On 3/27/22 1:49 AM, Ron Murray wrote:
I have the same problem here. Removing lmms and fluidsynth-dssi made no
difference.
I did a debug build and ran it with gdb. The resullt is attached.
Thanks. In your case it looks like something called "whysynth" is
the issue:
OP informs me that he tracked the problem down to the ZynAddSubFX
DSSI plugin. He switched to the ZynAddSubFX standalone app and all is
well for him.
This bug can be closed for rosegarden.
I just pushed improved plugin logging as [3a07309c]. If you can,
please test latest git:
https://sourceforge.net/p/rosegarden/git/ci/master/tree/
A debug build should now let you know when it is trying to load a
plugin. And that should be the last thing you see from rosegarden when
a
On 3/19/22 7:08 AM, Rainer Hans Liffers wrote:
> I compiled the latest version of rosegarden with debugging enabled
and subsequently ran gdb as shown in attached file. Is this information
sufficient?
Yes. Thanks.
The first line relevant to rg is this:
#19 0x55e3188e in
malloc_consolidate(): unaligned fastbin chunk detected Aborted
We've not seen this upstream. Best bet would be to do a debug build
from source and run under gdb. Then do a backtrace when it crashes.
That should help us determine whether it is Rosegarden or one of the
many libraries it
https://sourceforge.net/p/rosegarden/code/15845/
Patch attached.
Ted.
>From bdb6c75b733d4868e350af4baeeb15e309c7b62a Mon Sep 17 00:00:00 2001
From: Ted Felix
Date: Tue, 16 Jun 2020 00:48:04 -0400
Subject: [PATCH] Fix QPainterPath compilation error
---
src/gui/general/ThornStyle.cpp | 1 +
src/gu
Confirmed. This is likely working as designed.
In the Bass_sample.mid case, the note beginnings have been quantized,
but the note durations have not been quantized. Rosegarden seems to
prefer exact note durations when doing ties. If you double-click the
note that you want to tie and set
On 7/27/19 4:21 PM, ano...@users.sourceforge.net wrote:
The attached patch...
Patch has been applied and pushed. Test upstream at:
https://sourceforge.net/p/acpid2/code/ci/master/tree/
Planning a release for 8/15 or 9/15. Thanks again.
On 7/27/19 4:21 PM, ano...@users.sourceforge.net wrote:
On my system, this race happens frequently for /dev/input/event8 "Video
Bus" during boot, preventing acpid from handling my system's brightness
keys.
Thanks for the patch. I've posted the patch and this report upstream:
On 11/6/18 7:24 PM, Ted Felix wrote:
Will be included in the 2.0.31 release on November 15 or so.
A new upstream version (2.0.31) is now available that addresses this
issue.
https://sourceforge.net/projects/acpid2/files/
On 9/22/18 6:55 PM, zieg...@uni-freiburg.de wrote:
starting kacpimon fails because of "too many connections".
I've bumped the limit from 20 to 100. See upstream commit [9bc1579]:
https://sourceforge.net/p/acpid2/code/ci/9bc15796a5aed24719312a1505eb5d09965eef4d/
Will be included in the
Ah, now I get it. It's kacpimon that is failing. That's just a
debugging tool. I wouldn't even classify this as "important". "minor"
or "wishlist" at best.
kacpimon is still using the older version of connection_list.c. It's
a bit of a mess. And that one still has a 20 connection
On 09/22/2018 06:55 PM, zieg...@uni-freiburg.de wrote:
Version: 1:2.0.28-1+b1
This is odd since this exact problem was fixed in 2.0.20. It was
fixed in response to #719659. As a result, we should be able to handle
1024 connections. Do we somehow have a regression? Or was this not
On 12/15/2015 11:19 AM, Chris Lamb wrote:
Yep, works great. (Both in ./libnetlink.c and ./kacpimon/libnetlink.c)
Ok, this should be fixed upstream in [32f291].
https://sourceforge.net/p/acpid2/code/ci/32f291
I'll do an official release 1/15/2016. Unless you'd like it sooner.
Ted.
