synergy-core 🚀 - Copy-paste causes 'server is dead' error on switching

I am seeing this as well since upgrading to 1.7.3. I see no errors on the server, but I get this on the client, similar to above:

OTE: server is dead
WARNING: cursor may not be visible
WARNING: failed to connect to server: server is not responding
NOTE: connecting to '192.168.1.5': 192.168.1.5:24800
ERROR: passive ssl error limit exceeded: 100001
ERROR: failed to connect secure socket

Just like the person above wrote, it happens to me several times a day, and I have to restart the server and then client to get it working again.

markrcote on 8 Jun 2015

Unlike the reporter, though, I see "NOTE: disconnected from server" before I stop the server (I left it out above since I wanted to check when it happens).

The last message on the server is about the client connecting, even after the client errors out. Nothing further until I stop the server.

Finally, it doesn't seem that the client retries on its own, or at least I haven't had the patience to wait long enough to see it retry.

markrcote on 8 Jun 2015

In my case it really seems related to copy operations, even it is not easily reproducible.
This morning, after a few hours of properly working, I tried to copy little text (something like 500 characters) on the Windows based client and when I moved the pointer (but only after that) into the Linux based server screen the messages appeared in the client GUI (nothing in the server GUI) and the application stopped working.

This time I tried restarting only the server and it worked.

sthor69 on 9 Jun 2015

@XinyuHou , Does this look like a duplicate of #4712?

speaker on 17 Jun 2015

Please take into account that in my case I'm not using SSL and my steps are a little different than those in #4712 as the copy operation is done in the client and the server does not crash (at least this is not evident from server's GUI). Simply I see a "server dead" message in the client's GUI

sthor69 on 17 Jun 2015

@sthor69

NOTE: server is dead
WARNING: failed to connect to server: server is not responding
NOTE: connecting to '...'
WARNING: failed to connect to server: Timed out

Then the client attempts to connect to the server but fails with the 'Timed out' message.
If, during these attempts, I stop the server, in the client I see the message "NOTE: disconnected from server".

This looks like client is trying to establish a second connection without cleaning up the first one. If this is true, it would a new bug to us.

Since you are not using SSL, it is probably some generic clipboard issue. Would mind send us some logging in Debug level? You can change logging level in Edit->Settings->Logging level

XinyuHou on 17 Jun 2015

@markrcote

It looks like in your case you are using SSL which could be related to #4712.

ERROR: passive ssl error limit exceeded: 100001

This is a separate issue. Our next release that should in this week would fix it. Please sit tight.

XinyuHou on 17 Jun 2015

This time the energyc.exe process crashed.
I don't know whether it is related to this bug or to #4769 (or maybe they are the same bug).
I have debug level activated but I don't know how to attach the log to this thread, as I see only the way to attach images.

BTW, ti is really difficult to understand which logs relate to an issue, as there are no timestamps for each log. Why didn't you insert them?

sthor69 on 18 Jun 2015

I found that typically if I open my PC in the morning and I try a copy & paste operation in an Excel file, this issue comes back. This is the log from the WindowsXP (client) GUI from when I move the pointer from the server to the client up to the server dead message:

INFO: entering screen
DEBUG: clipboard changed: lost ownership
INFO: leaving screen
DEBUG: open clipboard
DEBUG: close clipboard
DEBUG: sending clipboard 0 seqnum=69
DEBUG: sent clipboard size=4
DEBUG: open clipboard
DEBUG: close clipboard
DEBUG: sending clipboard 1 seqnum=69
DEBUG: sent clipboard size=4
INFO: server is dead
WARNING: failed to connect to server: server is not responding
DEBUG: retry in 1 seconds
INFO: connecting to '172.20.1.39': 172.20.1.39:24800
WARNING: failed to connect to server: Timed out
DEBUG: retry in 1 seconds

This is the log in the server from when I move the pointer from the server to the client up to when I stop the server

INFO: switch from "" to "" at 0,559
INFO: leaving screen
DEBUG: open clipboard 0
DEBUG: close clipboard 0
DEBUG: ignored screen "" update of clipboard 0 (unchanged)
DEBUG: open clipboard 1
DEBUG: close clipboard 1
DEBUG: ignored screen "" update of clipboard 1 (unchanged)
DEBUG: received client "" grabbed clipboard 0 seqnum=61
DEBUG: received client "" grabbed clipboard 1 seqnum=61
INFO: screen "" grabbed clipboard 0 from ""
DEBUG: open clipboard 0
DEBUG: empty clipboard 0
DEBUG: grabbed clipboard 0
DEBUG: close clipboard 0
INFO: screen "" grabbed clipboard 1 from ""
DEBUG: open clipboard 1
DEBUG: empty clipboard 1
DEBUG: grabbed clipboard 1
DEBUG: close clipboard 1
INFO: switch from "" to "" at 1896,647
INFO: entering screen
DEBUG: open clipboard 0
NOTE: stopping synergy desktop process

sthor69 on 22 Jun 2015

Same problem here. My server is a MBP Yosemite, and the client is Ubuntu 14.04.
I tried synergy today, during 2 hours: 4 or 5 crashes with Server is dead message.
What information do you need to help me?

nicosomb on 23 Jun 2015

@XinyuHou Do you need other info? The issue is becoming really annoying. Dozens of crashes a day.

sthor69 on 29 Jun 2015

@sthor69
I need to look into the code and get better understanding.

Could you send us some log info with DEBUG2 this time? Thank you.

XinyuHou on 29 Jun 2015

And what about my problem?

nicosomb on 30 Jun 2015

@nicosomb

DEBUG2 log would be helpful. Thank you.

XinyuHou on 30 Jun 2015

This is the DEBUG2 log after the last crash (I included the last few lines before server dead and a couple of attempts to connect). If you need more let me know

DEBUG1: sending clipboard chunk
DEBUG: open clipboard
DEBUG2: sending clipboard chunk data: size=4
DEBUG2: writef(DCLP%1i%4i%1i%s)
DEBUG: close clipboard
DEBUG: sending clipboard 1 seqnum=1
DEBUG2: wrote 18 bytes
DEBUG1: sending clipboard chunk
DEBUG2: sending clipboard finished
DEBUG2: writef(DCLP%1i%4i%1i%s)
DEBUG2: wrote 14 bytes
DEBUG1: sending clipboard chunk
DEBUG2: sending clipboard chunk start: size=4
DEBUG2: writef(DCLP%1i%4i%1i%s)
DEBUG2: wrote 15 bytes

DEBUG2: writef(ILOG%s)
DEBUG2: wrote 523 bytes
DEBUG: sent clipboard size=4

DEBUG2: writef(ILOG%s)
DEBUG1: thread 0x000009f8 exit
DEBUG1: sending clipboard chunk
DEBUG2: sending clipboard chunk data: size=4
DEBUG2: writef(DCLP%1i%4i%1i%s)
DEBUG2: wrote 18 bytes
DEBUG1: sending clipboard chunk
DEBUG2: sending clipboard finished
DEBUG2: writef(DCLP%1i%4i%1i%s)
DEBUG2: wrote 14 bytes

DEBUG2: wrote 39 bytes
DEBUG2: writef(ILOG%s)
DEBUG2: wrote 303 bytes
INFO: server is dead

DEBUG2: writef(ILOG%s)
DEBUG2: wrote 31 bytes
DEBUG1: thread 0x000002d0 exit

DEBUG2: writef(ILOG%s)
WARNING: failed to connect to server: server is not responding
DEBUG: retry in 1 seconds

DEBUG2: wrote 41 bytes
DEBUG2: writef(ILOG%s)
DEBUG2: wrote 100 bytes
INFO: connecting to '172.20.1.39': 172.20.1.39:24800

DEBUG2: writef(ILOG%s)
DEBUG1: connecting to server

DEBUG2: wrote 63 bytes
DEBUG2: writef(ILOG%s)
DEBUG1: connected; wait for hello

DEBUG2: wrote 76 bytes
DEBUG2: writef(ILOG%s)
out

WARNING: failed to connect to server: Timed out
DEBUG: retry in 1 seconds

On 29/06/2015 23:51, Jerry (Xinyu Hou) wrote:

@sthor69 https://github.com/sthor69
I need to look into the code and get better understanding.

Could you send us some log info with DEBUG2 this time? Thank you.

—
Reply to this email directly or view it on GitHub
https://github.com/synergy/synergy/issues/4768#issuecomment-116857690.

sthor69 on 1 Jul 2015

@sthor69

Can you remember what your last couple of operations before the crash?

Please send us the log from both server and client sides. You can use log to file feature in settings.

XinyuHou on 1 Jul 2015

Typically it happens when I do a lot of copy&paste activity in Excel on
the WindowsXP client.
It is difficult for me to grasp the part of the server log related to
the issue I saw yesterday.
I'll try to do that today, but the lack of timestamping in logs does not
help. Why don't you (as synergy project team I mean) introduce this?

On 01/07/2015 21:53, Jerry (Xinyu Hou) wrote:

@sthor69 https://github.com/sthor69

Can you remember what your last couple of operations before the crash?

Please send us the log from both server and client sides. You can use
log to file feature in settings.

—
Reply to this email directly or view it on GitHub
https://github.com/synergy/synergy/issues/4768#issuecomment-117806597.

sthor69 on 2 Jul 2015

@sthor69
We are planning to implement it in 1.7.5.

XinyuHou on 2 Jul 2015

Seeing this as well. How does one turn off SSL in the textual config? I have no need for it.

Naatan on 6 Jul 2015

@XinyuHou Unfortunately, these days 'm not working much on Excel and I don't perform copy&paste operations, so I couldn't reproduce the bug. I'll send you the log as soon as I experienced it again.

Further, today I found what it seems a different bug, more related to mouse movements, but the problem is always how to filter the log without the timestamps, taking into account that enabling DEBUG2 logs in the server there is a huge amount of MotionNotify events, so the interesting logs are difficult to filter.

When is 1.7.5 planned?

sthor69 on 6 Jul 2015

@sthor69

We are working on it. Right now there is a hugh list of issues in 1.7.5 milestone. We might need to prioritize it first. Here is the link

https://github.com/synergy/synergy/milestones/v1.7.5-stable

XinyuHou on 6 Jul 2015

@sthor69

I only be able to reproduce it once but I will keep testing.

Did you mean this issue was probably application specific? Do you generally have this issue when you copy paste in other applications other than Excel?

XinyuHou on 6 Jul 2015

I typically work on Excel.
But yesterday I did a lot of copy & paste in Word and nothing happened.

On 06/07/2015 23:54, Jerry (Xinyu Hou) wrote:

@sthor69 https://github.com/sthor69

I only be able to reproduce it once but I will keep testing.

Did you mean this issue was probably application specific? Do you
generally have this issue when you copy paste in other applications
other than Excel?

—
Reply to this email directly or view it on GitHub
https://github.com/synergy/synergy/issues/4768#issuecomment-119007923.

sthor69 on 8 Jul 2015

It seems really related to Excel.
They are a few days I stop working on Excel and the issue has never come up again so far.

sthor69 on 15 Jul 2015

I use OSX and Linux and I have this issue, I never use any MS applications. So while Excel may stimulate the issue, its definitely not solely responsible.

Naatan on 15 Jul 2015

I found that when the issue is present, if I try to restart synergyd.exe service, a "memory reference" error is displayed (similar to the ones very often displayed when a Windows based system is shut down). I have the DEBUG2 logs related to this case, but they are pretty long. How can I provide you with them?

sthor69 on 29 Jul 2015

I put the log in the following gist

https://gist.github.com/sthor69/02390fe62bda2901c4c0#file-server-20dead-20log

sthor69 on 29 Jul 2015

Having same issue on 1.7.4

Server: Arch Linux 64 Bit Gnome 3
Client: Manjaro Linux 64 Bit XFCE
SSL Disabled

Last entry on client before disconnect:
INFO: leaving screen

Last entry on server:
INFO: entering screen

4ndv on 27 Sep 2015

I have had the same problem ever since version 1.7.3 and it continues on 1.7.4. The client is generally running linux mint. Both version are completely unusable for me, as the connection invariably drops within an hour or two leaving me unable to control my client. Some people have suggested that this might be due to heavy copy/paste operations, but I do not believe this to be the case. In my case, it seems only a matter of time before the client connection is dropped. Most of the time when this happens, the client is simply idling. I have reverted back to version 1.7.2 on all machines and the issue does not present itself.

coreyo on 29 Sep 2015

It keeps stopping even when i don't do ANY copy-paste operations.

