Re: Issue 439818 in chromium: Samsung 303: frequent dropping of wifi

276 views
Skip to first unread message

chro...@googlecode.com

unread,
Apr 15, 2015, 11:18:05 PM4/15/15
to chromi...@chromium.org

Comment #183 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
dropping of wifi
https://code.google.com/p/chromium/issues/detail?id=439818

Samsung Series 3 Chromebooks 303C-A12 (XE303C12-A01US)
Google Chrome 43.0.2357.19 (Official Build) dev (32-bit)
Revision 295a04f7fb673164cd5e82635b94c43fee69c5dd
Platform 6946.10.0 (Official Build) dev-channel daisy_freon

#wifi43 reproduced wifi drop by opening 10 youtube videos, after
alt-tabbing between videos the wifi turned off, lost ssid, had to rescan
before reconnecting. Took more than a few seconds to recover.

After recovering, I successfully watched 10 youtube videos simultaneously,
closed all 10 tabs, and then while watching a single youtube video
reproduced #wifi43 wifi drop but this time connection was regained after
just a couple seconds.

Afterward I submitted the log with tag #wifi43.

--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

chro...@googlecode.com

unread,
Apr 16, 2015, 12:22:05 AM4/16/15
to chromi...@chromium.org

Comment #184 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
The firmware for #183 is: Google_Snow_Rev4.2695.128.0

I also reproduced #wifi43 on: Google_Snow.2695.132.0
Samsung Series 3 Chromebooks 303C-A12 (XE303C12-A01US)
Google Chrome 43.0.2357.19 (Official Build) dev (32-bit)
Revision 295a04f7fb673164cd5e82635b94c43fee69c5dd
Platform 6946.10.0 (Official Build) dev-channel daisy_freon

but with many subtle differences.

I'll submitting logs for all four of my Samsung Series 3 Chromebooks
303C-A12 (XE303C12-A01US) with the tag "#wifi43" over the next hour.

chro...@googlecode.com

unread,
Apr 16, 2015, 1:03:07 AM4/16/15
to chromi...@chromium.org

Comment #185 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
#wifi43 drops wifi on all four: Samsung Series 3 Chromebooks 303C-A12
(XE303C12-A01US)

On the Chromebook that makes #crackling noises (even when the speakers are
muted) the dropping of wifi seems to correspond to the speakers emitting
static, popping, and crackling noises (I'm guessing both the #wifi43
dropping and the #crackling are a result of the same memory contention
scenario).

https://code.google.com/p/chromium/issues/detail?id=471506

On the Chromebook I'm testing now wifi dropped 4 times in 3 minutes while
watching Netflix.

I also experienced wifi dropping simply from cut and paste of URL links
about Big Government and from using gmail to write about Big Government. A
paranoid person might mistake this coincidence as indicative that the
Samsung firmware is falling out of sync with Chrome OS because it has
active keystroke listeners.

Wifi dropped permanently while I was typing that last sentence. Signal to
noise? I'm not kidding I had to plug in usb3-to-ethernet to submit this.

I submitted logs for all four with the tag #wifi43.

I submitted multiple logs to just now to demonstrate that the wifi dropped
while I was typing this and I had to connect to ethernet to submit this
comment.

chro...@googlecode.com

unread,
Apr 16, 2015, 1:56:48 AM4/16/15
to chromi...@chromium.org

Comment #186 on issue 439818 by julien.puydt: Samsung 303: frequent
I'm using the beta channel 6812.75.0 (the dev channel had a broken french
keyboard in addition to the wifi issues, so I switched since my last
message here), and this morning, I opened my chromebook, opened a first
page, which loaded, then tried a second... no network, again. Back to the
phone and cable! :'-(

chro...@googlecode.com

unread,
Apr 16, 2015, 3:44:21 PM4/16/15
to chromi...@chromium.org

Comment #187 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
If you are seeing issues please continue to submit feedback. We hoped that
patch would fix it but we are continuing to read all reports to continue
debugging.

For users getting Wi-Fi drop: can you indicate whether the Wi-Fi icon
starts to a) pulse, b) turn into an X, c) show an empty Wi-Fi cone, d) show
a Wi-Fi cone with one or more bars?

chro...@googlecode.com

unread,
Apr 16, 2015, 3:58:32 PM4/16/15
to chromi...@chromium.org

Comment #189 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
#188 cbuck...@gmail.com: CBuck's report may not read as scientific but I
think his description most accurately describes my own. I've been testing
the latest fix on four Chromebooks and I can plus-one all the observations
in #188 for each of my Samsung Series 3 Chromebooks 303C-A12
(XE303C12-A01US).

I'll continue to submit reports:

alt-shift-i

#wifi43

#pulse_cone | #x_cone | #empty_cone | #bar_cone

I just had a #wifi43 #x_cone while I was typing this. So now I'm
alt-shift-i and submitting with just those two tags in the description.

Hope this helps.

chro...@googlecode.com

unread,
Apr 17, 2015, 3:56:14 AM4/17/15
to chromi...@chromium.org

Comment #195 on issue 439818 by Pw.Rob...@gmail.com: Samsung 303: frequent
just to note the 6812.75.2 build is now on stable channel: my Samsung has
just been updated to this.

chro...@googlecode.com

unread,
Apr 17, 2015, 7:34:42 AM4/17/15
to chromi...@chromium.org

Comment #196 on issue 439818 by julien.puydt: Samsung 303: frequent
Im still on 6812.75.0 and I have something new : after yet another crash (I
just opened the lid... my mistake!), I logged in, the wifi icon said
everything was fine... but nothing would get through! After plugging the
phone in, I was still having connection issues until I disabled wifi. Then
I still had connection issues and I was still seeing both ethernet and my
home wifi in the list of connections ; the wifi one disappeared when I
clicked on it to have more information, and I was finally really online!

chro...@googlecode.com

unread,
Apr 17, 2015, 9:37:48 AM4/17/15
to chromi...@chromium.org

Comment #197 on issue 439818 by giro...@gmail.com: Samsung 303: frequent
Using 6812.75.2 since yesterday on HP11 (arm) with free.fr V5 modem router
(in france). Still getting dropped connections but only 2 or 3 times per
day. No longer need to reboot - connection comes back on its own after
about 20 secs.

Ran Chrome Connectivity Diagnostics V1.1.3 while the wifi was down and had
the following output. This is repeatable. (No other devices are having
issues through the same modem/router)

SUMMARY
Test Name: Resolver Present Test
Test ID: 3
Test verdict: Problem detected

CRITERIA
DNS resolver test performs a single DNS query and returns a failing result
if a valid response is not returned.
DNS resolver test performs a single DNS query and returns a passing result
if a valid response is returned.


DETAILS
Fri Apr 17 2015 14:23:01 GMT+0200 (CEST) - Attempting to resolve hostname:
ccd-testing-v4.gstatic.com
Fri Apr 17 2015 14:23:39 GMT+0200 (CEST) - DNS server returned DNS query
results.
Fri Apr 17 2015 14:23:40 GMT+0200 (CEST) - DNS server returned problem
status code: -21

chro...@googlecode.com

unread,
Apr 18, 2015, 7:36:07 AM4/18/15
to chromi...@chromium.org

Comment #201 on issue 439818 by reis...@gmail.com: Samsung 303: frequent
#wifi42

Issue timeline with #wifi42:
- Everytime I start using my Samsung 303, I start a chrosh session pinging
my router
- Start browsing as usual.
- Any delay, check ping: will see "buffer overload" message
- Wifi icon goes up to top four bars, no matter signal strength
- What happens after seems to be revision related: until last update, it
was reseting the adaptor, now it freezes (I have to switch off and on).

Comments: Why the signal icon goes up? How can it be related to i/o buffers?

Versão 42.0.2311.87 beta
Plataforma 6812.75.2 (Official Build) beta-channel daisy
Firmware Google_Snow.2695.132.0

chro...@googlecode.com

unread,
Apr 18, 2015, 2:58:30 PM4/18/15
to chromi...@chromium.org

Comment #202 on issue 439818 by edtay...@citcom.net: Samsung 303: frequent
303C Belkin router

Version 42.0.2311.87
Platform 6812.75.2 (Official Build) stable-channel daisy
Firmware Google_Snow.2695.117.0

Website slow loading, err address not found, then wifi dropout and
re-established with website displayed.