On 12/15/2015 06:00 AM, Chris Lamb wrote:
acpid fails to build from source in unstable/amd64:
libnetlink.c: In function 'addattr_l':
libnetlink.c:497:54: error: comparison between signed and unsigned integer
expressions [-Werror=sign-compare]
if ((int)NLMSG_ALIGN(n->nlmsg_len) +
This has now been fixed upstream in Rosegarden. The relevant commits
are r13910 and r13911.
https://sourceforge.net/p/rosegarden/code/13910/
https://sourceforge.net/p/rosegarden/code/13911/
Ted.
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of
On 04/19/2015 09:10 AM, Adrian Knoth wrote:
The cflags from pkg-config for jack are incorrect which breaks makefiles
and configure scripts that depend on them (e.g. those for rosegarden).
I tend to disagree - more likely, rosegarden is broken, because a ton of
other software that relies on
Package: libjack-jackd2-dev
Version: 1.9.10+20140719git3eb0ae6a~dfsg-2
The cflags from pkg-config for jack are incorrect which breaks makefiles
and configure scripts that depend on them (e.g. those for rosegarden).
The culprit is the Cflags line at the end of
On 10/27/2014 09:05 AM, Norbert Preining wrote:
The reason is that acpid is not started automatically anymore, but
some systemd tells me that something about socket activation etc.
I've not used systemd yet, but I'm pretty sure acpid works fine when
the socket is sent in as the stdin fd (or
On 06/22/2014 02:47 AM, Gijs Hillenius wrote:
Sometime following a recent update, closing the lid of this
Thinkpad T400s no longer suspends.
This is not likely to be an acpid issue. It could be a window
manager issue, an acpi-support issue, or a kernel issue. Here are some
things
On 12/16/2013 04:03 AM, Pigeon wrote:
event=battery.*
action=DIEDIEDIE
Thanks for the patch. I will review it today. It sounds like an
interesting feature.
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact
This has been fixed in acpid 2.0.20 just released today.
http://sourceforge.net/projects/acpid2
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
. Here's the patch in case anyone needs
it immediately:
commit 903271fb387ad4cc630c29810eecaacfb3e21080
Author: Ted Felix t...@tedfelix.com
Date: Wed Aug 14 19:02:33 2013 -0400
Increase MAX_CONNECTIONS to 100
This is an interim fix for Debian 719659.
diff --git a/connection_list.c b
The full fix for this is now available in the git repo on sourceforge.
http://sourceforge.net/projects/acpid2/
It now expands the connection list in increments of 20 connections as
needed up to 1024. add_connection() also has better error handling, so
problems in there will now be
I just released acpid 2.0.19 in response to the elevation to
serious severity. It should fix the compilation issue.
Ted.
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
On 04/24/2013 04:21 AM, Marek Elias wrote:
Input Layer: Type: 1 Code: 28 Value: 0
That's the Enter key. Nothing to worry about.
Input Layer: Type: 1 Code: 248 Value: 1
This is KEY_MICMUTE. That can easily be added to acpid by adding
the following lines to input_layer.c:
On 04/24/2013 10:19 AM, Marek Elias wrote:
I did the changed you suggested and rebuilt the acpid package.
Now I can see the events and run scripts with it.
Great. Then I'll go ahead and add the lines to the official source.
It will appear in the next version. Since there haven't been any
On 04/24/2013 10:19 AM, Marek Elias wrote:
I did the changed you suggested and rebuilt the acpid package.
Now I can see the events and run scripts with it.
I just added this change to the official source. If you wouldn't
mind retesting with the official source, I'd appreciate it. You can
On 02/22/2013 10:00 PM, Yevgeny Kosarzhevsky wrote:
when I run acpi_listen and press MicMute button I get only ^@ output, no codes
recognized.
What does kacpimon report for that button? You'll need to kill acpid
for kacpimon to work. Check the man page for more.
Ted.
--
To
Thanks for the bug report.
r12743 (Jan 8, 2012) of rosegarden fixes a similar bug. If you can
attach a sample MIDI file that causes the problem, I can test it against
the latest source.
Ted.
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of
On 3/19/2012 9:18 PM, Salvo Tomaselli wrote:
after the upgrade
srwxr-xr-x 1 root root 0 mar 20 02:15 /var/run/acpid.socket
before the upgrade
srw-rw-rw- 1 root root 0 mar 20 02:15 /var/run/acpid.socket
Thanks for the bug report.