4ndv on 1 Oct 2015

just installed v.1.8.0 and the issue persists.
I noticed that simply clicking "Apply" in the server (after I modified the log level) solved it.
Log in the client since when the client was not yet able to connect up to the successful connection (after the click on Apply):

[2015-10-06T12:00:33] DEBUG1: connected; wait for hello

[2015-10-06T12:00:33] DEBUG2: writef(ILOG%s)
[2015-10-06T12:00:33] DEBUG2: wrote 120 bytes
[2015-10-06T12:00:48] DEBUG1: connection timed out

[2015-10-06T12:00:48] WARNING: failed to connect to server: Timed out
[2015-10-06T12:00:48] DEBUG: retry in 1 seconds

[2015-10-06T12:00:48] DEBUG2: writef(ILOG%s)
[2015-10-06T12:00:48] DEBUG2: wrote 61 bytes
[2015-10-06T12:00:48] DEBUG2: writef(ILOG%s)
[2015-10-06T12:00:48] DEBUG2: wrote 129 bytes
[2015-10-06T12:00:49] NOTE: connecting to '172.20.1.243': 172.20.1.243:24800

[2015-10-06T12:00:49] DEBUG2: writef(ILOG%s)
[2015-10-06T12:00:49] DEBUG1: connecting to server

[2015-10-06T12:00:49] DEBUG2: wrote 87 bytes
[2015-10-06T12:00:49] DEBUG1: connected; wait for hello

[2015-10-06T12:00:49] DEBUG2: writef(ILOG%s)
[2015-10-06T12:00:49] DEBUG2: readf(Synergy%2i%2i)

[2015-10-06T12:00:49] DEBUG2: wrote 120 bytes
[2015-10-06T12:00:49] DEBUG2: readf: read 2 byte integer: 1 (0x1)
[2015-10-06T12:00:49] DEBUG2: readf: read 2 byte integer: 6 (0x6)
[2015-10-06T12:00:49] DEBUG1: got hello version 1.6
[2015-10-06T12:00:49] DEBUG1: say hello version 1.6
[2015-10-06T12:00:49] DEBUG2: writef(Synergy%2i%2i%s)
[2015-10-06T12:00:49] DEBUG2: wrote 29 bytes
[2015-10-06T12:00:49] DEBUG1: registered event type ClipboardEvents::clipboardSending as 22
[2015-10-06T12:00:49] DEBUG1: registered event type IScreenEvents::shapeChanged as 23
[2015-10-06T12:00:49] DEBUG1: registered event type ClipboardEvents::clipboardGrabbed as 24

[2015-10-06T12:00:49] DEBUG2: writef(ILOG%s)
[2015-10-06T12:00:49] DEBUG2: wrote 676 bytes
[2015-10-06T12:00:49] DEBUG2: msg from server: QINF

sthor69 on 6 Oct 2015

With the nightly installed on October 6th, the server dead message appears together with the GNOME shell freeze. I opened #5049 because I cannot say it is the same issue

sthor69 on 7 Oct 2015

I am regularly getting these server is dead messages. Windows 8.1 Pro server, Ubuntu 14.04 LTS Client. Both on 1.7.4.

[2015-10-09T09:18:28] INFO: leaving screen
[2015-10-09T09:18:28] DEBUG: open clipboard 1
[2015-10-09T09:18:28] INFO: entering screen
[2015-10-09T09:18:28] DEBUG: ICCCM fill clipboard 1
[2015-10-09T09:18:28] DEBUG:   available targets: STRING (31)
[2015-10-09T09:18:29] INFO: leaving screen
[2015-10-09T09:18:29] DEBUG: added format 0 for target text/plain (567) (146 bytes)
[2015-10-09T09:18:29] DEBUG: close clipboard 1
[2015-10-09T09:18:29] DEBUG: sending clipboard 1 seqnum=61
[2015-10-09T09:18:29] DEBUG: sent clipboard size=158
[2015-10-09T09:18:29] DEBUG: open clipboard 1
[2015-10-09T09:18:29] DEBUG: ICCCM fill clipboard 1
[2015-10-09T09:18:40] NOTE: server is dead

In fact the server is NOT dead, simply Stopping and Starting the client is enough to resolve the problem (although the clipboard continues to be broken I think until the server is restarted).

Could the client not simply restart itself when it gets that error - I appreciate that is not a true fix, but user experience is everything imho. The problem occurs about every hour during the day, so the client _has_ to have a keyboard and mouse connected!

I was not to my knowledge using clipboard at the time - despite the evidence to teh contrary in the logs. Of course Linux has two separate clipboards - the traditional one and the selection/middle-click clipboard - so possibly I was unaware. Could this difference in the way Windows and Linux clipboards work be an issue?

chrisjrob on 9 Oct 2015

Last log excerpt on server that seems to be time related to the server dead message in the client log is a list of (I replaced the name of the server screen with ):

[2015-10-13T16:10:01] DEBUG2: find neighbor on up of
[2015-10-13T16:10:01] DEBUG2: no neighbor on up of
[2015-10-13T16:10:01] DEBUG1: try to leave on up
[2015-10-13T16:10:01] DEBUG1: no neighbor up

sthor69 on 13 Oct 2015

Tested in 1.7.5 RC1 on both server and client and can confirm issue is still happening.

server-is-dead

[2015-11-02T13:47:07] DEBUG: starting process
[2015-11-02T13:47:07] INFO: starting client
[2015-11-02T13:47:07] INFO: command: /usr/bin/synergyc -f --no-tray --debug DEBUG --name asusi5 --enable-crypto 192.168.0.49:24800
[2015-11-02T13:47:07] INFO: config file: /tmp/qt_temp.TJ6304
[2015-11-02T13:47:07] INFO: log level: DEBUG
[2015-11-02T13:47:07] DEBUG: plugins dir: /home/chris/.synergy/plugins
[2015-11-02T13:47:07] DEBUG: loading plugin: libns.so
[2015-11-02T13:47:07] DEBUG: plugin loaded: libns.so (version 1.3)
[2015-11-02T13:47:07] DEBUG: XOpenDisplay(":0")
[2015-11-02T13:47:07] DEBUG: xscreensaver window: 0x00000000
[2015-11-02T13:47:07] DEBUG: screen shape: 0,0 1920x1080 
[2015-11-02T13:47:07] DEBUG: window is 0x03a00004
[2015-11-02T13:47:07] DEBUG: adopting new buffer
[2015-11-02T13:47:07] DEBUG: opened display
[2015-11-02T13:47:07] NOTE: started client
[2015-11-02T13:47:07] NOTE: connecting to '192.168.0.49': 192.168.0.49:24800
[2015-11-02T13:47:07] DEBUG: event queue is ready
[2015-11-02T13:47:07] NOTE: server fingerprint: 63:3B:76:E9:3D:C3:F8:AA:83:38:91:FA:FA:FE:5A:59:59:67:DB:4C
[2015-11-02T13:47:07] INFO: connected to secure socket
[2015-11-02T13:47:07] INFO: server ssl certificate info: /CN=Synergy
[2015-11-02T13:47:07] DEBUG: open clipboard 0
[2015-11-02T13:47:07] DEBUG: empty clipboard 0
[2015-11-02T13:47:07] DEBUG: grabbed clipboard 0
[2015-11-02T13:47:07] DEBUG: close clipboard 0
[2015-11-02T13:47:07] DEBUG: open clipboard 1
[2015-11-02T13:47:07] DEBUG: empty clipboard 1
[2015-11-02T13:47:07] DEBUG: grabbed clipboard 1
[2015-11-02T13:47:07] DEBUG: close clipboard 1
[2015-11-02T13:47:07] NOTE: connected to server
[2015-11-02T13:47:22] INFO: entering screen
[2015-11-02T13:47:22] DEBUG: start receiving clipboard data
[2015-11-02T13:47:22] DEBUG: receiving clipboard 0 size=55
[2015-11-02T13:47:22] DEBUG: received clipboard 0 size=55
[2015-11-02T13:47:22] DEBUG: open clipboard 0
[2015-11-02T13:47:22] DEBUG: empty clipboard 0
[2015-11-02T13:47:22] DEBUG: grabbed clipboard 0
[2015-11-02T13:47:22] DEBUG: add 43 bytes to clipboard 0 format: 0
[2015-11-02T13:47:22] DEBUG: close clipboard 0
[2015-11-02T13:47:22] INFO: clipboard was updated
[2015-11-02T13:47:22] DEBUG: start receiving clipboard data
[2015-11-02T13:47:22] DEBUG: receiving clipboard 1 size=55
[2015-11-02T13:47:22] DEBUG: received clipboard 1 size=55
[2015-11-02T13:47:22] DEBUG: open clipboard 1
[2015-11-02T13:47:22] DEBUG: empty clipboard 1
[2015-11-02T13:47:22] DEBUG: grabbed clipboard 1
[2015-11-02T13:47:22] DEBUG: add 43 bytes to clipboard 1 format: 0
[2015-11-02T13:47:22] DEBUG: close clipboard 1
[2015-11-02T13:47:22] INFO: clipboard was updated
[2015-11-02T13:47:24] INFO: leaving screen
[2015-11-02T13:47:29] INFO: entering screen
[2015-11-02T13:47:41] DEBUG: lost clipboard 1 ownership at 3710638
[2015-11-02T13:48:56] INFO: leaving screen
[2015-11-02T13:48:56] DEBUG: open clipboard 1
[2015-11-02T13:48:56] INFO: entering screen
[2015-11-02T13:48:56] DEBUG: start receiving clipboard data
[2015-11-02T13:48:56] DEBUG: receiving clipboard 0 size=4
[2015-11-02T13:48:57] DEBUG: ICCCM fill clipboard 1
[2015-11-02T13:48:58] DEBUG:   available targets: STRING (31)
[2015-11-02T13:48:58] DEBUG: received clipboard 0 size=4
[2015-11-02T13:48:58] DEBUG: open clipboard 0
[2015-11-02T13:48:58] DEBUG: empty clipboard 0
[2015-11-02T13:48:58] DEBUG: grabbed clipboard 0
[2015-11-02T13:48:58] DEBUG: close clipboard 0
[2015-11-02T13:48:58] INFO: clipboard was updated
[2015-11-02T13:48:58] INFO: leaving screen
[2015-11-02T13:48:58] DEBUG: added format 0 for target UTF8_STRING (284) (18 bytes)
[2015-11-02T13:48:58] DEBUG: close clipboard 1
[2015-11-02T13:48:58] DEBUG: sending clipboard 1 seqnum=18
[2015-11-02T13:48:58] DEBUG: sent clipboard size=30
[2015-11-02T13:48:58] INFO: entering screen
[2015-11-02T13:48:58] DEBUG: open clipboard 1
[2015-11-02T13:48:58] DEBUG: ICCCM fill clipboard 1
[2015-11-02T13:48:58] DEBUG:   available targets: TIMESTAMP (442), TARGETS (441), MULTIPLE (438), UTF8_STRING (284), COMPOUND_TEXT (521)
[2015-11-02T13:49:00] DEBUG: close clipboard 1
[2015-11-02T13:49:00] DEBUG: sending clipboard 1 seqnum=20
[2015-11-02T13:49:00] DEBUG: sent clipboard size=4
[2015-11-02T13:49:34] INFO: leaving screen
[2015-11-02T13:49:34] DEBUG: open clipboard 1
[2015-11-02T13:49:35] DEBUG: ICCCM fill clipboard 1
[2015-11-02T13:49:47] NOTE: server is dead

chrisjrob on 2 Nov 2015

Glad to know I'm not alone. I had synergy 1.5 working well (on Linux, as a server) but then when I installed 1.7.x on my Macs, I was told it wasn't compatible, so I had to upgrade the Linux server. It usually stops responding during a copy operation, but not always.

When it stops responding, I have to kill -9 it to get it to die.

