Skip to contentSkip to conversation

Hammerfall DSP System Problems

DISCUSSION [email protected] 6 participants 8 messages Jun 10, 2003 – Jun 22, 2003 1 message not in the archive

lore
Share & tools
lore
#0 lore raw

I recompiled a new kernel 2.4.19 with the latency patch and preemptive kernel patch. I still got the same unresolved symbol, so I loaded the modules that it was giving errors for individually. Like this:

/sbin/modprobe snd-rawmidi
/sbin/modprobe snd-hammerfall-mem
/sbin/modprobe snd-hwdep
/sbin/modprobe snd-hdsp

After I issue the last command I get the following errors:

/lib/modules/2.4.19/kernel/sound/pci/rme9652/snd-hdsp.o: init_module:
No such device
Hint: insmod errors can be caused by incorrect module parameters, including invalid IO or IRQ parameters

I know that I have the latest firmware. However, I am using the lastest version of ALSA 0.9.4 from the ftp site. I also tried the lastest CVS and I get the same error. When I check the hdsp.c it has the code for my firmware version 0x0b.

What is the status of this driver on the newest firmware? I am sorry if I am asking questions that have already been answered.

I am okay with C programming if you need more help. The main thing is I don't understand the status of the driver with the newest firmware and what do I use the hdsploader for, or how. The documentation I have found for the setup of this card doesn't mention this program.

Someone please help or direct me were I can get it.

Thanks,
Frank

Quoted 29 lines · Mark Knecht

Hi Frank,

I'm an HDSP 9652 non-user for many of the same reasons. A couple of

comments below.

Mark

Quoted 19 lines

Hello,

I recently purchased a Hammerfall DSP system to use on Windows and

especially LINUX for use on Ardour. I have the multiface system. I

must say I am confused on the support of the newest version of this

pci card (I know that I have firmware version 11). I downloaded the

current CVS on alsa-dev and compiled it fine. However when I run

modprobe I get the following error:

Quoted 8 lines
Quoted 5 lines

/sbin/modprobe snd-hdsp

/lib/modules/2.4.18-3/kernel/sound/pci/rme9652/snd-hdsp.o:

init_module: No such device

[MWK] This is sometimes caused by a mismatch between between the

firmware

Quoted 2 lines

revision in your card/box and those firmware revisions that the

driver knows

Quoted 2 lines

about. Run lspci to see what your card has, and read through the

driver code

Quoted 5 lines

to see what the driver knows about.

NOTE TO DEVELOPERS - Can't you make this message "init_module: No

such

Quoted 2 lines

device" more descriptive, such as "init_module: Checked for devices

with PCI

Quoted 2 lines

Device ID's XXX, firmware revisions AA, BB, CC, DD & EE - No devices

found"?

Quoted 2 lines

Many people, including me, don't know how to read source code and

can't give

Quoted 31 lines

you as much direct information. Just a thought.

Quoted 25 lines

Hint: insmod errors can be caused by incorrect module parameters,

including invalid IO or IRQ parameters

/lib/modules/2.4.18-3/kernel/sound/pci/rme9652/snd-hdsp.o: insmod

/lib/modules/2.4.18-3/kernel/sound/pci/rme9652/snd-hdsp.o failed

/lib/modules/2.4.18-3/kernel/sound/pci/rme9652/snd-hdsp.o: insmod

snd-hdsp failed

Then when I just offer the command insmod snd-hdsp I get the

following:

/sbin/insmod snd-hdsp

Using /lib/modules/2.4.18-3/kernel/sound/pci/rme9652/snd-hdsp.o

/lib/modules/2.4.18-3/kernel/sound/pci/rme9652/snd-hdsp.o:

unresolved

Quoted 4 lines
Quoted 4 lines

symbol snd_hammerfall_get_buffer_R5bab1fdf

/lib/modules/2.4.18-3/kernel/sound/pci/rme9652/snd-hdsp.o:

unresolved

Quoted 8 lines
Quoted 2 lines

symbol snd_rawmidi_receive_R1e37534c

[MWK] I have seen these sorts of errors when:

1) The kernel source code that the driver was compiled against is not

the

Quoted 4 lines

same as the kernel that's running. For instanace, you have a standard

distribution kernel and a custom kernel from kernel.org. If you are

running

Quoted 2 lines

the standard kernel, but the Alsa driver was compiled with

/usr/src/linux

Quoted 5 lines

pointing at the custom kernel, then problems ensue.

