FS#46550 - [lirc] remotes are too sensitive under 0.9.3

Attached to Project: Arch Linux
Opened by John (graysky) - Saturday, 03 October 2015, 14:50 GMT
Last edited by Doug Newgard (Scimmia) - Saturday, 07 November 2015, 14:54 GMT
Task Type Bug Report
Category Upstream Bugs
Status Closed
Assigned To Lukas Fleischer (lfleischer)
Architecture All
Severity Low
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 3
Private No

Details

Upon updating to 1:0.9.3-2, my streamzap remote[1] has become too sensitive. A key press is often echoed twice to applications or more simplistically, to /usr/bin/irw. Under previous versions, it worked flawlessly. Downgrading to 1:0.9.2a-1 (which I have rebuilt from ABS) solves the problem.

I am unsure if this is an upstream bug or something unique to our package, but at least one other user with a different remote/hardware has confirmed this bug and that downgrading fixes it[2]. I have posted to the lirc ML[3] but haven't gotten much back.

1. http://www.streamzap.com/consumer/pc_remote/index.php
2. https://bbs.archlinux.org/viewtopic.php?id=203177
3. https://sourceforge.net/p/lirc/mailman/message/34506523

Steps to reproduce:
1) Install lirc package and configure remote (mine is using https://github.com/graysky2/streamzap/blob/master/00-streamzap.conf)
2) Start lircd
2) Run irw and hit keys. Notice duplicated entries.
This task depends upon

Closed by  Doug Newgard (Scimmia)
Saturday, 07 November 2015, 14:54 GMT
Reason for closing:  Fixed
Additional comments about closing:  lirc 1:0.9.3.a-1
Comment by Jon Bergli Heier (snakebite) - Saturday, 03 October 2015, 15:11 GMT
Having the same issue here. According to git bisect commit 78a2e110a355849d7fa0a51225e26ae9501d3285 is the bad commit. Sounds like an upstream bug.
Comment by Andrew Gaydenko (student975) - Saturday, 03 October 2015, 15:15 GMT
https://bugs.archlinux.org/task/46457 - probably is related.
Comment by John (graysky) - Saturday, 03 October 2015, 18:44 GMT
@Jon - I reverted that commit against 0.9.3 but I am still observing the sensitivity issue.
Comment by patrick (potomac) - Saturday, 03 October 2015, 21:04 GMT
@graysky : you can try to create a new config file by using irrecord and see if the bug is still here
Comment by John (graysky) - Saturday, 03 October 2015, 23:40 GMT
@patrick - I believe this is completely independent of the conf file.
Comment by Jon Bergli Heier (snakebite) - Sunday, 04 October 2015, 09:03 GMT
Seconded, I tried generating a new config file yesterday, but still had the same issue. I also tried emulating a different remote (I have a Harmony 200) but no go.
Comment by Lukas Fleischer (lfleischer) - Sunday, 04 October 2015, 09:11 GMT
Did any of you file a upstream report? If yes, could you please provide a link?
Comment by Andrew Gaydenko (student975) - Sunday, 04 October 2015, 09:15 GMT
@lfleischer, probably one can try to check a state of the current upstream repository if the issue is already fixed. I'm ready to try, but can not prepare an appropriate PKGBUILD.
Comment by John (graysky) - Sunday, 04 October 2015, 09:54 GMT
Upstream bug: https://sourceforge.net/p/lirc/tickets/148
Comment by John (graysky) - Sunday, 04 October 2015, 12:45 GMT
@Lukas - It seems from Alec (Lirc dev) that the offending commit was added to allow older kernels (early 4.1 series) to work properly. I have applied the attached to our 1:0.9.3-2 and it solved the double-keypress for me.
   revert_219b01.patch (1.1 KiB)
Comment by Jon Bergli Heier (snakebite) - Sunday, 04 October 2015, 19:37 GMT
Apparently on my end the issue was caused of a different bug, which is reported (with patch, accept pending) upstream:

https://sourceforge.net/p/lirc/tickets/149/

I don't know how https://bugs.archlinux.org/task/46457 relates, maybe it's the same bug I'm seeing? (It's not quite clear to me from the description exactly what the problem there is)
Comment by John (graysky) - Monday, 05 October 2015, 20:03 GMT
Please disregard my comment/suggestion above about patching our package as a long term solution (I have edited out the comment). I think it's prudent to patch our package now, but long-term, lirc upstream should revert the offending commit. See: https://sourceforge.net/p/lirc/tickets/148/#6630
Comment by John (graysky) - Saturday, 10 October 2015, 21:29 GMT
@Likas - Any opinion on the value of including the patch as an interim fix for Arch users?
Comment by John (graysky) - Thursday, 22 October 2015, 05:07 GMT
@Likas - This is fixed in 0.9.3a released on 14-Oct.
Comment by John (graysky) - Saturday, 07 November 2015, 12:15 GMT
Probably ok to close: https://projects.archlinux.org/svntogit/packages.git/commit/trunk?h=packages/lirc&id=50174302f8f0ec8fbe68547cfb7c69a010b2b353
Comment by Andrew Gaydenko (student975) - Saturday, 07 November 2015, 13:40 GMT
It's OK for mew to close, thanks!

Loading...