Also, another detail, when it stops responding, I can connect to it on port 24800 (via telnet), but it doesn't print out the "Synergy" greeting like it does when it is working properly.

jkugler on 7 Nov 2015

Today it worked properly for the whole morning, when I didn't do any copy&paste operation.
At the first copy&paste it stopped working.
Also for me Synergy greeting is not printed when connected on port 24800. After a click on "Apply" in the Synergy GUI the Synergy greeting appears again and the client is connected

sthor69 on 9 Nov 2015

Just another piece of info.

The same is happening with Synergy 1.7.5 last night build on Linux 14.04 on server and Synergy 1.7.4 from Linux software centre on Linux 14.10 on client. In this case a restart of the server does not solve and I needed to restart the client.

sthor69 on 17 Nov 2015

Just bought a synergy license yesterday, got to say, this bug has left me feeling really disappointed in the software as a whole.

Any workaround/update to sort this issue out would be much appreciated as currently it's near unusable.

Server is running Windows 7 (64 bit, Synergy 1.7.4), client is running Linux Mint 17.2 Rafaela (64 bit, synergyc 1.7.4, protocol version 1.6).

kjbweb on 18 Nov 2015

Same. Bought license yesterday, not too happy.
Linux Mint 17.2 on both machines, client just drops to "Server is Dead"

1.7.4 on both.

Clicking "Apply" on the client causes it to reconnect. Why doesn't it just retry it's self?

rickoneeleven on 18 Nov 2015

@rickoneeleven @kjbweb I'm really sorry you guys are having issues with this. If this is causing Synergy to be unusable, please email [email protected] and I'll refund your purchase.

Malcolmlowe on 18 Nov 2015

It seems related only to linux.
I tried with Linux on both client and server and it's happening again.
The difference with the Windows + Linux case is that the restart of the server does not help and also the client must be restarted

sthor69 on 18 Nov 2015

+1, Is still happening in 2 machines with Ubuntu 14.04, I can restart Synergy every time when it happen but is really upset, I bought the license yesterday and the Synergy version is 1.7.4.

woohoou on 18 Nov 2015

@Malcolmlowe I downgraded to 1.6.3 and it's working as expected now, much better.

@sthor69 I think you're right. In my case, with the server being on Windows, I always need to restart the client to get it working again after throwing something in the clipboard, whereas my colleague has the server on his Linux machine, and he always needs to restart the server instead.

kjbweb on 18 Nov 2015

@woohoou Under the downloads section in your account, head to "Alternate Downloads" and grab version 1.6.3 for now instead.

kjbweb on 18 Nov 2015

1.7.1 works fine for me if you don't want to go back a full major version.

markrcote on 18 Nov 2015

It seems that with two Linux machines (server and client) the occurrences are much more than with Linux + Windows. Today I'm working in this environment and Synergy is almost unusable

sthor69 on 18 Nov 2015

Thanks guys, 1.7.1 works fine for me as well.

woohoou on 18 Nov 2015

When I try to install 1.7.1 I fall into #3329

sthor69 on 19 Nov 2015

Can anyone test this nightly to see if the problem persists?
http://synergy-project.org/nightly?filter=synergy-v1.7.6-rc1-dcc1c

XinyuHou on 23 Nov 2015

Just installed the nightly on osx as server and fedora as client. Will let you know if the bug is still there

achassagne on 23 Nov 2015

linux mint 17.2 - linux mint 17.2.
tried all versions including synergy-v1.7.6-rc1-dcc1c. This problem is since version 1.7.3. Version 1.7.2 is working fine. I put every update - make sure it doesn't work and rolled back to 1.7.2.

weduser on 24 Nov 2015

Just installed Linux Ubuntu 14.04 (server) - Ubuntu 15.04 (client)

sthor69 on 24 Nov 2015

Argh! I paid for an account, and then I go to download, and 1.7.2 isn't
there!

http://synergy-project.org/download/?alt

1.7.1, then 1.7.3

Siiiiiigh.

j

On Tuesday, November 24, 2015 03:07:52 weduser wrote:

linux mint 17.2 - linux mint 17.2.
tried all versions including synergy-v1.7.6-rc1-dcc1c. This problem is since
version 1.7.3. Version 1.7.2 is working fine. I put every update - make
sure it doesn't work and rolled back to 1.7.2.

Reply to this email directly or view it on GitHub:
https://github.com/synergy/synergy/issues/4768#issuecomment-159231575

Joshua J. Kugler - Fairbanks, Alaska
Azariah Enterprises - Programming and Website Design
[email protected] - Jabber: [email protected]
PGP Key: http://pgp.mit.edu/ ID 0x73B13B6A

jkugler on 24 Nov 2015

I just noticed that the protocol version was updated (1.5 to 1.6) somewhere between 1.7.1 and 1.7.5. This came to light when I tried to use a 1.7.1 server with 1.7.5 clients. Is it possible this bug is because of the protocol.

Side note: that is a really bad practice to introduce a break/non-backward-compatible change in a point release.

jkugler on 24 Nov 2015

Bug still here with nightly ?

GeoHolz on 25 Nov 2015

Yes.
The bug not only <=1.7.2

weduser on 25 Nov 2015

On my end, didn’t have synergy crash since the installation of the nightly.
So far so good

achassagne on 25 Nov 2015

Copy/paste works ?

GeoHolz on 25 Nov 2015

The same for me: no crash since the installation.
I didn't perform any copy&paste operation so far.
Trying with WindowsXP on client...

sthor69 on 25 Nov 2015

Yes, copy/paste is working for me and is more reliable than ever :)

achassagne on 25 Nov 2015

Which version of nightly you installed ? :)

GeoHolz on 25 Nov 2015

synergy-v1.7.6-rc1-dcc1c

On 25 Nov 2015, at 10:18, GeoHolz [email protected] wrote:

Which version of nightly you installed ? :)

—
Reply to this email directly or view it on GitHub https://github.com/synergy/synergy/issues/4768#issuecomment-159638296.

achassagne on 25 Nov 2015

I'll await a stable release, but fingers crossed this is finally it.

I reverted to 1.6 for the time being, which has me dealing with the sticky modifier issue again. If 1.7.6 works I will just be sticking with that version until someone finally creates a competent alternative or the developers do a rewrite that doesn't break every other release.

Naatan on 25 Nov 2015

Using 1.7.6 rc1, so far working great, copy paste working, no crashes yet, OSX 10.11 Server, Client Ubuntu 12.04

krewmarco on 25 Nov 2015

@weduse
Do you use SSL? If yes, did you update the plugin when you install a new version?

XinyuHou on 26 Nov 2015

Yes, I use SSL. I uninstall then re-install the new version. I believe
I remember it installing a new plugin on the client and server.

Marco

On 11/25/2015 03:41 PM, Jerry (Xinyu Hou) wrote:

@weduse
Do you use SSL? If yes, did you update the plugin when you install a
new version?

—
Reply to this email directly or view it on GitHub
https://github.com/synergy/synergy/issues/4768#issuecomment-159756022.

krewmarco on 26 Nov 2015

I use Debian Client 1.7.6 rc1 and Windows 7 Server 1.7.5
Works fine

GeoHolz on 26 Nov 2015

I experienced again a lot of "server dead" with Ubuntu 15.04 in client and Ubuntu 14.04 in server with the synergy-v1.7.6-rc1-dcc1c doing some copy&paste.

one thing: when I installed the nightly in the client, I was not presented with the configuration Wizard, even if I removed the previous Synergy app from the machine. How can I be sure the nightly is the one used in the machine? In the Help only the v1.7.6 is presented

sthor69 on 26 Nov 2015

Hey guys, could you test this nightly?

http://synergy-project.org/nightly?filter=synergy-v1.7.6-rc1-5287c6

If the problem persists, please leave comment with your platform information and which way you copy and paste, for example copy from linux server to mac client.

XinyuHou on 4 Dec 2015

I try with 1.7.6.rc1-5287c6 on my Debian Jessie and 1.7.5 on Windows 7 Synergy Server

GeoHolz on 4 Dec 2015

Hi guys,

I don't know if this is relevant for this, but I think that can help.

I'm using v1.7.6-rc1-c97bd2c on Ubuntu 14.04 x64 as server and Windows 8 as client.

https://gist.github.com/joubertredrat/3bc24bdf488891a926a6

sy_1

joubertredrat on 4 Dec 2015

@joubertredrat

Please try this nightly
http://synergy-project.org/nightly?filter=synergy-v1.7.6-rc1-5287c62

XinyuHou on 4 Dec 2015

I am experiencing this issue as well. Having tried the linked nightly from the comment above, it appears the issue is solved (although it is difficult to say definitively, as the issue only occurred sporadically before, perhaps 2 or 3 times a day)

However, this nightly seems quite unstable in my environment. From limited tests, it seems to consistently crash on the server the first time the cursor is moved to the client, and sometimes on a few subsequent attempts. (Initial clipboard contents when client connects may be the cause?)

Incidentally, after crashing, the cursor and all other input becomes very unresponsive on the server until I dismiss the Windows crash dialog box.

Server: Windows 7 (not elevated)
Client: Ubuntu 15.10
SSL off

Chippit on 5 Dec 2015

+1 same issue with same config

GeoHolz on 7 Dec 2015

After change to nightly 5287c62, sometimes when I try to go to slave, mouse return to center of screen on server, after tries a lot times, works.

[2015-12-07T14:08:52] DEBUG1: thread 0x00000003 exit
[2015-12-07T14:08:52] DEBUG1: sending clipboard chunk
[2015-12-07T14:08:52] DEBUG1: sending clipboard chunk
[2015-12-07T14:08:52] NOTE: client "server-note" has disconnected
[2015-12-07T14:08:52] DEBUG1: registered event type disconnected as 46
[2015-12-07T14:08:52] INFO: jump from "server-note" to "devel01" at 683,384
[2015-12-07T14:08:52] INFO: entering screen
[2015-12-07T14:08:53] NOTE: accepted client connection
[2015-12-07T14:08:53] DEBUG1: saying hello
[2015-12-07T14:08:53] DEBUG1: parsing hello reply
[2015-12-07T14:08:53] DEBUG1: querying client "server-note" info
[2015-12-07T14:08:53] DEBUG1: created proxy for client "server-note" version 1.6
[2015-12-07T14:08:53] DEBUG: received client "server-note" info shape=0,0 1366x768 at 0,0
[2015-12-07T14:08:53] DEBUG1: send info ack to "server-note"
[2015-12-07T14:08:53] NOTE: client "server-note" has connected
[2015-12-07T14:08:53] DEBUG1: send reset options to "server-note"
[2015-12-07T14:08:53] DEBUG1: send set options to "server-note" size=24

joubertredrat on 7 Dec 2015

@Chippit
Could you send us some log? If it's a crash, please use Debug2 level logging and log to file feature in settings.

XinyuHou on 7 Dec 2015

@XinyuHou
The issue has not recurred after a fresh reboot of both machines. I shall keep the log level set to D2 in case it does, however. This nightly appears to have otherwise resolved the disconnection issues. I have not had them since upgrading.

Chippit on 7 Dec 2015

Hi guys,

Again, when I try to run to slave, mouse return to center screen on master, I will test nightly 7a207b4 to see if persists this.

https://gist.github.com/joubertredrat/388d047ecfd91d945453

joubertredrat on 9 Dec 2015

@XinyuHou I've tried nightly 5287c62 and problem still prevails.
The only change is when I try to copy something larger from server to client it will keep crashing (but fortunately reconnecting) until the whole clipboard is copied to other machine.
Before it just crashed.
Server log: http://pastebin.com/5DwvySdx
Client log: http://pastebin.com/c1z3uuKX

Darkless012 on 9 Dec 2015

@joubertredrat
Please try 1.7 nightly, 1.8 is our branch for new features or enhancement while 1.7 is for bug fixing.

Unfortunately the changes I did didn't fix the issue so I have backed them out.

XinyuHou on 9 Dec 2015

@Darkless012
What content did you copy, image or text? How big is it roughly?

I found sometimes some data, not necessarily big, can corrupt the clipboard and cause Synergy disconnected when leave the screen. Work around is copy something else to refresh the clipboard.

XinyuHou on 9 Dec 2015

@XinyuHou
I installed the last nightly as requested, but I still have the problem to understand whether the installed version is the last one. I mean, when I install a nightly sometimes the configuration wizard starts and sometimes not, as the last installation found some configuration files somewhere (I removed both ~/synergy and ~/.synergy directory after removing the app via Ubuntu Software Centre).