Unfortunately I broke out with shingles around left eye and can't give this
a rigorous workout.

chro...@googlecode.com

unread,
Apr 18, 2015, 3:09:55 PM4/18/15
to chromi...@chromium.org

Comment #203 on issue 439818 by edtay...@citcom.net: Samsung 303: frequent
dropping of wifi
https://code.google.com/p/chromium/issues/detail?id=439818

#wifi42
303C Belkin router

Version 42.0.2311.87
Platform 6812.75.2 (Official Build) stable-channel daisy
Firmware Google_Snow.2695.117.0

Similar occurrence process takes approx 2 minutes total from page change
initiated until wifi has dropped and re-established. Display
ERR_NETWORK_CHANGED.

chro...@googlecode.com

unread,
Apr 18, 2015, 10:35:50 PM4/18/15
to chromi...@chromium.org

Comment #204 on issue 439818 by terri192...@gmail.com: Samsung 303:
Version 41.0.2272.118
Platform 6680.81.0 (Official Build) stable-channel daisy
Firmware Google_Snow.2695.132.0

I'm using my schools wifi (away at college) and it happens really randomly
throughout the day. Sometimes it'll happen very frequently, about every
10-15 minutes. Other times I can go hours without wifi drop out. I know
it's not the wifi because my streaming will go completely uninterrupted.
The only way I can get it to even recognize any sort of network, despite
there being about 3 around me, is to completely shutdown and reboot. It's
causing a lot of issues with schoolwork!

chro...@googlecode.com

unread,
Apr 19, 2015, 3:17:40 PM4/19/15
to chromi...@chromium.org

Comment #205 on issue 439818 by giro...@gmail.com: Samsung 303: frequent
Report back

Version 42.0.2311.87 beta
Platform 6812.75.2 (Official Build) beta-channel daisy_spring
Firmware Google_Spring.3824.129.0

Since post #198 I have changed my DNS servers to google servers 8.8.8.8 and
8.8.4.4 and have not have a drop out since. Fingers crossed.

chro...@googlecode.com

unread,
Apr 19, 2015, 3:20:37 PM4/19/15
to chromi...@chromium.org

Comment #206 on issue 439818 by giro...@gmail.com: Samsung 303: frequent
dropping of wifi
https://code.google.com/p/chromium/issues/detail?id=439818

Report back

Version 42.0.2311.87 beta
Platform 6812.75.2 (Official Build) beta-channel daisy_spring
Firmware Google_Spring.3824.129.0

Since post #197 I have changed my DNS servers to google servers 8.8.8.8 and

chro...@googlecode.com

unread,
Apr 19, 2015, 3:21:38 PM4/19/15
to chromi...@chromium.org

Comment #207 on issue 439818 by giro...@gmail.com: Samsung 303: frequent

chro...@googlecode.com

unread,
Apr 20, 2015, 3:59:17 PM4/20/15
to chromi...@chromium.org

Comment #210 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
"Please note we're leaving out a lot of technical commentary."

Please post more technical commentary. The information in #140 and #290 is
crucial for technical users and Chromebook developers to know which
symptoms are attributable to which issues and how we can help reproduce
them.

If it weren't for the positive feedback from the Chromium Team I would have
stopped using WiFi a long time ago and I certainly wouldn't be tolerating
the plethora of other dev-channel stability issues just so I could produce
more crash logs.

Many Chromebook developers are already supporting their user-bases for
their own apps. The more accurate technical information from the very top
means less emotional frustration at the very bottom.

chro...@googlecode.com

unread,
Apr 21, 2015, 10:24:17 AM4/21/15
to chromi...@chromium.org

Comment #212 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
#211 Thanks for correcting me on #210. I've deleted and revised below:

Thank you for the detailed technical commentary. We realize that
maintaining a public forum can be a distraction and the Chromium developer
community is grateful for Google's commitment to open source transparency.

With the help of #140 and #209 the community is better able to focus their
testing and feedback. Detailed analysis such as this is crucial for
encouraging the community to upgrade to dev-channel and help isolate which
symptoms involve which specific issues.

Many Chromebook developers are now supporting clientele Chromebooks and the
more technical information from the very top means less emotional

chro...@googlecode.com

unread,
Apr 21, 2015, 5:25:40 PM4/21/15
to chromi...@chromium.org

Comment #214 on issue 439818 by FloridaG...@gmail.com: Samsung 303:
#213, I've contemplated going back to the supplier and asking for a
replacement of another brand, but assumed it was an issue across all
manufacturers, having seen comments on here about Asus and other brands.

Is this issue exclusive to the Samsung Chromebook? Maybe Chromium staff can
clarify. If so, I at least have recourse to address the issue for my
mother, although I appreciate all the effort to correct the problem on the
Chrome side.

chro...@googlecode.com

unread,
Apr 21, 2015, 5:26:40 PM4/21/15
to chromi...@chromium.org

Comment #215 on issue 439818 by FloridaG...@gmail.com: Samsung 303:
frequent dropping of wifi
https://code.google.com/p/chromium/issues/detail?id=439818

#213, I've contemplated going back to the supplier and asking for a
replacement of another brand, but assumed it was an issue across all
manufacturers, having seen comments on here about Asus and other brands.

Is this issue exclusive to the Samsung Chromebook? Maybe Chromium staff can
clarify. If so, I at least have recourse to address the issue for my
mother, although I appreciate all the effort to correct the problem on the
Chrome side. #wifi40

chro...@googlecode.com

unread,
Apr 21, 2015, 5:45:43 PM4/21/15
to chromi...@chromium.org

Comment #216 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
The chromium team has no evidence of the theory posted in #213 and are
continuing investigation into the issue.

This specific disconnection issue in this bug is specific to the Samsung
303 series, HP 11" series systems. We do not believe this issue occurs on
other Chromebooks at this time.

However if you are basing purchasing decisions on this we highly recommend
you consider testing a different router first-- most users have router
issues rather than system issues.

chro...@googlecode.com

unread,
Apr 21, 2015, 6:51:42 PM4/21/15
to chromi...@chromium.org

Comment #220 on issue 439818 by reis...@gmail.com: Samsung 303: frequent
!!! What??

As far as I know, chromebook means portability. Specific routers, dump our
chromebooks... Sounds funny and tragic.

Please Google, there is no reason to give up on a platform strictly under
your control. Hardware, firmware, O.S. and Browser are totally under your
discretion, by design.

We the chromebook users have been tolerant and supportive about this
glitch, up to a point. I am getting disappointed due to the impossibility
to take my self declared "trouble free" chromebook anywhere and perform
simple tasks like browsing.

What I would really expect from google was some respect to the clients.
Whatever the issue is, please make it work or replace it, whatever.

Regards,

chro...@googlecode.com

unread,
Apr 21, 2015, 7:06:43 PM4/21/15
to chromi...@chromium.org

Comment #221 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
ch...@chromium.org in #193 you told us you were going to moderate this
forum:

"we will be removing comments that do not directly pertain to engineering:
please take discussions to the forums. Thank you."

The time for moderation is now.

chro...@googlecode.com

unread,
Apr 21, 2015, 10:51:44 PM4/21/15
to chromi...@chromium.org

Comment #223 on issue 439818 by FloridaG...@gmail.com: Samsung 303:
#218 I apologize, I am on the road for work and do not have ready access to
this right now, but will follow up as soon as possible (next day or two)
with the information you have requested. Thank you, again, for your ongoing
communications and efforts with this matter.

chro...@googlecode.com

unread,
Apr 23, 2015, 6:27:15 AM4/23/15
to chromi...@chromium.org

Comment #224 on issue 439818 by lesgetsc...@gmail.com: Samsung 303:
Further to #207 After a couple of days its dropping out again. About
twice per day, two dropouts together then all OK for hours.

