Skip to contentSkip to conversation

config bug with sbawe and --with-isapnp=no

DISCUSSION Mario Loritz 3 participants 10 messages Jan 22, 2004 – Jan 29, 2004

lore
Share & tools
lore
#0 lore raw

Hello List,

Since about Alsa version 0.9.8, the compilation of the alsa-driver for the Sound Blaster AWE (non-pnp) results in a non-usable driver.

The driver fails to load with a message about a symbol error. (Sorry, lost the output, have to regenerate it.)

If I compile the driver with pnp turned on, then I can load the driver with the option "isapnp=no", which didn't work BEFORE alsa 0.9.8.

So we replaced one bug with another, didn't we?

Thanks for reading,
Mario

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

#1 lore raw

At Thu, 22 Jan 2004 12:49:07 +0100,

Quoted 6 lines · Mario Loritz

Hello List,

Since about Alsa version 0.9.8, the compilation of the alsa-driver for the Sound Blaster AWE (non-pnp) results in a non-usable driver.

Quoted 7 lines

The driver fails to load with a message about a symbol error. (Sorry, lost the output, have to regenerate it.)

If I compile the driver with pnp turned on, then I can load the driver with the option "isapnp=no", which didn't work BEFORE alsa 0.9.8.

So we replaced one bug with another, didn't we?

likely.

please give more information: which kernel and how is the kernel config, which symbols are undefined, which configure options you gave, etc. more detail is more helpful.

Takashi

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

#2 lore raw
In reply to Mario Loritz

Hello list,

I reproduced the error message while the driver is loading:

# rcalsasound restart

Shutting down sound driver: done
Starting sound driver: snd-sbawe /lib/modules/2.4.21-166-default/kernel/sound/ isa/sb/snd-sbawe.o: unresolved symbol pnp_activate_dev_R9e329e8c /lib/modules/2.4.21-166-default/kernel/sound/isa/sb/snd-sbawe.o: unresolved symbol pnp_register_card_driver_R14e159bf /lib/modules/2.4.21-166-default/kernel/sound/isa/sb/snd-sbawe.o: unresolved symbol pnp_manual_config_dev_R480f17aa /lib/modules/2.4.21-166-default/kernel/sound/isa/sb/snd-sbawe.o: unresolved symbol pnp_release_card_device_R0d53db9b /lib/modules/2.4.21-166-default/kernel/sound/isa/sb/snd-sbawe.o: unresolved symbol pnp_init_resource_table_R16084899 /lib/modules/2.4.21-166-default/kernel/sound/isa/sb/snd-sbawe.o: unresolved symbol pnp_unregister_card_driver_R82704514 /lib/modules/2.4.21-166-default/kernel/sound/isa/sb/snd-sbawe.o: unresolved symbol pnp_request_card_device_Rc3534026 /lib/modules/2.4.21-166-default/kernel/sound/isa/sb/snd-sbawe.o: insmod /lib/ modules/2.4.21-166-default/kernel/sound/isa/sb/snd-sbawe.o failed /lib/modules/2.4.21-166-default/kernel/sound/isa/sb/snd-sbawe.o: insmod snd-sbawe failed
done
Starting sound driver: snd-ice1712 done

What is strange is that the drivers try to use symbols like "pnp_activate" that they dont need and dont use, as the drivers were compiled WITHOUT pnp support.

Thanks for the answers so far!
Mario

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

#3 lore raw
Quoted 3 lines · Mario Loritz

Hello list,

I reproduced the error message while the driver is loading:

Could you try the patch bellow?

Jaroslav

Index: pnp.h
===================================================================
RCS file: /cvsroot/alsa/alsa-driver/support/pnp/pnp.h,v retrieving revision 1.4

pnp.h +1 -1
diff -u -r1.4 pnp.h
@@ -186,7 +186,7 @@
 
 #define pnp_resource_change(resource, start, size) isapnp_resource_change(resource, start, size)
 
-#if defined(CONFIG_PNP)
+#if defined(CONFIG_PNP) && (!defined(ALSA_BUILD) || defined(CONFIG_PNP_KERNEL))
 
 struct pnp_dev * pnp_request_card_device(struct pnp_card_link *clink, const char * id, struct pnp_dev * from);
 void pnp_release_card_device(struct pnp_dev * dev);

-----
Jaroslav Kysela <[email protected]>
Linux Kernel Sound Maintainer
ALSA Project, SuSE Labs

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

#4 lore raw

At Fri, 23 Jan 2004 11:26:38 +0100 (CET),

Quoted 8 lines · Jaroslav
Quoted 3 lines · Mario Loritz

Hello list,

I reproduced the error message while the driver is loading:

Could you try the patch bellow?

i found the bug in adriver.h, which always defines CONFIG_PNP when CONFIG_ISAPNP is set.

the attached patches are the fix and clean-up for pnp. please give a try.

