Closed Bug 1717644 Opened 3 years ago Closed 2 months ago

Firefox 89 won't load Google, won't search Google in the address bar, won't load any google related website or load google font into any website

Categories

(Core :: Networking, defect, P2)

Product:
Core
Component:
Networking
Version:
Firefox 89
Type:
defect

Tracking

()

Status:
RESOLVED INCOMPLETE
Tracking Flags:
Tracking Status
firefox-esr91 --- affected
firefox94 --- wontfix
firefox95 --- affected
firefox96 --- affected

People

(Reporter: eladrahmin, Unassigned)

References

Details

(Whiteboard: [necko-triaged])

Attachments

(3 files)

Screen Shot 2021-06-23 at 14.50.15.png
48.63 KB, image/png
Details
HTTP log of issue on Debian (gzipped)
1.18 MB, application/gzip
Details
SSL issue with google.com
5.04 KB, application/octet-stream
Details

User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:89.0) Gecko/20100101 Firefox/89.0

Steps to reproduce:

Firefox 89.0.1 on a Mac:

  1. search something in address or search bar
  2. go to any google website (docs, slides, etc..)
  3. Go to any website that has a Google font embedded in it.

Actual results:

  1. Firefox stays on a blank page, doesn't load anything.
  2. Nothing is loaded, stays on a blank page.
  3. No fonts are loaded, it's a system font.

Expected results:

  1. Google search results should have been loaded.
  2. The website should have loaded.
  3. A Google Font was supposed to be in place of the system font.

The Bugbug bot thinks this bug should belong to the 'Core::Layout: Text and Fonts' component, and is moving the bug to that component. Please revert this change in case you think the bot is wrong.

Component: Untriaged → Layout: Text and Fonts
Product: Firefox → Core

This doesn't sound like a Core Layout bug, it sounds like any kind of access to google domains is failing for the reporter. (Might turn out to be a temporary DNS or network communication issue?)

Component: Layout: Text and Fonts → Untriaged
Product: Core → Firefox

Hi Elad, Thank you for taking the time to report this issue, I also tried to reproduce this issue on our side but without success, can you please try creating a fresh profile and see if the issue still occurs? just to rule out any Addons or maybe some preferences set, here are the steps on how to create a fresh profile:
https://support.mozilla.org/en-US/kb/profile-manager-create-and-remove-firefox-profiles?redirectlocale=en-US&redirectslug=Managing-profiles#w_starting-the-profile-manager

Also can you please try our lates Nightly build ? you can find it here https://nightly.mozilla.org/

and just to rule out any Network issues does this issue occur on Google Chrome ? or Edge ?

Component: Untriaged → Networking
Flags: needinfo?(eladrahmin)
Product: Firefox → Core
Flags: needinfo?(eladrahmin)

(In reply to Rares Doghi from comment #3)

Hi Elad, Thank you for taking the time to report this issue, I also tried to reproduce this issue on our side but without success, can you please try creating a fresh profile and see if the issue still occurs? just to rule out any Addons or maybe some preferences set, here are the steps on how to create a fresh profile:
https://support.mozilla.org/en-US/kb/profile-manager-create-and-remove-firefox-profiles?redirectlocale=en-US&redirectslug=Managing-profiles#w_starting-the-profile-manager

Also can you please try our lates Nightly build ? you can find it here https://nightly.mozilla.org/

and just to rule out any Network issues does this issue occur on Google Chrome ? or Edge ?

Hi, checked Chrome and it works without any problem.
I Attached an image to this issue showing the security error I get.