In WindowsXP I removed the app via the installation manager and then I removed the registry and it seems to work (the configuration wizard starts as expected), but in Ubuntu I'm still facing the issue.

sthor69 on 10 Dec 2015

@sthor69

Sorry for the inconvenience. We have a feature in 1.8 that show the git revision in about page. But we haven't merged the change downwards to 1.7, and we probably never will.

The reason you have wizard on start is because you delete the settings files on Linux or the registry on Windows. You only need to do that if you think your old settings cause some conflict. But most of the time it's not necessary.

The trick I use sometimes to tell if I have installed the correct version is running the installer again. On Windows it would say repair instead of installing, and on Linux I think it's reinstall instead of install.

XinyuHou on 10 Dec 2015

@XinyuHou, sorry, I'm using nightly 7a207b4 since yesterday and so far is working fine, without random server dead and another crashes.

Then I continue testing with this nightly or you can that I downgrade to f551d36 to continue test?

joubertredrat on 10 Dec 2015

@joubertredrat
This is our latest nightly for 1.7.6
http://synergy-project.org/nightly?filter=synergy-v1.7.6-rc2-a5acb08

7a207b4 is in 1.8 which includes new features and all bug fixes from 1.7.5 and before.

So you want to test this 4768 particular issue, 1.7.6 would be a better option.

XinyuHou on 10 Dec 2015

@XinyuHou

When I run the installer on Linux it says "Reinstall", but the date is not correct. Isn't Linux looking only at the major.minor.maintenance number?

I attach as an example the screenshot when I launch the Linux installer of the a5acb08 nightly. As you can see the date is on November.

swcentre

sthor69 on 11 Dec 2015

@sthor69
On my Ubuntu, it doesn't show the installed date. But it shows the version number as 1.7.6.

XinyuHou on 11 Dec 2015

@XinyuHou
I've been using rc2 for few hours and it seems that problem was fixed (at least it didn't happened so far)

But rc2 requires to remove (and it indeed removed) 52 of my packages for instalation. ?!?! It removes (among the others) libraries like postgres openssl oracle-jdk google-chrome. That's horrific. Or am I doing something wrong?
And a while happened problem with synergy rc2 (no executable found) - there is something I'm missing.

Darkless012 on 15 Dec 2015

Linux Ubuntu client Linux Ubuntu server v1.7.6-rc2-a5acb08. To copy from the client to the server drops.

firefox client to qutim server copy in paste(shift + insert) at this point, the synergy does not work
[2015-12-15T14:20:58] DEBUG1: sending clipboard chunk
[2015-12-15T14:20:58] DEBUG2: sending clipboard finished
[2015-12-15T14:20:58] DEBUG2: writef(DCLP%1i%4i%1i%s)
[2015-12-15T14:20:58] DEBUG2: wrote 14 bytes
[2015-12-15T14:20:58] DEBUG2: writing secure socket:0x17a4360
[2015-12-15T14:20:58] DEBUG: sent clipboard size=28
[2015-12-15T14:20:58] DEBUG1: send took 0.131885s
[2015-12-15T14:20:58] DEBUG1: thread 0x00000003 exit
[2015-12-15T14:21:11] NOTE: server is dead
[2015-12-15T14:21:12] WARNING: failed to connect to server: server is not responding
[2015-12-15T14:21:12] DEBUG: retry in 1 seconds
[2015-12-15T14:21:13] NOTE: connecting to '192.168.0.107': 192.168.0.107:24800

weduser on 15 Dec 2015

So far no "server dead" event since last nightly a5acb08.
The only annoying issue is a multiple "go to the centre of screen" events when pointer is kept for a long time in the server screen, even without copy & paste operations

sthor69 on 15 Dec 2015

That's incredible!! Just happened after I sent the comment, on WindowsXP (client) and Ubuntu 14.04 (server) after a copy & paste (not too big, only an URL). Unfortunately I didn't have DEBUG2 active on client

sthor69 on 15 Dec 2015

We backed out some changes related to this issue in 1.7.6. It didn't fix the issue and cause more connection drop :(

XinyuHou on 15 Dec 2015

@Darkless012
Please check if you installed a 32 bit package on a 64 bit machine or visa versa.

XinyuHou on 15 Dec 2015

@XinyuHou oh, you are completely right. I didn't think about that :( Thank you.

Darkless012 on 16 Dec 2015

Hi guys,

Here problem persists, I will update to nightly 51705f7 to test again.

joubertredrat on 7 Jan 2016

@joubertredrat
Did you test 1.7.6-rc2 before?

XinyuHou on 7 Jan 2016

Not real sure, but I think I have the same problem as listed in this thread, I am using nightly 51705f7 on a Linux client and server. Due to the clipboard, it seems like the server eventually dies, requiring a kill -9 from the GUI to return. Here is the client output:

[2016-01-09T13:39:37] DEBUG2: writef(DCLP%1i%4i%1i%s)
        /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ProtocolUtil.cpp,36
[2016-01-09T13:39:37] DEBUG2: wrote 18 bytes
        /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ProtocolUtil.cpp,87
[2016-01-09T13:39:37] DEBUG1: sending clipboard chunk
        /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ClipboardChunk.cpp,132
[2016-01-09T13:39:37] DEBUG2: sending clipboard finished
        /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ClipboardChunk.cpp,151
[2016-01-09T13:39:37] DEBUG2: writef(DCLP%1i%4i%1i%s)
        /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ProtocolUtil.cpp,36
[2016-01-09T13:39:37] DEBUG2: wrote 14 bytes
        /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ProtocolUtil.cpp,87
[2016-01-09T13:39:37] DEBUG2: can't read property 517 on window 0x01800004
        /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/platform/XWindowsUtil.cpp,1378
[2016-01-09T13:39:46] NOTE: server is dead
        /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/client/ServerProxy.cpp,337
[2016-01-09T13:39:46] WARNING: failed to connect to server: server is not responding
        /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ClientApp.cpp,312
[2016-01-09T13:39:46] DEBUG: retry in 1 seconds
        /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ClientApp.cpp,284
[2016-01-09T13:39:46] DEBUG2: can't read property 517 on window 0x00a00003

Here is the final output from the server before its untimely death:

[2016-01-09T13:39:37] DEBUG2: readf: read 1 byte integer: 2 (0x2)
    /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ProtocolUtil.cpp,124
[2016-01-09T13:39:37] DEBUG2: readf: read 4 byte string: 
    /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ProtocolUtil.cpp,236
[2016-01-09T13:39:37] DEBUG2: msg from "HomeDesktop": DCLP
    /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/server/ClientProxy1_0.cpp,150
[2016-01-09T13:39:37] DEBUG2: readf(%1i%4i%1i%s)
    /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ProtocolUtil.cpp,52
[2016-01-09T13:39:37] DEBUG2: readf: read 1 byte integer: 1 (0x1)
    /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ProtocolUtil.cpp,124
[2016-01-09T13:39:37] DEBUG2: readf: read 4 byte integer: 193 (0xc1)
    /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ProtocolUtil.cpp,143
[2016-01-09T13:39:37] DEBUG2: readf: read 1 byte integer: 3 (0x3)
    /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ProtocolUtil.cpp,124
[2016-01-09T13:39:37] DEBUG2: readf: read 0 byte string: 
    /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ProtocolUtil.cpp,236
[2016-01-09T13:39:37] DEBUG: received client "HomeDesktop" clipboard 1 seqnum=193, size=4
    /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/server/ClientProxy1_6.cpp,88

And here is the immediate log after hitting the "Apply" button in the GUI on the server (which I use to force restart the server):

[2016-01-09T13:48:45] DEBUG: stopping process
[2016-01-09T13:48:45] INFO: stopping synergy desktop process
[2016-01-09T13:48:45] ERROR: process exited with error code: 9
[2016-01-09T13:48:45] DEBUG: starting process
[2016-01-09T13:48:46] INFO: starting server
[2016-01-09T13:48:46] INFO: command: /usr/bin/synergys -f --no-tray --debug DEBUG2 --name HomeLaptop --log /var/log/synergy.log -c /tmp/Synergy.yJ2191 --address :24800
[2016-01-09T13:48:46] INFO: config file: /tmp/Synergy.FW2191
[2016-01-09T13:48:46] INFO: log level: DEBUG2
[2016-01-09T13:48:46] INFO: log file: /var/log/synergy.log
[2016-01-09T13:48:46] DEBUG1: logging to file (/var/log/synergy.log) enabled
    /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/App.cpp,155
[2016-01-09T13:48:46] DEBUG: opening configuration "/tmp/Synergy.yJ2191"
    /tmp/pkgbuild-1000/synergy-git/src/synergy/src/lib/synergy/ServerApp.cpp,229

More than likely, I was copying something from a terminal or Google Chrome when this happened. It doesn't happen consistently, so I can't seem to figure out exactly what is causing it. Also, this doesn't seem to happen on my Windows running synergy server, master 7a207b4 with Linux clients running 1.7.6 a5acb08. This only happens on my Linux running synergy server (previously was running a5acb08, but same behavior on 51705f7). None of my clients in either scenario were running Windows, only Linux.

ramesh45345 on 9 Jan 2016

@vinadoros
Thank you for all the information.

Could you please try to change the checking client time in server configuration. You can do that by:

Click Configure Server
Advanced Server Settings
In the Options section, tick "Check clients every"
Change the time

XinyuHou on 11 Jan 2016

@XinyuHou I have ticked and changed the setting in the server configuration to 1000ms. Haven't seen the server die yet, but I have only used it in that configuration for about an hour. May take a few more days of use to see what the effects are of this.

ramesh45345 on 12 Jan 2016

@XinyuHou I made the changes in the Check Client setting (1000ms) and I didn't see a single 'server dead' since this morning, even if I must say that I rarely used the clipboard

sthor69 on 12 Jan 2016

@XinyuHou I have been using the Check Client setting with 1000ms for a few days now. I have counted one server death in the manner I mentioned above since then. Will continue testing like this. I would say that the issue is cropping up less, but that could be my imagination too.

ramesh45345 on 15 Jan 2016

@vinadoros @sthor69
Thanks for letting me know. I'm not expecting this workaround solving the problem completely. But it could help us narrow down the causes.

XinyuHou on 15 Jan 2016

I have the same issue. Occurs roughly once every hour (or less) Setup is:

OSX Yosemite (client) -- Windows 10 (server) -- Ubuntu 15.04 (client).

The Ubuntu 15.04 client regularly sees dead server, and does not seem to attempt a re-connect. The OSX client does not have the dead server issue.

I tried setting "check clients every 1000ms" with no affect.

Using latest Synergy 1.7.5 with SSL disabled (as it seems to cause issues).

vegardbystrom on 16 Jan 2016

tested on synergy-v1.7.6-rc3-51705f7-Linux-x86_64.deb Ubuntu client and Ubuntu server. When you try to copy from the client to the server, the client is dead. 1000ms is not helping.

weduser on 16 Jan 2016

@XinyuHou Still happened today with Check set to 1000ms. I don't know whether it is related, but this time ti occurred concurrently with a complete GNOME freeze (but Linux still working, as I heard notifications arriving). I had to restart my PC (server on Ubuntu 14.04)

sthor69 on 18 Jan 2016

installed version synergy-v1.7.6-stable-51705f7-Linux-x86_64.deb, you just 2-3 times copy, then the server dies. If you restart the server. But again, 2-3 copy-and-die. Went back to version synergy-v1.7.2-stable-728e9cd-Linux-x86_64.deb

weduser on 17 Mar 2016

I have been running the new version for three days and have only had one
failure. This compares with at least five failures per day prior to the
upgrade.

The one failure I did have was really bad - stopping the services on both
machines did not fix the problem, even checking that there were no
processes running. In the end it required a reboot of both machines - I
hope that was isolated incident.

Interestingly the clipboard has also become usable between the two machines
for the first time.

Early days, but so far - bliss.

Thank you.

Regards,
Chris.

chrisjrob on 22 Mar 2016

I installed the new version and after the first copy&paste the "server dead" issue raised again.
I'll go on testing it in the next days.

sthor69 on 23 Mar 2016

I am running 1.7.6 and I can lock the server running on Ubuntu 15.10 on demand. I get multiple messages like this

[2016-04-13T10:13:16] INFO: switch from "ubuntu" to "Laptop" at 1919,1079
[2016-04-13T10:13:16] INFO: leaving screen
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:16] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:17] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:17] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:17] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:17] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:17] DEBUG2: waiting to grab keyboard
[2016-04-13T10:13:17] DEBUG2: grab keyboard timed out
[2016-04-13T10:13:17] WARNING: can't leave screen

Eventually the server unlocks itself and my 2 clients can re-connect.

