7 May
2020
7 May
'20
12:07 p.m.
Hi,
A user reports that Radium prints an error when it tries to connect
input ports. The user has a USB soundcard with just one input channel.
Here's the code:
https://github.com/kmatheussen/radium/blob/master/audio/KillJackd.cpp#L128
(If code has changed, search for "// Connect input ports")
When running the program, the user gets this error message:
"KillJackd.cpp: Could not connect input port 2"
If there is only one input port, this message should not be printed,
because portname for port 2 should be NULL. I don't see anything wrong
with the code... Could it be a bug in Jack?
Full discussion:
https://users.notam02.no/~kjetism/radium/forum/viewtopic.php?f=7&t=265
7 May
7 May
12:16 p.m.
I can't reproduce the problem with the dummy driver, and my alsa
soundcard won't start if I give "-i 1" option.
On Thu, May 7, 2020 at 10:06 AM Kjetil Matheussen
<k.s.matheussen(a)gmail.com> wrote:
Hi,
A user reports that Radium prints an error when it tries to connect
input ports. The user has a USB soundcard with just one input channel.
Here's the code:
https://github.com/kmatheussen/radium/blob/master/audio/KillJackd.cpp#L128
(If code has changed, search for "// Connect input ports")
When running the program, the user gets this error message:
"KillJackd.cpp: Could not connect input port 2"
If there is only one input port, this message should not be printed,
because portname for port 2 should be NULL. I don't see anything wrong
with the code... Could it be a bug in Jack?
Full discussion:
https://users.notam02.no/~kjetism/radium/forum/viewtopic.php?f=7&t=265
8 May
8 May
12:13 a.m.
You should get the output of 'jack_lsp' from the user. Who knows what
qjackctl is doing between calling 'jack_get_ports' and the patchbay display.
Another thing to examine, 'jack_connect' may also return EEXIST, if the
connection is already made, but your code doesn't distinguish between
this and other errors. Maybe there's a race condition where some other
program (qjackctl) has already connected the ports before your code gets
there?
Am 07.05.20 um 10:06 schrieb Kjetil Matheussen:
BTW, while not relevant to your problem
const char *portname2 = portname1==NULL ? NULL : portnames[1];
portname1 will never be NULL in this block, because it is checked to be
!= NULL in line 135, won't it?
Chris
12:52 p.m.
On Thu, May 7, 2020 at 10:13 PM Christopher Arndt <chris(a)chrisarndt.de> wrote:
You should get the output of 'jack_lsp' from the user. Who knows what
qjackctl is doing between calling 'jack_get_ports' and the patchbay display.
Thanks, I've requested that on the forum now.
Another thing to examine, 'jack_connect' may
also return EEXIST, if the
connection is already made, but your code doesn't distinguish between
this and other errors. Maybe there's a race condition where some other
program (qjackctl) has already connected the ports before your code gets
there?
Good suggestion. I've asked the user to run jackd without qjackctl.
I'm also going to check for EEXIST in the code.
Am 07.05.20 um 10:06 schrieb Kjetil Matheussen:
BTW, while not relevant to your problem
const char *portname2 = portname1==NULL ? NULL : portnames[1];
portname1 will never be NULL in this block, because it is checked to be
!= NULL in line 135, won't it?
Yes. :-) Guess the second block was outside if-block originally, and
when I moved inside I forgot to remove the portname1==NULL test.
Doesn't make any difference though.
2:58 p.m.
On Fri, May 8, 2020 at 10:52 AM Kjetil Matheussen
<k.s.matheussen(a)gmail.com> wrote:
On Thu, May 7, 2020 at 10:13 PM Christopher Arndt <chris(a)chrisarndt.de> wrote:
We figured it out. Radium didn't check if the physical port was an
audio port, and the user has physical MIDI ports. So the error message
was caused by failing to connect an audio port to a midi port.
You should get the output of 'jack_lsp' from the user. Who knows what
qjackctl is doing between calling 'jack_get_ports' and the patchbay display.
Thanks, I've requested that on the forum now.
7:31 p.m.
Am 08.05.20 um 12:58 schrieb Kjetil Matheussen:
We figured it out. Radium didn't check if the
physical port was an
audio port, and the user has physical MIDI ports. So the error message
was caused by failing to connect an audio port to a midi port.
Ah, a classcial case of getting what you asked for, not what you thought
you'd asked for ;)
Glad you got to the bottom of this and JACK is not to blame.
Chris
1689
days inactive
1690
days old
jackaudio@lists.linuxaudio.org
5 comments
2 participants
participants (2)
-
Christopher Arndt -
Kjetil Matheussen