Can you make a HTTP log:
https://firefox-source-docs.mozilla.org/networking/http/logging.html

Please note that log may contain cookie and URL you are visiting. You can also send the log to my email address. Use a clean profile if possible (for you privacy, because logs may contain cookies)

Flags: needinfo?(eladrahmin)

Can confirm bug on Debian 11 (Bullseye, testing)
Useragent: Mozilla/5.0 (X11; Linux x86_64; rv:89.0) Gecko/20100101 Firefox/89.0
Firefox version: 89.0.2
Sometimes google and it's services (i.e. youtube) works, sometimes they don't. No such issue on Chromium.
I did HTTP log. I have compressed it from 26MB, (only 8 secs of logging), where I tried to search something in google (via search bar) and open google.com address (via url bar). I have attached logs to issue.

I also noticed that if I try to search in duckduckgo using google results (you can do that by adding !g at beginning in search bar) it won't work as well.

can confirm on arch linux and firefox 89.0.2
i will be logged into google services, and then all get requests (including just opening google.com) will fail
no response logged in the developer tools
i need to clear google cookies to temporarily resolve this, but will re-appear ~2hours later
this is absolutely KILLING my work productivity
PLEASE FIX THIS ASAP!!!!!!!!!
if this isnt resolved soon, i will need to drop firefox permanently

After browsing some bugs reports I tried to disable http3 in about:config (network.http.http3.enabled), but without luck - FF still doesn't open any google sites. It worked for a while after FF restart, but now it doesn't again.
Can you assign this issue to someone? Do you need more confirmations? How can I help you with resolving this bug? It's really annoying (I'm working alongside with Chromium at this moment), so I'll do my best to help you with this.

Flags: needinfo?(jstutte)

Can confirm this happens on Windows 10 (10.0.19042 Build 19042) as well, both in 89.0.2 and Nightly. Still haven't found a way to reproduce this. It seems to happen at random.

Google apps in general become unresponsive to anything that requires a network call. Not even refreshing the page works (meaning: the page doesn't even refresh). Closing the tab and reopening the app in another tab doesn't work either. The only solution is restarting the browser.

If further info is needed I can provide it.

Assignee: nobody → dd.mozilla
Severity: -- → S3
Status: UNCONFIRMED → ASSIGNED
Ever confirmed: true
Priority: -- → P2
Whiteboard: [necko-triaged]

On this connection 0RTT was used. and the 0RTT-data has been accepted:

2021-06-24 12:26:59.986088 UTC - [Parent 1434226: Socket Thread]: V/nsHttp nsHttpConnection::EnsureNPNComplete [this=7f996139f400] - early data that was sent during 0RTT has been accepted [rv=0].
2021-06-24 12:26:59.986097 UTC - [Parent 1434226: Socket Thread]: I/nsHttp Http2Session::Finish0RTT 7f995a9d2800 aRestart=0 aAlpnChanged=0
2021-06-24 12:26:59.986106 UTC - [Parent 1434226: Socket Thread]: V/nsHttp nsHttpTransaction::Finish0RTT 7f995a9b9800 0 0

but it looks like the NSS is already considering the TLS connection to be in error state. This can be seen when changing keep-live flags fails:

021-06-24 12:26:59.986457 UTC - [Parent 1434226: Socket Thread]: D/nsSocketTransport nsSocketTransport::SetKeepaliveEnabled [7f99613a1000] disabled, idle time[10s] retry interval[1s] packet count[4]: globally enabled.
2021-06-24 12:26:59.986473 UTC - [Parent 1434226: Socket Thread]: D/nsSocketTransport ErrorAccordingToNSPR [in=-12190 out=805a2f9e]
2021-06-24 12:26:59.986481 UTC - [Parent 1434226: Socket Thread]: D/nsSocketTransport SetKeepaliveEnabled failed rv[0x805a2f9e]
2021-06-24 12:26:59.986489 UTC - [Parent 1434226: Socket Thread]: D/nsSocketTransport SetKeepaliveEnabledInternal failed rv[0x805a2f9e]

Necko performs TLS verification for HTTP2 and the checks pass:

