26 Feb
2005
26 Feb
'05
10:56 p.m.
So my US428 was working pretty well on my rackmount computer, but
with that motherboard fried, I'm now back to using my laptop. For
some reason my US428 doesn't like my laptop very much. One thing
that seemed to be a problem was that here was some sort of
qjackctl conflict with the non-multithreaded QT libs, so I
installed the multithreaded ones, recompiled, and at least my
computer is no longer crashing whenever I start start jackd from
qjackctl with my us428. However, I'm getting more xruns than I
think I should, and it might be beacuse I think I'm not using the
rawusb mode. Seems that I need to use -dhw0,2 to use rawusb mode,
but when I try that, I get
ALSA/USX2Y: Cannot open hwdep device "hw:0,1"
It should be located at hw:0,0, I think, but I'm not sure how to
specify that. Any ideas?
thanks,
spencer
26 Feb
26 Feb
11:35 p.m.
Ok, scratch what I said about it not crashing anymore. It's still
quite less than stable. I was using qjackctl, and it was fine
until I started pd, at which point my computer froze. So I
rebooted, and started jackd without qjackctl
sudo jackd -R -dalsa -dhw:0,0 -p1024 -n2 -i2 -o2
and was using it fine for a while, I started pd without problems,
then I was trying to figure out how to connect to the midi port
of PD when the following appeared on /var/log/mesages:
Feb 26 22:26:10 slingshot kernel: ALSA /usr/src/alsa/ \
alsa-driver-1.0.8/alsa-kernel/usb/usx2y/usbusx2yaudio.c:199: \
usb_submit_urb() returned -11
and at about the same time, jackd reported a bunch of xruns and
then:
subgraph starting at pure_data_0 timed out (subgraph_wait_fd=9, \
status = 0, state = Triggered)
**** alsa_pcm: xrun of at least 17.389 msecs
jackd watchdog: timeout - killing jackd
Killed
apon closer reading of the responses to my earlier question about
getting capture channels 3 and 4 to work, it looks like there're
2 different usx2y drivers? I'm using whatever came with alsa
1.0.8, and the changelog says that they include rawusb support,
so I think I'm using the correct one. But it'd be kinda nice if
I'm not, cause that's an easy fix.
-spencer
On Sat, Feb 26, 2005 at 09:55:59PM -0500, Spencer Russell wrote:
So my US428 was working pretty well on my rackmount
computer, but
with that motherboard fried, I'm now back to using my laptop. For
some reason my US428 doesn't like my laptop very much. One thing
that seemed to be a problem was that here was some sort of
qjackctl conflict with the non-multithreaded QT libs, so I
installed the multithreaded ones, recompiled, and at least my
computer is no longer crashing whenever I start start jackd from
qjackctl with my us428. However, I'm getting more xruns than I
think I should, and it might be beacuse I think I'm not using the
rawusb mode. Seems that I need to use -dhw0,2 to use rawusb mode,
but when I try that, I get
ALSA/USX2Y: Cannot open hwdep device "hw:0,1"
It should be located at hw:0,0, I think, but I'm not sure how to
specify that. Any ideas?
thanks,
spencer
27 Feb
27 Feb
8:16 a.m.
Spencer Russell wrote:
apon closer reading of the responses to my earlier question about
getting capture channels 3 and 4 to work, it looks like there're
2 different usx2y drivers? I'm using whatever came with alsa
1.0.8, and the changelog says that they include rawusb support,
so I think I'm using the correct one. But it'd be kinda nice if
I'm not, cause that's an easy fix.
The rawusb support is only available if you load the snd-usb-usx2y module
with nrpacks=1. Have you ruled this in?
Cheers.
--
rncbc aka Rui Nuno Capela
rncbc(a)rncbc.org
12:05 p.m.
Ahh, that worked. It also apparantly lets me access channels 3
and 4! Unfortunately, I can't keep it running long enough to play
with it. Jackd exits after a short time, at the same time that
usb_submit_urb() returns -11, on line 199 of usbusx2yaudio.c.
Running jackd again after it's given that error seems to pretty
consistantly hang my system completely. Is there any more
advanced debugging I can do to try to track down the problem?
-spencer
On Sun, Feb 27, 2005 at 12:15:33PM +0000, Rui Nuno Capela wrote:
The rawusb support is only available if you load the snd-usb-usx2y module
with nrpacks=1. Have you ruled this in?
Cheers.
--
rncbc aka Rui Nuno Capela
rncbc(a)rncbc.org
3 Mar
3 Mar
12:17 a.m.
On reading some older emails about this problem, I realized that
Karsten asked me earlier to post the output of lspci and lsusb,
and I don't think I ever did, so here they are:
srussell@slingshot:~$ lsusb
Bus 001 Device 001: ID 0000:0000
Bus 001 Device 002: ID 07a6:0986 ADMtek, Inc. AN986 Pegasus Ethernet
Bus 001 Device 005: ID 1604:8001 Tascam US-428 Audio/Midi Controller
srussell@slingshot:~$ lspci
0000:00:00.0 Host bridge: VIA Technologies, Inc. VT8363/8365 [KT133/KM133] (rev 80)
0000:00:01.0 PCI bridge: VIA Technologies, Inc. VT8363/8365 [KT133/KM133 AGP]
0000:00:0a.0 CardBus bridge: O2 Micro, Inc. OZ6912 Cardbus Controller
0000:00:11.0 ISA bridge: VIA Technologies, Inc. VT8231 [PCI-to-ISA Bridge] (rev 10)
0000:00:11.1 IDE interface: VIA Technologies, Inc. VT82C586A/B/VT82C686/A/B/VT823x/A/C
PIPC Bus Master IDE (rev 06)
0000:00:11.2 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev
1e)
0000:00:11.4 Bridge: VIA Technologies, Inc. VT8235 ACPI (rev 10)
0000:00:11.5 Multimedia audio controller: VIA Technologies, Inc. VT82C686 AC97 Audio
Controller (rev 40)
0000:00:11.6 Communication controller: VIA Technologies, Inc. Intel 537 [AC97 Modem] (rev
20)
0000:01:00.0 VGA compatible controller: S3 Inc. VT8636A [ProSavage KN133] AGP4X VGA
Controller (TwisterK) (rev 01)
I know I've heard pretty bad things about these VIA chipsets, especially for audio
stuff, because they tend to hog the bus, but hopefully I can at least make this usable.
Thanks,
-spencer
On Sun, Feb 27, 2005 at 11:05:45AM -0500, Spencer Russell wrote:
Ahh, that worked. It also apparantly lets me access
channels 3
and 4! Unfortunately, I can't keep it running long enough to play
with it. Jackd exits after a short time, at the same time that
usb_submit_urb() returns -11, on line 199 of usbusx2yaudio.c.
Running jackd again after it's given that error seems to pretty
consistantly hang my system completely. Is there any more
advanced debugging I can do to try to track down the problem?
-spencer
On Sun, Feb 27, 2005 at 12:15:33PM +0000, Rui Nuno Capela wrote:
>
> The rawusb support is only available if you load the snd-usb-usx2y module
> with nrpacks=1. Have you ruled this in?
>
> Cheers.
> --
> rncbc aka Rui Nuno Capela
> rncbc(a)rncbc.org
>
12:52 a.m.
And in latest news, I disabled the normal UHCI driver, and
instead enabled the alternate driver. Now I'm getting pretty much
the same behavior, but different messages on /var/log/messages.
Mar 2 23:48:31 slingshot kernel: ALSA usbusx2yaudio.c:297: Sequence Error!(hcd_frame=1662
ep=10out;wait=636,frame=637).
Mar 2 23:48:31 slingshot kernel: Most propably some urb of usb-frame 636 is still
missing.
Mar 2 23:48:31 slingshot kernel: Cause could be too long delays in usb-hcd interrupt
handling.
Thanks,
spencer
On Sun, Feb 27, 2005 at 11:05:45AM -0500, Spencer Russell wrote:
> Ahh, that worked. It also apparantly lets me access channels 3
> and 4! Unfortunately, I can't keep it running long enough to play
> with it. Jackd exits after a short time, at the same time that
> usb_submit_urb() returns -11, on line 199 of usbusx2yaudio.c.
> Running jackd again after it's given that error seems to pretty
> consistantly hang my system completely. Is there any more
> advanced debugging I can do to try to track down the problem?
>
> -spencer
>
>
> On Sun, Feb 27, 2005 at 12:15:33PM +0000, Rui Nuno Capela wrote:
> >
> > The rawusb support is only available if you load the snd-usb-usx2y module
> > with nrpacks=1. Have you ruled this in?
> >
> > Cheers.
> > --
> > rncbc aka Rui Nuno Capela
> > rncbc(a)rncbc.org
> >
7:40 a.m.
Am Sonntag 27 Februar 2005 04:35 schrieb Spencer Russell:
Ok, scratch what I said about it not crashing anymore.
It's still
quite less than stable. I was using qjackctl, and it was fine
until I started pd, at which point my computer froze. So I
rebooted, and started jackd without qjackctl
sudo jackd -R -dalsa -dhw:0,0 -p1024 -n2 -i2 -o2
and was using it fine for a while, I started pd without problems,
then I was trying to figure out how to connect to the midi port
of PD when the following appeared on /var/log/mesages:
Feb 26 22:26:10 slingshot kernel: ALSA /usr/src/alsa/ \
alsa-driver-1.0.8/alsa-kernel/usb/usx2y/usbusx2yaudio.c:199: \
usb_submit_urb() returned -11
can you check if this error doesn't happen, when modprobing like
$ modprobe snd_usb_usx2y nrpacks=4
?
nrpacks=1 and "rawusb mode" are irrelevant if you run jack with moderate
latencies
as above "-p1024 -n2" and 2 inputs.
