453305 - Notify user when out of disk space

Status: RESOLVED DUPLICATE of bug 346003

(Toolkit :: Places, defect)

Description

[timeless]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1b1pre) Gecko/20080901033305 Minefield/3.1b1pre
Build Identifier: Build identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1b1pre) Gecko/20080901033305 Minefield/3.1b1pre

visiting urls that i haven't been to before doesn't result in them being entered into my history.

for example. before today, i'd never visited https://bugzilla-stage-tip.mozilla.org/



Reproducible: Always

Steps to Reproduce:
1. so, i visited it.
- load https://bugzilla-stage-tip.mozilla.org/

2. then i went to another window or tab and tried to write:
- bugzilla stage tip
Actual Results:  
the spinner spun and gave me no results

Expected Results:  
since i've visited bugzilla-stage-tip.mozilla.org a moment ago, it should be in my found history :)

i've always wanted to file a useless bug like this.

notes:
this morning i had <1mb of free space on my laptop
i now have ~3gb of free space on my laptop

this error appears many times in my js console:
Error: [Exception... "Component returned failure code: 0x80520010 (NS_ERROR_FILE_NO_DEVICE_SPACE) [nsINavBookmarksService.getBookmarkIdsForURI]"  nsresult: "0x80520010 (NS_ERROR_FILE_NO_DEVICE_SPACE)"  location: "JS frame :: file:///C:/home/Desktop/firefox-3.1b1pre.en-US.win32/firefox/modules/utils.js :: PU_getMostRecentBookmarkForURI :: line 872"  data: no]
Source File: chrome://browser/content/tabbrowser.xml
Line: 446

Comment 1

[Shawn Wilsher :sdwilsh]

Out of curiosity, does restarting the browser fix this?

Comment 2

[timeless]

um, i don't intend to restart the browser until the live process is investigated. if you give me instructions for copying the files into another profile, i can test them there....

Comment 3

[Shawn Wilsher :sdwilsh]

What I'm curious about is if the service manager is caching the fact that places failed to init once.  I don't think it does that however...

Comment 4

[Shawn Wilsher :sdwilsh]

No reply to comment 3 - closing as INCOMPLETE.

Comment 5

[timeless]

sdwilsh: so.. the service manager didn't fail....

2599 nsNavBookmarks::GetBookmarkIdsForURITArray(nsIURI *aURI,
2600                                          nsTArray<PRInt64> *aResult) 

2604   nsresult rv = BindStatementURI(mDBFindURIBookmarks, 0, aURI);
2605   NS_ENSURE_SUCCESS(rv, rv);
2609   while (NS_SUCCEEDED((rv = mDBFindURIBookmarks->ExecuteStep(&more))) && 
2615   NS_ENSURE_SUCCESS(rv, rv);

these are the two possible paths.

those would only be reached if

107 nsNavBookmarks::Init()

didn't fail

Comment 6

[Marco Bonardo [:mak]]

could just be places had another free space to init, but next calls to sqlite did not have enough space for the journal.

not sure what we can do about that, handling the disk space error in all storage calls does not sound applicable.

Comment 7

[Marco Bonardo [:mak]]

no reasons to stay UNCO, full disk conditions can happen, clearly.