Version 42.0.2311.87 beta
Platform 6812.75.2 (Official Build) beta-channel daisy_spring
Firmware Google_Spring.3824.129.0

chro...@googlecode.com

unread,
Apr 23, 2015, 11:55:16 AM4/23/15
to chromi...@chromium.org

Comment #225 on issue 439818 by jpala...@gmail.com: Samsung 303: frequent
#216 as well as #209... These issues do affect other chromebooks but some
of the updates youve been pushing out for this issue seem to be making it
less of an issue for them (seems to be occur more when streaming. used to
get it constantly when using spotify or google play music) The issue isn't
as prevalent, esp with chromebooks with a little higher specs
(cpu/memory).. but it does occur.. Iv reported similar drop issues and not
being able to see ap's without a reboot with a Dell Chromebook 11.. it just
isn't as frequent or severe.. (esp after first round of updates for this
issue)... Iv been refraining from commenting since its more of an issue
with the Samsungs.. its still here but just not happening as often..

chro...@googlecode.com

unread,
Apr 23, 2015, 7:54:27 PM4/23/15
to chromi...@chromium.org

Comment #226 on issue 439818 by edtay...@citcom.net: Samsung 303: frequent
Version 42.0.2311.87
Platform 6812.75.2 (Official Build) stable-channel daisy
Firmware Google_Snow.2695.117.0

Samsung not working well at home so I took it to Starbucks where it
acquired the network. I closed the lid , did some shopping then sat down to
try some surfing. It had dropped WiFi while sleeping and would not
re-aquire. I performed a hard reset and it still would not acquire so I
gave up and came home. I had to do another reset to acquire my Belkin and
it is working for the moment though it erased my report once and I had to
start over. I am hopeful that Google will do the right thing and provide me
a working Chromebook. If and when this mess is ever resolved Google can
reposess if it wishes but I just don't feel like suffering through this
problem

chro...@googlecode.com

unread,
Apr 23, 2015, 10:36:38 PM4/23/15
to chromi...@chromium.org

Comment #227 on issue 439818 by edtay...@citcom.net: Samsung 303: frequent
dropping of wifi
https://code.google.com/p/chromium/issues/detail?id=439818

Version 42.0.2311.87
Platform 6812.75.2 (Official Build) stable-channel daisy
Firmware Google_Snow.2695.117.

Samsung 303 Belkin router

I tried to switch from pages of thumbnail images to view all and can't get
there, various errors and wifi dropouts.

chro...@googlecode.com

unread,
Apr 26, 2015, 4:14:26 AM4/26/15
to chromi...@chromium.org

Comment #228 on issue 439818 by p.voorth...@gmail.com: Samsung 303:
Version 43.0.2357.32 beta
Platform 6946.20.0 (Official Build) beta-channel daisy
Firmware Google_Snow.2695.117.0

I'm also having random dropping connections, these are pretty
annoying.These disconnect are caused by resets of the mmc wifi network
controller. A request to the network controller firmware times out and
after a while the entire controller is reset and the connection is
re-established.

I think this issue is related to power management in combination with the
firmware of the network controller. It occurs during runtime use but also
during wake-up (opening lid). I'll attach some kernel logs.

Attachments:
runtime_disconnect.txt 5.2 KB
runtime_disconnect2.txt 6.6 KB
startup_disconnect.txt 5.7 KB

chro...@googlecode.com

unread,
Apr 26, 2015, 9:42:54 AM4/26/15
to chromi...@chromium.org

Comment #229 on issue 439818 by rlag...@gmail.com: Samsung 303: frequent
Dear debuggers:

I use my xe303c12-ad1br to watch videos and listen to music with a Logitech
bluetooth audio receiver (the one with rounded corners). I too have
suffered from the #wifi40 crashes, and would like to add that my bluetooth
audio always crashes at the same time as the wifi, when both are active.

Until version 41, the audio would reconnect and resume playing at about the
same time as the wifi came back, BUT on versions 42 and 43 beta, the audio
goes into a limbo and won't come back unless I reboot. This also halts any
tab that tries to play audio, preventing me from watching whatever video I
was, or opening new ones.

Please consider this information.

chro...@googlecode.com

unread,
Apr 26, 2015, 11:57:35 AM4/26/15
to chromi...@chromium.org

Comment #230 on issue 439818 by OreoR...@gmail.com: Samsung 303: frequent
This last Chromebook update fixed whatever my dis-connectivity issue was.

chro...@googlecode.com

unread,
Apr 27, 2015, 8:16:10 PM4/27/15
to chromi...@chromium.org

Comment #233 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
We have gotten a further analysis in order and we believe we have traced
the root cause of the issue to a spinlock being held by kswapd.

kswapd is holding a spinlock for around 10 seconds which is causing loads
of subsystems to fall behind. We have now replicated this issue using
ethernet as well. High memory contention causes this problem to occur more
as kswapd will attempt to swap to disk/flash and run into this problem.

Our previous fix attempt for this problem was to recheck for interrupts
(you can think of this as a "service request) which essentially piled up
behind kswapd when this occurred and process them: at that point we had not
yet understood what was causing requests to pile up.

We are now bringing in kernel experts to investigate the problem in the
swap subsystem. We appreciate the feedback from contributors which assisted
in debugging this issue thus far.

We will continue updating this bug when more information is available.

chro...@googlecode.com

unread,
Apr 28, 2015, 4:18:08 AM4/28/15
to chromi...@chromium.org

Comment #234 on issue 439818 by julien.puydt: Samsung 303: frequent
I had wifi this night, closed the lid ; after I opened it again this
morning, it had dropped and come back a few times, until now I finally need
to use my phone and a usb cable as no wifi network is detected anymore --
thinking back to other occurrences, I'm wondering if suspending wouldn't in
fact "help" trigger the problem. I'm not 100% convinced but still wanted to
share.

In fact, I'm not convinced by the kswapd idea either, since the network is
the only thing failing, and we should also notice other things go sour if
it were kswapd, wouldn't we ?

Still, another clue and another lead to follow is nice... Thanks for the
feedback.

chro...@googlecode.com

unread,
Apr 28, 2015, 5:39:36 PM4/28/15
to chromi...@chromium.org

Comment #237 on issue 439818 by allatse...@gmail.com: Samsung 303: frequent
Not sure what changes you've made, but by far the worst day today for a
long time. Constantly dropping, takes a few minutes when page freezes
before cross appears with err messages as above. Would not re connect
without powering down just now.

Version 42.0.2311.87
Platform 6812.75.2 (Official Build) stable-channel daisy
Firmware Google_Snow.2695.132.0

chro...@googlecode.com

unread,
Apr 29, 2015, 4:10:43 PM4/29/15
to chromi...@chromium.org

Comment #240 on issue 439818 by klfoxwor...@gmail.com: Samsung 303:
I'm not savvy as to figuring out what is going on with my Samsung CB but
whatever was done with the update I did on April 28 2015 is a disaster! I
get unable to use Google App msgs while it shows I am online. Also my
cursor is freezing at random times, pages are freezing and it's taking a
full minute to reboot.This device is useless. Who is responsible for this
epic fail? Samsung or Google?

chro...@googlecode.com

unread,
Apr 29, 2015, 4:19:56 PM4/29/15
to chromi...@chromium.org

Comment #241 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
Preliminary testing indicates #wifi40 is FIXED as of:

Google Chrome 42.0.2311.134 (Official Build) (32-bit)
Revision 2917a27a2af596a58a4c21d3527bcb2c02426d5c
Platform 6812.83.0 (Official Build) stable-channel daisy
Firmware Google_Snow.2695.132.0

chrome://system/ - netlog shows no errors under normal use. I generated a
few [ERROR:http_request.cc(176)] by forcing memory contention, but the
errors appear to be seamlessly and transparently resolved by the
[INFO:connection_tester.cc(65)] ConnectivityTester.

Unless advised otherwise I'll continue to test stable for a couple days
then progress to beta and dev to help generate more feedback reports for
the kswapd component of this issue.

It is my understanding that because Chromebooks use SSD kswapd is always
caching and especially active during memory contention and wake/sleep
(open/close). Unless someone suggests otherwise I will focus my testing on
those conditions.