My Ubuntu setup consists of 2 screen and I have one screen running a Windows 10 VM (using virtualbox). If I am working in the VM and then attempt to move the cursor to another computer the server will lockup. If I leave the VM and click on a window on the Ubuntu side then the problem does not occur.

ZippyManiac on 13 Apr 2016

I observe something similar, I also have 2 linux machines and one of them is running virtualbox with windows. Disconnections additionally randomly arise when copying text from the windows on virtualbox, especially from msoffice apps.

CyberKiller40 on 13 Apr 2016

I'm getting this when SSL is enabled. I'm running v1.7.6 server on Fedora23 and v1.7.6 client on Windows 7. The fingerprint exchange happens and everything appears to work but then I get Server is dead message and no keyboard/mouse capture works.

Disable SSL and everything works fine.

jclabeaux on 13 Jun 2016

I was brought here from #5393 and shall await further instructions. If there is anything I can do to help diagnose the issue please let me know.

TafThorne on 27 Jun 2016

Hi guys,

Mine synergy continues with this problem, however, very decreased the number of errors, mostly, when I copy links from browsers.

joubertredrat on 27 Jun 2016

This usually happens multiple times a day, but I can't predict when it happens. I have an Ubuntu (12.0.4 LTS) synergy host that has a single Windows 7 client connected to it. From time to time, when I copy a URL on the Windows screen and move to my Ubuntu host, the connection between them will 'drop' without warning. Enabling debugging when this happens yields this:

[2016-06-29T16:07:31] INFO: switch from "[my windows box]" to "[my ubuntu box]" at 6730,592
[2016-06-29T16:07:31] INFO: entering screen
[2016-06-29T16:07:40] DEBUG: ssl connection closed

On the client side:

[2016-06-29T16:07:14] INFO: entering screen
[2016-06-29T16:07:15] INFO: clipboard was updated
[2016-06-29T16:07:15] INFO: clipboard was updated
[2016-06-29T16:07:20] INFO: leaving screen
[2016-06-29T16:07:29] NOTE: server is dead
[2016-06-29T16:07:30] WARNING: failed to connect to server: server is not responding

Ubuntu version: 1.7.6
Windows version: 1.7.6

If it makes a difference, I drive 3 monitors on my Ubuntu box. Once this happens, my Windows client continues to try and reconnect, but constantly fails to do so. Only when I explicitly stop and start my Ubuntu synergy server does it allow a new connection.

NOTE: Occasionally, I also see this in the host logs:

[2016-06-30T11:12:18] ERROR: ssl error occurred (system call failure)
[2016-06-30T11:12:18] ERROR: eof violates ssl protocol

XinyuHou on 1 Jul 2016

Can anyone test this nightly?
http://symless.com/nightly?filter=85bbbd4

XinyuHou on 11 Jul 2016

hi @XinyuHou for both server and client?

joubertredrat on 11 Jul 2016

I am installing the 64-bit version on an Ubuntu server and Windows 7 client now. I will let you know if I get any further 'server dead' messages from either machine in the debug logs.

TafThorne on 11 Jul 2016

@joubertredrat
Yes, please. Thank you.

@TafThorne
Thank you.

XinyuHou on 11 Jul 2016

Installing 32bit version on WindowsXP client and 64bit version on Ubuntu 14.04.03 server

sthor69 on 11 Jul 2016

Server on Ubuntu 14.04.4 x64 - done
Client on Windows 8 x64 - done

@XinyuHou are you want DEBUG1 enabled?

joubertredrat on 11 Jul 2016

@joubertredrat
Debug should be enough.

@sthor69
Thank you

XinyuHou on 11 Jul 2016

Installed on ubuntu 16.04 + 2x windows 10 machines

ZippyManiac on 12 Jul 2016

@XinyuHou since installed nightly I have problems only if I copy from browser, as it was before.

But I saw a new strange issue, now on client, randomly caps lock light on my notebook
turns on and off light when I type on a alpha keyboard, if I enable caps lock, same problem, only I press Shift and type, works normally.

On stable 1.7.3 this doesn't happen, works normally. If you want, I can provide small video with this.

I noticed this while playing, because I couldn't move my tank.

My log with this test is here, https://gist.github.com/joubertredrat/189aea98ef6490e1f7e6b4b58258c395

joubertredrat on 13 Jul 2016

@joubertredrat
Could you send us the log when you have clipboard issue? Debug level should be fine.

What content did you try to copy from browser, text, formatted text or image?

The caps lock might be the side effect from #3044 fix. Could you test this nightly to narrow down the cause?
http://symless.com/nightly?filter=0034ca4

XinyuHou on 13 Jul 2016

@XinyuHou I try to copy url address from address bar, plain text, or with right click on link and click on 'copy url', but isn't all time, is randomly too, but when it happens, I can't go to client, display no neighbor left (in my case client is on left), then I stop and start server and return to work normally.

I will provide log with this issue and install next nightly to test #3044

joubertredrat on 13 Jul 2016

@joubertredrat
Thank you.

Could you post clipboard in this issue and the caps lock one into #3044 if it's doesn't happen in 0034ca4?

XinyuHou on 13 Jul 2016

Hi @XinyuHou

Random dead client with https://github.com/symless/synergy/commit/0034ca4b7666f5d570eab4a9c275e373008b7c64, was after open World of Tanks game client to test about caps lock. Now return only if I stop and start client, but, 2 or 4 minutes after, stop again.

https://gist.github.com/joubertredrat/a990637e4323ce94bf934bda991ea4c2

Note: now is stable, I will try more tests here

joubertredrat on 13 Jul 2016

I do not see a server dead message but my Windows 7 client stopped responding to the server until I stopped and re-started the Ubuntu server. That is much like the problem I have been seeing for the past few months.

Client Log

https://gist.github.com/XinyuHou/badc4dc1c9aedf95ec0490f16b63027a

Server Log

https://gist.github.com/XinyuHou/5809d13730563d1063ea7a71f601d271

Activity At Time Of Problem

I had moved my mouse from the client to the server then copied a URL from the Chrome address bar on the server. When I attempted to move the mouse back to the client I noticed that I could not do so. I wasted about one minute to see if things would recover but they did not. I clicked the Stop button on the server GUI to stop the server and then clicked the Start button a few seconds later. After that I was able to move the mouse to the client machine again. Copying the URl a second time did not cause the problem to repeat.

TafThorne on 14 Jul 2016

So far not a single event, but I didn't use a lot of copy&paste yet...
keep monitoring

sthor69 on 14 Jul 2016

I have been trying 0034ca4 today on my Win 7 server/Ubuntu 16.04 client (x64 on both sides). While the issue seems to happen less often than previously, I am still getting "NOTE: server is dead".

thalin on 18 Jul 2016

@thalin
Could you send us some logging when the issue happens? Debug2 log level should be fine.

XinyuHou on 18 Jul 2016

@TafThorne
Thanks for the logging.

Could you send us another one with log level set to Debug2 please?

XinyuHou on 18 Jul 2016

OK @XinyuHou I have turned my logging up to Debug2 on both the Client and the Server. I will post back here if I see a repeat of the issue. (I will try and make Gist of the logs too rather than spam the thread).

TafThorne on 18 Jul 2016

@TafThorne
Thanks for the help.

XinyuHou on 18 Jul 2016

@XinyuHou
Overnight I had two crash reports on the Windows client and a crash on the Ubuntu server. Is it usual to see some instability when the very verbose DEBUG2 mode is set? If it is not then how can I report what has happened in a more helpful manner than it fell down? Do I need to be logging everything to files that might contain some last gasp clues?

TafThorne on 19 Jul 2016

Windows client had a hickup just now and wanted me to re-affirm the server finger print too. I do not know what prompted that.
Windows Client Log from time of failure: https://gist.github.com/TafThorne/ba222c094fe0b9b1fcdab6d00251cde2
Ubuntu Server Log from time of failure: https://gist.github.com/TafThorne/4c8bd42ed3b5231f0b0a45bad18fa739

TafThorne on 19 Jul 2016

@TafThorne
We just found a bug in our logging system when users set log level to Debug2. There would be potentially a buffer overflow crash, mainly on Linux. Here is an experimental branch that try to fix it and also includes all the clipboard fixes we currently have. Please try it out.

http://symless.com/nightly?filter=3f8ec6a

XinyuHou on 19 Jul 2016

@XinyuHou
Now installed.

TafThorne on 19 Jul 2016

@TafThorne
Rerun wizard would cause GUI to reinstall plugins and generate a new fingerprint. So if you have reran the wizard, that would cause clients to show that pop up window to accept server's new fingerprint.

XinyuHou on 19 Jul 2016

@XinyuHou It seems that the Windows client will not connect to the Ubuntu server. I have stopped, Quit and restarted the application on both machines to ensure they are no running version 1.8.2-beta-3f8ec6a. The client machine keeps saying:https://gist.github.com/TafThorne/4ce6584f142f2f14e210637c80dc989f

TafThorne on 19 Jul 2016

@TafThorne
If you use SSL, please rerun the wizard. I know this is annoying. We need to polish SSL after we stabilize clipboard.

XinyuHou on 19 Jul 2016

@XinyuHou
Thank you. I think it was a combination of a new finger print and an old server still running in the background even though I had used the GUI to stop it. Working again now.

TafThorne on 19 Jul 2016

Looks like the connection just got dropped. On the server end:

