Bug 516274
Summary: | Keyboard does not work | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | Alexey Kuznetsov <axet> | ||||||||
Component: | tigervnc | Assignee: | Adam Tkac <atkac> | ||||||||
Status: | CLOSED ERRATA | QA Contact: | Fedora Extras Quality Assurance <extras-qa> | ||||||||
Severity: | medium | Docs Contact: | |||||||||
Priority: | low | ||||||||||
Version: | 11 | CC: | atkac, ovasik, sascha-web-bugzilla.redhat.com | ||||||||
Target Milestone: | --- | ||||||||||
Target Release: | --- | ||||||||||
Hardware: | All | ||||||||||
OS: | Linux | ||||||||||
Whiteboard: | |||||||||||
Fixed In Version: | 1.0.0-3.fc12 | Doc Type: | Bug Fix | ||||||||
Doc Text: | Story Points: | --- | |||||||||
Clone Of: | Environment: | ||||||||||
Last Closed: | 2009-10-27 06:34:22 UTC | Type: | --- | ||||||||
Regression: | --- | Mount Type: | --- | ||||||||
Documentation: | --- | CRM: | |||||||||
Verified Versions: | Category: | --- | |||||||||
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||||||
Cloudforms Team: | --- | Target Upstream Version: | |||||||||
Embargoed: | |||||||||||
Attachments: |
|
Description
Alexey Kuznetsov
2009-08-07 18:04:20 UTC
Would it be possible to attach the server log, please (located in ~/.vnc/*.log file). Thanks. Created attachment 356862 [details]
mini:1.log
yep.
Are you still able to reproduce this issue? I tested fully up2date F11 powerpc and everything works fine. If it still doesn't work in your case please append "-log *:stderr:100" to Xvnc parameters and then attach vnc log again, please. Created attachment 363363 [details]
mini:1.log
yep, problem still here.
Interesting. Could you please tell me which client are you using? Sure. I tried booth, vinagre, and vncview (last one started from remote station (mini.local) with X forwarding). Same issue. [axet@axet-laptop ~]$ ssh -X mini.local reverse mapping checking getaddrinfo for mini.tax.local [192.168.54.3] failed - POSSIBLE BREAK-IN ATTEMPT! Last login: Thu Oct 1 19:40:10 2009 from 192.168.54.5 [axet@mini ~]$ vncviewer localhost:1 It's weird, from log it seems your server receives no key event from client. More interesting lines are: 01/10/2009 07:41:56 PM rfbProcessClientNormalMessage: ignoring unknown encoding type 7 01/10/2009 07:41:56 PM rfbProcessClientNormalMessage: ignoring unknown encoding type -258 01/10/2009 07:41:56 PM Enabling NewFBSize protocol extension for client ::ffff:192.168.54.5 01/10/2009 07:41:56 PM rfbProcessClientNormalMessage: ignoring unknown encoding type 1464686185 01/10/2009 07:41:56 PM rfbProcessClientNormalMessage: ignoring unknown encoding type -257 It seems that you are not connecting to Xvnc but to vino server. Could you please check that you are really connecting to Xvnc? For example, when keyboard doesn't work when you start vncviewer with "localhost:1" parameter, run `netstat -apn |grep Xvnc` and check where Xvnc is listenning. If you are connecting to vino server please go to system->preferences->remote desktop and disable it. Or simply connect to Xvnc, not to vino server. Does keyboard work now? Yes. When i did connect to vncserver (localhost:1) i got keyboard working. But vncserver miss mdns extension, which allow to broadcast server address. I prefer to use vino. When i did connect to vncviewer localhost:0 (vncserver) it wont to recognize my keyboard and i cant type anyting. Same issue with OSX native VNC viewer. After inspection this is really a bug in Xvnc but it's very hard to fix. Basically it is more design bug than a code bug. I found a temporary workaround which works: 1. connect to Xvnc with any viewer 2. press any key 3. restart vino server (make sure vino process is really restarted) 4. connect to vino server Don't expect fix for this issue soon, it probably won't be fixed in TigerVNC 1.0 series at all, not sure yet. interesting. how can i restart vino server? 1. go to preferences->remote desktop and disable it 2. check that vino-server is really down (killall vino-server, for example) 3. go to preferences->remote desktop and enable it Created attachment 364724 [details]
Simple example to show the bug
This script fetches the keymap from the VNC server using XGetKeyboardMapping and prints the first 20 values. On a server affected by this bug the output will be all-zero.
While trying to get Sugar to run inside TigerVNC, I hit the same issue. For us the symptoms are that the hotkeys are not working because no key names can be parsed (because the keymap is empty). I've attached a small script to reproduce the issue. On RealVNC (Debian) it works fine: [array('L', [0L, 0L]), array('L', [65307L, 0L]), array('L', [49L, 33L]), array('L', [50L, 64L]), array('L', [51L, 35L]), array('L', [52L, 36L]), array('L', [53L, 37L]), array('L', [54L, 94L]), array('L', [55L, 38L]), array('L', [56L, 42L]), array('L', [57L, 40L]), array('L', [48L, 41L]), array('L', [45L, 95L]), array('L', [61L, 43L]), array('L', [65288L, 0L]), array('L', [65289L, 0L]), array('L', [113L, 81L]), array('L', [119L, 87L]), array('L', [101L, 69L]), array('L', [114L, 82L])] On TigerVNC (Fedora) we'll get an empty keymap until a key gets pressed: [array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L]), array('I', [0L, 0L, 0L, 0L])] tigervnc-1.0.0-2.fc11 has been pushed to the Fedora 11 stable repository. If problems still persist, please make note of it in this bug report. tigervnc-1.0.0-3.fc12 has been submitted as an update for Fedora 12. http://admin.fedoraproject.org/updates/tigervnc-1.0.0-3.fc12 tigervnc-1.0.0-3.fc12 has been pushed to the Fedora 12 stable repository. If problems still persist, please make note of it in this bug report. |