960113 - JavaScript Error: "XPCOMUtils is not defined" {file: "chrome://satchel/content/formSubmitListener.js"

Status: RESOLVED WORKSFORME

(Firefox OS Graveyard :: Gaia::System, defect)

Description

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

Attachment: adb logcat output

E/GeckoConsole(  468): [JavaScript Error: "XPCOMUtils is not defined" {file: "chrome://satchel/content/formSubmitListener.js" line: 15}]

This happens after starting up the emulator, tested with the jb version of the emulator in pvt builds, with timestamp 20140114154437, then waiting awhile to wait for the splash screen then the lockscreen to appear, unlock the lockscreen, then drag right to unlock the lockscreen. Somehow a crash is seen, I manually click to send report, then the homescreen appears. I check the logs and this issue is seen.

Not sure how to move this forward, setting needinfo? from Felix Fung and Felipe, who seem to have touched files in toolkit/components/satchel before. Are these references to "satchel" the same set of files?

Comment 1

[:Felipe Gomes (needinfo for replies!)]

Hm, I have no idea how XPCOMUtils could be undefined there, given how it was just imported a few lines above. CC'ing bill to see if he knows what's up. I wonder if it's caused to the difference on how jsms are loaded in b2g. Note that formSubmitListener.js is a content script.

Gary: do you think that this message has anything to do with the crash? I'm thinking it might be unrelated.

But now I'm wondering if FormHistory is at all used on b2g, or if this is just code wasted there

Comment 2

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

> Gary: do you think that this message has anything to do with the crash?

I don't know one way or other, sorry.

Comment 3

[Bill McCloskey [inactive unless it's an emergency] (:billm)]

That sounds a lot like bug 952183. Does your tree have the patch in that bug, Gary?

Comment 4

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

Attachment: sources.xml

I used the jb version of the emulator in pvt builds, with timestamp 20140114154437.

Not sure what Gecko rev that has. John, do you happen to know?

Comment 6

[John Ford [:jhford] CET/CEST Berlin Time]

that's odd that sources.xml doesn't contain a copy of the gecko information.  The Gecko revision *should* be stored on the device, baked into the gecko image.  I wrote this script for QA a while back, not sure if it still is working:

#!/bin/bash
set -e
echo "Plug in your device"
adb wait-for-device
echo "Found device"
dir=$(mktemp -d -t revision)
cd $dir 
adb pull /system/b2g/omni.ja &>/dev/null || echo "Error"
adb pull /system/b2g/webapps/settings.gaiamobile.org/application.zip &> /dev/null || echo "Error"
adb pull /system/b2g/application.ini &> /dev/null || echo "Error pulling application.ini"
unzip omni.ja chrome/toolkit/content/global/buildconfig.html > /dev/null
unzip application.zip resources/gaia_commit.txt > /dev/null 
echo 'Gecko ' $(grep "Built from" ./chrome/toolkit/content/global/buildconfig.html \
 | sed "s,.*<a href=\",,g ; s,\">.*,,g")
echo 'Gaia  ' $(head -n 1 ./resources/gaia_commit.txt)

for i in BuildID Version ; do
    echo $i      $(grep "^ *$i" application.ini | sed "s,.*=,,g")
done
rm -rf $dir

Comment 7

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

> that's odd that sources.xml doesn't contain a copy of the gecko information.
> The Gecko revision *should* be stored on the device, baked into the gecko
> image.

This sources.xml file was obtained from pvtbuilds, I have no idea what device it came from.

Comment 8

[John Ford [:jhford] CET/CEST Berlin Time]

so you only have access to this sources.xml file and nothing else?  Can you link me to this file?

Comment 9

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

I sent a mail to John to follow up.

Comment 10

[John Ford [:jhford] CET/CEST Berlin Time]

Chris, do you know why we aren't including gecko/gaia in the sources.xml files for this?  I remember that there was an issue similar to this with the XML munger you guys use, but I don't know what's up here.

Comment 11

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

> Chris, do you know why we aren't including gecko/gaia in the sources.xml
> files for this?

Gaia version is present in the case of the sources.xml in this bug - it's d452351697699ce35d2fc4d79a8298ab3a73d9f3

I think the Gecko version is missing.

Comment 12

[Aus Lacroix [:aus]]

This occurs in Gaia v1.3 / b2g28 as well.

Comment 13

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

Nominating for tracking-b2g-v1.3? and blocking-b2g1.3? (because for whatever reason I cannot tell the difference between the 2).

This seems to block bug 966027 which is tagged as [mwcdemo2014] in the whiteboard.

Comment 14

[Jason Smith [:jsmith]]

I don't think we've proven that bug 966027 is fully dependent on this bug yet per a discussion with aus, so I'm pulling the noms here. If it ends up being a hard dependency, then we'll reconsider.

Comment 15

[Chris AtLee [:catlee]]

(In reply to John Ford [:jhford] -- please use 'needinfo?' instead of a CC from comment #8)
> so you only have access to this sources.xml file and nothing else?  Can you
> link me to this file?

We don't include the gecko hash in sources.xml for per-checkin builds. The reason is that we're pulling gecko from hg, and we don't want to wait for the hg -> git -> mapper process to complete for per-checkin builds.

We could include the hg revision information as a comment if that would be useful?

Comment 16

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

(In reply to Chris AtLee [:catlee] from comment #15)
> We could include the hg revision information as a comment if that would be
> useful?

As a comment in which location? A text file with the hash would definitely work too, if that's what you mean, so long as we can retrieve it for bug filing purposes. Do we need a new bug for this?

Comment 17

[Chris AtLee [:catlee]]

could go in sources.xml as an XML comment?

Comment 18

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

(In reply to Chris AtLee [:catlee] from comment #17)
> could go in sources.xml as an XML comment?

WFM.

Comment 19

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

(In reply to Chris AtLee [:catlee] from comment #17)
> could go in sources.xml as an XML comment?

Spun this off as bug 970601.

Comment 20

[Aus Lacroix [:aus]]

We figured out a different way to fix the issue and this bug no longer blocks 966027.

Comment 21

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

Attachment: screenshot of crash dialog from last month's build

Comment 22

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

I haven't seen this recently with the Flame. (will file a new bug should it happen again)

$ ./check_versions.sh
Gaia      0f9f11d0a6dadb3ea27160204bbe911c1ad69a6f
Gecko     https://hg.mozilla.org/mozilla-central/rev/196d05832e12
BuildID   20140709040203
Version   33.0a1
ro.build.version.incremental=109
ro.build.date=Mon Jun 16 16:51:29 CST 2014
B1TC00011220