Deprecated: Methods with the same name as their class will not be constructors in a future version of PHP; ADOConnection has a deprecated constructor in /srv/http/bugs/adodb/adodb.inc.php on line 237 Deprecated: Methods with the same name as their class will not be constructors in a future version of PHP; ADORecordSet has a deprecated constructor in /srv/http/bugs/adodb/adodb.inc.php on line 2714 Deprecated: Methods with the same name as their class will not be constructors in a future version of PHP; ADORecordSet_array has a deprecated constructor in /srv/http/bugs/adodb/adodb.inc.php on line 3719 Deprecated: Methods with the same name as their class will not be constructors in a future version of PHP; Project has a deprecated constructor in /srv/http/bugs/includes/class.project.php on line 3 Deprecated: Methods with the same name as their class will not be constructors in a future version of PHP; Url has a deprecated constructor in /srv/http/bugs/includes/class.tpl.php on line 825 FS#2 : console-client locking after inactivity
Please read the Arch Linux guidelines before reporting a bug, they're very good:
http://wiki.archlinux.org/index.php/Reporting_Bug_Guidelines

Do NOT report bugs when a package is just outdated. Use the "Mark out of date" link on the package page, or the mailing list.
Tasklist

FS#2 - console-client locking after inactivity

Attached to Project: Arch Hurd
Opened by S├ębastien Duquette (ekse) - 2010-04-22 23:14:01
Last edited by Matthias Lanzinger (melpo) - 2011-08-15 14:34:01
Task Type Bug Report
Category Backend / Core
Status Researching
Assigned To No-one
Operating System All
Severity High
Priority Normal
Reported Version 1.0
Due in Version Undecided
Due Date Undecided
Percent Complete 0%
Votes 1
Private No

Details

The system is becoming unresponsive after a certain amount of inactivity. Currently, a work-around is to run the "vmstat 5" command in a terminal to generate activity.
This task depends upon

Comment by Matthias Lanzinger (melpo) - 2010-04-25 18:38:36
This issue has been reported on different Virtual Machines but does not seem to exist when running on a real machine.
Comment by Diego Martin Nieto Cid (diego) - 2010-07-28 22:51:24
Yes it does block on real machines. (well.. mine at least :( )
Comment by Diego Martin Nieto Cid (diego) - 2010-07-29 18:26:44
The serial console* doesn't block.

* http://article.gmane.org/gmane.linux.debian.ports.hurd/5320
(in quemu Ctrl+Alt+2 shows the serial line on screen)
Comment by Diego Martin Nieto Cid (diego) - 2010-07-29 22:26:52
By killing the console process the machine, now at the mach console, starts reacting to the keyboard. However, the screen is no longer refreshed.

Pretty useless, but at least it's possible to use the machine by heart. :P
Comment by Matthias Lanzinger (melpo) - 2010-08-13 16:11:46
In my opinion it looks like the console drivers (pc_kbd, vga) stop reacting.
Even when the system locks I can perfectly continue using it via ssh.
Any ideas on how to debug this?
Comment by Diego Martin Nieto Cid (diego) - 2010-08-13 18:28:52
If my last comment were reproducible I'd say vga is causing this. But it isn't :P

I tried using gdb from the serial console but after telling gdb to continue, the console hangs. Maybe through ssh it works fine.

Here are some notes about the hurd console: http://wiki.github.com/diegonc/arch-hurd-xkb-driver/notes-hurd-console

Unfortunately I couldn't complete the data flow. I fear glibc is in the middle. :(

BTW, Debian doesn't appear to block *shrug*.
Comment by Diego Martin Nieto Cid (diego) - 2010-08-13 23:38:48
For some reason the thread started by the input driver is dying.

* http://archhurd.pastebin.com/MaEKtjz8
* http://archhurd.pastebin.com/75dHpg1a
Comment by Alexander Preisinger (Giselher) - 2010-08-26 12:33:00
It seems to be better now with the new packages.
Comment by Matthias Lanzinger (melpo) - 2010-11-21 16:07:22
I don't get any more freezes with hurd-20101030-2 (which includes Sergio Lopez' tmpfs patches from bug-hurd)
Can anybody confirm this?
Comment by Matthias Lanzinger (melpo) - 2011-08-15 14:34:01
  • Field changed: Summary (system locking after inactivity → console-client locking after inactivity)
changed name to better explain the actual problem. The system continues to run fine, only the console-client freezes
Comment by Diego Martin Nieto Cid (diego) - 2011-11-03 00:01:37
Hi,

There's a patch[1] Debian applies to Hurd which prevents the locking. Also, I sent another patch[2] to bug-hurd to fix just this problem. But other issues in libports may remain.

So I guess it's safer to follow Debian here. At least until it's reviewed.


[1] http://patch-tracker.debian.org/patch/series/view/hurd/20110519-3/libports_stability.patch
[2] http://lists.gnu.org/archive/html/bug-hurd/2011-10/msg00087.html
Comment by henrique conzatti (conzatti) - 2012-06-27 18:00:58
yes, there is this in real machine, at least in my machine.

Loading...