credentials
please disregard

Comment 17

[Natalya Kot [:nkot]]

Attachment: logcat with unsuccessful setup

(In reply to Andrew Sutherland (:asuth) from comment #15)
> logcat, please

having problem only with one specific account..

Comment 18

[Andrew Sutherland [:asuth] (he/him)]

Thanks for the log.  This indicates a catastrophic failure of the IMAP timezone logic probing code.  I'll file a new bug with your log.

Comment 19

[Andrew Sutherland [:asuth] (he/him)]

Bug 865786 filed for IMAP timezone logic probing failure.

Comment 20

[Tracy Walker [:tracy]]

verified with 1.1 build 20130425070204