2021-06-24 12:26:59.986556 UTC - [Parent 1434226: Socket Thread]: I/nsHttp Http2Session::ConfirmTLSProfile 7f995a9d2800 mConnection=7f99654cd580
2021-06-24 12:26:59.986565 UTC - [Parent 1434226: Socket Thread]: V/nsHttp nsHttpConnection::GetSecurityInfo trans=7f995a9d2800 tlsfilter=0 socket=7f99613a1018
2021-06-24 12:26:59.986574 UTC - [Parent 1434226: Socket Thread]: I/nsHttp Http2Session::ConfirmTLSProfile 7f995a9d2800 sslsocketcontrol=7f995ead1948
2021-06-24 12:26:59.986582 UTC - [Parent 1434226: Socket Thread]: I/nsHttp Http2Session::ConfirmTLSProfile 7f995a9d2800 version=304
2021-06-24 12:26:59.986590 UTC - [Parent 1434226: Socket Thread]: I/nsHttp Http2Session::ConfirmTLSProfile 7f995a9d2800 MAC Algortihm (aead==6) 6

but then reading form the socket returns the error:

2021-06-24 12:26:59.986598 UTC - [Parent 1434226: Socket Thread]: I/nsHttp Http2Session::ReadSegments 7f995a9d2800
2021-06-24 12:26:59.986616 UTC - [Parent 1434226: Socket Thread]: I/nsHttp Http2Session 7f995a9d2800 could not identify a stream to write; suspending.
2021-06-24 12:26:59.986626 UTC - [Parent 1434226: Socket Thread]: V/nsHttp nsHttpConnection::OnSocketWritable 7f996139f400 ReadSegments returned [rv=80470007 read=0 sock-cond=0 again=1]
2021-06-24 12:26:59.986634 UTC - [Parent 1434226: Socket Thread]: D/nsSocketTransport nsSocketInputStream::OnSocketReady [this=7f99613a1270 cond=0]
2021-06-24 12:26:59.986642 UTC - [Parent 1434226: Socket Thread]: V/nsHttp nsHttpConnection::OnSocketReadable [this=7f996139f400]
2021-06-24 12:26:59.986650 UTC - [Parent 1434226: Socket Thread]: I/nsHttp Http2Session::WriteSegments 7f995a9d2800 InternalState 0
2021-06-24 12:26:59.986658 UTC - [Parent 1434226: Socket Thread]: D/nsSocketTransport nsSocketInputStream::Read [this=7f99613a1270 count=9]
2021-06-24 12:26:59.986665 UTC - [Parent 1434226: Socket Thread]: D/nsSocketTransport calling PR_Read [count=9]
2021-06-24 12:26:59.986673 UTC - [Parent 1434226: Socket Thread]: D/nsSocketTransport PR_Read returned [n=-1]
2021-06-24 12:26:59.986684 UTC - [Parent 1434226: Socket Thread]: D/nsSocketTransport ErrorAccordingToNSPR [in=-12190 out=805a2f9e]

0x805a2f9e is SSL_ERROR_PROTOCOL_VERSION_ALERT and firefox closes the connection.

bbeurdouche, can you help figuring out why this is happening?
nss would set SSL_ERROR_DOWNGRADE_WITH_EARLY_DATA when connection is downgraded from 1.3 to 1.2.

Flags: needinfo?(bbeurdouche)

And also this line in the log:
2021-06-24 12:26:59.986582 UTC - [Parent 1434226: Socket Thread]: I/nsHttp Http2Session::ConfirmTLSProfile 7f995a9d2800 version=304

shows that TLS version is 1.3. (maybe it is only not updated because TLS is already in the error state)

I am trying to reproduce but I didn't manage to do so yet. I'll keep having a look.
Maybe @Martin has some ideas here...

Flags: needinfo?(mt)

Here's my attempt to reproduce this:

while SSLTRACE=50 ../nss-tools/nss-run.sh tstclnt -b -D -h www.google.com -p 443 -V :tls1.3 -v -s -u -Z -B -A ~/GET -G -L 2 >log 2>&1; do sleep 1; done

~/GET contains:

GET / HTTP/1.1
Host: www.google.com
Connection: close

So far, all I've likely done is trip DoS protections on Google's servers. It doesn't look like this is failing.

