[bug] Inter-VM Copy/Paste Broken After Long Uptime

[Andrew]

Hello,

I have one particular Qubes instance with an uptime >60 days. Aside
from needing to occasionally restart Qubes Manager to deal with the
memory leak problem, this works fine.

However, recently, I have become unable to copy and paste between VMs.
Pressing the key combination appears to do nothing. FWIW,
/var/run/qubes/qubes-clipboard.bin and
/var/run/qubes/qubes-clipboard.bin.source are both 0-sized files
(presumably) dated to back when inter-VM copy/paste stopped working.

Any suggestions on how to debug?

Andrew

[Drew White]

What version of Qubes?

[Drew White]

On Tuesday, 19 July 2016 23:13:28 UTC+10, Andrew wrote:

By the way,m the system isn't designed to be up for that amount of time, or so I've been told recently.
Mine normally breaks after a week. Then I have to reboot.
It typicaly says it's a permisison issue with the file, but ssetting the permisisons doesn't fix it, no matter what you do it can't seem to access the files.. so only a reboot works.

Qubes, they say, is designed for "normal users" not people like us who don't turn off our computers.

[Marek Marczykowski-Górecki]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Yes, remove /var/run/qubes/qubes-clipboard.bin.xevent

The file contains X event timestamp of last copy/paste operation (to
avoid some race conditions). But this counter overflows after some time
(a month or so) and our code do not handle this case.

- --
Best Regards,
Marek Marczykowski-Górecki
Invisible Things Lab
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2

iQEcBAEBCAAGBQJXjoVrAAoJENuP0xzK19csJOwH/iXTl3LwfmhG756g8IwhBSs8
uaoSAh3LIU46P1Fjy0f4+0hZnl6EHfvmKY0YkcX2Kq6ZqBlHXdfkcooCGGqLYilX
5NKD5hmGcVPtAcgZ1eRADXxyUBoMgeCEoQo95M7/Y4dQFsvJbp+pyX+zASRL9YZf
G7LN9yRG29EjRnWSi2briSKYyoTP4Wn3euF4USlPZ67k01uVNLigCcurzPGKw2HT
z+UHtqMPJYFQ/eajJBjENwyozLRRNa3G2wgte7M8eU3rWAbQmIeJlY6QLo7E4Tba
oYve8pRem8/vD4mIKBavfc6wzaG9MvdpUlc3iPWt9+W3FEXbk5rLkBzI0Y9/mhw=
=qQCd
-----END PGP SIGNATURE-----

[Andrew]

Marek Marczykowski-Górecki:

> On Tue, Jul 19, 2016 at 01:13:22PM +0000, Andrew wrote:
>> Hello,
>
>> I have one particular Qubes instance with an uptime >60 days.
>> Aside from needing to occasionally restart Qubes Manager to deal
>> with the memory leak problem, this works fine.
>
>> However, recently, I have become unable to copy and paste between
>> VMs. Pressing the key combination appears to do nothing. FWIW,
>> /var/run/qubes/qubes-clipboard.bin and
>> /var/run/qubes/qubes-clipboard.bin.source are both 0-sized files
>> (presumably) dated to back when inter-VM copy/paste stopped
>> working.
>
>> Any suggestions on how to debug?
>
> Yes, remove /var/run/qubes/qubes-clipboard.bin.xevent
>
> The file contains X event timestamp of last copy/paste operation
> (to avoid some race conditions). But this counter overflows after
> some time (a month or so) and our code do not handle this case.
>
>

I see, thanks.

Andrew

[Holger Levsen]

Hi,

On Tue, Jul 19, 2016 at 09:54:19PM +0200, Marek Marczykowski-Górecki wrote:
> > I have one particular Qubes instance with an uptime >60 days. Aside

[...]

> > Any suggestions on how to debug?
>
> Yes, remove /var/run/qubes/qubes-clipboard.bin.xevent
>
> The file contains X event timestamp of last copy/paste operation (to
> avoid some race conditions). But this counter overflows after some time
> (a month or so) and our code do not handle this case.

Is this a known issue tracked in github? If not I think it should be…

Also, are there other known problems with running Qubes with a such a
"long" uptime?


--
cheers,
Holger

[Manuel Amador (Rudd-O)]

My machine becomes slower and slower, but that may be an artifact of how
ZFS cache fragmentation affects the system and its interplay with memory
requests from VMs.


--
Rudd-O
http://rudd-o.com/

[Andrew David Wong]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

On 2016-07-20 01:23, Holger Levsen wrote:
> Hi,
>
> On Tue, Jul 19, 2016 at 09:54:19PM +0200, Marek Marczykowski-Górecki
> wrote:
>>> I have one particular Qubes instance with an uptime >60 days. Aside
> [...]
>>> Any suggestions on how to debug?
>>
>> Yes, remove /var/run/qubes/qubes-clipboard.bin.xevent
>>
>> The file contains X event timestamp of last copy/paste operation (to
>> avoid some race conditions). But this counter overflows after some time
>> (a month or so) and our code do not handle this case.
>
> Is this a known issue tracked in github? If not I think it should be…
>

Thanks. Tracking:

https://github.com/QubesOS/qubes-issues/issues/2189

> Also, are there other known problems with running Qubes with a such a
> "long" uptime?
>

This and the Qubes Manager memory leak are the only two I know of.

https://github.com/QubesOS/qubes-issues/issues/860

(Note that the latter was recently fixed, but the package is in the R3.2
dom0 testing repo.)

- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----
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=qHcC
-----END PGP SIGNATURE-----

[Drew White]

On Thursday, 21 July 2016 08:28:29 UTC+10, Andrew David Wong wrote:
> > Also, are there other known problems with running Qubes with a such a
> > "long" uptime?
> >
>
> This and the Qubes Manager memory leak are the only two I know of.
>
> https://github.com/QubesOS/qubes-issues/issues/860
>
> (Note that the latter was recently fixed, but the package is in the R3.2
> dom0 testing repo.)

"Testing REPO" != "fixed"
"Testing REPO" == "is it fixed?"

I hope that it does get fixed sometime soon.