[2016-07-19T11:05:03] DEBUG2: can't read property 572 on window 0x01000009
[2016-07-19T11:05:03] DEBUG1: thread 0x00000013 entry
[2016-07-19T11:05:03] DEBUG2: event: MotionNotify 1651,514
[2016-07-19T11:05:03] DEBUG2: event: MotionNotify 2531,546
[2016-07-19T11:05:03] DEBUG2: event: MotionNotify 1651,514
[2016-07-19T11:05:03] DEBUG2: event: MotionNotify 2289,519
[2016-07-19T11:05:03] DEBUG: open clipboard 0
[2016-07-19T11:05:03] DEBUG2: event: MotionNotify 1651,514
[2016-07-19T11:05:03] DEBUG2: event: MotionNotify 2056,519
[2016-07-19T11:05:03] DEBUG2: event: MotionNotify 1651,514
[2016-07-19T11:05:03] DEBUG2: event: MotionNotify 1849,519
[2016-07-19T11:05:03] DEBUG2: event: MotionNotify 1651,514
[2016-07-19T11:05:03] DEBUG2: event: MotionNotify 1651,514
[2016-07-19T11:05:03] DEBUG2: msg from "THORNE-W7-LT": CNOP
[2016-07-19T11:05:03] DEBUG2: no-op from
[2016-07-19T11:05:03] DEBUG2: msg from "THORNE-W7-LT": DCLP
[2016-07-19T11:05:03] DEBUG2: readf(%1i%4i%1i%s)
[2016-07-19T11:05:03] DEBUG2: readf: read 1 byte integer: 0 (0x0)
[2016-07-19T11:05:03] DEBUG2: readf: read 4 byte integer: 16 (0x10)
[2016-07-19T11:05:03] DEBUG2: readf: read 1 byte integer: 1 (0x1)
[2016-07-19T11:05:03] DEBUG2: readf: read 2 byte string: 51
[2016-07-19T11:05:03] DEBUG: start receiving clipboard data
[2016-07-19T11:05:03] DEBUG: receiving clipboard 0 size=51
[2016-07-19T11:05:03] DEBUG2: msg from "THORNE-W7-LT": DCLP
[2016-07-19T11:05:03] DEBUG2: readf(%1i%4i%1i%s)
[2016-07-19T11:05:03] DEBUG2: readf: read 1 byte integer: 0 (0x0)
[2016-07-19T11:05:03] DEBUG2: readf: read 4 byte integer: 16 (0x10)
[2016-07-19T11:05:03] DEBUG2: readf: read 1 byte integer: 2 (0x2)
[2016-07-19T11:05:03] DEBUG2: readf: read 51 byte string: 
[2016-07-19T11:05:03] DEBUG2: msg from "THORNE-W7-LT": DCLP
[2016-07-19T11:05:03] DEBUG2: readf(%1i%4i%1i%s)
[2016-07-19T11:05:03] DEBUG2: readf: read 1 byte integer: 0 (0x0)
[2016-07-19T11:05:03] DEBUG2: readf: read 4 byte integer: 16 (0x10)
[2016-07-19T11:05:03] DEBUG2: readf: read 1 byte integer: 3 (0x3)
[2016-07-19T11:05:03] DEBUG2: readf: read 0 byte string: 
[2016-07-19T11:05:03] DEBUG: received client "THORNE-W7-LT" clipboard 0 seqnum=16, size=51
[2016-07-19T11:05:03] DEBUG2: msg from "THORNE-W7-LT": DCLP
[2016-07-19T11:05:03] DEBUG2: readf(%1i%4i%1i%s)
[2016-07-19T11:05:03] DEBUG2: readf: read 1 byte integer: 1 (0x1)
[2016-07-19T11:05:03] DEBUG2: readf: read 4 byte integer: 16 (0x10)
[2016-07-19T11:05:03] DEBUG2: readf: read 1 byte integer: 1 (0x1)
[2016-07-19T11:05:03] DEBUG2: readf: read 2 byte string: 51
[2016-07-19T11:05:03] DEBUG: start receiving clipboard data
[2016-07-19T11:05:03] DEBUG: receiving clipboard 1 size=51
[2016-07-19T11:05:03] DEBUG2: msg from "THORNE-W7-LT": DCLP
[2016-07-19T11:05:03] DEBUG2: readf(%1i%4i%1i%s)
[2016-07-19T11:05:03] DEBUG2: readf: read 1 byte integer: 1 (0x1)
[2016-07-19T11:05:03] DEBUG2: readf: read 4 byte integer: 16 (0x10)
[2016-07-19T11:05:03] DEBUG2: readf: read 1 byte integer: 2 (0x2)
[2016-07-19T11:05:03] DEBUG2: readf: read 51 byte string: 
[2016-07-19T11:05:03] DEBUG2: msg from "THORNE-W7-LT": DCLP
[2016-07-19T11:05:03] DEBUG2: readf(%1i%4i%1i%s)
[2016-07-19T11:05:03] DEBUG2: readf: read 1 byte integer: 1 (0x1)
[2016-07-19T11:05:03] DEBUG2: readf: read 4 byte integer: 16 (0x10)
[2016-07-19T11:05:03] DEBUG2: readf: read 1 byte integer: 3 (0x3)
[2016-07-19T11:05:03] DEBUG2: readf: read 0 byte string: 
[2016-07-19T11:05:03] DEBUG: received client "THORNE-W7-LT" clipboard 1 seqnum=16, size=51
[2016-07-19T11:05:10] DEBUG2: reading secure socket
[2016-07-19T11:05:10] DEBUG: ssl connection closed

On the client end:

#Working on it

TafThorne on 19 Jul 2016

@TafThorne
Please use gist and send us the whole log. Debug2 is verbose, so 30 lines of logging may not cover the whole time frame when issue happens. Thanks.

XinyuHou on 19 Jul 2016

@XinyuHou As reuqested.
Ubuntu Server: https://gist.github.com/TafThorne/f59f2cc5b9d66819d8575459e382f144
Windows Client (had to be truncated): https://gist.github.com/TafThorne/b08d49ce8a2dec6e511c6e6a5d99038c

TafThorne on 19 Jul 2016

@TafThorne
Thank you!

XinyuHou on 19 Jul 2016

@TafThorne
Is that the log after you get server and client connected? I can't find anything about connection or clipboard in those two logging.

XinyuHou on 19 Jul 2016

@TafThorne
The log you posted before looks promising. We just need a more complete log around that.
https://github.com/symless/synergy/issues/4768#issuecomment-233588300

XinyuHou on 19 Jul 2016

@XinyuHou It was the complete log from my getting the new binary connected and working up until I have the failure and including getting a connection working again by stopping and restarting the client. I think it was just before [2016-07-19T11:06:36] DEBUG2: want to read, error=2, attempt=2 in the client log that things went wrong. My client and server seem to be around 30s out of step today for some reason. I am reviewing the NTP settings to see if I can identify where the drift has come from.

TafThorne on 19 Jul 2016

Server - Ubuntu 16.04

Repeatable failure. I have a virtualbox windows 10 instance running on my ubuntu box. I also have 2x windows 10 PCs (StreamPC & Laptop). If I am working in the virtualbox instance and attempt to move the cursor to either of the windows 10 PCs then I get the following (unless I get focus away from virtualbox by selecting another app on the ubuntu box).

https://gist.github.com/XinyuHou/a8b950ae0d1e89497718a249ea055a05

Client - Windows 10

https://gist.github.com/XinyuHou/06b6145217c0e11a784123b9dab3d148

ZippyManiac on 19 Jul 2016

I didn't think it's relevant before, but I also had a virtualbox running on the linux machine, I had:
server opensuse with vbox guest windows7 and client opensuse (without vbox, but with rdp client connected to another machine).

CyberKiller40 on 19 Jul 2016

I installed Virtual Box on the Ubuntu server yesterday and have had it running an Ubuntu VM for most of that time. Before that I did not have Virtual Box installed on the server. I have had Virtual Box installed on the client machine for years but it is only run every few months.

TafThorne on 19 Jul 2016

I observed the same (?) problem here. I tracked it down to a locking problem in Xlib. I just sent an email to [email protected]. The text follows FYR, please check if this may match the problem you're encountering.

I encounter the following situation on my system reproduceably when I
run the synergy KVM switching software (server side).

Thread A is stuck in this call path:

XSync()
   _XReply()
    // line 625 ff: wait for event_waiter to process first event
        while(dpy->xcb->event_waiter)
        { /* need braces around ConditionWait */
             ConditionWait(dpy, dpy->xcb->event_notify);
        }

I observe that this thread never leaves this while loop any more.

Thread B executes:

XIfEvent()
   _XReadEvents()

Until the event it's waiting for arrives, XIfEvent() calls
_XReadEvents() in a loop with the display lock held. _XReadEvents() only
releases the lock after setting dpy->xcb->event_waiter. Only after
re-acquiring the lock, it clears event_waiter and sends a broadcast for
the condition variable dpy->xcb->event_notify.

In this situation, thread A is effectively blocked: ConditionWait()
receives the broadcast, but before it can acquire the display lock and
return, _XReadEvents() will have set event_waiter once more, so it needs
to wait again.

This goes on forever unless the event XIfEvent() is waiting for arrives.
In some situations, this seems to take a forever too, I'm not sure why
(maybe we are looking at a deadlock situation here in the sense that the
expected condition can't happen unless thread A gets a chance to run).

I am attaching a tentative patch that I've come up with. Please review
it. The rationale is to yield to a request waiter after grabbing one
event. I am currently testing it on my system (too early to come to a
conclusion though, at least it doesn't seem to have caused a regression
so far).

I also came up with a tentative patch for libX11 which I'm currently testing. So far no deadlock, but in my case the time to failure was several hours, so no conclusion yet.

mwilck on 20 Jul 2016

👍1

As the synergy developers are reading this, I should include more context here. Maybe that gives you an idea about the synergy side of the problem (after all, it seems to have occured only with recent versions).

It seems that the trick used by XWindowsUtil::getCurrentTime() to receive a time stamp may sometimes fail (maybe the expected event is consumed by another thread before XIfEvent() is called?).

Thread A:

#0  pthread_cond_wait@@GLIBC_2.3.2 () at ../sysdeps/unix/sysv/linux/x86_64/pthread_cond_wait.S:185
#1  0x00007f475c4202b5 in _XReply (dpy=dpy@entry=0xd8e7a0, rep=rep@entry=0x7ffffbbfd4f0, extra=extra@entry=0, discard=discard@entry=1) at xcb_io.c:624
#2  0x00007f475c41bc3d in XSync (dpy=0xd8e7a0, discard=discard@entry=0) at Sync.c:44
#3  0x000000000046621c in XWindowsUtil::ErrorLock::install (this=0x7ffffbbfd5e0, handler=0x465e30 <XWindowsUtil::ErrorLock::ignoreHandler(_XDisplay*, XErrorEvent*, void*)>, data=0x0) at /usr/src/debug/synergy-1.7.6-stable/src/lib/platform/XWindowsUtil.cpp:1754
#4  0x00000000004662ec in XWindowsUtil::getWindowProperty (display=0xd8e7a0, window=18874372, property=487, data=data@entry=0x0, type=type@entry=0x7ffffbbfd660, format=format@entry=0x0, deleteProperty=false) at /usr/src/debug/synergy-1.7.6-stable/src/lib/platform/XWindowsUtil.cpp:1306
#5  0x0000000000464bad in XWindowsScreenSaver::isXScreenSaver (this=<optimized out>, w=<optimized out>) at /usr/src/debug/synergy-1.7.6-stable/src/lib/platform/XWindowsScreenSaver.cpp:381
#6  0x0000000000465b60 in XWindowsScreenSaver::handleXEvent (this=0xda3490, xevent=xevent@entry=0xdad598) at /usr/src/debug/synergy-1.7.6-stable/src/lib/platform/XWindowsScreenSaver.cpp:189
#7  0x0000000000461ad8 in XWindowsScreen::handleSystemEvent (this=0xd8d250, event=...) at /usr/src/debug/synergy-1.7.6-stable/src/lib/platform/XWindowsScreen.cpp:1261
#8  0x00000000004487aa in EventQueue::dispatchEvent (this=0x7ffffbbfdaf0, event=...) at /usr/src/debug/synergy-1.7.6-stable/src/lib/base/EventQueue.cpp:271

Thread B:

#4  0x00007f475c41fed8 in _XReadEvents (dpy=dpy@entry=0xd8e7a0) at xcb_io.c:401
#5  0x00007f475c407809 in XIfEvent (dpy=dpy@entry=0xd8e7a0, event=event@entry=0x7f4743ff6c00, predicate=predicate@entry=0x465dc0 <X
WindowsUtil::propertyNotifyPredicate(_XDisplay*, _XEvent*, char*)>, arg=arg@entry=0x7f4743ff6b60 "\004") at IfEvent.c:68
#6  0x0000000000465f8b in XWindowsUtil::getCurrentTime (display=0xd8e7a0, window=18874372) at /usr/src/debug/synergy-1.7.6-stable/src/lib/platform/XWindowsUtil.cpp:1455
#7  0x000000000046755c in XWindowsClipboard::motifUnlockClipboard (this=this@entry=0xdaafb0) at /usr/src/debug/synergy-1.7.6-stable/src/lib/platform/XWindowsClipboard.cpp:646
#8  0x0000000000467842 in XWindowsClipboard::open (this=0xdaafb0, time=54750371) at /usr/src/debug/synergy-1.7.6-stable/src/lib/platform/XWindowsClipboard.cpp:338
#9  0x00000000004997b5 in IClipboard::copy (dst=0xdaafb0, src=0xdae120, time=54750371) at /usr/src/debug/synergy-1.7.6-stable/src/lib/synergy/IClipboard.cpp:132
#10 0x000000000048876a in Server::sendClipboardThread (this=0xdae040) at /usr/src/debug/synergy-1.7.6-stable/src/lib/server/Server.cpp:1866

mwilck on 20 Jul 2016

👍1

@mwilck
Awesome spot!

Would you mind create a pull request for your patch?

I found Synergy client would send clipboard to server when it detected clipboard changes and the cursor was not on client screen. So I reckon in the event queue something like this would cause race condition or even dead lock, clipboard changed -> leave (this is where client creates a separate thread to send clipboard data) -> clipboard grab (this is where clipboard sends newly detected clipboard on main thread). I'm not 100% sure if this is the cause. Please check my later comment to test the nightly builds.

XinyuHou on 20 Jul 2016

@mwilck @TafThorne @CyberKiller40 @ZippyManiac
Please test this nightly. This might not solve the problem but could help me narrow down the cause. Thanks.

http://symless.com/nightly?filter=fc98966

XinyuHou on 20 Jul 2016

@XinyuHou

Would you mind create a pull request for your patch?

