830462 - OTA update should remove update.mar file from sdcard

Status: VERIFIED FIXED

(Firefox OS Graveyard :: General, defect)

Description

[Dave Hylands [:dhylands]]

Currently, if the OTA update places the update.mar file on the sdcard, then it doesn't get removed.

I'm not aware of any issues caused by this, other than it consumes space on the sdcard.

Comment 1

[Dave Hylands [:dhylands]]

It seems that if the update.mar file is corrupt, then this can prevent further updates from occurring.

Comment 2

[Dave Hylands [:dhylands]]

Attachment: Remove update.mar from sdcard when finished (or when starting a new download)

Now removes stale update.mar file from sdcard at the start of a new download, and after an update was applied.

Comment 3

[Dave Hylands [:dhylands]]

https://tbpl.mozilla.org/?tree=Try&rev=20180251f907

Comment 4

[Dave Hylands [:dhylands]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/c7572ad0ef03

Comment 6

[Tony Chung [:tchung]]

Hey Dave, can you not dupe older bugs to newer bugs?  that is inconsiderate to original bug filers as they view their bug becomes useless.   (you duped my bug that was filed as 830126 :( )  

that said, i'm still seeing this problem on today's build, when attempting to update from 20130115070201 -> 20130116070204; and watching it fail.

Logcat:
01-16 10:41:20.335: I/Gecko(108): *** AUS:SVC Creating Downloader
01-16 10:41:20.335: E/GeckoConsole(108): AUS:SVC Creating Downloader
01-16 10:41:20.335: I/Gecko(108): *** AUS:SVC UpdateService:_downloadUpdate
01-16 10:41:20.335: E/GeckoConsole(108): AUS:SVC UpdateService:_downloadUpdate
01-16 10:41:20.335: I/Gecko(108): *** AUS:SVC readStringFromFile - file doesn't exist: /data/local/updates/0/update.status
01-16 10:41:20.335: E/GeckoConsole(108): AUS:SVC readStringFromFile - file doesn't exist: /data/local/updates/0/update.status
01-16 10:41:20.335: I/Gecko(108): *** AUS:SVC readStatusFile - status: null, path: /data/local/updates/0/update.status
01-16 10:41:20.335: E/GeckoConsole(108): AUS:SVC readStatusFile - status: null, path: /data/local/updates/0/update.status
01-16 10:41:20.335: I/Gecko(108): *** AUS:SVC Downloader:_selectPatch - found existing patch with state: null
01-16 10:41:20.335: E/GeckoConsole(108): AUS:SVC Downloader:_selectPatch - found existing patch with state: null
01-16 10:41:20.335: I/Gecko(108): *** AUS:SVC Downloader:_selectPatch - failed to apply complete patch!
01-16 10:41:20.335: E/GeckoConsole(108): AUS:SVC Downloader:_selectPatch - failed to apply complete patch!
01-16 10:41:20.365: D/memalloc(108): /dev/pmem: Allocated buffer base:0x4b200000 size:81920 offset:8273920 fd:144
01-16 10:41:20.365: D/memalloc(360): /dev/pmem: Mapped buffer base:0x44100000 size:8355840 offset:8273920 fd:28
01-16 10:41:20.395: I/Gecko(108): *** AUS:SVC Downloader:downloadUpdate - no patch to download
01-16 10:41:20.395: E/GeckoConsole(108): AUS:SVC Downloader:downloadUpdate - no patch to download
01-16 10:41:20.395: I/Gecko(108): *** AUS:SVC readStatusFile - status: null, path: /data/local/updates/0/update.status
01-16 10:41:20.395: E/GeckoConsole(108): AUS:SVC readStatusFile - status: null, path: /data/local/updates/0/update.status
01-16 10:41:20.405: I/Gecko(108): UpdatePrompt: Error downloading update Firefox 18.0: 0
01-16 10:41:20.405: I/Gecko(108): UpdatePrompt: Update error, state: download-failed, errorCode: 0
01-16 10:41:20.415: I/Gecko(108): UpdatePrompt: Setting gecko.updateStatus: The integrity of the update could not be verified
01-16 10:41:20.415: I/GeckoDump(108): XXX FIXME : Got a mozContentEvent: update-available-result