I've verified this and should have a fix by the end of the day.
Patch to 2.0.15 attached. This fixes the fchmod() issue and should
give the correct permissions.
The -g option was a bit more tricky to fix. I ended up having to
go back to chown() and that seems to work ok. No matter where I put
fchown(), it just didn't work.
If you can give it
On Mon, Mar 12, 2012 at 10:28:14PM +0100, Luciano Bello wrote:
Is it a security problem?
Given that F_SETFD != F_SETFL, I would say that this is a security
problem. A userspace program can cause acpid to stop processing by
blocking on a socket.
Ted.
--
To UNSUBSCRIBE, email to
On 1/20/2012 3:39 PM, Adam Heath wrote:
The manpage for acpid(8) mentions /proc/acpi/event. That file doesn't
exist in 3.2(probably 3.1, maybe others). I'm guessing that the file
hasn't actually existed for a large number of kernel revisions.
Since debian is going to be shipping a 3.x
), sh(1), socket(2), connect(2)
+acpid(8)
.SH AUTHORS
+Ted Felix (www.tedfelix.com)
+.br
Tim Hockin thoc...@hockin.org
.br
Luca Capello l...@pca.it
On 7/23/2011 6:34 PM, Samuel Thibault wrote:
Shouldn't acpid somehow avoid calling the same script for the same
event over and over?
acpid's only purpose is to call scripts in response to acpi events.
It is up to the script to avoid doing unnecessary things that result in
writing to the
On 4/5/2011 12:41 AM, Hannes Calitz wrote:
It seems the only problem is
detecting the hotkeys on F1 to F9. Maybe then not an acpi problem but
rather a hal or xorg problem?
Usually, this is a kernel driver issue. It's easy to eliminate acpid
as the cause simply by running a window manager
Are you running a window manager like KDE, GNOME, Xfce... when this
happens?
Ted.
On 4/1/2011 12:01 PM, HCalitz wrote:
Package: acpid
Version: 1:2.0.8-3
Severity: normal
The Fn-F1 to Fn-F9 keys does not work anymore. They used to work fine until
the last update.
There may be a problem
Thanks for this. I'll get it into my next release.
Ted.
On 10/18/2010 4:18 AM, Stanislav Maslovski wrote:
Package: acpid
Version: 1:2.0.6-1
Severity: wishlist
Tags: upstream patch
Hello,
I think that it is reasonable to add support for the very common CD
play/pause and related buttons
Try the attached patch.
diff -N -r -U 10 -x '*.o' -x '*~' -x '.*' -x '*.out' -x CVS -x '*.moc'
acpid-2.0.6/event.c acpid/event.c
--- acpid-2.0.6/event.c 2010-03-28 10:00:04.0 -0400
+++ acpid/event.c 2010-09-27 18:26:48.0 -0400
@@ -115,20 +115,26 @@
}
/*
My power button stops working after resume from suspend due to some
kernel issue.
To work around this issue I reload the button module in my suspend
script and after reloading the module kacpimon shows events from the button.
However, acpid still does not react to the button until restarted.
If you don't mind grabbing the code and building it, you can go into
the file connection_list.c and change this line:
#define MAX_CONNECTIONS 10
To this:
#define MAX_CONNECTIONS 20
That will get rid of the Too many connections. message in the log.
Whether it will solve your problem, I'm
Thanks for the patch. It looks good and appears to work for me. I'll
get it into my next build, 2.0.2 scheduled for 2/15.
Ted.
Willi Mann wrote:
Ted Felix schrieb:
If you don't mind grabbing the code and building it, you can go into
the file connection_list.c and change this line
Thanks for the bug report.
This is working as designed. acpid ignores any files with dots in
their names. .sh files aren't config files, they are shell scripts
that are run by the config files. Give your config files a name without
a .sh then have the config file point to a shell script
Looks like the problem is in this line from open_logs():
logfd = open(logfile, O_WRONLY|O_CREAT|O_APPEND);
It should be this:
logfd = open(logfile, O_WRONLY|O_CREAT|O_APPEND, 0640);
And (theoretically, as I've not tested it) the problem is solved.
As mentioned, this doesn't fix any
45 matches
Mail list logo