Thought about it, but I didn't have time for that yet. For now, I prefer to wait if anyone from the freedesktop ML responds to my posting.

mwilck on 20 Jul 2016

@XinyuHou, do you have positive evidence that your change actually fixes the issues we were seeing?

Do you still want us to test the nightly build referenced above? If yes, if I understand correctly, I'd need to install the test package on both server and client? Should I run it with our without my libX11 patch?

Btw I ran synergys (Fedora 1.7.6-1) all day yesterday without a hang together with the libX11 patch. That's not a proof but it isn't bad either.

mwilck on 21 Jul 2016

Running this nightly: http://symless.com/nightly?filter=fc98966 with SSL
Linux Mageia 5 server, Windows 8.1 client.
See Debug1 Level logs attached for recent server stop.

Hans
Synergy_log_2016-07-21.zip

hscheff on 21 Jul 2016

@mwilck
Yes, we will soon push a nightly into our beta testers. Here is the link.
http://symless.com/nightly?filter=12e0120

Please test without the patch and create a pull request for our review

XinyuHou on 21 Jul 2016

Just happened again on 85bbbd4

Linux 16.04.1 server
Windows XP client

Server log: https://gist.github.com/sthor69/ce44898c01bc6dbd4194e0a41f918952
Client log: https://gist.github.com/sthor69/eba38836aa8c868eaa2aee9588d9e1d1

sthor69 on 21 Jul 2016

XinyuHou: do you provide linux .debuginfo packages for the nightlys? Just in case I encounter a hang again, I'd like to be able to debug it.

mwilck on 25 Jul 2016

Since I applied my libX11 patch, I have indeed not observed the original hang situation any more. But I have observed other situation where the sendClipboardThread was hung in XWindowsUtil::getCurrentTime.

Honestly, the implementation of XWindowsUtil::getCurrentTime seems awkward to me. I can see no guarantee whatsoever that the code will catch the event it is waiting for (the one it just generated before entering XIfEvent()). Maybe it would help to lock the display first, but why aren't you simply using gettimeofday()?

mwilck on 25 Jul 2016

Nope. Still happens on 1.8.2.RC1

server: Ubuntu 16.04 ('Kris-PC')
client: Ubuntu 16.04 ('Kris-Laptop')

client.log.txt
server.log.txt

Note that I do not have to restart client or server to reconnect. The only thing I need to do is click the 'Apply' button on the server and wait for a second or two for the client to reconnect.

Kris-LIBIS on 25 Jul 2016

@XinyuHuo: I observed a hang again with build 12e0120.
Attaching a DEBUG2 server log of the last seconds the server lived, and a gdb session log of the hang situation. You can see thread 2 again trapped in the GetCurrentTime->XIfEvent->_XreadEvents stack. Running the gdb command "finish" in XIfEvent() never completes. Thread 1 is in a different state this time as on my original report; it seems to be explicitly waiting for Thread 2.
synergys2.log.txt
synergys2_gdb.txt

mwilck on 27 Jul 2016

The last message in synergys2.log.txt "no clipboard to interrupt" looks suspicious.

mwilck on 27 Jul 2016

@XinyuHou, please consider the attached patch. It's untested, it just illustrates what I think could be done to avoid hangs in getCurrentTime().

mwilck on 27 Jul 2016

And here comes the attachment: getcurrenttime.diff.txt

mwilck on 27 Jul 2016

After the last installation of 85bbbd4, besides the caps-lock issue already mentioned in this thread, I also saw a strange behavior: pressing the tab key the order the items are selected in the screen is reversed. As an example, in a form on a webpage, the fields are selected from bottom to top

sthor69 on 28 Jul 2016

@XinyuHou , I have run synergys with my getCurrentTime() patch (see above) for 2 days now and haven't seen a single hang-up. That's strong evidence in favor of that patch, although no proof yet.

mwilck on 29 Jul 2016

Crash details follow. At the time of the crash, I had just copy-pasted something from somewhere (don't remember what precisely now).

Process: synergys [21108]
Path: /Applications/Synergy.app/Contents/MacOS/synergys
Identifier: synergy
Version: 0
Code Type: X86 (Native)
Parent Process: Synergy [552]
Responsible: synergys [21108]
User ID: 501

Date/Time: 2016-07-28 11:55:49.487 -0700
OS Version: Mac OS X 10.11.6 (15G7b)
Report Version: 11
Anonymous UUID: 6F090FE4-7AA9-BE1F-C847-344172499A72

Time Awake Since Boot: 430000 seconds

System Integrity Protection: disabled

Crashed Thread: 5

Exception Type: EXC_BAD_ACCESS (SIGSEGV)
Exception Codes: KERN_INVALID_ADDRESS at 0x0000000000000068
Exception Note: EXC_CORPSE_NOTIFY

VM Regions Near 0x68:
-->
__TEXT 00000000000d5000-000000000016c000 [ 604K] r-x/rwx SM=COW /Applications/Synergy.app/Contents/MacOS/synergys

Thread 5 Crashed:
0 synergys 0x00110e13 ClientProxy1_6::handleClipboardSendingEvent(Event const&, void) + 19
1 synergys 0x00111326 TMethodEventJob::run(Event const&) + 54
2 synergys 0x000e3b02 EventQueue::dispatchEvent(Event const&) + 98
3 synergys 0x000e2c8c EventQueue::loop() + 348
4 synergys 0x001329f4 App::runEventsLoop(void) + 20
5 synergys 0x00149cae TMethodJob::run() + 46
6 synergys 0x000f7b3c Thread::threadFunc(void) + 140
7 synergys 0x000dd21c ArchMultithreadPosix::doThreadFunc(ArchThreadImpl) + 76
8 synergys 0x000dc86b ArchMultithreadPosix::threadFunc(void*) + 75
9 libsystem_pthread.dylib 0x98820780 _pthread_body + 138
10 libsystem_pthread.dylib 0x988206f6 _pthread_start + 155
11 libsystem_pthread.dylib 0x9881df7a thread_start + 34

XinyuHou on 2 Aug 2016

May be this will help fix the problem?
https://github.com/symless/synergy/pull/4883

weduser on 3 Aug 2016

@mwilck Me and Jerry ( @XinyuHou ) have investigated this issue and agree with you on what is going on here. We are considering alternatives to just locking the X Server. This is old and bad code.

Do you have an update for us on the stability of the version you are running with your patch?

nlyan on 8 Aug 2016

This is not fixed yet. The ticket needs opening. I just hit it at fd3ce6c:

[2016-08-10T14:51:57] DEBUG2: can't read property 524 on window 0x04c00001
[2016-08-10T14:51:57] DEBUG2: can't read property 524 on window 0x04c00028
[2016-08-10T14:51:57] DEBUG2: can't read property 524 on window 0x04c00054
[2016-08-10T14:51:57] DEBUG2: can't read property 524 on window 0x042000c2
[2016-08-10T14:51:57] DEBUG2: can't read property 524 on window 0x03800004
[New Thread 0x7fff93f67700 (LWP 1060)]
[2016-08-10T14:51:57] DEBUG1: thread 0x000000a8 entry
[2016-08-10T14:51:57] DEBUG2: event: MotionNotify 2520,748
[2016-08-10T14:51:57] DEBUG2: event: MotionNotify 2586,739
[2016-08-10T14:51:57] DEBUG2: event: MotionNotify 2520,748
^C
Thread 1 "synergys" received signal SIGUSR1, User defined signal 1.
pthread_cond_wait@@GLIBC_2.3.2 () at ../sysdeps/unix/sysv/linux/x86_64/pthread_cond_wait.S:185
185 62: movl    (%rsp), %edi
(gdb) where
#0  pthread_cond_wait@@GLIBC_2.3.2 ()
    at ../sysdeps/unix/sysv/linux/x86_64/pthread_cond_wait.S:185
#1  0x00007ffff701a430 in _XReply (dpy=dpy@entry=0x6f5010, rep=rep@entry=0x7fffffffd2b0, 
    extra=extra@entry=0, discard=discard@entry=1) at xcb_io.c:624
#2  0x00007ffff7015cf0 in XSync (dpy=0x6f5010, discard=0) at Sync.c:44
#3  0x0000000000465f0c in XWindowsUtil::ErrorLock::install(void (*)(_XDisplay*, XErrorEvent*, void*), void*) ()
#4  0x0000000000465fc0 in XWindowsUtil::getWindowProperty(_XDisplay*, unsigned long, unsigned long, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, unsigned long*, int*, bool) ()
#5  0x0000000000470ead in XWindowsScreenSaver::isXScreenSaver(unsigned long) const ()
#6  0x0000000000471c40 in XWindowsScreenSaver::handleXEvent(_XEvent const*) ()
#7  0x0000000000462ddd in XWindowsScreen::handleSystemEvent(Event const&, void*) ()
#8  0x000000000044a06a in EventQueue::dispatchEvent(Event const&) ()
#9  0x000000000044bf5c in EventQueue::loop() ()
#10 0x000000000045764a in ServerApp::mainLoop() ()
#11 0x00000000004551ca in ServerApp::runInner(int, char**, ILogOutputter*, int (*)(int, char**))
    ()
#12 0x000000000045c679 in App::run(int, char**) ()
#13 0x000000000044107f in main ()

brianjmurrell on 10 Aug 2016

Can we get this ticket reopened please, due to the comment above?

brianjmurrell on 11 Aug 2016

👍2

Unfortunately, I was testing on the wrong commit. I asked 3 days ago in #5447 when that was closed as a duplicate of this ticket what I should check out and build to test if this issue is gone and there was no answer. At all. All I had to go on as to what to build and test was there was a v1.8.2 branch but strangely, all of the fixes intended for 1.8.2 weren't actually on the v1.8.2 branch. And there was no communication about that either.

If somebody had just simply answered my question then all of this confusion could have been avoided and my time testing the wrong commit for the last 3 days would not have been wasted.

It does also strike me odd that a branch named for a release not actually contain all of the fixes for that release.

Some better communication with community would certainly help this project I think.

brianjmurrell on 11 Aug 2016

@brianjmurrell

Sorry for the late reply.

We made some changes before we released 1.8.2. Please checkout our 1.8.2 tag or master. With the last change, 1.8.2 should solve the issue you reported. Please let us know if the problem persists.

XinyuHou on 12 Aug 2016

I should add here that I've been running the native package synergy-1.7.6-1.fc24 with native libX11 (i.e. neither my libX11 patch from July 20, nor my synergy patch from July 27 applied) since my update to Fedora 24 on Aug 11th, and haven't encountered a single hangup as I had described previously.
I have no idea why this is so, as there is no indication whatsoever that the installed packages would contain appropriate fixes for the problems I had encountered previously. But that's how it is sometimes - after the F24 update the problem isn't reproducible (by me) any more.

mwilck on 28 Aug 2016

Sorry for the delayed response. I was away from the office for most of August.
I did manage to download and install Synergy 1.8.2-stable-36cd521, so some feedback is due.

So far I have not experienced the 'dead server' issue on copy and paste running this version! :)
Thanks guys!

Copy and paste still fails occasionally from the client to the server, but all other functionality keeps working. A stop and start of Synergy on the client restores copy & paste functionality.
I have not had an opportunity to gather log files when this happens, but will do so if it happens at a "convenient time".

Something else that I noticed, is that the build date reported in the Synergy About window is always the current date instead of the build date.

Regards
Hans

hscheff on 30 Aug 2016

Something else that I noticed, is that the build date reported in the
Synergy About window is always the current date instead of the build date.
Well spotted. I have just checked and my 1.8.2-stable-36cd521 install
is also presenting with today's date.

On 30/08/16 16:39, Hans wrote:

Sorry for the delayed response. I was away from the office for most of
August.
I did manage to to download and install Synergy 1.8.2-stable-36cd521,
so some feedback is due.

So far I have not experienced the 'dead server' issue on copy and
paste running this version! :)
Thanks guys!

Copy and paste still fails occasionally from the client to the server,
but all other functionality keeps working. A stop and start of Synergy
on the client restores copy & paste functionality.
I have not had an opportunity to gather log files when this happens,
but will do so if it happens at a "convenient time".

Something else that I noticed, is that the build date reported in the
Synergy About window is always the current date instead of the build date.