Comment 7

[Dave Hylands [:dhylands]]

(In reply to Tony Chung [:tchung] from comment #6)
> Hey Dave, can you not dupe older bugs to newer bugs?  that is inconsiderate
> to original bug filers as they view their bug becomes useless.   (you duped
> my bug that was filed as 830126 :( )  

So how would you like it handled then? The patch is in this bug, which is currently on inbound. I expect it will land on b2g18 once it lands on central.

> that said, i'm still seeing this problem on today's build, when attempting
> to update from 20130115070201 -> 20130116070204; and watching it fail.

> 01-16 10:41:20.415: I/Gecko(108): UpdatePrompt: Setting gecko.updateStatus:
> The integrity of the update could not be verified

So that looks like the same problem, and should be addressed once this lands.

Comment 8

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/c7572ad0ef03

Comment 9

[Tony Chung [:tchung]]

(In reply to Dave Hylands [:dhylands] from comment #7)
> (In reply to Tony Chung [:tchung] from comment #6)
> > Hey Dave, can you not dupe older bugs to newer bugs?  that is inconsiderate
> > to original bug filers as they view their bug becomes useless.   (you duped
> > my bug that was filed as 830126 :( )  
> 
> So how would you like it handled then? The patch is in this bug, which is
> currently on inbound. I expect it will land on b2g18 once it lands on
> central.

It's okay Dave, im just giving a headsup for next time.  Thanks for understanding.

> 
> > that said, i'm still seeing this problem on today's build, when attempting
> > to update from 20130115070201 -> 20130116070204; and watching it fail.
> 
> > 01-16 10:41:20.415: I/Gecko(108): UpdatePrompt: Setting gecko.updateStatus:
> > The integrity of the update could not be verified
> 
> So that looks like the same problem, and should be addressed once this lands.

cool, thanks for verifying.  happy to reproduce if you need more information.

Comment 10

[Dave Hylands [:dhylands]]

(In reply to Tony Chung [:tchung] from comment #9)
> (In reply to Dave Hylands [:dhylands] from comment #7)
> > (In reply to Tony Chung [:tchung] from comment #6)
> > > Hey Dave, can you not dupe older bugs to newer bugs?  that is inconsiderate
> > > to original bug filers as they view their bug becomes useless.   (you duped
> > > my bug that was filed as 830126 :( )  
> > 
> > So how would you like it handled then? The patch is in this bug, which is
> > currently on inbound. I expect it will land on b2g18 once it lands on
> > central.
> 
> It's okay Dave, im just giving a headsup for next time.  Thanks for
> understanding.

No problem. I normally would have dup'd the other way, but the patch and everything was all done before the dup was discovered.

Comment 11

[Dave Hylands [:dhylands]]

If you have adb access, you should be able to recover by doing:

adb shell rm /sdcard/updates/0/update.mar

and then doing a "Check Now" in the settings app (under Device Information). Preferrably from someplace that you get good wifi.

Comment 12

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-b2g18/rev/2103501b2fd1

Comment 14

[Lukas Blakk [:lsblakk] use ?needinfo]

Landed on mozilla-b2g18/gaia master prior to the 1/25 branching to mozilla-b2g18_v1_0_0/v1.0.0, updating status-b2g-v1.0.0 to fixed.

Comment 15

[mlevin]

Issue is verified fixed on:

Gecko: http://hg.mozilla.org/releases/mozilla-b2g18_v1_0_1/rev/98354c0298ab
Gaia: edaca00b1eb7534120b6255db5d5200fb1d86d65   
build: 20130219070200
kernel date: Dec 5