If the "wifi drop" symptom of this issue is fixed, perhaps the kswapd
should be isolated in a separate issue for the few kernel diehards who are
still interested.

chro...@googlecode.com

unread,
Apr 30, 2015, 1:35:41 AM4/30/15
to chromi...@chromium.org

Comment #248 on issue 439818 by julien.puydt: Samsung 303: frequent
I'm running version 43.0.2357.32 beta, the connection problem is still here
-- I haven't quite noticed any improvement. On the contrary since the last
upgrade, things got *worse* : sometimes the chromebook mostly freezes : the
mouse cursor is still moving, but nothing else is.

In fact, I was typing the above paragraph when it happened again : I had to
reboot. And when I started anew, I immediately lost the wifi connection --
so the memory contention theory seems dubious, given the box was just
restarted -- I only have two tabs : a crosh one with nothing, and this one!

chro...@googlecode.com

unread,
Apr 30, 2015, 1:33:10 PM4/30/15
to chromi...@chromium.org

Comment #250 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
The slowing that people are experiencing is symptomatic of the hang we are
seeing in kswapd. If you are seeing slowing before a hang or connection
drop, we are actively debugging this problem. We will contribute additional
news when available.

chro...@googlecode.com

unread,
May 1, 2015, 4:25:04 AM5/1/15
to chromi...@chromium.org

Comment #253 on issue 439818 by julien.puydt: Samsung 303: frequent
I have seen the problems occur within a single minute after a fresh start
with very few opened tabs (see my comment #248) : it's not a leak, it's a
waterfall!

chro...@googlecode.com

unread,
May 1, 2015, 2:31:32 PM5/1/15
to chromi...@chromium.org

Comment #254 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
I've produced a couple feedback reports where immediately after reboot the
first URL I type in causes the wifi cone to go from 2-bars to 4-bars,
stalls, then goes #cone_x, then reconnects after a few seconds.
Unfortunately, when I try to repeat the steps I cannot reproduce the
behavior. I'm still very happy with how quickly version 42.0.2311.134
recovers the network, usually less than 20 seconds.

chro...@googlecode.com

unread,
May 1, 2015, 11:10:46 PM5/1/15
to chromi...@chromium.org

Comment #255 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
WARNING: Samsung 303 is not stable on Beta channel. I had wifi disabled
immediately after power wash and all four of my Samsung 303 Chromebooks
crashed continuously on v43. I suspect the way the patches for #wifi40
interact with v43 graphics causes the frequent system freezing. First the
screen freezes and only the mouse cursor is responsive but after enough
mouse and keyboard events the screen goes black. I submitted plenty of
feedback reports with #sod (screen of death) and #bsod (black screen of
death) before I suspected the freezes were isolated to Samsung 303
Chromebooks and I should wait for #wifi40 to resolve before further
testing.

All the chromebooks crashed repeatedly while trying to upgrade to Dev
Channel, so I would recommend all Samsung 303 users hang-tight on Stable
Channel for now.

chro...@googlecode.com

unread,
May 1, 2015, 11:11:46 PM5/1/15
to chromi...@chromium.org

Comment #256 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent

chro...@googlecode.com

unread,
May 4, 2015, 1:24:42 PM5/4/15
to chromi...@chromium.org

Comment #257 on issue 439818 by aashuto...@chromium.org: Samsung 303:
Issue 312719 has been merged into this issue.

chro...@googlecode.com

unread,
May 4, 2015, 2:28:07 PM5/4/15
to chromi...@chromium.org

Comment #260 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
Flashing an older version of ChromeOS is possible but please consult the
product forums. This bug thread is generally used for reports on activity
and collecting technical information and isn't meant as a general
discussion forum. Please understand that we are trying to keep this thread
as concise as reasonable.

chro...@googlecode.com

unread,
May 4, 2015, 5:44:40 PM5/4/15
to chromi...@chromium.org
Updates:
Cc: albe...@chromium.org

Comment #263 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
(No comment was entered for this change.)

chro...@googlecode.com

unread,
May 6, 2015, 3:40:51 PM5/6/15
to chromi...@chromium.org

Comment #265 on issue 439818 by cbuck...@gmail.com: Samsung 303: frequent
Similarly, after I replaced my ASUS RT-N13U Rev. B1 with the Archer C7 in
late April, the connectivity problem disappeared. That is what I expected
would happen, but it's not the primary reason I got the new router.

I've been using the Dev channel during this time.

chro...@googlecode.com

unread,
May 7, 2015, 3:40:45 PM5/7/15
to chromi...@chromium.org

Comment #267 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
This bug is extraordinarily complicated to reproduce and involves subtle
interactions between the Samsung 303 chipset and each individual user's
network.

If you want to help fix this problem you need to submit feedback reports by
pressing:

alt-shift-i and typing "#wifi40" in the description.

At first submit a report each time you lose your connection, then submit
reports whenever it's convenient. Those reports are the only way the
Chrome team can view the problems you're having and the only way those
problems can get fixed.

When you post on this forum be sure to include the version of the system
you're using by cut and pasting the first three lines from the
web-page: "chrome://version"

chro...@googlecode.com

unread,
May 8, 2015, 11:46:09 AM5/8/15
to chromi...@chromium.org

Comment #268 on issue 439818 by p...@equalitycharterschool.org: Samsung
We had the same problem on Samsung E303C12-A01US Chromebooks running
version 42.

We have a Ruckus wireless network.

We turned off 5GHz on our APs to force the Chromebooks to 2.4GHz and have
been ok for 2 days.

chro...@googlecode.com

unread,
May 8, 2015, 5:15:55 PM5/8/15
to chromi...@chromium.org

Comment #270 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
We are making more progress debugging the swap issue. At this time we are
investigating a condition occurring between our memory compression system
and our swap. ChromeOS uses zram which is a memory compression system to
improve utilization of available system ram. When the system is low on
available memory our swap system attempts to find a free section of memory
and move things around.

We are encountering a condition which is exacerbated by video playback or
other programs which result in difficult to compress memory chunks.
Essentially when low on ram and trying to free up a contiguous block the
system tries to compress pieces of ram and fails to compress it much. When
this happens it tries to find another block and compress that as well. When
playing back video it turns out video is already very compressed and will
not compress well further, which means that the memory compression
subsystem is spending a lot of time trying to find available memory.

This results in the hangs that you are seeing prior to Wi-Fi dropping.
Wi-Fi drops because the Wi-Fi driver was not able to operate while the
system was busy trying to find free memory, and the Wi-Fi card is not given
enough system time to maintain a connection.

You can tell if you are reaching this problem because it will also happen
on ethernet. Our engineers are continuing to try to craft a solution for
this issue. Some improvements to the situation were incorporated in M44 as
mitigation measures but we are still working to address the root problem.

chro...@googlecode.com

unread,
May 9, 2015, 2:20:09 AM5/9/15
to chromi...@chromium.org

Comment #271 on issue 439818 by julien.puydt: Samsung 303: frequent
I just restarted my chromebook ; the only opened tabs in chrome where a
crosh, a doodle, and gmail. I lost connection within a minute. No high
memory pressure, no streamed video. I sent a report with a #wifi40 tag.

I'm running 43.0.2357.46 (Build officiel) beta (32 bits) revision
e9c2697a07ce72635910285c5addb2e3d9fa2e3b.

That makes me think not all dropping wifi is explained by comment #270 : I
generally see wifi drop either very fast after starting/resuming or much
longer after (only those latter could be memory-related) ; I don't have the
impression there is much in-between.

Is it possible to write a script which will go through the logs and extract
data about the durations between shutdown, startups, sleep, resume, wifi
up/lost ?

chro...@googlecode.com

unread,
May 10, 2015, 6:51:41 PM5/10/15
to chromi...@chromium.org

Comment #272 on issue 439818 by darren.m...@gmail.com: Samsung 303:
After reading through this chain, I am glad I'm not the only one having
issues over the last few weeks...

To be specific:
- I have NOT seen the system-tray network icon drop.
- I have experienced plenty of dropouts where my 303C was unable to
completely load pages, often with the "circle" icon continuously circling.
At first, I thought it may be a DNS issue, but it could also simply be
dropping packets or missing interrupts.