Regards
Hans

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/symless/synergy/issues/4768#issuecomment-243482381,
or mute the thread
https://github.com/notifications/unsubscribe-auth/AAjzNpqQ3f_kAYxlz-YQzCtJiC3bbQjpks5qlE6fgaJpZM4E5qJ-.

Thomas A. F. Thorne [email protected]

ASCII ribbon campaign ( )

against HTML email X
& vCards / \
--

TafThorne on 30 Aug 2016

I'm having this exact same issue. I'm running 1.8.5 both on linux server and windows client.

I can reproduce it 100% of the time, just need to do a screenshot and copy to the clipboard.
Here is debug2 output from the time of the event. Neither the server nor client crash, just hangs. I have to ssh in and SIGKILL the server to get control again.

[2016-12-02T11:52:20] DEBUG2: find neighbor on up of "arch" /build/synergy/src/synergy-1.8.5-stable/src/lib/server/Server.cpp,614 [2016-12-02T11:52:20] DEBUG2: "wintendo" is on up of "arch" at 0.471558 /build/synergy/src/synergy-1.8.5-stable/src/lib/server/Server.cpp,637 [2016-12-02T11:52:20] DEBUG1: try to leave "arch" on up /build/synergy/src/synergy-1.8.5-stable/src/lib/server/Server.cpp,811 [2016-12-02T11:52:20] DEBUG2: mapping state: 16 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsKeyState.cpp,115 [2016-12-02T11:52:20] DEBUG2: |= modifier: 4 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsKeyState.cpp,120 [2016-12-02T11:52:20] INFO: switch from "arch" to "wintendo" at 1931,2159 /build/synergy/src/synergy-1.8.5-stable/src/lib/server/Server.cpp,472 [2016-12-02T11:52:20] INFO: leaving screen /build/synergy/src/synergy-1.8.5-stable/src/lib/synergy/Screen.cpp,131 [2016-12-02T11:52:20] DEBUG2: grabbed keyboard /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsScreen.cpp,2021 [2016-12-02T11:52:20] DEBUG1: grabbed pointer and keyboard /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsScreen.cpp,2040 [2016-12-02T11:52:20] DEBUG2: warped to 2048,1080 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsScreen.cpp,1960 [2016-12-02T11:52:20] DEBUG2: mapping state: 16 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsKeyState.cpp,115 [2016-12-02T11:52:20] DEBUG2: |= modifier: 4 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsKeyState.cpp,120 [2016-12-02T11:52:20] DEBUG1: modifiers on update: 0x2000 /build/synergy/src/synergy-1.8.5-stable/src/lib/synergy/KeyState.cpp,519 [2016-12-02T11:52:20] DEBUG: open clipboard 0 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,323 [2016-12-02T11:52:20] DEBUG1: locked motif clipboard /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,631 [2016-12-02T11:52:20] DEBUG2: can't read property 656 on window 0x000001dc /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsUtil.cpp,1378 [2016-12-02T11:52:20] DEBUG1: motif does not own clipboard /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,337 [2016-12-02T11:52:20] DEBUG1: unlocked motif clipboard /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,638 [2016-12-02T11:52:20] DEBUG1: request selection=CLIPBOARD (451), target=TIMESTAMP (460), window=6600004 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,1287 [2016-12-02T11:52:20] DEBUG2: read property 530 on window 0x06600004: bytes=4 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsUtil.cpp,1374 [2016-12-02T11:52:20] DEBUG1: target INTEGER (19) /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,1481 [2016-12-02T11:52:20] DEBUG1: got data, 4 bytes /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,1487 [2016-12-02T11:52:20] DEBUG1: request succeeded after 0.000002s /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,1367 [2016-12-02T11:52:20] DEBUG1: got ICCCM time 231020889 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,599 [2016-12-02T11:52:20] DEBUG: ICCCM fill clipboard 0 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,503 [2016-12-02T11:52:20] DEBUG1: request selection=CLIPBOARD (451), target=TARGETS (459), window=6600004 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,1287 [2016-12-02T11:52:20] DEBUG2: read property 530 on window 0x06600004: bytes=40 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsUtil.cpp,1374 [2016-12-02T11:52:20] DEBUG1: target ATOM (4) /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,1481 [2016-12-02T11:52:20] DEBUG1: got data, 40 bytes /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,1487 [2016-12-02T11:52:20] DEBUG1: request succeeded after 0.000002s /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,1367 [2016-12-02T11:52:20] DEBUG: available targets: TARGETS (459), MULTIPLE (456), image/jpeg (649), image/tiff (640), image/x-MS-bmp (639) /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,522 [2016-12-02T11:52:20] DEBUG1: request selection=CLIPBOARD (451), target=text/html (597), window=6600004 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,1287 [2016-12-02T11:52:20] DEBUG2: can't read property 530 on window 0x06600004 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsUtil.cpp,1378 [2016-12-02T11:52:20] DEBUG1: request failed after 0.000003s /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,1367 [2016-12-02T11:52:20] DEBUG1: can't get data for selection target text/html (597) /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,580 [2016-12-02T11:52:20] DEBUG1: no data for target text/html (597) /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,557 [2016-12-02T11:52:20] DEBUG1: request selection=CLIPBOARD (451), target=image/bmp (637), window=6600004 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsClipboard.cpp,1287 [2016-12-02T11:52:20] DEBUG2: read property 530 on window 0x06600004: bytes=4 /build/synergy/src/synergy-1.8.5-stable/src/lib/platform/XWindowsUtil.cpp,1374 [2016-12-02T11:52:28] DEBUG2: reading secure socket /build/synergy/src/synergy-1.8.5-stable/src/lib/net/SecureSocket.cpp,254 [2016-12-02T11:52:28] DEBUG: ssl connection closed /build/synergy/src/synergy-1.8.5-stable/src/lib/net/SecureSocket.cpp,548 [2016-12-02T11:52:28] DEBUG1: registered event type stopRetry as 48 /build/synergy/src/synergy-1.8.5-stable/src/lib/base/EventQueue.cpp,142 [2016-12-02T11:52:28] DEBUG1: registered event type disconnected as 49 /build/synergy/src/synergy-1.8.5-stable/src/lib/base/EventQueue.cpp,142 [2016-12-02T11:52:38] ERROR: process exited with error code: 9

discoltk on 2 Dec 2016

👍1

I see the same behavior as @discoltk running ubuntu 14.04 linux as server, gnome ubuntu 16.04 as client. If I copy an image to clipboard, like from a screenshot app, as soon as I drag my mouse over to the other system, mouse locks up. need to ssh in to server and kill server process to get mouse back. running latest stable versions, 1.8.7

joeroback on 30 Jan 2017

@joeroback Not sure if it's going to actually get attention, but there's another issue open for clipboard issues and it's at least marked as priority. Really hope it does as having to remember to disable synergy when I need to screenshot something is a pain!

https://github.com/symless/synergy/issues/5786

discoltk on 31 Jan 2017

I'm having the same issue. I've already wasted a lot of my morning trying to debug this with no success. My setup is an Ubuntu 16.04 Server and an OSX Sierra client. The OSX client dies VERY frequently. However, it appears to be exiting gracefully, which makes this especially puzzling. Looking at the packets, the server does appear to send a [FIN, ACK] packet as this happens, followed by some RST packets before the client is determined dead and restarted on the OSX laptop. In this setup I had the machines directly connected to each other (i.e., Ethernet cable directly connected to both). I've tried with/without SSL, with/without clipboard, and with/without drag-drop. All I have to do to trigger this, is start interacting with the client. It will crash within 2 minutes every time.

Server (Ubuntu 16.04 / synergy-master-stable-a5140aa)

[2017-02-01T11:09:12] DEBUG1: got ICCCM time 158565327
[2017-02-01T11:09:12] DEBUG: close clipboard 1
[2017-02-01T11:09:12] DEBUG: ignored screen "Shortman" update of clipboard 1 (unchanged)
[2017-02-01T11:09:12] DEBUG1: send enter to "Chads-MacBook-Pro.local", -1920,291 57 0000
[2017-02-01T11:09:13] DEBUG1: try to leave "cspensky-mbp" on left
[2017-02-01T11:09:13] INFO: switch from "cspensky-mbp" to "Shortman" at 3797,554
[2017-02-01T11:09:13] DEBUG1: send leave to "Chads-MacBook-Pro.local"
[2017-02-01T11:09:13] INFO: entering screen
[2017-02-01T11:09:19] NOTE: client "Chads-MacBook-Pro.local" is dead
[2017-02-01T11:09:22] INFO: OpenSSL 1.0.2g  1 Mar 2016

Client (OSX Sierra / synergy-master-stable-a5140aa)

...
[2017-02-01T11:09:16] DEBUG2: want to read, error=2, attempt=1
[2017-02-01T11:09:19] DEBUG2: reading secure socket
[2017-02-01T11:09:19] DEBUG2: reading secure socket
[2017-02-01T11:09:19] DEBUG2: want to read, error=2, attempt=1
[2017-02-01T11:09:19] DEBUG2: reading secure socket
[2017-02-01T11:09:19] DEBUG: ssl connection closed
[2017-02-01T11:09:21] DEBUG2: writef(CNOP)
[2017-02-01T11:09:21] DEBUG2: wrote 4 bytes
[2017-02-01T11:09:21] DEBUG2: writing secure socket:0x7a130310
[2017-02-01T11:09:21] WARNING: cursor may not be visible
[2017-02-01T11:09:21] WARNING: cursor may not be visible
[2017-02-01T11:09:21] DEBUG2: reading secure socket
[2017-02-01T11:09:21] DEBUG: ssl connection closed
[2017-02-01T11:09:21] DEBUG2: msg from server: DMMV
...
[2017-02-01T11:09:21] DEBUG2: msg from server: CALV
[2017-02-01T11:09:21] DEBUG2: writef(CALV)
[2017-02-01T11:09:21] DEBUG2: wrote 4 bytes
[2017-02-01T11:09:21] DEBUG2: writef(CNOP)
[2017-02-01T11:09:21] DEBUG2: wrote 4 bytes
[2017-02-01T11:09:21] DEBUG: showing cursor
[2017-02-01T11:09:21] INFO: process exited normally
[2017-02-01T11:09:21] INFO: detected process not running, auto restarting
[2017-02-01T11:09:22] DEBUG: starting process
[2017-02-01T11:09:22] INFO: starting client

UPDATE: Rebooting the client machine seemed to solve the problem. Still no clue why it was happening in the first place.

cspensky on 1 Feb 2017

I found this bug still happens on synergy 1.10.3-stable-ca35737a, Server OS is Win10 ver:18362.30 , Client OS is Ubuntu18.04.3.

It happens only on cllient ,but Not only use copy-paste. Sometimes when I use alt+tab to switch apps, it also happened. Just like the keys are pressed and hold on.

When it happend ,Client Log is like this:

[2019-11-03T21:09:20] DEBUG2: readf: read 1 byte integer: 1 (0x1)
[2019-11-03T21:09:20] DEBUG1: recv mouse down id=1
[2019-11-03T21:09:20] DEBUG2: writef(CNOP)
[2019-11-03T21:09:20] DEBUG2: wrote 4 bytes
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00014
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00014
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00014
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00014
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00014
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00014
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00026
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00026
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00026
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00026
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00026
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00026
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00026
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00026
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00026
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00014
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x00a00003
[2019-11-03T21:09:20] DEBUG2: can't read property 569 on window 0x03a00014
[2019-11-03T21:09:20] DEBUG2: msg from server: DMMV
[2019-11-03T21:09:20] DEBUG2: readf(%2i%2i)
[2019-11-03T21:09:20] DEBUG2: readf: read 2 byte integer: 2824 (0xb08)
[2019-11-03T21:09:20] DEBUG2: readf: read 2 byte integer: 1782 (0x6f6)
[2019-11-03T21:09:20] DEBUG2: recv mouse move 2824,1782
[2019-11-03T21:09:20] DEBUG2: writef(CNOP)
[2019-11-03T21:09:20] DEBUG2: wrote 4 bytes
[2019-11-03T21:09:20] DEBUG2: msg from server: DMMV

zlphoenix on 3 Nov 2019

Synergy-core: Copy-paste causes 'server is dead' error on switching

Most helpful comment

All 186 comments

Thank you.

Client Log

Server Log

Activity At Time Of Problem

Thomas A. F. Thorne [email protected]

Related issues