I know that Google only recently enabled 0-RTT on their servers and this could be a bug on their end. We've certainly never had any problems using 0-RTT with other servers. I've asked their developers about this problem.

Flags: needinfo?(mt)

I should add that people who are stuck dealing with this problem can disable TLS early data by going to about:config and flipping security.tls.enable_0rtt_data to false. (Edit: ...probably. If my theory about 0-RTT is correct.)

May I ask someone who can reproduce to make a wireshark capture?

Thank you.

Flags: needinfo?(bbeurdouche)
See Also: → 1719312

A couple more notes:

  1. I went on and installed Wireshark to provide a dump for when it happens, but then I realized I don't want to accidentally (even though it's SSL) share my google.com cookies, so I deleted them beforehand - and the issue doesn't reproduce after deleting them... That coincides with me being able to google freely when I start a new Private browsing window. Can I really share my Wireshark dump without worrying it may leak unwanted information?
  2. Other SSL sites work fine, only google.com / gmail.com don't work.

Ok, without logging in to my gmail account, it now happens again... Attached SSL pcap.

(In reply to blueblue from comment #19)

Ok, without logging in to my gmail account, it now happens again... Attached SSL pcap.

Flags: needinfo?(jstutte) → needinfo?(dd.mozilla)

We are still looking into this. We do not know what is causing. It is probably caused by some middleboxes (rewriting something in the TLS record layer).

Do you know if there are any firewalls in the network where this is happening?

Flags: needinfo?(dd.mozilla)

(In reply to Dragana Damjanovic [:dragana] from comment #21)

Do you know if there are any firewalls in the network where this is happening?

This is happening from my home connection. The only thing between me and the internet is my home router, which I have limited access to (it's the one leased from the provider so the interface is pretty barebones). Is there anything I can try and check about the router configuration which can help?

(In reply to Elad from comment #5)

Hi, checked Chrome and it works without any problem.
I Attached an image to this issue showing the security error I get.

Can you try the Firefox Nightly build available at https://nightly.mozilla.org/ and check if the issue still happens?

Bug 1718520 may have fixed it.

I am using Firefox 99.0.1 and I still see this issue

Same problem when using Firefox 99.0.1 (64-bit) on Windows 11

Could you try to record a http log?
Thanks.

Flags: needinfo?(renatoyamane)

(In reply to Kershaw Chang [:kershaw] from comment #27)

Could you try to record a http log?

Logging 2 minutes resulted in a ~200Mb file.
How can I filter this for you?

Workaround on Comment #15 fixed the problem.

You could zip it and send the log to necko at mozilla dot com.
Thanks.

(In reply to Kershaw Chang [:kershaw] from comment #29)

You could zip it and send the log to necko at mozilla dot com.
Thanks.

Sent 30 seconds of the log to your email as requested.
Workaround on Comment #15 fixed the problem.

Flags: needinfo?(renatoyamane)

My brother reached out to me and explained he has the same issue. I just recently came across this bug in bugzilla so wanted to share the info I had too. My brother wrote:

"The 99 release caused me to not be able to use Google at least once a week. Had to find a post online where people had similar problem and a workaround for it - https://www.reddit.com/r/firefox/comments/pi9rn0/google_stops_working_in_firefox_only/. I almost had to quit using Firefox as my primary browser because of it. People noticed it since 91. I had sporadic problem with accessing Google services (mail, news, voice, hangouts/chat) on the web browser. One workaround was to close all tabs/windows, but that wasn't a good workaround for me since I tend to have a lot of them. Somehow changing security.tls.enable_0rtt_data to false allowed Google to be more reliable again."

I am now following this bug to see the latest status.

Redirect a needinfo that is pending on an inactive user to the triage owner.
:dragana, since the bug has high priority and recent activity, could you have a look please?

