Linksys WMP11 + Soundcard conflict

phoenixx · 09-09-2005, 02:36 AM

Ok...I'm completely lost. I have a Linksys wireless PCI card (model WMP11) in my Slackware system that is having a crazy fit with my sound card. The wireless card works perfectly when my sound is off. However, when I turn the sound playback on, my card seems to stop working. If I stop playback, my card instantly starts working again. (By start/stop playback, I actually mean pressing play and stop/pause in xmms ... nothing more complicated than that...). The strangest part is that I get no error messages (dmesg shows nothing). It's almost as if the card was designed to work in this way. I believe it's a strict hardware issue with the card because i've tried this particular card with several different PCs all with different soundcards/motherboards/OS's and the problem is common to all. Anybody have any clue as to what is happening? Thanks in advance.

-phoenixx

Electro · 09-11-2005, 12:47 AM

I suggest taking out the PCI wireless NIC and use a wired to wireless bridge instead. I heard that any PCI wireless NIC models and brands have the same problems. By using a wired to wireless bridge it gives you a few options to choose for a wired network. One is connecting it to a hub or switch. Another option is connecting it straight to a wired NIC. Third option is providing wireless networking to other devices such as printers.

I recommend people use wired networks instead of wireless networks because it is a lot less risky, provides a reliable connection, and its bandwidth is several times faster. The cost of a wired network is a lot cheaper than a wireless network. An admin will have an endless battle of securing a wireless network.

Snowbat · 09-11-2005, 04:10 PM

It sounds like an IRQ conflict. Check in /proc/interrupts if both devices are sharing a single interrupt. On many motherboards you can solve this by moving one of the cards to a different slot. Your BIOS may offer IRQ assignment options too. Check the motherboard manual for interrupt assignment details.

phoenixx · 09-11-2005, 10:23 PM

I don't think so...unless I'm missing some small detail

Code:

$ cat /proc/interrupts:

           CPU0
  0:   32764748    IO-APIC-edge  timer
  1:       9182    IO-APIC-edge  i8042
  7:          0    IO-APIC-edge  parport0
  9:          0   IO-APIC-level  acpi
 14:     990650    IO-APIC-edge  ide0
 15:         89    IO-APIC-edge  ide1
 16:     590420   IO-APIC-level  ehci_hcd:usb1, uhci_hcd:usb2, uhci_hcd:usb3, uhci_hcd:usb4
 17:    5198152   IO-APIC-level  eth0
 18:     482877   IO-APIC-level  EMU10K1
 19:    9259161   IO-APIC-level  radeon@pci:0000:01:00.0
NMI:          0
LOC:   33135687
ERR:          0
MIS:          0

Snowbat · 09-12-2005, 08:24 PM

Ah, you have APIC interrupts. You may find that APIC 17 and APIC 18 share a single physical interrupt on the PCI bus but I don't know where you can query this. Try moving one of the cards to a different PCI slot and see if matters improve.