Anyway, I'm on a XE303C. I was running 2311 stable, but then switched to
43.0.2357.46 beta (today) to see if the problem went away. Unfortunately,
it didn't seem to make a difference.

Platform 6946.31.0 (Official Build) beta-channel daisy_freon
Firmware Google_Snow.2695.117.0

My router is a DD-WRT flashed Netgear N300.

My Win7 laptop doesn't seem to have any issues. The chromebook is
near-worthless until fixed.

I've power-washed and still see the issues after reboot, even with just
Gmail open.

chro...@googlecode.com

unread,
May 10, 2015, 7:22:50 PM5/10/15
to chromi...@chromium.org

Comment #273 on issue 439818 by bobtuc...@gmail.com: Samsung 303: frequent
As a Samsung 303 user that had months of problems I would like to just add
my experience with the wifi#40 issue.
Over the last months I have changed my router to a

chro...@googlecode.com

unread,
May 10, 2015, 7:43:04 PM5/10/15
to chromi...@chromium.org

Comment #274 on issue 439818 by bobtuc...@gmail.com: Samsung 303: frequent
dropping of wifi
https://code.google.com/p/chromium/issues/detail?id=439818

As a Samsung 303 user that had months of problems I would like to just add
my experience with the wifi#40 issue.
Over the last months I have changed my router to a Dlink 889L and continued
to have trouble with the 5ghz band and switched back to 2.4. I am currently
at ver:
Google Chrome 42.0.2311.87 (Official Build) (32-bit)
Revision b8296729c0097ac3aba988052b6855a274607ec7
Platform 6812.75.2 (Official Build) stable-channel daisy

I am back running perfectly w/o any problems and have been for awhile. I
believe I have 18 tabs open at the moment. I am looking at an update arrow
on the bottom right and refuse to do another update after leaving #39. I
can do the sh/alt i report if it would be of interest to engineers. Good
luck everyone.

chro...@googlecode.com

unread,
May 10, 2015, 7:46:05 PM5/10/15
to chromi...@chromium.org

Comment #275 on issue 439818 by bobtuc...@gmail.com: Samsung 303: frequent
dropping of wifi
https://code.google.com/p/chromium/issues/detail?id=439818

As a Samsung 303 user that had months of problems I would like to just add
my experience with the wifi#40 issue.
Over the last months I have changed my router to a Dlink 880L and continued

chro...@googlecode.com

unread,
May 11, 2015, 11:06:01 AM5/11/15
to chromi...@chromium.org

Comment #276 on issue 439818 by julien.puydt: Samsung 303: frequent
I sent two feedbacks today with a #wifi40 tag. I must say I'm pondering
just dumping my cb to the trash : I spend more time playing with cables to
get some data than actually using it!

chro...@googlecode.com

unread,
May 11, 2015, 11:31:01 AM5/11/15
to chromi...@chromium.org

Comment #277 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
julien.puydt: Don't give up. They're getting closer to isolating the
source of this behavior. Of all the posts on this forum yours have best
described the same problems I'm having. Now I have the distinct advantage
of having one network where my Chromebooks loses connectivity persistently
and one network where this issue is scantly detectable. Once the
kswapd-zram issue (detailed in #270) is fixed we'll be better able to
isolate the root of the wifi dropping. If you're tenacious enough to
continue submitting feedback logs and testing your Chromebook on other
networks I predict you will look back at this later and be very glad you
stuck with the debugging process. Hint: you might end up fixing a problem
outside your Chromebook as well.

chro...@googlecode.com

unread,
May 12, 2015, 10:00:17 PM5/12/15
to chromi...@chromium.org

Comment #278 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
My four Samsung 303's updated today to:

Google Chrome 42.0.2311.153 (Official Build) (32-bit)
Revision 539eb2520880a611cf331fafdc5cab5d23ece4ef
Platform 6812.88.0 (Official Build) stable-channel daisy

Using usb3-to-ethernet Linksys usb3gig adapters I tested two ethernet
switches:

Cisco SF100-24 24 port 10/100 Switch Unmanaged:
Didn't witness any connectivity losses but a few flash plugin crashes on
Snow Rev4 (after a power wash those seemed to stop).

Cisco SF100D-05 5 port 10/100 Switch Unmanaged:
immediately began losing connectivity with wifi-cone-x icon popping up
despite the wifi being turned off and the ethernet being connected.

I submitted feedback logs with "#wifi40" for all four Chromebooks.

Why would the newer 24 port switch be immune to #wifi40 and the older 5
port switch not?

chro...@googlecode.com

unread,
May 13, 2015, 7:13:37 AM5/13/15
to chromi...@chromium.org

Comment #279 on issue 439818 by 2452s...@gmail.com: Samsung 303: frequent
Version 42.0.2311.134
Platform 6812.83 (Official Build) stable-channel daisy
Firmware Google_Snow.2695.117.0

I have two XE303C12 CB's. Both manufactured Jan 2013. Both on same OS
version and using a BT Home hub 5.
One CB works with no problems, the other suffers from the wifi drop problem.

This has been an ongoing problem for months now. ChromeOS has basically
bricked one of my CB's. I know that you are working on a fix....but 4
months, come on guys. If Apple or MS had bricked everyone's PC's with an
update, there would have been a fix by now.
Even if you just gave us the ability to roll back to an OS Version previous
to the wifi drops until the problem is solved, surely it can't be that hard
to post an R39 Recovery somewhere to allow us to roll back. Not happy.....

chro...@googlecode.com

unread,
May 13, 2015, 9:27:38 AM5/13/15
to chromi...@chromium.org

Comment #280 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
Frustration is increasing again. The activity on this forum gives the
impression progress is stalling. I can now recite this #wifi40 issue id
number ("439818") as easily as my childhood phone number. I purchased four
Samsung 303's six months ago so that I could demonstrate how simple Web
Store Apps could be developed on off-the-shelf ARM Chromebooks. I planned
to launch before Google I/O 2015. That's two weeks away now and I still
don't have stable Chromebooks!

Not to theory-craft, but if this issue is triggered by something like a
Stuxnet-style-worm in networking firmware then Chrome OS needs to "set
thine house in order" so the true culprit can be identified. As of now,
Chrome OS simply drops it's connection without explanation.

Example: using usb3-to-ethernet adapter specific unmanaged switches with
specific network usage triggers a connectivity loss and Chrome OS just
shows "cone x" even when wifi is disabled! Users observing this behavior
assume Chrome OS dropped it's connection.

Example: using corrupted firmware on a wireless router with specific
network usage triggers a connectivity loss and Chrome OS stalls, shows full
bars, then no bars, then stops scanning for networks. Users observing this
behavior assume discounted Chromebooks are generally unreliable.

If specific network behavior is triggering the connection loss then Chrome
OS should identify it as a network error. For example: "Access Point blah
violated protocol blah and was disconnected by Chrome OS to protect your
Chromebook from blah."

I've exhausted my spare time testing this. Preliminary tests of Stable
Build 42.0.2311.153 connected to Cisco SF100-24 via Linksys USB3GIG show
good stability. I'm restricting myself to that configuration until someone
on this forum can inspire some hope for change.

chro...@googlecode.com

unread,
May 13, 2015, 9:43:36 AM5/13/15
to chromi...@chromium.org

Comment #281 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
I call shenanigans on Comment #279. "2452s...@gmail.com" doesn't have any
other posts and doesn't participate in any projects. If he hasn't
submitted any feedback reports he should be moderated. "439818" has become
a festering sore on the open source community and the vultures are
circling. Why are people who don't participate in code.google.com projects
allowed to post here?

I've been analysing this bug for six months now and there are ZERO reports
of two identical Chromebooks on the same network behaving differently. The
problem is between the network and the Samsung 303 Chipset. (Since no one
fixed the kswapd-vram bug yet it's harder to prove that but I'm convinced).