Takashi

#5 lore raw

Am Freitag, 23. Januar 2004 12:49 schrieb Takashi Iwai:

Quoted 8 lines

i found the bug in adriver.h, which always defines CONFIG_PNP when CONFIG_ISAPNP is set.

the attached patches are the fix and clean-up for pnp. please give a try.

Takashi

After I applied the patches, something strange is happening:
The module option "isapnp=yes/no" is not recognized anymore, regardless of the configure option "--with-isapnp=yes" or "no".

The driver itself works, even if I compile with pnp. But why? I cannot tell the driver, that it shouldn't try pnp. Do you think the driver will still work when a pnp-card is used? I can't check this, because my sbawe32 is non-pnp.

When those questions are answered and the module option "isapnp=yes/no" really isn't needed anymore, then I think the driver should be changed in the proposed way.

Thanks for the solution so far.
Mario

Some parts of the output of "./configure --with-cards=sbawe,ice1712":
...
checking for directory with kernel source... /lib/modules/2.4.21-166-default/ build
checking for kernel version... 2.4.21-166-default
checking for GCC version... Kernel compiler: gcc 3.3.1 (SuSE Linux) Used
compiler: gcc (GCC) 3.3.1 (SuSE Linux)
checking for built-in ALSA... "no"
checking for existing ALSA module... "yes"
checking for Red Hat kernel... "auto"
checking for Red Hat kernel... "no"
...
checking for kernel module symbol versions... "yes"
checking for PCI support in kernel... "yes"
checking for SGI/MIPS (HAL2) architecture... "no"
checking for directory to store kernel modules... /lib/ modules/2.4.21-166-default/kernel/sound
checking for verbose printk... on
checking for debug level... none
checking for ISA support in kernel... "yes"
checking for processor type... "unknown"
checking for SMP... "no"
checking for ISA PnP driver in kernel... yes
checking for PnP driver in kernel... no
checking for Kernel ISA-PnP support... "no"
checking for Kernel ISA-PnP module support... "yes"
checking for ISA PnP support... yes
....

Output of "rcalsasound restart"

Shutting down sound driver: done
Starting sound driver: snd-sbawe Warning: ignoring isapnp=0, no such parameter in this module
Module snd-sbawe loaded, with warnings
done
Starting sound driver: snd-ice1712 done

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

#6 lore raw

At Sat, 24 Jan 2004 15:57:21 +0100,

Quoted 15 lines · Mario Loritz

Am Freitag, 23. Januar 2004 12:49 schrieb Takashi Iwai:

Quoted 8 lines

i found the bug in adriver.h, which always defines CONFIG_PNP when CONFIG_ISAPNP is set.

the attached patches are the fix and clean-up for pnp. please give a try.

Takashi

After I applied the patches, something strange is happening:
The module option "isapnp=yes/no" is not recognized anymore, regardless of the configure option "--with-isapnp=yes" or "no".

Quoted 8 lines

The driver itself works, even if I compile with pnp. But why? I cannot tell the driver, that it shouldn't try pnp. Do you think the driver will still work when a pnp-card is used? I can't check this, because my sbawe32 is non-pnp.

When those questions are answered and the module option "isapnp=yes/no" really isn't needed anymore, then I think the driver should be changed in the proposed way.

did you run cvscompile? otherwise it won't work properly.

Takashi

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

#7 lore raw

Am Montag, 26. Januar 2004 12:15 schrieb Takashi Iwai:

Quoted 1 line

did you run cvscompile? otherwise it won't work properly.

No, I didn't.
Mario

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

#8 lore raw

Thanks to all!

The sbawe driver is working in both configurations now. (alsa 1.0.2)

Mario

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

#9 lore raw
In reply to Mario Loritz

Thanks to all!

The sbawe driver is working in both configurations now. (alsa 1.0.2)

Mario

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn