Harmony on 710 working?
Hi,
I recently downloaded and compiled the latest pa-kernel. I also included harmony sound support and now I would like to know if and how I can access the sound device. Serching on the web was no success.
Chris
Signature · 1 line
Linux - Cause life is too short for reboots
Quoted 3 lines · Christian Schmitt
I recently downloaded and compiled the latest pa-kernel. I also included harmony sound support and now I would like to know if and how I can access the sound device. Serching on the web was no success.
insmod harmony
aumix
mpg123 <your-favorite-mp3-file>
Greetings,
Helge
Hi Christian,
at least I got the Harmony driver working on my c110. I used the Kernel driver first, but found out that kmixer wasn't working (obviously expected alsa)
I could however cat *.au files to /dev/audio making a lot of sound. After that, I compiled the alsa stuff -there are two sets of drivers, but one didn't work as expected. I cannot look things up, currently I am on holliday in Ireland.. You need to insmod the driver as far as I remember. I'll be able to tell you more after the linux beer hike... ;-)
cheers
Juergen
=========================================== __ _
Juergen Braukmann [email protected]| -o)/ / (_)__ __ ____ __
Tel: 0201-743648 dk4jb@db0qs.#nrw.deu.eu | /\\ /__/ / _ \/ // /\ \/ / ===========================================_\_v __/_/_//_/\_,_/ /_/\_\
Quoted 16 lines · Christian Schmitt
Hi,
I recently downloaded and compiled the latest pa-kernel. I also included harmony sound support and now I would like to know if and how I can access the sound device. Serching on the web was no success.
Chris
--
Linux - Cause life is too short for reboots_______________________________________________
parisc-linux mailing list
[email protected]
http://lists.parisc-linux.org/mailman/listinfo/parisc-linux
Quoted 5 lines · Juergen Braukmann
Hi Christian,
at least I got the Harmony driver working on my c110. I used the Kernel driver first, but found out that kmixer wasn't working (obviously expected alsa)
A great many Linux sound programs seem to assume that the machine is little-endian and don't work on HP IME. mpg321 seems OK to me as is sox but anything more serious (particularly on the recording side) tends to have problems.
Maybe I should file some bugs in the Debian BTS.
Signature · 1 line
patrick
Quoted 11 lines · Helge Deller
Quoted 3 lines · Christian Schmitt
I recently downloaded and compiled the latest pa-kernel. I also included harmony sound support and now I would like to know if and how I can access the sound device. Serching on the web was no success.
insmod harmony
aumix
mpg123 <your-favorite-mp3-file>
I recompiled the kernel with harmony as a module. I get unresolved symbols when loading the harmony driver:
ccio_get_fake
Any suggestions?
Chris
Signature · 1 line
Linux - Cause life is too short for reboots
I have a theory on this .... but will probably be dis-reguarded by someone more knowledgable :)
I will try to explain what I believe the problem is, and provid a quick hack/work around for you to get it working.
First can you tell me if CONFIG_IOMMU_CCIO is defined in your kernel as this will affect my answer.
My suspicion is that it is not defined, thus my answer makes this assumption.
Problem:
The CONFIG_IOMMU_CCIO is not defined in your kernel, so the ccio_get_fake() is defined to be NULL in include/asm-parisc/pci.h. This usually works fine if the drivers are compiled into the kernel ... but you are loading this as a module.... My suspicion is that the harmony module is trying to find the ccio_get_fake() entry point and your kernel does not export that function. (CONFIG_IOMMU_CCIO is not defined).
What confuses my about this problem is why ccio_get_fake() is not defined to NULL by include/asm-parisc/pci.h. Is there a mis-match between your kernel and module CONFIG_* options?
Solution:
Since the 710 does not have ccio, the cico_get_fake() is not needed. The ccio_get_fake() only needed for ccio machines on busses below the ccio that are not PCI (ie, PIC, GSC). So for you temp hack, just assign harmony.fake_pci_dev to NULL. Once we have a better picture of what is happening, we will insert a better fix if appropriate :)
Thanks,
- Ryan
Quoted 32 lines · Christian Schmitt
Quoted 11 lines · Helge Deller
Quoted 3 lines · Christian Schmitt
I recently downloaded and compiled the latest pa-kernel. I also included harmony sound support and now I would like to know if and how I can access the sound device. Serching on the web was no success.
insmod harmony
aumix
mpg123 <your-favorite-mp3-file>I recompiled the kernel with harmony as a module. I get unresolved symbols when loading the harmony driver:
ccio_get_fakeAny suggestions?
Chris
Signature · 7 lines
Linux - Cause life is too short for reboots _______________________________________________ parisc-linux mailing list [email protected] http://lists.parisc-linux.org/mailman/listinfo/parisc-linux
Quoted 3 lines · Christian Schmitt
ccio_get_fakeAny suggestions?
harmony.c needs to include pci.h :^(
./include/asm-parisc/pci.h:#define ccio_get_fake(dev) NULL
that's where the pci_dma_xxx interface definitions live. Or those definitions need to move to a seperate file that harmony.c and pci.h can include. Maybe create a pci-dma.h?
helge, rbrad, willy?
grant
Quoted 4 lines · Ryan Bradetich
What confuses my about this problem is why
ccio_get_fake()is not defined to NULL byinclude/asm-parisc/pci.h. Is there a mis-match between your kernel and module CONFIG_* options?
I think this is right. I overlooked that pci.h was getting included...even though we should move pci_dma* to a different file like I suggested.
grant
Quoted 10 lines · Ryan Bradetich
I have a theory on this .... but will probably be dis-reguarded by someone more knowledgable :)
I will try to explain what I believe the problem is, and provid a quick hack/work around for you to get it working.
First can you tell me if
CONFIG_IOMMU_CCIOis defined in your kernel as this will affect my answer.
Hello,
CONFIG_IOMMU_CCIO is set to "y" in the kernel config. I set it to "n", recompiled the modules and "modprobe harmony" works. However, aumix doesn't find a mixer device. Is there any way to find out where the problem is?
Signature · 1 line
Linux - Cause life is too short for reboots
You should not have to compile the modules differently then from the kernel. What happens when you compile both the kernel and modules with CONFIG_IOMMU_CCIO=y?
[Note: The ccio_get_fake will return the appropriate NULL if your system does not have any ccio devices.]
Sorry, I don't know anything about aumix. Do you have a /dev/mixer?
sane permissions? etc?
Thanks,
- Ryan
Quoted 26 lines · Christian Schmitt
Quoted 10 lines · Ryan Bradetich
I have a theory on this .... but will probably be dis-reguarded by someone more knowledgable :)
I will try to explain what I believe the problem is, and provid a quick hack/work around for you to get it working.
First can you tell me if
CONFIG_IOMMU_CCIOis defined in your kernel as this will affect my answer.Hello,
CONFIG_IOMMU_CCIOis set to "y" in the kernel config. I set it to "n", recompiled the modules and "modprobe harmony" works.
However, aumix doesn't find a mixer device. Is there any way to find out where the problem is?Signature · 1 line
Linux - Cause life is too short for reboots
As a side note (and interesting test) this should work properly when bot the kernel and modules are compiled with CONFIG_IOMMU_CCIO not set either.
I believe the problem you are seeing is when the kernel is compiled with CONFIG_IOMMU_CCIO is not set, and the modules are compiled with CONFIG_IOMMU_CCIO=y.
- Ryan
Quoted 49 lines · Ryan Bradetich
You should not have to compile the modules differently then from the kernel. What happens when you compile both the kernel and modules with
CONFIG_IOMMU_CCIO=y?[Note: The
ccio_get_fakewill return the appropriate NULL if your system does not have any ccio devices.]Sorry, I don't know anything about aumix. Do you have a
/dev/mixer?
sane permissions? etc?Thanks,
- Ryan
Quoted 13 lines · Christian Schmitt
Quoted 10 lines · Ryan Bradetich
I have a theory on this .... but will probably be dis-reguarded by someone more knowledgable :)
I will try to explain what I believe the problem is, and provid a quick hack/work around for you to get it working.
First can you tell me if
CONFIG_IOMMU_CCIOis defined in your kernel as this will affect my answer.Quoted 13 lines
Hello,
CONFIG_IOMMU_CCIOis set to "y" in the kernel config. I set it to "n", recompiled the modules and "modprobe harmony" works.
However, aumix doesn't find a mixer device. Is there any way to find out where the problem is?Signature · 1 line
Linux - Cause life is too short for reboots_______________________________________________
parisc-linux mailing list
[email protected]
http://lists.parisc-linux.org/mailman/listinfo/parisc-linux