2) The C compiler used to compile Alsa does not match the compiler

used to

Quoted 12 lines

compile the kernel.

<SNIP>

Quoted 5 lines

When I run hdsploader it completes but says nothing like this:

Quoted 8 lines
Quoted 5 lines

hdsploader - firmware loader for RME Hammerfall DSP cards

Looking for HDSP + Multiface or Digiface cards :

[MWK] Again, this strikes me that Alsa does not understand your

firmware

Quoted 10 lines

revision since it isn't even finding a card.

Good luck!!

--------------------------------------------------------------------------
Looking for a roommate? Let CollegeClub.com help you out!
<http://navisite.collegeclub.com/channels/college/>

CollegeClub.com makes college life even better.

#1 lore raw
Quoted 6 lines

/sbin/modprobe snd-rawmidi
/sbin/modprobe snd-hammerfall-mem
/sbin/modprobe snd-hwdep
/sbin/modprobe snd-hdsp

Quoted 7 lines

After I issue the last command I get the following errors:

/lib/modules/2.4.19/kernel/sound/pci/rme9652/snd-hdsp.o: init_module:
No such device
Hint: insmod errors can be caused by incorrect module parameters, including invalid IO or IRQ parameters

I am having the very same problems with my HDSP 9652. What's even stranger is, in my case, that for one evening the card worked, at least as far as loading the driver. (No sound, but the driver loaded.) I was able to look at /proc/asound/card0 (or whatever the directory/device was at the time...) and see sensible stuff. I even took some screen shots and sent email to another individual as I was debugging why I had no sound.

Later that evening I powered down. Upon rebooting the next morning it failed, and has failed ever since. That was last Thursday evening.

I too am using 0.9.4 as provided in the Gentoo emerge, along with a custom compiled kernel.

Like you I think the HDSP driver understands my firmware revision. I have also downgraded and upgraded my firmware without success.

This whole Alsa/HDSP thing is amazingly weird...

- Mark

-------------------------------------------------------
This SF.net email is sponsored by: Etnus, makers of TotalView, The best thread debugger on the planet. Designed with thread debugging features you've never dreamed of, try TotalView 6 free at www.etnus.com.

#2 lore raw
Quoted 9 lines

I am having the very same problems with my HDSP 9652. What's even stranger is, in my case, that for one evening the card worked, at least as far as loading the driver. (No sound, but the driver loaded.) I was able to look at /proc/asound/card0 (or whatever the directory/device was at the time...) and see sensible stuff. I even took some screen shots and sent email to another individual as I was debugging why I had no sound.

Later that evening I powered down. Upon rebooting the next morning it failed, and has failed ever since. That was last Thursday evening.

the same thing happened to me a week ago. i have, as usual, not had time to investigate further.

Quoted 7 lines

I too am using 0.9.4 as provided in the Gentoo emerge, along with a custom compiled kernel.

Like you I think the HDSP driver understands my firmware revision. I have also downgraded and upgraded my firmware without success.

This whole Alsa/HDSP thing is amazingly weird...

at the moment, i regret that we made the move to user-space loading of the "firmware" when we did. i think now that it was a mistake to do it at that time. it has caused a fundamental change in the way the driver sets up, but it came at a time when we weren't confident that setup was working for various cases.

-------------------------------------------------------
This SF.net email is sponsored by: Etnus, makers of TotalView, The best thread debugger on the planet. Designed with thread debugging features you've never dreamed of, try TotalView 6 free at www.etnus.com.