"2452s" is a proponent of closed source netbooks and should be dismissed.
No one in their right mind would want to rollback to R39. Someone just
needs to explain to us how we get this fixed by R44.

chro...@googlecode.com

unread,
May 13, 2015, 10:22:38 AM5/13/15
to chromi...@chromium.org

Comment #282 on issue 439818 by dsm...@fortosage.net: Samsung 303: frequent
We are a school district with 3700 chromebooks deployed right now, all the
same model of Samsung XE303c12 but with different manuf dates and we have
about 10 on our bench right now that have this issue. Why a few and not
all? I was actually led here a few months back about another known issue
and just happened across this post, so I have not shared reports, because
we cannot get back on the network with those devices to do so. If there is
a back door to do this, please share and we will post. Short of buying a
new motherboard and replacing it (that does seem to fix the issue, at least
for 2 of ours) we would prefer to find a more feasible solution. I can
understand folks being frustrated, as are we, that when we called into both
Google and Samsung, both are pointing their fingers at each other. Sorry
to say to #281, we do have identical chromebooks, on the same network
behaving differently, we just have not had or were aware of this known
issue until Monday.

chro...@googlecode.com

unread,
May 13, 2015, 10:48:32 AM5/13/15
to chromi...@chromium.org

Comment #283 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
#282 There are numerous parts to this issue. If a Chromebook is unable to
stay online long enough to submit a feedback report then it is likely a
hardware defect. If less than %1 of your Chromebooks are unable to
connect, and when you replace the motherboards they work again, then you
are posting on the wrong issue. Samsung replaced the motherboard in one of
my Chromebooks for free and now it behaves exactly like the other three -
specific networking firmware triggers intermittent dropping of all network
connectivity.

A 00.3% defect rate for silicon boards in plastic cases used by public
school children is a pretty good showing for Samsung. I'm eager to let
Samsung off the hook once we're able to identify the network problems that
are still triggering this.

This is another example of why Chrome OS needs to put it's house in order.
When users don't get clear error messages then the community at large sees
one thing "Samsung 303 drops Wifi" and everything gets blamed on Chromium.

The vultures are circling.

chro...@googlecode.com

unread,
May 13, 2015, 12:12:46 PM5/13/15
to chromi...@chromium.org

Comment #284 on issue 439818 by dsm...@fortosage.net: Samsung 303: frequent
We have a much higher defect rate for other stuff on those chrome books,
actually at 20% right now so we are moving to another brand. Just to
clarify, you cannot get on the network if the wifi is not working...so that
is this issue. We just replaced the motherboard internally to see if it
would fix it, but safe to say once it all updates, and it is doing that
now, it may not reconnect.

Like #283, we have had Samsung replace motherboards OOW when there was a
known issue with another update that corrupted it, but this one no one is
taking responsibility for. I feel fortunate in that we only started seeing
this in the last 4 weeks, looks like others have experienced it much longer.

chro...@googlecode.com

unread,
May 13, 2015, 1:52:16 PM5/13/15
to chromi...@chromium.org

Comment #287 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
#284 dsm...@fortosage.net "We have a much higher defect rate for other
stuff on those chrome books, actually at 20% right now so we are moving to
another brand."

Okay so maybe Samsung 303 is unreliable compared to other chipsets. You
can view a system log offline by going to the URL "chrome://system". If
you click "expand all" then you can save the page as "About System.mhtml".
The logs are truncated but at least you can save them to a thumbdrive and
attach them to an alt-shift-i feedback report from a Chromebook that will
connect. A more detailed log of network connectivity can be exported
from "chrome://net-internals".

This forum needs to focus on differentiating which connectivity drops are
caused by bad routers and which drops are caused by Chromium OS. Until the
swap-zram issue is fixed I suggest everyone (with drop behavior specific to
#wifi40) try to put their router in rescue mode and manually upload the
latest firmware while connected directly to the router offline.

chro...@googlecode.com

unread,
May 13, 2015, 1:55:17 PM5/13/15
to chromi...@chromium.org
Updates:
Labels: Restrict-AddIssueComment-EditIssue

Comment #288 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
We are temporarily restricting comments as we sort out the zram issue. We
are continuing to analyze logs tagged as #wifi40 . We will resume comments
as we have a solution. This bug is CCed to over 100 people so we are trying
to reduce email churn.

chro...@googlecode.com

unread,
May 13, 2015, 1:58:19 PM5/13/15
to chromi...@chromium.org

Comment #289 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
We recommend the community forum for continued discussion while this bug
tracker is in restricted mode-- we are attempting to lower SNR for anyone
wanting to follow just engineering commentary.

chro...@googlecode.com

unread,
May 28, 2015, 8:51:17 AM5/28/15
to chromi...@chromium.org

Comment #292 on issue 439818 by cereb...@gmail.com: Samsung 303: frequent
Is the fix across all daisy derivative boards, esp. daisy_spring?

-- T

chro...@googlecode.com

unread,
May 28, 2015, 2:07:22 PM5/28/15
to chromi...@chromium.org

Comment #293 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
It should be deployed across boards, yes. This fix should reduce the system
hang and restart reports we have been getting.

chro...@googlecode.com

unread,
May 29, 2015, 8:59:51 PM5/29/15
to chromi...@chromium.org

Comment #296 on issue 439818 by molly....@gmail.com: Samsung 303: frequent
I am having same issue:

Have a WPA network with and 2.4GHz and 5GHz networks; DIR 857 D-Link router
randomly chromecast will disconnect from wifi
super frustrating

chro...@googlecode.com

unread,
May 29, 2015, 9:09:52 PM5/29/15
to chromi...@chromium.org

Comment #297 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
Version 44.0.2403.19 dev

I'm having so many Flash crashes, so many Chromecast disconnects, and so
many problems when waking from sleep, I'm going to try the following to
better isolate testing for #wifi40:

1. disabling Flash on chrome://plugins
2. uninstall Chromecast extension
3. close all windows before closing the lid or letting Chromebook fall
asleep

chro...@googlecode.com

unread,
May 30, 2015, 12:46:14 AM5/30/15
to chromi...@chromium.org

Comment #298 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
Updated just now to:

Version 44.0.2403.25 dev
Platform 7077.20.0 (Official Build) dev-channel daisy_freon
Firmware Google_Snow.2695.132.0

I tested four Samsung 303 Chromebooks on 3 switches and 2 wifi routers.

sf100-24, sg100d-05, and e1200 all work perfectly.

sf100d-05 drops usb3gig connectivity occasionally, probably because it's a
cheap old desktop switch, but Chrome OS gives no error except a momentary
#cone_x icon, so users will reasonably assume Chrome OS has unreliable
network connectivity and blame Chromium.

Linksys ea4500 was the worst. I suspect it's gone rogue. Other users with
iOS and Win8 are using the ea4500 wifi without complaints but Samsung 303's
connectivity is awful and under network congestion breaks everything from
GMail to this forum. The latest update shows a cone with the same number
of bars consistently even when the connection is unusable.

Other details I won't mention here suggest this problem involves hacked
routers probing for Chromebook vulnerabilities.

chro...@googlecode.com

unread,
May 30, 2015, 8:40:00 AM5/30/15
to chromi...@chromium.org

Comment #299 on issue 439818 by bill.wil...@gmail.com: Samsung 303:
So, this issue has been going on for the last year through many versions of
Chrome. I have not seen any suggestions or inputs from Samsung/ Chrome
yet. Mine has been happening since I bought my Chrome, same symptoms as
the hundred others on this thread. I have a Netgear router, but by now it
is obviously not a router problem since it happens on every version of
router while other machines have no issues. Has anyone seen either Samsung
or Chrome even acknowledge the issue and that they are working on it?

chro...@googlecode.com

unread,
May 30, 2015, 10:36:10 AM5/30/15
to chromi...@chromium.org

Comment #300 on issue 439818 by p.voorth...@gmail.com: Samsung 303:
The problem is the following:

mwifiex_sdio mmc2:0001:1: Firmware wakeup failed

During idle connections, the WIFI firmware falls asleep|is put to sleep.
When you start requeusting new data, the firmware needs to wake up. In
my/our case it does not wake up and the WIFI chipset is reset (last
resort). The result is that the entire connection needs to be
re-established.

