823517 - Internet connection breaks after a crash in Firefox

Status: RESOLVED INVALID

(Firefox :: Shell Integration, defect)

Description

[Mihaela Velimiroviciu (:mihaelav)]

Mozilla/5.0 (Windows NT 6.1; rv:20.0) Gecko/20121219 Firefox/20.0
Mozilla/5.0 (Windows NT 6.1; rv:18.0) Gecko/20100101 Firefox/18.0 - beta 5

After Firefox crashes, the internet connection doesn't work anymore, so you cannot load any site/use any application which need internet connection (Skype, Yahoo Messenger, etc)

Steps to reproduce:
1. Start Firefox and navigate to some pages
2. Use crash me add-on (http://ted.mielczarek.org/mozilla/crashme.html) to trigger a crash
3. Enter a comment in the crash report and submit
4. Restore session after crash and try to navigate to some pages

Expected result:
1. The report should be submitted successfully and Session Restore should be working as expected. 
2. Comment appears in the crash report. 
3. I can use Firefox normally after the crash (can load sites successfully)

Actual result:
1. Breakpad remains more time than usually in submitting report state.
2. Crash report doesn't contain the user comment (or sometimes it doesn't even get submitted)
3. Network connection breaks: "!" mark on network icon in taskbar and "No Internet access". To fix the problem I need to disable and re-enable the network adapter.


Notes:
1. I can reproduce this only on one machine with the following configuration
OS: Win 7 Ultimate 32bit
Motherboard: Asus P5B-VM SE 
Netwok Adapter: Atheros L1 Gigabit Ethernet 10/100/100Base-T Controller
  - Driver version: 2.4.7.29
  - Driver Date: 10/13/2009

Info from about:support:
  Application Basics

        Name
        Firefox

        Version
        18.0

        User Agent
        Mozilla/5.0 (Windows NT 6.1; rv:18.0) Gecko/20100101 Firefox/18.0

        Build Configuration

          about:buildconfig

  Extensions

        Name

        Version

        Enabled

        ID

        Crash Me Now! Advanced
        0.3
        true
        [e-mail]

        Feedback
        1.2.2
        true
        [e-mail]

        AVG Safe Search
        12.0.0.2191
        false
        {1E73965B-8B48-48be-9C8D-68B920ABC1C4}

  Important Modified Preferences

      Name

      Value

        browser.cache.disk.capacity
        358400

        browser.cache.disk.smart_size.first_run
        false

        browser.cache.disk.smart_size.use_old_max
        false

        browser.cache.disk.smart_size_cached_value
        358400

        browser.places.smartBookmarksVersion
        4

        browser.startup.homepage_override.buildID
        20121219074241

        browser.startup.homepage_override.mstone
        18.0

        extensions.lastAppVersion
        18.0

        gfx.blacklist.direct2d
        3

        gfx.blacklist.layers.direct3d9
        3

        gfx.direct3d.prefer_10_1
        true

        network.cookie.prefsMigrated
        true

        places.history.expiration.transient_current_max_pages
        80510

        privacy.sanitize.migrateFx3Prefs
        true

  Graphics

        Adapter Description
        AMD Radeon HD 6450

        Adapter Drivers
        aticfx32 aticfx32 aticfx32 atiumdag atidxx32 atiumdva

        Adapter RAM
        1024

        Device ID
        0x6779

        Direct2D Enabled
        Blocked for your graphics driver version.

        DirectWrite Enabled
        false (6.1.7600.16972)

        Driver Date
        7-27-2012

        Driver Version
        8.982.0.0

        GPU #2 Active
        false

        GPU Accelerated Windows
        0/1 Basic Blocked for your graphics driver version.

        Vendor ID
        0x1002

        WebGL Renderer
        Google Inc. -- ANGLE (AMD Radeon HD 6450)

        AzureCanvasBackend
        cairo

        AzureContentBackend
        none

        AzureFallbackCanvasBackend
        none

  JavaScript

        Incremental GC
        true

  Accessibility

        Activated
        false

        Prevent Accessibility
        0

  Library Versions

        Expected minimum version

        Version in use

        NSPR
        4.9.4
        4.9.4

        NSS
        3.14.0.1 Basic ECC
        3.14.0.1 Basic ECC

        NSSSMIME
        3.14.0.1 Basic ECC
        3.14.0.1 Basic ECC

        NSSSSL
        3.14.0.1 Basic ECC
        3.14.0.1 Basic ECC

        NSSUTIL
        3.14.0.1
        3.14.0.1

Comment 1

[u279076]

This is a very common test scenario, in that we test it every beta and release, so I would assume this to be a regression since the bug was not filed earlier. Please find a regression range.

CCing Alex Keybl to see if we have any outreach channels with Atheros.

Comment 2

[juan becerra [:juanb]]

I haven't been able to reproduce this problem yet, using a Win7 32bit machine (running on an old Mac Mini).

Comment 3

[juan becerra [:juanb]]

I've since seen the problem described in comment #0 where the crash reporter takes a long time trying to submit the report and at the end it says: "there was a problem trying to submit your report." The crash report is in about:crashes though.

Comment 4

[Matthias Versen [:Matti]]

That looks like a bug in the networking driver. No application should be able to "crash" the connection on the system

Comment 5

[juan becerra [:juanb]]

I've tried some more stress testing, triggering several crash reports using fx18b5 on a Win7 32bit system, so far I've only been able to see the "there was a problem submitting your report" message once.

Then I tried unplugging the ethernet cable while the crash reporter dialog was up, before submitting, and I was able to recreate the message.

Perhaps Mihaela's machine was experiencing network issues while testing?

Comment 6

[Mihaela Velimiroviciu (:mihaelav)]

The issue is reproducible on Firefox 4.0, as well. The reason why this issue was not spotted before may be because of the specific configuration of this machine (it is a temporary machine I got these days). This morning some Windows updates installed and the issue is no longer reproducible after that (I had to restore the machine to yesterday configuration to be able to reproduce the issue again on 4.0).

Comment 7

[u279076]

So at least this isn't a regression then. Given this is "fixed" by a windows update it seems like this isn't a Firefox bug. It would be good to know specifically which updates you installed that fixed this though. In particular if there was an update to any network related firmware or drivers.