#3 lore raw
Quoted 10 lines · Paul Davis
Quoted 4 lines

Like you I think the HDSP driver understands my firmware revision. I have also downgraded and upgraded my firmware without success.

This whole Alsa/HDSP thing is amazingly weird...

at the moment, i regret that we made the move to user-space loading of the "firmware" when we did. i think now that it was a mistake to do it at that time. it has caused a fundamental change in the way the driver sets up, but it came at a time when we weren't confident that setup was working for various cases.

I have now joined the ranks of troubleshooting HDSP owner. I have a cardbus HDSP (rev 0b) with a Multiface. IBM Thinkpad A31. Kernel 2.4.20 LL+preemp with cardbus support in kernel. Multiface connected and powered via AC adaptor.
I've been trying alsa-driver-0.9.4, and I detail my setup and problems below:

snd-hammerfall-mem loads normally

snd-hdsp first stage loads fine (dmesg output snip):
PCI: Setting latency timer of device 06:00.0 to 64
ALSA ../../alsa-kernel/pci/rme9652/hdsp.c:4110: card initialization pending : waiting for firmware

------------------

# hdsploader
hdsploader - firmware loader for RME Hammerfall DSP cards
Looking for HDSP + Multiface or Digiface cards :
Card 0 : RME Hammerfall DSP at 0x31000000, irq 9
Upload firmware for card hw:0
Hwdep ioctl error on card hw:0 : Input/output error.
------------------------

dmesg snippet while hdsploader ran with extra debugging added by
me:
ALSA ../../alsa-kernel/pci/rme9652/hdsp.c:644: wait for FIFO status <= 0 failed after 30 iterations ALSA ../../alsa-kernel/pci/rme9652/hdsp.c:3804: initializing firmware upload
ALSA ../../alsa-kernel/pci/rme9652/hdsp.c:522: loading firmware
....
the code in snd_hdsp_create_alsa_devices() is executed here,
successfully up until snd_hdsp_set_defaults() is called. In that
function, the loop where the gains for the matrix mixer are
silenced executes hdsp_write_gain 128 times, and on the next
iteration the following message is thrown:
ALSA ../../alsa-kernel/pci/rme9652/hdsp.c:644:
wait for FIFO status <= 127 failed after 5000 iterations
causing the function to return -EIO hence the ioctl error passed
back to hdsploader.

Any attempts to run hdsploader again result in more errors
involving FIFO timeouts, getting worse until the following
message is thrown every time:
ALSA ../../alsa-kernel/pci/rme9652/hdsp.c:3804: initializing firmware upload
ALSA ../../alsa-kernel/pci/rme9652/hdsp.c:508: Hammerfall-DSP: no Digiface or Multiface connected!

Unloading the modules, and ejecting and reinserting the card doesn't help matters, nor does a warm reboot. I do not have windows on the laptop (yet) so I have not attempted to get it working there (to load the multiface firmware, then warm boot) or to downgrade the hdsp firmware to 0a. Ideally, I can help to debug the primary problems with recent HDSP and recent ALSA.... I don't *need* to use the hardware right away (but I would like to :)

If there is any offlist discussion about the HDSP debugging, please include me in those messages.

D R Holsbeck, what version of ALSA, and what HDSP firmware are you successfully using on your A31? I know you have the Digiface though... I have the multiface.

Paul, any info you have on the initialization steps or iobox identification would be useful. The code in hdsp_get_iobox_version() seems a bit odd to me. I always get
the: ALSA ../../alsa-kernel/pci/rme9652/hdsp.c:644:
wait for FIFO status <= 0 failed after 30 iterations
in the primary conditional where it picks multiface or digiface.
I suppose that is intentional, but I wonder if that technique masks some other problem, even though I do have a multiface. How does that sequence work?

I want to get this crap working so I can test ardour better :)

jlc

-------------------------------------------------------
This SF.NET email is sponsored by: eBay
Great deals on office technology -- on eBay now! Click here:
http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5

#4 lore raw
Quoted 7 lines · Jesse Chappell

doesn't help matters, nor does a warm reboot. I do not have windows on the laptop (yet) so I have not attempted to get it working there (to load the multiface firmware, then warm boot) or to downgrade the hdsp firmware to 0a. Ideally, I can help to debug the primary problems with recent HDSP and recent ALSA.... I don't *need* to use the hardware right away (but I would like to :)

Woohoo!

Well, there is no time like the present. I installed windows XP and the RME drivers. No problems there, HDSP works in windows, red light gone. Warm reboot into Linux, everything loads and works like a champ! I think that this solution was mentioned by someone not too long ago on the list archives.

I am still interested in solving the problem so that the firmware loading to the multiface/digiface all works within linux. I will try my hand at fixes if folks have any ideas.

jlc

-------------------------------------------------------
This SF.NET email is sponsored by: eBay
Great deals on office technology -- on eBay now! Click here:
http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5

#5 lore raw
In reply to Jesse Chappell
Quoted 1 line · Jesse Chappell

Hwdep ioctl error on card hw:0 : Input/output error.

[...]

Quoted 4 lines

ALSA ../../alsa-kernel/pci/rme9652/hdsp.c:644:
wait for FIFO status <= 127 failed after 5000 iterations
causing the function to return -EIO hence the ioctl error passed back to hdsploader.

This is exactly what I'm seeing on my hp xe4100
(with multiface+cardbus). Thank you Jesse for your explanation.

Quoted 2 lines

If there is any offlist discussion about the HDSP debugging, please include me in those messages.

Yes, please, me too.

Signature · 9 lines
Jaakko Prättälä
[email protected]



-------------------------------------------------------
This SF.NET email is sponsored by: eBay
Great deals on office technology -- on eBay now! Click here:
http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5
#7 lore raw
In reply to a message not in the archive
Quoted 8 lines · D R Holsbeck

The newer Alsa code (with the new HDSP firmware code), has some new device nodes, that dont get handled correctly if you are using the devfs. I havent worked it all out yet, but running the snddevices script will create them. Here is my list of my /dev/snd directory

Do you have all the device nodes setup? Especially the hwC*D* nodes? I had problems running the hdsploader at first, because these nodes are not being handled by the devfsd correctly.

No luck, running snddevices created the extra hw nodes but I got the same behavior when running hdsploader. The only node I had before running the script was hwC1D0.

jlc

-------------------------------------------------------
This SF.Net email is sponsored by: INetU
Attention Web Developers & Consultants: Become An INetU Hosting Partner. Refer Dedicated Servers. We Manage Them. You Get 10% Monthly Commission! INetU Dedicated Managed Hosting http://www.inetu.net/partner/index.php

#8 lore raw
Quoted 7 lines · Jesse Chappell
Quoted 5 lines · D R Holsbeck

The newer Alsa code (with the new HDSP firmware code), has some new device nodes, that dont get handled correctly if you are using the devfs. I havent worked it all out yet, but running the snddevices script will create them. Here is my list of my /dev/snd directory

Quoted 8 lines
Quoted 3 lines

Do you have all the device nodes setup? Especially the hwC*D* nodes? I had problems running the hdsploader at first, because these nodes are not being handled by the devfsd correctly.

No luck, running snddevices created the extra hw nodes but I got the same behavior when running hdsploader. The only node I had before running the script was hwC1D0.

Are any of you DSP users having any luck with the MIDI interface? I'm finding it still drops lots of notes, which has so far stimied any possibilities of my doing soft synths using Linux.

Thanks,
Mark

-------------------------------------------------------
This SF.Net email is sponsored by: INetU
Attention Web Developers & Consultants: Become An INetU Hosting Partner. Refer Dedicated Servers. We Manage Them. You Get 10% Monthly Commission! INetU Dedicated Managed Hosting http://www.inetu.net/partner/index.php