This issue is still present in 43.0.2357.81 beta, I've posted multiple
#wifi40 's. I have kernel and driver experience. Let me know if I can
assist in solving/reproducing the problem next to sending crash-log
information.

chro...@googlecode.com

unread,
May 30, 2015, 12:09:04 PM5/30/15
to chromi...@chromium.org

Comment #301 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
p.voorth...@gmail.com The new fix is in M44 currently on the Dev Channel
(ref #291). Beta Channel (M43) doesn't include the fix yet. If you have
experience with kernels read #270 for a specific explanation. I had the
same problem as you and it is fixed in R44.

chro...@googlecode.com

unread,
May 30, 2015, 12:26:44 PM5/30/15
to chromi...@chromium.org

Comment #302 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
Cheap wifi routers configured from infected Win8 machines is how this
problem began for me. The annoying thing is that the rogue routers only
cause obvious connectivity problems for older ARM Chromebooks (a result of
old ARM SoC issues I guess). I had multiple wifi routers that were causing
Samsung 303 connectivity issues. After taking each offline (put into
recovery mode) and manually uploading the latest firmware from an offline
Chromebook, all of them are fixed except for the one router that is also
administered from a Win8 box. Samsung hardware is working fine. I had one
Samsung Chromebook that would reset with this issue and Samsung replaced
the motherboard for free in a period of just a few days. The Samsung
Chromebook 2 (with a slightly newer ARM SoC) works perfectly even on the
rogue routers. Other Chrome OS bugs (zram) complicated isolating this
issue on Samsung 303 but I expect those bugs will be fixed by R44. The
real remaining problem is that Chrome OS needs much more precise error
notification so users know where the problem is coming from... otherwise
end users naturally assume Samsung and Chromium are the root of the problem
(which I assumed for months as well).

chro...@googlecode.com

unread,
May 31, 2015, 5:16:52 AM5/31/15
to chromi...@chromium.org

Comment #303 on issue 439818 by p.voorth...@gmail.com: Samsung 303:
Thanks John, I missed that one. Just moved to the dev channel and
unfortunatly it also disconnects -> See my last #wifi40. My router is a
zyxel P-2812HNU-F1 VDSL modem.

Can somebody explain to me what the exact issue is and the relation with
zram? Zram eventually also makes use of SDIO (EMMC) and so is the Marvell
SDIO wifi chip. However these are seperate busses and even controllers. So,
why is the wifi not waking up? In my (current) opinion this is a
power-management / wifi firmware issue.

chro...@googlecode.com

unread,
Jun 1, 2015, 6:11:26 AM6/1/15
to chromi...@chromium.org

Comment #304 on issue 439818 by bill.wil...@gmail.com: Samsung 303:
Thanks P.voorth - you sound like you have troubleshot it down to at least
proximate cause. This has been going on since long before I owned this
Chromebook, but I have yet to see anything from Samsung or Google that is
going to address it, have you or anyone else?

chro...@googlecode.com

unread,
Jun 1, 2015, 12:22:54 PM6/1/15
to chromi...@chromium.org

Comment #305 on issue 439818 by julien.puydt: Samsung 303: frequent
Comment #304 -- did you notice you're discussing this on a thread where
google is asking for feedback about various fix attempts ? There is no
question they're trying to do something.

I'm much more annoyed by the fact that they don't manage to get anything
really done : the usb port on my cb is starting to show signs of wear.
USB-tethering to my phone since *months* is taking its toll...

chro...@googlecode.com

unread,
Jun 1, 2015, 12:34:53 PM6/1/15
to chromi...@chromium.org

Comment #306 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
bill.wil...@gmail.com: Bill you don't seem very interested in fixing this
problem, but instead you seem very interested in "Samsung and Google
addressing" this issue. I've been dealing with this issue for months now.
Samsung has responsibly addressed this issue as it related to the hardware
I purchased from them (honored their warranty with expediency). Chromium
(an open source software consortium that includes Google) has responsibly
addressed this issue as well. It may appear that because there are still
connectivity issues that "someone isn't fixing it," but quite the opposite
is true. A variety of connectivity issues were being caused by a
confluence of networking events and Chromium has been wildly successful at
sorting them all out. If you've been following this forum then you know
most of the symptomatic Chromebooks are working now and that R44 contains
fixes for many of the remaining (more subtle) problems.

You may have also noticed most users on this forum were able to fix their
problems by either recovering or securing their WiFi AP. Just because your
Samsung 303 Chromebook is your only symptomatic WiFi device doesn't mean
your WiFi isn't messing with your other devices just the same.

So yes, I've seen lots of things from Samsung and Google that address
this. In particular, I've seen Chromium go above and beyond the call of
duty by addressing issues for people who don't take responsibility for
their own WiFi routers.

chro...@googlecode.com

unread,
Jun 1, 2015, 6:04:44 PM6/1/15
to chromi...@chromium.org

Comment #309 on issue 439818 by tushar.h...@gmail.com: Samsung 303:
I have an asus rt-ac52u. Still getting the occasional dropped connection
with rev44 - but it is nowhere near as severe as it was a couple of months
ago

Version 44.0.2403.25 dev
Platform 7077.20.0 (Official Build) dev-channel daisy_freon
Firmware Google_Snow.2695.117.0

chro...@googlecode.com

unread,
Jun 1, 2015, 11:02:16 PM6/1/15
to chromi...@chromium.org

Comment #310 on issue 439818 by bill.wil...@gmail.com: Samsung 303:
#308 - Fairly common one: Netgear Router:

Hardware Version WNDR3400v2
Firmware Version V1.0.0.38_1.0.61
GUI Language Version V1.0.0.36_2.1.28.1

Chrome version: 42.0.2311.153

chro...@googlecode.com

unread,
Jun 5, 2015, 12:26:42 AM6/5/15
to chromi...@chromium.org

Comment #311 on issue 439818 by edtay...@citcom.net: Samsung 303: frequent
Samsung 303C and Belkin router

Version 43.0.2357.81
Platform 6946.55.0 (Official Build) stable-channel daisy
Firmware Google_Snow.2695.117.0

Not under test conditions by any means but it seems that drop-outs are
fewer and automatic recovery is reasonably fast since recent update to
v43. The best news is that after recovery pages load normally without
necessity of a hard reset.

chro...@googlecode.com

unread,
Jun 5, 2015, 2:04:53 AM6/5/15
to chromi...@chromium.org

Comment #312 on issue 439818 by p.voorth...@gmail.com: Samsung 303:
Hi,

This is what chrome says:
Version 44.0.2403.25 dev
Platform 7077.20.0 (Official Build) dev-channel daisy_freon
Firmware Google_Snow.2695.117.0

Its a XE303C12, UK model

chro...@googlecode.com

unread,
Jun 5, 2015, 4:13:54 AM6/5/15
to chromi...@chromium.org

Comment #313 on issue 439818 by n4b...@gmail.com: Samsung 303: frequent
Glad to see comments have been re-enabled. I stumbled across a fix for my
Samsung 303C12 and I thought it might help with the trouble shooting. I've
gone 14 days without a wifi drop.

I was experimenting with different router settings and happened upon a
combination that has worked wonders. To begin with I was on :

Version 42.0.2311.153
Platform 6812.88.0 (Official Build) stable-channel daisyFirmware
Google_Snow.2695.117.0

After setting my CB to use 5 Ghz band only and restricting the protocol
on the router to 802.11A only (i.e. 802.11N disabled), the wifi drops that
I was having since Feb. have stopped completely. I'm now getting the
reliable connectivity I had in the 1.5 yrs. before V40. Of course this
doesn't help when I'm away from home, but it's something. Just thought the
info might be of interest to those doing the troubleshooting.

Oh yeah, my router is a Linksys EA4500 N900 Dual Band running the latest
firmware.

chro...@googlecode.com

unread,
Jun 5, 2015, 4:34:57 AM6/5/15
to chromi...@chromium.org

Comment #314 on issue 439818 by n4b...@gmail.com: Samsung 303: frequent
Additional router info for post #313 -

Firmware version : 2.1.41.162351
Security Mode : WPA2 Personal
Channel Width : 20MHz only
Channel : Auto

chro...@googlecode.com

unread,
Jun 5, 2015, 2:26:13 PM6/5/15
to chromi...@chromium.org

Comment #315 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
Update: We believe that our current set of fixes in R44 fix a lot of issues
for many users. However we are aware that some compatibility issues
persist. We have been debugging this issue with our Wi-Fi chipset vendors
and as a result of a lot of the feedback submitted by users (thank you!) we
have isolated around 30 patches that need to be added.

