Re: Failed to compile HDSP driver from the CVS

Re: Failed to compile HDSP driver from the CVS

[Ivica Bukvic]

Apologies if this thing got posted twice. My mail has been acting up
lately.

I am a bit confused. It is not the fact that I do have this file, but
that the compile was looking for it. It must have been that issue you
mentioned about CVS being not up-to-date. However, here's the error when
I try to compile the driver now, after updating it with 

cvs -z3 -d ':pserver:anonymous@cvs.alsa-project.org:/cvsroot/alsa'
update -P -d 

I get the following error: 

make[3]: Leaving directory
`/LocalDump/Install/Alsa/CVS/alsa-driver/pci/pdplus' 
make  -C rme9652 modules 
make[3]: Entering directory
`/LocalDump/Install/Alsa/CVS/alsa-driver/pci/rme9652' 
gcc -DALSA_BUILD -D__KERNEL__ -DMODULE=1
-I/LocalDump/Install/Alsa/CVS/alsa-driver/include
-I/lib/modules/2.4.8-26mdk/build/include -O2
-mpreferred-stack-boundary=2 -march=i586 -DLINUX -Wall
-Wstrict-prototypes -fomit-frame-pointer -pipe   -DEXPORT_SYMTAB -c
hammerfall_mem.c 
In file included from hammerfall_mem.c:1: 
../../alsa-kernel/pci/rme9652/hammerfall_mem.c:58: parse error before
string constant 
../../alsa-kernel/pci/rme9652/hammerfall_mem.c:58: warning: type
defaults to `int' in declaration of `MODULE_LICENSE' 
../../alsa-kernel/pci/rme9652/hammerfall_mem.c:58: warning: function
declaration isn't a prototype 
../../alsa-kernel/pci/rme9652/hammerfall_mem.c:58: warning: data
definition has no type or storage class 
make[3]: *** [hammerfall_mem.o] Error 1 
make[3]: Leaving directory
`/LocalDump/Install/Alsa/CVS/alsa-driver/pci/rme9652' 
make[2]: *** [_modsubdir_rme9652] Error 2 
make[2]: Leaving directory `/LocalDump/Install/Alsa/CVS/alsa-driver/pci'
make[1]: *** [compile] Error 1 
make[1]: Leaving directory `/LocalDump/Install/Alsa/CVS/alsa-driver' 
make: *** [build] Error 2 

Any help is greatly appreciated! 

Sincerely, 

Ico 




_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: bandwidth@sourceforge.net

Re: Failed to compile HDSP driver from the CVS

[Paul Davis]

>I am a bit confused. It is not the fact that I do have this file, but
>that the compile was looking for it. It must have been that issue you
>mentioned about CVS being not up-to-date. 

its a combination of the "fastdep" target looking at *all* existing
sources, whether they are meant to be there or not, plus the CVS sync issue.

					   However, here's the error when
>I try to compile the driver now, after updating it with 

use cvscompile to build it, i think.

--p


_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: bandwidth@sourceforge.net

unresolved symbols, cvs, hammerfall

[Fernando Pablo Lopez-Lezcano]

Just downloaded from cvs, I'm getting this unresolved symbols:

# depmod -a -e
depmod: *** Unresolved symbols in  
/lib/modules/2.4.18-2.ll/kernel/drivers/sound/pci/rme9652/snd-hammerfall-mem.o
depmod: 	pci_free_consistent
depmod: 	pci_alloc_consistent
depmod: 	pci_devices
depmod: 	mem_map
depmod: 	printk
depmod: *** Unresolved symbols in  
/lib/modules/2.4.18-2.ll/kernel/drivers/sound/pci/rme9652/snd-rme9652.o
depmod: 	snd_hammerfall_get_buffer_R511cb671
depmod: 	snd_hammerfall_free_buffer_R4a11134f

Maybe something's missing in the recent Hammerfall updates?
This is being compiled under 2.4.18 + 19-pre7 + low latency,
full sources for the kernel are properly installed.

-- Fernando

_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: bandwidth@sourceforge.net

RE: Failed to compile HDSP driver from the CVS

[Ivica Bukvic]

> -----Original Message-----
> From: Paul Davis [mailto:pbd@op.net]
> Sent: Tuesday, May 14, 2002 8:02 PM
> To: Ivica Bukvic
> Cc: alsa-devel@lists.sourceforge.net
> Subject: Re: Failed to compile HDSP driver from the CVS
> 
> >I am a bit confused. It is not the fact that I do have this file, but
> >that the compile was looking for it. It must have been that issue you
> >mentioned about CVS being not up-to-date.
> 
> its a combination of the "fastdep" target looking at *all* existing
> sources, whether they are meant to be there or not, plus the CVS sync
> issue.
> 
> 					   However, here's the error
when
> >I try to compile the driver now, after updating it with
> 
> use cvscompile to build it, i think.
> 
> --p

But all this still comes up with the same error, so this does not seem
to be the solution. The compiling still stops with the same error which
I am unable to figure out since it has to do with some define statements
at the beginning of the hammerfall_mem.c file:

make  -C rme9652 modules
make[2]: Entering directory
`/LocalDump/Install/Alsa/CVS/alsa-driver/pci/rme9652'
gcc -DALSA_BUILD -D__KERNEL__ -DMODULE=1
-I/LocalDump/Install/Alsa/CVS/alsa-driver/include
-I/lib/modules/2.4.8-26mdk/build/include -O2
-mpreferred-stack-boundary=2 -march=i586 -DLINUX -Wall
-Wstrict-prototypes -fomit-frame-pointer -pipe   -DEXPORT_SYMTAB -c
hammerfall_mem.c
In file included from hammerfall_mem.c:1:
../../alsa-kernel/pci/rme9652/hammerfall_mem.c:58: parse error before
string constant
../../alsa-kernel/pci/rme9652/hammerfall_mem.c:58: warning: type
defaults to `int' in declaration of `MODULE_LICENSE'
../../alsa-kernel/pci/rme9652/hammerfall_mem.c:58: warning: function
declaration isn't a prototype
../../alsa-kernel/pci/rme9652/hammerfall_mem.c:58: warning: data
definition has no type or storage class
make[2]: *** [hammerfall_mem.o] Error 1
make[2]: Leaving directory
`/LocalDump/Install/Alsa/CVS/alsa-driver/pci/rme9652'
make[1]: *** [_modsubdir_rme9652] Error 2
make[1]: Leaving directory `/LocalDump/Install/Alsa/CVS/alsa-driver/pci'
make: *** [compile] Error 1

It is obvious that it is having a problem with the hammerfall_mem.c
file, not the addition or removal of rme9652_mem.c file any more.

Does anyone else seem to have a problem with this one or is it just me?

Ico



_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: bandwidth@sourceforge.net

RE: Failed to compile HDSP driver from the CVS -- solution?

[Ivica Bukvic]

Hmmm, when I comment out the line 58 in the hammerfall_mem.c which says

MODULE_LICENSE("GPL");

Then everything compiles fine.

?

Ico




_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: bandwidth@sourceforge.net

HDSP -- how to configure modules.conf?

[Ivica Bukvic]

Sorry Paul for cluttering your inbox :-(

Just a quick question:

Do I use snd-rme9652.o module for hdsp or which one?

Obviously, I also need to modprobe snd-hammerfall-mem.o.

Could you please just post your example of the modules.conf for this
beast so that I know for sure that there are no extra flags I need to
worry about.

Thanks a lot!

Ico




_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: bandwidth@sourceforge.net

Re: HDSP -- how to configure modules.conf?

[Paul Davis]

>Sorry Paul for cluttering your inbox :-(
>
>Just a quick question:
>
>Do I use snd-rme9652.o module for hdsp or which one?

snd-hdsp.o

>Obviously, I also need to modprobe snd-hammerfall-mem.o.

yes, and almost certainly early in the boot sequence. i use
/etc/rc.d/rc.sysinit, and do it before running fsck.

>Could you please just post your example of the modules.conf for this
>beast so that I know for sure that there are no extra flags I need to
>worry about.

this is the audio-related section. the only option you need worry
about is the snd-enable option to snd-hammerfall-mem.o, and even that
is optional because by default the module will just allocate memory
for every Hammerfall or Hammerfall DSP device that it finds.

--p

----------------------------------------------------------------------
# Audio

# soundcore is the multiplexer for any and all sound drivers
# and ALSA hangs from major device 116

alias char-major-14 soundcore
alias char-major-116 snd

# soundcore/OSS service number notes:
# 0: /dev/mixer
# 1: /dev/sequencer
# 2: /dev/midi
# 3: /dev/dsp
# 4: /dev/audio
# 5: /dev/dspW
# 6: sndstat
# 8: /dev/music
# 9: /dev/synth
# 10: /dev/dmfm
# 12: /dev/adsp

# Four soundcards. 
# OSS-style opens look for sound-slot-N
# ALSA-style opens look for snd-card-N

alias sound-slot-0 snd-card-0
alias sound-slot-1 snd-card-1
alias sound-slot-2 snd-card-2
alias sound-slot-3 snd-card-3

alias snd-card-0 snd-wavefront
alias snd-card-1 snd-trident
alias snd-card-2 snd-rme9652
alias snd-card-3 snd-hdsp

# First soundcard is a Turtle Beach Tropez+, which has a CS4232 (with
# an ad1848 CODEC), the usual OPL3, and an ICS2115 wavetable
# synthesizer with its own MIDI interface.  

# Here, we leave the CS4232 MIDI interface undefined, since it uses a
# physical interface that is only accessible from within the case of
# the machine and I'm not interested in using it. To use it, set:
#
#	snd_cs4232_mpu_port
#	snd_cs4232_mpu_irq

options snd-wavefront snd_id="Tropez+"
options snd-wavefront-synth \
	reset_time=100 \
	debug_default=0xffff \
	wf_raw=1

alias sound-service-0-0 snd-mixer-oss
alias sound-service-0-3 snd-pcm-oss
alias sound-service-0-12 snd-pcm-oss

# Second soundcard is HoonTech SoundWave 4D-NX, based on
# the Trident 4D-NX chip. 

options snd-trident snd_id="4D"

alias sound-service-1-0  snd-mixer-oss
alias sound-service-1-3  snd-pcm-oss
alias sound-service-1-12 snd-pcm-oss

# Both the RME9652 and Hammerfall DSP drivers use
# the snd-hammerfall_mem module.

options snd-hammerfall_mem snd_enable=1,1

# Third card is an RME Digi9652 (Hammerfall)
# It offers nothing but pure digital PCM 

options snd-rme9652 snd_id="9652"

# Fourth card is an RME Hammerfall DSP card
# It offers variety of things

options snd-hdsp snd_id="hdsp"

_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: bandwidth@sourceforge.net

RE: HDSP -- how to configure modules.conf?

[Ivica Bukvic]

> >Do I use snd-rme9652.o module for hdsp or which one?
> 
> snd-hdsp.o

Problem is there is no such module that gets compiled. Tried compiling
on two different boxes with multiple CVS updates and complete nukes.

After installing the whole thing, in the alsa-driver/modules/ there is
no such module, nor in the /lib/modules/...

Again, has anyone else tried compiling the driver from the CVS and had a
success?

Ico



_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: bandwidth@sourceforge.net

Re: HDSP -- how to configure modules.conf?

[Paul Davis]

>> >Do I use snd-rme9652.o module for hdsp or which one?
>> 
>> snd-hdsp.o
>
>Problem is there is no such module that gets compiled. Tried compiling
>on two different boxes with multiple CVS updates and complete nukes.
>
>After installing the whole thing, in the alsa-driver/modules/ there is
>no such module, nor in the /lib/modules/...
>
>Again, has anyone else tried compiling the driver from the CVS and had a
>success?

How did you compile? I mean right from the very beginning ... I've
obviously been building from CVS all along, and I assume that Jaroslav
did a rebuild after he committed the changes.

--p

_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: bandwidth@sourceforge.net

RE: HDSP -- how to configure modules.conf?

[Ivica Bukvic]

> How did you compile? I mean right from the very beginning ... I've
> obviously been building from CVS all along, and I assume that Jaroslav
> did a rebuild after he committed the changes.
> 
> --p

Thank you for your continuous help!

Here's what I tried so far:

Did a clean cvs checkout on several occasions (last being today
afternoon):

1) used "build" script in the root dir of the cvs:
./build prep
./build config
./build all

2) used simple "make" in the root dir

3) did "cvscompile in the alsa-driver/ dir

The following 2 I did after the above one (once I got all the configure
files made with the above process and saw that it did not compile
properly)

4) did "./config --use-sequencer=yes" in the alsa-driver/ dir

5) did "./config --with-cards=hdsp"

All of them gave me the same error (the one I sent you) except the last
one which reported that there is no such card.

In between I did clean CVS-checkouts and/or updates just to make sure it
worked. Finally, I messed with the code since the error always reported
error in the hammerfall_mem.c file on the line 58. I commented that line
out and suddenly it compiled miraculously. I can even modprobe
hammerfall_mem.o without any problem and the built-in soundcard (Dell
Inspiron 8000) maestro3.o starts up just fine, but neither in the
/lib/modules/... nor in the <CVS>/alsa-driver/modules/ there is no sign
of snd-hdsp.o. These are the files I do have there:

CVS/                  snd-es1688.o@          snd-pdplus.o@
snd-ac97-codec.o@     snd-es18xx.o@          snd-rawmidi.o@
snd-ad1816a-lib.o@    snd-es1938.o@          snd-rme32.o@
snd-ad1816a.o@        snd-es1968.o@          snd-rme9652.o@
snd-ad1848-lib.o@     snd-es968.o@           snd-rme96.o@
snd-ad1848.o@         snd-fm801.o@           snd-sb16-csp.o@
snd-ainstr-fm.o@      snd-gusclassic.o@      snd-sb16-dsp.o@
snd-ainstr-gf1.o@     snd-gusextreme.o@      snd-sb16.o@
snd-ainstr-iw.o@      snd-gus-lib.o@         snd-sb8-dsp.o@
snd-ainstr-simple.o@  snd-gusmax.o@          snd-sb8.o@
snd-ak4531-codec.o@   snd-gus-synth.o@       snd-sbawe.o@
snd-ali5451.o@        snd-hammerfall-mem.o@  snd-sb-common.o@
snd-als100.o@         snd-hwdep.o@           snd-seq-device.o@
snd-als4000.o@        snd-i2c.o@             snd-seq-instr.o@
snd-azt2320.o@        snd-ice1712.o@         snd-seq-midi-emul.o@
snd-cmi8330.o@        snd-intel8x0.o@        snd-seq-midi-event.o@
snd-cmipci.o@         snd-interwave.o@       snd-seq-midi.o@
snd-cs4231-lib.o@     snd-interwave-stb.o@   snd-seq.o@
snd-cs4231.o@         snd-korg1212.o@        snd-seq-oss.o@
snd-cs4232.o@         snd-maestro3.o@        snd-seq-virmidi.o@
snd-cs4236-lib.o@     snd-mixer-oss.o@       snd-serialmidi.o@
snd-cs4236.o@         snd-mpu401.o@          snd-serial-u16550.o@
snd-cs4281.o@         snd-mpu401-uart.o@     snd-sgalaxy.o@
snd-cs46xx.o@         snd-mtpav.o@           snd-sonicvibes.o@
snd-cs8427.o@         snd-nm256.o@           snd-tea6330t.o@
snd-dt0197h.o@        snd.o@                 snd-timer.o@
snd-dummy.o@          snd-opl3-lib.o@        snd-trident.o@
snd-emu10k1.o@        snd-opl3sa2.o@         snd-trident-synth.o@
snd-emu10k1-synth.o@  snd-opl3-synth.o@      snd-util-mem.o@
snd-emu8000-synth.o@  snd-opti92x-ad1848.o@  snd-via686.o@
snd-emux-synth.o@     snd-opti92x-cs4231.o@  snd-via8233.o@
snd-ens1370.o@        snd-opti93x.o@         snd-virmidi.o@
snd-ens1371.o@        snd-pcm.o@             snd-wavefront.o@
snd-es1688-lib.o@     snd-pcm-oss.o@         snd-ymfpci.o@

(all of these are sym-links to different folders where these files
reside i.e. ../driver/ etc.)


ALSA-CVS/alsa-driver/pci/rme9652 folder has the following entries, and
as you can see all the files have the .o counterparts except for the
hdsp.c:

CVS/              hdsp.c     rme9652.o
hammerfall_mem.c  Makefile   snd-hammerfall-mem.o
hammerfall_mem.o  rme9652.c  snd-rme9652.o

So, I have no clue as to what is going on. I will check the makefile in
the rme9652 folder to see if it even mentions building hdsp.c file.

Finally, my system is MDK 8.1 (slightly modified) Kernel 2.4.8 (I know,
I've been lazy with my updating of the kernel).

Thank you for your help! Sincerely,

Ico




_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: bandwidth@sourceforge.net

Re: HDSP -- how to configure modules.conf?

[Paul Davis]

>> How did you compile? I mean right from the very beginning ... I've
>> obviously been building from CVS all along, and I assume that Jaroslav
>> did a rebuild after he committed the changes.
>> 
>> --p
>
>Thank you for your continuous help!

well, after some more digging, it seems that the problem is the
acinclude file in alsa-driver, which is what defines the list of cards
that exist and sets them up to be compiled.

i believed that it was stored in CVS, but it apparently is not. it
needs modifications that i believe i sent to jaroslav, but because its
not in CVS, i did not ("cvs -z3 diff" didn't notice it).

instead, it seems that acinclude is built from several files called
Module.deps by a utility call mod-deps. i don't understand when its
built. 

jaroslav or takashi or someone who understands how this works: would
you please modify Modules.dep or whatever needs to be done to fix this
issue? the existing Modules.dep file doesn't contain entries for the
H-DSP, and it continue to name snd-rme9652-mem, rather than
snd-hammerfall-mem. 

it worked for me because i had edited acinclude by hand, and for some
reason that i don't understand it was never rebuilt by the automated
process (which would have removed the hdsp entries, i think).

--p


_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: bandwidth@sourceforge.net