For more information, please visit auto_nag documentation.

Flags: needinfo?(eladrahmin) → needinfo?(dd.mozilla)

I've started to experience these symptoms, sometime in the past few months. Sometimes I go days with no symptoms, other times they occur several times a day. I just discovered this bug, as I was about to open my own bug. Here is what I was getting ready to write in a new bug (until I found this one)

=================

After Firefox has been running for a while (sometimes minutes, sometimes days) Firefox "ignores" requests towards www.google.com. I type https://www.google.com into address bar, hit [Enter] but then nothing happens. No errors. It's like Firefox is ignoring the request.

Firefox 101.0 (20220526203855) on company-managed Windows 10 Enterprise 21H2 (19044.1706)

I can reproduce in a brand new profile with no modifications. See linked screen capture of brand new profile showing these symptoms. https://ufile.io/n5423om8

I can fix the symptoms by restarting Firefox. I've discovered that I can also fix the symptoms using History, Clear Recent History, then clear only the "Active Logins". As shown in linked screen capture. Which is strange, because this is a new profile, and I'm not "logged into" anything at all in this profile.

These same symptoms occur on my "real" profile. I noticed these symptoms sometime in the last couple of months on my "real" profile. The unpredictability is annoying. Sometimes days without this symptom, then sometimes symptoms several times a day.

I've experienced symptoms on www.google.com and maps.google.com. But not for news.google.com or mail.google.com. I've only experienced symptoms on google.com sites, never any other site with these symptoms.

I just tried the workaround in comment #15 to set security.tls.enable_0rtt_data to false in about:config in my fresh profile. This also resolved the symptoms.

I'm setting back in fresh profile, but then will try this in my "real" profile to see if it stops symptoms in my "real" profile.

This might be the same as bug 1753204.

(In reply to Mathew Hodson from comment #35)

This might be the same as bug 1753204.

FYI, bug 1753204 descriptions says an explicit error is displayed. The symptoms described in this bug, including mine, do not include any error displayed.

FYI: It's been 2 weeks with no symptoms after I changed security.tls.enable_0rtt_data to false in my real profile. This workaround is working for me. Previously the symptoms would occur at least every few days.

(In reply to netjeff from comment #37)

FYI: It's been 2 weeks with no symptoms after I changed security.tls.enable_0rtt_data to false in my real profile. This workaround is working for me. Previously the symptoms would occur at least every few days.

Yes, the workaround is also working for me.
I sent the log as requested on Comment #30 two months ago, but I never received any feedback, so looks like this 1 year old bug is on the limbo.

Hi Renato, netjeff, Rosanna,

Thank you both for the confirmation that disabling 0-rtt fixed the issue. We're looking at how we can overhaul the way we try to recover from 0-RTT errors to fix this class of problems, but it isn't a simple task. Please be assured this hasn't fallen off our radar and we're actively working on the issue.

I am having a similar issue on OpenSUSE 15.4 (and 15.3 and tumbleweed) and Firefox 91.12.0esr (64-bit).
The issue does not occur under OpenSUSE 15.2.

The issue also only occurs for me when connected to a work-related Fortinet VPN. When disconnected from the VPN, the issue does not occur.

*** The workaround of setting "security.tls.enable_0rtt_data" to "false" immediately resolved the issue. ***

The problem does not immediately occur, but happens after a few minutes of being connected to the vpn, At that point, when logged into my google profile, I cannot use the firefox search functionality (nothing happens on submit). and I cannot load gmail.com.

However, all of these work when I am using a private window or if I use chrome.

A couple of weeks back, I also verified that it occurred (or did not occur under 15.2) for various versions of firefox 0esr, which makes me think that it is not specific to firefox, per se, but is also operating-system and/or network-configuration related. It is definitely related to Google's websites, though.