Router compatibility is extremely challenging-- it is sort of whack-a-mole.
Fixes introduced to work better with one type of router often can introduce
issues with another type of router so it is a delicate balance as we patch
things in.

We will update after the 30 or so patches are sorted out. At this time we
are testing patches to ensure no side effects. At this time we have found
that many user issues are resolved with a router update or router change--
the routers that are giving people the most trouble are often in violation
of Wi-Fi protocol but carefully worked around.

chro...@googlecode.com

unread,
Jun 5, 2015, 2:28:12 PM6/5/15
to chromi...@chromium.org

Comment #316 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
EA4500 offers version 3.x for download:
http://www.linksys.com/ca/support-article?articleNum=148385

You may ask yourself why your EA4500 (set to auto update by default) never
updated to version 3.x.

You may ask yourself why your EA4500 is running version 2.x when the
support website has erased all history of the 2.x release.

You may ask yourself why your EA4500 (running rogue firmware) has led you
(of all places) to the Samsung 303 support forum.

chro...@googlecode.com

unread,
Jun 6, 2015, 5:31:04 AM6/6/15
to chromi...@chromium.org

Comment #317 on issue 439818 by n4b...@gmail.com: Samsung 303: frequent
johnjyates : Your snarkiness is noted.

You may ask yourself if perhaps there are two versions of the EA4500 (V1
and V3) that require two different versions of firmware.

You also may ask yourself why the support website lists 2.1.41 as the
latest release for my v1. And while you're at it, you might ask yourself
why the router's own user interface declares that there is no newer update
available for this router.

You may ask yourself why you, and not Linksys, consider this to be "rogue
software".

I think I can see why the developers chose to shut this thread down to
general comments. As a network engineer myself, I see how the SNR on here
can be more trouble than it is worth.

chro...@googlecode.com

unread,
Jun 6, 2015, 6:44:06 AM6/6/15
to chromi...@chromium.org

Comment #318 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
n4b: thanks for correcting me on the versioning. Of all the network
configurations I've tested 44.0.2403.25 with the EA4500 is the only one
that still shows #wifi40 symptoms... but the EA4500 doesn't just cause
connectivity problems, it seems to be triggering problems with sync and
making some extensions unresponsive. When I last tried to update the
firmware the admin console continually timed out so I naturally assumed
there was a vast rogue software conspiracy. It's informative to read that
other EA4500s are updating and allowing setting changes. I apologize for
the snark noise. Hopefully these behaviors are all a result of Wi-Fi
protocol violations.

chro...@googlecode.com

unread,
Jun 6, 2015, 9:16:58 AM6/6/15
to chromi...@chromium.org

Comment #319 on issue 439818 by bill.wil...@gmail.com: Samsung 303:
@315 - Should we try to download 44 or wait for the patches & testing. On
your router statement, I have a mass-market Netgear - I would assume they
are a fairly solid citizen & not one of the routers violating protocols?

chro...@googlecode.com

unread,
Jun 6, 2015, 2:56:22 PM6/6/15
to chromi...@chromium.org

Comment #320 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
#319: After reading #315 my plan is to wait until the "30 patches" are
merged into dev channel. The merges roll out incrementally but once
everybody has access to them this forum reliably announces their
availability. Once they announce the merges you can upgrade to dev channel
and begin testing against whichever routers are still causing problems.

chro...@googlecode.com

unread,
Jun 8, 2015, 6:18:12 PM6/8/15
to chromi...@chromium.org

Comment #321 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
R44 is in Beta at this point for many users. If you are experiencing
significant Wi-Fi problems we highly recommend trying R44 in Beta.

Unfortunately it is difficult for us to tell precisely which routers are in
protocol violation without knowing the exact model number and firmware
revisions. We do know we found violations in the EA4500 specifically as
this was one that was reported and we tested. Our Wi-fi team has actually
replaced our home routers with the EA4500 to try to reproduce this and have
discovered numerous issues with it we are working around.

chro...@googlecode.com

unread,
Jun 9, 2015, 4:50:47 AM6/9/15
to chromi...@chromium.org

Comment #322 on issue 439818 by nospo...@gmail.com: Samsung 303: frequent
Been using R44 for the past 36hours. 0 dropouts. Use to get at a minimum 2
dropouts per 4 hours of video watching.

Thanks alot to the chromium team for this.

Model: Samsung Chromebook 303 (Early 2013 model)
Router: Dlink DAP-1360

<3

chro...@googlecode.com

unread,
Jun 10, 2015, 8:18:40 AM6/10/15
to chromi...@chromium.org

Comment #323 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
For those of you dogfooding at home: please test configuring the routers
directly from Chrome OS. In particular, please test the admin consoles
when the routers are in recovery mode, offline, restricted to https, with
the chromebook connected via usb3gig.

chro...@googlecode.com

unread,
Jun 12, 2015, 5:13:27 PM6/12/15
to chromi...@chromium.org

Comment #326 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
New versions today fix my complaint in #c325.

Version 44.0.2403.44 beta
Platform 7077.34.0 (Official Build) beta-channel daisy_freon
Firmware Google_Snow.2695.132.0

Version 45.0.2427.10 dev
Platform 7155.1.0 (Official Build) dev-channel daisy_freon
Firmware Google_Snow.2695.132.0

chro...@googlecode.com

unread,
Jun 17, 2015, 7:24:48 AM6/17/15
to chromi...@chromium.org

Comment #327 on issue 439818 by anneke...@gmail.com: Samsung 303: frequent
Hello,
Found this group by accident - and wish to report that my Samsung 303C wifi
is still frequently dropping out, even after all the updates.

Google Chrome 43.0.2357.81 (Official Build) (32-bit)
Revision f873f0128382b8cf6b5c2a877d12491ca807748e
Platform 6946.55.0 (Official Build) stable-channel daisy

Any suggestions?
Thanks,

chro...@googlecode.com

unread,
Jun 18, 2015, 2:41:14 PM6/18/15
to chromi...@chromium.org

Comment #330 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
Your mileage may vary. Some users have reported significant improvements.
We have additional patches coming for users still reporting issues intended
to resolve router compatibility. Additionally we have ordered copies of
reported problem routers.

Again, if you are not an experienced tester, stay on 43 and wait to be
automatically updated.

chro...@googlecode.com

unread,
Jun 18, 2015, 2:56:21 PM6/18/15
to chromi...@chromium.org

Comment #331 on issue 439818 by tushar.h...@gmail.com: Samsung 303:
Still having issues with an asus rt-ac52u - the recent builds have helped
the chromebook to reconnect faster, but the disconnections do keep
happening.

Temporarily switching to a zyxel extender has resolved the disconnection
issue so it does suggest certain routers are a contributing factor. Having
said that, none of my other devices have ever had problems with my router.

chro...@googlecode.com

unread,
Jun 18, 2015, 6:01:09 PM6/18/15
to chromi...@chromium.org

Comment #332 on issue 439818 by johnj...@gmail.com: Samsung 303: frequent
#329 julien.puydt: The remnants of the #wifi40 connectivity problems are
network device specific. I provided tons of detailed reports on the ea4500
and as of today (Version 45.0.2431.0 dev) the ea4500 has been testing
flawlessly (and the performance is magnitudes imrpoved since R39). Please
request information from your network administrator and provide it to Team
Chromium so your configuration can be added to the list of devices to be
fixed.

chro...@googlecode.com

unread,
Jun 18, 2015, 6:14:10 PM6/18/15
to chromi...@chromium.org

Comment #333 on issue 439818 by ch...@chromium.org: Samsung 303: frequent
We are in contact with #329 and we are acquiring a router of the same make
as the customer. Thanks.
It is loading more messages.
0 new messages