What kernel are you using? If its not 2.6.11 you might give that a try.
apon closer reading of the responses to my earlier question about
getting capture channels 3 and 4 to work, it looks like there're
2 different usx2y drivers? I'm using whatever came with alsa
1.0.8, and the changelog says that they include rawusb support,
so I think I'm using the correct one. But it'd be kinda nice if
I'm not, cause that's an easy fix.
its 2 jack drivers for usx2y.
my first implementation doesn't really differ from whats in
jack's alsadriver now, if you apply the patch below.
kernel wise make shure that
$modinfo snd-usb-usx2y
gives you
"TASCAM US-X2Y Version 0.8.7.1"
Recommendation for basic setup:
===============================
When you use jack with "-p1024 -n2" and 2 ins are enough just ignore the
"rawusb hipe" ;-)
Do 1.
$ modprobe snd_usb_usx2y nrpacks=4
and then 2.
$ jackd -R -dalsa -dhw:0 -p1024 -n2 -i2 -o2
that will give you 2 ins & 2 outs. but channels 3+4 in will not work.
I strongly suggest you get this simpler case going first.
if that works reliably, proceed with more aggressiv nrpacks settings
======================
$ rmmod snd_usb_usx2y (will propably have to end some processes first)
$ modprobe snd_usb_usx2y nrpacks=2
check if its ok still......
and finally
$ rmmod snd_usb_usx2y (will propably have to end some processes first)
$ modprobe snd_usb_usx2y nrpacks=1
Only on success and if you want channels 3+4 or low latencies, proceed with the
following:
==========================================================================================
1. Patch jack's alsa driver, compile and install it:
>>>>>>
---
alsa_driver.c.~1.65.~ 2005-02-22 01:15:07.000000000 +0100
+++ alsa_driver.c 2005-02-22 01:16:32.000000000 +0100
@@ -450,7 +450,7 @@
}
if ((err = snd_pcm_hw_params_set_channels (handle, hw_params,
- *nchns)) < 0) {
+ 2 /* *nchns*/)) < 0) {
jack_error ("ALSA: cannot set channel count to %u for %s",
*nchns, stream_name);
return -1;
<<<<<<<
2.
$ modprobe snd_usb_usx2y nrpacks=1
3.
$ jackd -R -dalsa -dhw:0,2
have fun
Karsten
8:42 a.m.
Karsten,
1. Patch jack's alsa driver, compile and install it:
I once asked if this patch is really necessary. I think, but not sure (as
I gave no us428 here), it is quite equivalent to start jackd explicitly
with -i2 -o2 arguments.
Am I wrong?
Cheers.
--
rncbc aka Rui Nuno Capela
rncbc(a)rncbc.org
>>>>>>
---
alsa_driver.c.~1.65.~ 2005-02-22 01:15:07.000000000 +0100
+++ alsa_driver.c 2005-02-22 01:16:32.000000000 +0100
@@ -450,7 +450,7 @@
}
if ((err = snd_pcm_hw_params_set_channels (handle, hw_params,
- *nchns)) < 0) {
+ 2 /* *nchns*/)) < 0) {
jack_error ("ALSA: cannot set channel count to %u for %s",
*nchns, stream_name);
return -1;
<<<<<<<
2.
$ modprobe snd_usb_usx2y nrpacks=1
3.
$ jackd -R -dalsa -dhw:0,2
1:21 p.m.
Am Donnerstag 03 März 2005 13:41 schrieb Rui Nuno Capela:
Karsten,
partly. with -i2 us428 will use 2 in channels, but we want all 4.
without that option or with -i4 and
without patch
4 in channels but nothing is usable: sounds awful
with patch
4 in channels and works
The patch is needed to make alsa think that jackd wants only 2 in channels,
while jackd actually works with 4 in channels.
For us428 snd-usb-usx2y correctly reports "I have 4 in channels"
but there is still something wrong (in snd-usb-usx2y raw usb mode: have put some hours in
it already without success)
that makes it necessary to cheat alsa a little for the us428 4in channels case.
cheers
1. Patch jack's alsa driver, compile and install it:
I once asked if this patch is really necessary. I think, but not sure (as
I gave no us428 here), it is quite equivalent to start jackd explicitly
with -i2 -o2 arguments.
Am I wrong? >>>>>>
---
alsa_driver.c.~1.65.~ 2005-02-22 01:15:07.000000000 +0100
+++ alsa_driver.c 2005-02-22 01:16:32.000000000 +0100
@@ -450,7 +450,7 @@
}
if ((err = snd_pcm_hw_params_set_channels (handle, hw_params,
- *nchns)) < 0) {
+ 2 /* *nchns*/)) < 0) {
jack_error ("ALSA: cannot set channel count to %u for %s",
*nchns, stream_name);
return -1;
<<<<<<<
2.
$ modprobe snd_usb_usx2y nrpacks=1
3.
$ jackd -R -dalsa -dhw:0,2
11:32 a.m.
On Thu, Mar 03, 2005 at 12:41:17PM +0100, Karsten Wiese wrote:
can you check if this error doesn't happen, when modprobing like
$ modprobe snd_usb_usx2y nrpacks=4
?
nrpacks=1 and "rawusb mode" are irrelevant if you run jack with moderate
latencies
as above "-p1024 -n2" and 2 inputs.
With nrpacks=4, it seems to last slightly longer before giving
the "Sequence Error!" and "Most propably some urb of usb-frame
935 is still missing"
What kernel are you using? If its not 2.6.11 you might
give that a try.
Every time I've tried using a 2.6 kernel on this machine, I get
these random lockups that I can't figure out, so I'm using 2.4.26
with lowlatency, preempt, and givecap patches. I'll try 2.6.11,
and see how that goes.
Rui said a while ago that he was getting a similar message, until
he tuned the IRQ priorities, but I've been googling around and I
can't figure out how to do it. I tried
chrt -f -p 90 'pidof "IRQ 0"'
as directed on some website, but it just returns
execvp: No such file or directory
failed to execute pidof "IRQ 0"
I'm still not exactly sure what it is I'm trying to do with this,
I'm just sort of blindly following what I read on the
internet(always a good idea). The man page doesn't say anything
about IRQs, so if anyone wants to throw some help my way, that'd
be great.
have fun
Karsten
Thanks a lot for the help!
-spencer
1:06 p.m.
Spencer,
Rui said a while ago that he was getting a similar message, until
he tuned the IRQ priorities, but I've been googling around and I
can't figure out how to do it. I tried
chrt -f -p 90 'pidof "IRQ 0"'
as directed on some website, but it just returns
execvp: No such file or directory
failed to execute pidof "IRQ 0"
When I said that I was assuming that you have Ingo Molnar's
realtime-preempt patch, which applies to vanilla 2.6 kernels only (latest
is 2.6.11-rc4). The patch is here:
http://people.redhat.com/~mingo/reatime-preempt/
You can only do that IRQ tunning if you have this kernel configured with
PREEMPT_RT=y (i.e. complete realtime-preemption enabled).
Once you have this up and running, tune the IRQ threads (with chrt) like
this example:
timer: chrt -f -p 90 `pidof "IRQ 0"`
rtc: chrt -f -p 80 `pidof "IRQ 8"`
snd: chrt -f -p 70 `pidof "IRQ 5"`
usb: chrt -f -p 60 `pidof "IRQ 10"`
Please note you must check which IRQ numbers are the appropriate (cat
/proc/interrupts); the above is just an example taken from my laptop,
where the USB controller (ohci_hcd) is on IRQ 10 and the onboard sound is
on IRQ 5. YMMV.
Bye.
--
rncbc aka Rui Nuno Capela
rncbc(a)rncbc.org
1:17 p.m.
On Thu, 2005-03-03 at 17:05 +0000, Rui Nuno Capela wrote:
Spencer,
Also you can't change the priority of the timer IRQ thread (IRQ 0)
anymore, because it does not run in a thread. This is not a problem.
Lee
Rui said a while ago that he was getting a similar message, until
he tuned the IRQ priorities, but I've been googling around and I
can't figure out how to do it. I tried
chrt -f -p 90 'pidof "IRQ 0"'
as directed on some website, but it just returns
execvp: No such file or directory
failed to execute pidof "IRQ 0"
When I said that I was assuming that you have Ingo Molnar's
realtime-preempt patch, which applies to vanilla 2.6 kernels only (latest
is 2.6.11-rc4). The patch is here:
http://people.redhat.com/~mingo/reatime-preempt/
You can only do that IRQ tunning if you have this kernel configured with
PREEMPT_RT=y (i.e. complete realtime-preemption enabled).
Once you have this up and running, tune the IRQ threads (with chrt) like
this example:
timer: chrt -f -p 90 `pidof "IRQ 0"`
rtc: chrt -f -p 80 `pidof "IRQ 8"`
snd: chrt -f -p 70 `pidof "IRQ 5"`
usb: chrt -f -p 60 `pidof "IRQ 10"`
Please note you must check which IRQ numbers are the appropriate (cat
/proc/interrupts); the above is just an example taken from my laptop,
where the USB controller (ohci_hcd) is on IRQ 10 and the onboard sound is
on IRQ 5. YMMV.
1:22 p.m.
When I said that I was assuming that you have Ingo
Molnar's
realtime-preempt patch, which applies to vanilla 2.6 kernels only (latest
is 2.6.11-rc4). The patch is here:
http://people.redhat.com/~mingo/reatime-preempt/
Type. The real (obvious) URL is here:
http://people.redhat.com/~mingo/realtime-preempt/
Bye.
--
rncbc aka Rui Nuno Capela
rncbc(a)rncbc.org
1:57 p.m.
Am Donnerstag 03 März 2005 16:32 schrieb Spencer Russell:
also on 2.6.11 you'd disable apm & acpi like above to try to get rid
of those random lockups or "Sequence Error!"s.
On Thu, Mar 03, 2005 at 12:41:17PM +0100, Karsten
Wiese wrote:
try with apm and acpi disabled.
append "apm=off acpi=off" to the kernels boot parameter line.
on laptops lots of things happen "beghind the scenes" .
"apm=off acpi=off" made a big differenc here when I used 2.4 kernels.
can you check if this error doesn't happen, when modprobing like
$ modprobe snd_usb_usx2y nrpacks=4
?
nrpacks=1 and "rawusb mode" are irrelevant if you run jack with moderate
latencies
as above "-p1024 -n2" and 2 inputs.
With nrpacks=4, it seems to last slightly longer before giving
the "Sequence Error!" and "Most propably some urb of usb-frame
935 is still missing"
What kernel
are you using? If its not 2.6.11 you might give that a try.
Every time I've tried using a 2.6 kernel on this machine, I get
these random lockups that I can't figure out, so I'm using 2.4.26
with lowlatency, preempt, and givecap patches. I'll try 2.6.11,
and see how that goes.
Rui said a while ago that he was getting a similar message, until
he tuned the IRQ priorities, but I've been googling around and I
can't figure out how to do it. I tried
chrt -f -p 90 'pidof "IRQ 0"'
as directed on some website, but it just returns
execvp: No such file or directory
failed to execute pidof "IRQ 0"
this is for the realtime kernel patches by Ingo Molnar.
I may be wrong, but I'd suggest you get it running on 2.6.11 before you
proceed with the realtime kernels.
Plain 2.6.11 gives me better low latency (and better behavior with acpi enabled too) than
patched 2.4 kernels. lots of Ingo's enhancements are in 2.6.11 already.
Still the very basic mode with
$ modprobe snd_usb_usx2y nrpacks=4
must work flawlessly with 2 in / 2 out jack like
$ jackd -R -dalsa -dhw:0
on a plain kernel as first step.
srussell@slingshot:~$ lsusb
Bus 001 Device 001: ID 0000:0000
Bus 001 Device 002: ID 07a6:0986 ADMtek, Inc. AN986 Pegasus Ethernet
Bus 001 Device 005: ID 1604:8001 Tascam US-428 Audio/Midi Controller
Your ethernet is also on usb!
Usb audio is enough to fully occupy an USB 1.1 bus.
Try with us428 as the only usb client (without that usb ethernet).
cheers
Karsten
5 Mar
5 Mar
4:20 a.m.
OK, So as per these several suggestions, I've now tried the US428
with a vanilla 2.6.11, a 2.6.11 with Ingo's preempt patch and
full preemption, nrpacks=4, apm and acpi not compiled in the
kernel, and without my usb ethernet card plugged in. I'm still
getting the "Sequence Error!" messages, then a couple seconds
later qjackctl dissapears.
I still haven't tuned the IRQ stuff, because I'm still getting
the error of:
execvp: No such file or directory
failed to execute pidof "IRQ 8"
when I try to execute chrt -f -p 80 'pidof "IRQ 8"' I'm not sure
what do do
about that one, and google and man don't seem to help me much. Any thoughts?
Thanks,
-spencer
On Thu, Mar 03, 2005 at 06:58:08PM +0100, Karsten Wiese wrote:
Am Donnerstag 03 M?rz 2005 16:32 schrieb Spencer
Russell:
also on 2.6.11 you'd disable apm & acpi like above to try to get rid
of those random lockups or "Sequence Error!"s.
On Thu, Mar 03, 2005 at 12:41:17PM +0100, Karsten
Wiese wrote:
try with apm and acpi disabled.
append "apm=off acpi=off" to the kernels boot parameter line.
on laptops lots of things happen "beghind the scenes" .
"apm=off acpi=off" made a big differenc here when I used 2.4 kernels.
can you check if this error doesn't happen, when modprobing like
$ modprobe snd_usb_usx2y nrpacks=4
?
nrpacks=1 and "rawusb mode" are irrelevant if you run jack with moderate
latencies
as above "-p1024 -n2" and 2 inputs.
With nrpacks=4, it seems to last slightly longer before giving
the "Sequence Error!" and "Most propably some urb of usb-frame
935 is still missing"
What
kernel are you using? If its not 2.6.11 you might give that a try.
Every time I've tried using a 2.6 kernel on this machine, I get
these random lockups that I can't figure out, so I'm using 2.4.26
with lowlatency, preempt, and givecap patches. I'll try 2.6.11,
and see how that goes.
Rui said a while ago that he was getting a similar message, until
he tuned the IRQ priorities, but I've been googling around and I
can't figure out how to do it. I tried
chrt -f -p 90 'pidof "IRQ 0"'
as directed on some website, but it just returns
execvp: No such file or directory
failed to execute pidof "IRQ 0"
this is for the realtime kernel patches by Ingo Molnar.
I may be wrong, but I'd suggest you get it running on 2.6.11 before you
proceed with the realtime kernels.
Plain 2.6.11 gives me better low latency (and better behavior with acpi enabled too)
than
patched 2.4 kernels. lots of Ingo's enhancements are in 2.6.11 already.
Still the very basic mode with
$ modprobe snd_usb_usx2y nrpacks=4
must work flawlessly with 2 in / 2 out jack like
$ jackd -R -dalsa -dhw:0
on a plain kernel as first step.
srussell@slingshot:~$ lsusb
Bus 001 Device 001: ID 0000:0000
Bus 001 Device 002: ID 07a6:0986 ADMtek, Inc. AN986 Pegasus Ethernet
Bus 001 Device 005: ID 1604:8001 Tascam US-428 Audio/Midi Controller
Your ethernet is also on usb!
Usb audio is enough to fully occupy an USB 1.1 bus.
Try with us428 as the only usb client (without that usb ethernet).
cheers
Karsten
7:24 a.m.
Am Samstag 05 März 2005 09:20 schrieb Spencer Russell:
OK, So as per these several suggestions, I've now
tried the US428
with a vanilla 2.6.11, a 2.6.11 with Ingo's preempt patch and
full preemption, nrpacks=4, apm and acpi not compiled in the
kernel, and without my usb ethernet card plugged in. I'm still
getting the "Sequence Error!" messages, then a couple seconds
later qjackctl dissapears.
Please post the output of
$cat /proc/interrupts
maybe the usb-interrupt is also busy with something else...
Also make shure that any option in your bios sounding like "usb legacy support"
is disabled.
I still haven't tuned the IRQ stuff, because I'm still getting
the error of:
execvp: No such file or directory
failed to execute pidof "IRQ 8"
when I try to execute chrt -f -p 80 'pidof "IRQ 8"' I'm not sure
what do do
about that one, and google and man don't seem to help me much. Any thoughts?
Thanks,
-spencer
On Thu, Mar 03, 2005 at 06:58:08PM +0100, Karsten Wiese wrote:
Am Donnerstag 03 M?rz 2005 16:32 schrieb Spencer
Russell:
also on 2.6.11 you'd disable apm & acpi like above to try to get rid
of those random lockups or "Sequence Error!"s.
On Thu, Mar 03, 2005 at 12:41:17PM +0100, Karsten
Wiese wrote:
try with apm and acpi disabled.
append "apm=off acpi=off" to the kernels boot parameter line.
on laptops lots of things happen "beghind the scenes" .
"apm=off acpi=off" made a big differenc here when I used 2.4 kernels.
can you check if this error doesn't happen, when modprobing like
$ modprobe snd_usb_usx2y nrpacks=4
?
nrpacks=1 and "rawusb mode" are irrelevant if you run jack with moderate
latencies
as above "-p1024 -n2" and 2 inputs.
With nrpacks=4, it seems to last slightly longer before giving
the "Sequence Error!" and "Most propably some urb of usb-frame
935 is still missing"
What
kernel are you using? If its not 2.6.11 you might give that a try.
Every time I've tried using a 2.6 kernel on this machine, I get
these random lockups that I can't figure out, so I'm using 2.4.26
with lowlatency, preempt, and givecap patches. I'll try 2.6.11,
and see how that goes.
Rui said a while ago that he was getting a similar message, until
he tuned the IRQ priorities, but I've been googling around and I
can't figure out how to do it. I tried
chrt -f -p 90 'pidof "IRQ 0"'
as directed on some website, but it just returns
execvp: No such file or directory
failed to execute pidof "IRQ 0"
this is for the realtime kernel patches by Ingo Molnar.
I may be wrong, but I'd suggest you get it running on 2.6.11 before you
proceed with the realtime kernels.
Plain 2.6.11 gives me better low latency (and better behavior with acpi enabled too)
than
patched 2.4 kernels. lots of Ingo's enhancements are in 2.6.11 already.
Still the very basic mode with
$ modprobe snd_usb_usx2y nrpacks=4
must work flawlessly with 2 in / 2 out jack like
$ jackd -R -dalsa -dhw:0
on a plain kernel as first step.
srussell@slingshot:~$ lsusb
Bus 001 Device 001: ID 0000:0000
Bus 001 Device 002: ID 07a6:0986 ADMtek, Inc. AN986 Pegasus Ethernet
Bus 001 Device 005: ID 1604:8001 Tascam US-428 Audio/Midi Controller
Your ethernet is also on usb!
Usb audio is enough to fully occupy an USB 1.1 bus.
Try with us428 as the only usb client (without that usb ethernet).
cheers
Karsten
10:39 a.m.
On Sat, Mar 05, 2005 at 12:25:13PM +0100, Karsten Wiese wrote:
Am Samstag 05 M?rz 2005 09:20 schrieb Spencer
Russell:
srussell@slingshot:~$ cat /proc/interrupts
CPU0
0: 23968893 XT-PIC timer 0/68893
1: 5085 XT-PIC i8042 0/5085
2: 0 XT-PIC cascade 0/0
7: 0 XT-PIC parport0 0/0
8: 3 XT-PIC rtc 0/3
10: 0 XT-PIC VIA686A 0/0
11: 9538103 XT-PIC yenta, uhci_hcd 0/38103
12: 36200 XT-PIC i8042 0/36200
14: 101724 XT-PIC ide0 0/1723
15: 13 XT-PIC ide1 1/12
NMI: 0
LOC: 0
ERR: 0
MIS: 0
Hmm, it does look like my PCMCIA driver is also on the same
interrupt. I don't have any PCMCIA cards in now, but could that
still be causing problems?
OK, So as per these several suggestions, I've
now tried the US428
with a vanilla 2.6.11, a 2.6.11 with Ingo's preempt patch and
full preemption, nrpacks=4, apm and acpi not compiled in the
kernel, and without my usb ethernet card plugged in. I'm still
getting the "Sequence Error!" messages, then a couple seconds
later qjackctl dissapears.
Please post the output of
$cat /proc/interrupts
maybe the usb-interrupt is also busy with something else...
Also make shure that any option in your bios sounding
like "usb
legacy support" is disabled.
I have kind of a strange BIOS, where pretty much the only option
I have when I go into the setup is boot order, there isn't
anything about USB support.
Thanks for the help,
Spencer
3:43 p.m.
Am Samstag 05 März 2005 15:39 schrieb Spencer Russell:
srussell@slingshot:~$ cat /proc/interrupts
CPU0
0: 23968893 XT-PIC timer 0/68893
1: 5085 XT-PIC i8042 0/5085
2: 0 XT-PIC cascade 0/0
7: 0 XT-PIC parport0 0/0
8: 3 XT-PIC rtc 0/3
10: 0 XT-PIC VIA686A 0/0
11: 9538103 XT-PIC yenta, uhci_hcd 0/38103
12: 36200 XT-PIC i8042 0/36200
14: 101724 XT-PIC ide0 0/1723
15: 13 XT-PIC ide1 1/12
NMI: 0
LOC: 0
ERR: 0
MIS: 0
Hmm, it does look like my PCMCIA driver is also on the same
interrupt. I don't have any PCMCIA cards in now, but could that
still be causing problems?
only if yenta is in front of uhci_hcd interrupt handler chain wise.
just recheck after
$rmmod yenta
might make sure its gone with
$lsmod
post above.
it looks like something keeps the usb's interrupt from being handled right in time.
like
bios: no idea.....
harddisk: dma enabled?
$hdparm /dev/hda
post above also.
video: recheck in text mode:
$telinit 3
On Sat, Mar 05, 2005 at 12:25:13PM +0100, Karsten
Wiese wrote:
> Am Samstag 05 M?rz 2005 09:20 schrieb Spencer Russell:
> > OK, So as per these several suggestions, I've now tried the US428
> > with a vanilla 2.6.11, a 2.6.11 with Ingo's preempt patch and
> > full preemption, nrpacks=4, apm and acpi not compiled in the
> > kernel, and without my usb ethernet card plugged in. I'm still
> > getting the "Sequence Error!" messages, then a couple seconds
how long does it work until "Sequence Error!"?
any hints that it happens in regular intervals?
later qjackctl dissapears.
Please post the output of
$cat /proc/interrupts
maybe the usb-interrupt is also busy with something else...
Also make shure that any option in your bios
sounding like "usb
legacy support" is disabled.
I have kind of a strange BIOS, where pretty much the only option
I have when I go into the setup is boot order, there isn't
anything about USB support.
Thanks for the help,
Spencer
9:38 a.m.
Spencer Russell wrote:
OK, So as per these several suggestions, I've now
tried the US428
with a vanilla 2.6.11, a 2.6.11 with Ingo's preempt patch and
full preemption, nrpacks=4, apm and acpi not compiled in the
kernel, and without my usb ethernet card plugged in. I'm still
getting the "Sequence Error!" messages, then a couple seconds
later qjackctl dissapears.
I still haven't tuned the IRQ stuff, because I'm still getting
the error of:
execvp: No such file or directory
failed to execute pidof "IRQ 8"
when I try to execute chrt -f -p 80 'pidof "IRQ 8"' I'm not sure
what do
do about that one, and google and man don't seem to help me much. Any
thoughts?
Did you set PREEMPT_RT=y (complete real-time) on your kernel
configuration? It seems not... otherwise you would see those IRQ threads
around.
Bye now.
--
rncbc aka Rui Nuno Capela
rncbc(a)rncbc.org
10:40 a.m.
On Sat, Mar 05, 2005 at 01:36:59PM +0000, Rui Nuno Capela wrote:
Spencer Russell wrote:
I can see the IRQ threads in top, so they're there, but it's
still giving me the execvp error.
Thanks,
Spencer
OK, So as per these several suggestions, I've
now tried the US428
with a vanilla 2.6.11, a 2.6.11 with Ingo's preempt patch and
full preemption, nrpacks=4, apm and acpi not compiled in the
kernel, and without my usb ethernet card plugged in. I'm still
getting the "Sequence Error!" messages, then a couple seconds
later qjackctl dissapears.
I still haven't tuned the IRQ stuff, because I'm still getting
the error of:
execvp: No such file or directory
failed to execute pidof "IRQ 8"
when I try to execute chrt -f -p 80 'pidof "IRQ 8"' I'm not sure
what do
do about that one, and google and man don't seem to help me much. Any
thoughts?
Did you set PREEMPT_RT=y (complete real-time) on your kernel
configuration? It seems not... otherwise you would see those IRQ threads
around.
11:33 a.m.
On Sat, 05 Mar 2005 09:40:45 -0500, Spencer Russell
<Spencer.Russell(a)oberlin.edu> wrote:
I can see the IRQ threads in top, so they're there, but it's
still giving me the execvp error.
It really looks like you simply don't have the 'pidof' program
installed. On Debian it's in the sysvinit package, but that may be
different on RedHat.
Hope this helps,
Antoine
3:27 p.m.
On Sat, 2005-03-05 at 07:33, Antoine Busch wrote:
On Sat, 05 Mar 2005 09:40:45 -0500, Spencer Russell
<Spencer.Russell(a)oberlin.edu> wrote:
Yeah, that is what is probably happening. In Fedora Core 3 it is part of
the SysVinit package and most probably it is installed, it is a very
core package. I would suspect that the path does not include /sbin
(where pidof is located). So try replacing pidof in the script with
/sbin/pidof
-- Fernando
I can see the IRQ threads in top, so they're there, but it's
still giving me the execvp error.
It really looks like you simply don't have the 'pidof' program
installed. On Debian it's in the sysvinit package, but that may be
different on RedHat.
7 Mar
7 Mar
12:05 p.m.
On Sat, Mar 05, 2005 at 10:33:31AM -0500, Antoine Busch wrote:
On Sat, 05 Mar 2005 09:40:45 -0500, Spencer Russell
<Spencer.Russell(a)oberlin.edu> wrote:
heh heh, turns out that instead of using the ` character, I was
using a regular apostrophe, ' to inclose pidof "IRQ 10", which is
why it didn't work.
thanks,
spencer
I can see the IRQ threads in top, so they're there, but it's
still giving me the execvp error.
It really looks like you simply don't have the 'pidof' program
installed. On Debian it's in the sysvinit package, but that may be
different on RedHat.
Hope this helps,
Antoine
28 Feb
28 Feb
2:47 a.m.
Just out of curiosity. What laptop are you using? Mine is an ACER
Ferrari 3000 (1800+ Athlon-XP-M) and it's crap, for sound. I've tried
a Tascam-USB and it was no substantial improvement on the system. The
card was definitely better than the built in (I suppose that need no
explanation). I've installed a P4 (no laptop) in my small studio and
that beats that hell out of the laptop.
regards,
/bengan
11:57 a.m.
I'm using an HP Pavilion ze1110. It's a couple years old, but the
price was right at about $250. It's a AMD Duron 1Ghz with 128 MB
of RAM(I really need to upgrade that). VIA Chipset, which is I
think where a lot of my problems are. My new "studio" computer is
a huge improvement, but I need to get the MoBo replaced, after a
snapping sparky sound and slight aroma of burning.
-spencer
On Mon, Feb 28, 2005 at 07:46:41AM +0100, Bengt G?rd?n wrote:
Just out of curiosity. What laptop are you using? Mine is an
ACER Ferrari 3000 (1800+ Athlon-XP-M) and it's crap, for sound.
I've tried a Tascam-USB and it was no substantial improvement
on the system. The card was definitely better than the built in
(I suppose that need no explanation). I've installed a P4 (no
laptop) in my small studio and that beats that hell out of the
laptop.
regards,
/bengan
6983
days inactive
6992
days old
linux-audio-user@lists.linuxaudio.org
24 comments
7 participants
participants (7)
-
Antoine Busch -
Bengt Gördén -
Fernando Lopez-Lezcano -
Karsten Wiese -
Lee Revell -
Rui Nuno Capela -
Spencer Russell