Using Firefox 104 on Windows 10 and having this exact same issue, which is causing me to use Chrome most of the times at work because I can no longer rely on this browser for something so basic yet so frequently used. Issue extends beyond google.com pages, for example when it happens all thumbnails on YouTube will show gray (failed to load) and no amount of force refresh fixes it.

I already uninstalled Firefox completely, deleted all remnants of its files in AppData as well, and tried again with no profile logged in, eventually the same problem returns.

There are 2 things that kinda help work around this issue, but yet again just for a limited time:

  • Open about:config and change any setting, doesn't matter which setting it is, then google.com loads just fine
  • Open google.com in private mode at least once and then the regular mode starts working again

This has been happening to me for the past 3 months now, and I keep the browser up to date with whichever latest version becomes available automatically.

I hope this issue gets fixed soon because I am tired of switching to Chrome every single time this browser just fails to be a basic browser, if this keeps going on without a fix for longer I will eventually drop it completely and switch to Chrome permanently during work hours.

(In reply to particlecore from comment #41)

Using Firefox 104 on Windows 10 and having this exact same issue, which is causing me to use Chrome most of the times at work because I can no longer rely on this browser for something so basic yet so frequently used. Issue extends beyond google.com pages, for example when it happens all thumbnails on YouTube will show gray (failed to load) and no amount of force refresh fixes it.

I already uninstalled Firefox completely, deleted all remnants of its files in AppData as well, and tried again with no profile logged in, eventually the same problem returns.

There are 2 things that kinda help work around this issue, but yet again just for a limited time:

  • Open about:config and change any setting, doesn't matter which setting it is, then google.com loads just fine
  • Open google.com in private mode at least once and then the regular mode starts working again

This has been happening to me for the past 3 months now, and I keep the browser up to date with whichever latest version becomes available automatically.

I hope this issue gets fixed soon because I am tired of switching to Chrome every single time this browser just fails to be a basic browser, if this keeps going on without a fix for longer I will eventually drop it completely and switch to Chrome permanently during work hours.

I want to add that this happens at home network, I work remotely with and without VPN, makes no difference whatsoever.

This may be fix with bug 1768250.

Flags: needinfo?(dd.mozilla)

Just got this error when re-logging (for checkout) on decathlon.ro/.
I think i fixed this by restarting the local router. The system used to access the website was a laptop that was running for about 12h, with intermittent sleep/wakeup cycles, and Firefox was running with the website open this whole time.

Got the idea from https://bugzilla.mozilla.org/show_bug.cgi?id=1768250#c15 where a middle box is mentioned. I should mention that I didn't disable enable_0rtt_data for this.

Disconnecting and reconnecting to the wi-fi didn't do anything, but restarting the router fixed it, so it does seems like a middle box issue.

Using 105.0.2 (64-bit), on Linux Mint.

I will not have time to work on this bug, so I am unassigning myself.

Assignee: dd.mozilla → nobody
Status: ASSIGNED → NEW

Putting back into the priority process

Whiteboard: [necko-triaged] → [necko-triaged][necko-priority-new]

We're going to try to land bug 1753204, and then ask the reporters to see if they can reproduce and verify if it's fixed.

Whiteboard: [necko-triaged][necko-priority-new] → [necko-triaged][necko-priority-review]

We'll just land bug 1753204, though this was probably already fixed by bug 1718520.
Please reopen if you're still seeing this issue.

Status: NEW → RESOLVED
Closed: 2 months ago
Resolution: --- → INCOMPLETE
Whiteboard: [necko-triaged][necko-priority-review] → [necko-triaged][reminder-test 2024-04-15]

a month ago, valentin placed a reminder on the bug using the whiteboard tag [reminder-test 2024-04-15] .

jesup, please refer to the original comment to better understand the reason for the reminder.

Flags: needinfo?(rjesup)
Whiteboard: [necko-triaged][reminder-test 2024-04-15] → [necko-triaged]

Bug 1753204 landed successfully and is riding the trains with Firefox 125.

Flags: needinfo?(rjesup)
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: