Skip to contentSkip to conversation

Odd behavior of ssb, b43, b43legacy, and b44

DISCUSSION Larry Finger 2 participants 5 messages Jan 6, 2011 – Jan 7, 2011 Sonics silicon backplane driver (ssb)

lore
Share & tools
lore
#0 lore raw

Michael,

On one of my boxes, I have installed two PCI-format BCM43xx cards for testing. One is a BCM4306 Rev. 3, which uses b43. The other is a BCM4303, which uses b43legacy. The output of lspci -nn for these devices is

01:09.0 Network controller [0280]: Broadcom Corporation BCM4306 802.11b/g Wireless LAN Controller [14e4:4320] (rev 03)
01:0a.0 Network controller [0280]: Broadcom Corporation BCM4303 802.11b Wireless LAN Controller [14e4:4301] (rev 02)

Upon booting, I noticed the following messages in the log:

b44: b44.c:v2.0
b44: Invalid MAC address found in EEPROM
b44 ssb1:1: Problem fetching invariants of chip, aborting

b44: probe of ssb1:1 failed with error -22

As this box does not have a b44 installed, I wondered why this was happening. When I unloaded all the drivers and used modprobe to load ssb, I found that b43, b43legacy and b44 were all loaded. The console output is

finger at pam:~> lsmod | grep b4 <== none loaded
finger@pam:~> sudo modprobe -v ssb <== load ssb
insmod /lib/modules/2.6.37-wl+/kernel/drivers/ssb/ssb.ko

The above looks normal, but look at what is now resident!

finger@pam:~> lsmod | grep b4

b43legacy             115302  0
b44                    28767  0
b43                   174321  0
ssb                    38157  3 b43legacy,b44,b43
mac80211              266240  2 b43legacy,b43
cfg80211              161930  3 b43legacy,b43,mac80211

Any idea why loading ssb should silently load b43legacy AND b44? Any ideas on where to look?

Thanks,

Larry

#1 lore raw
Quoted 37 lines · Larry Finger

Michael,

On one of my boxes, I have installed two PCI-format BCM43xx cards for testing. One is a BCM4306 Rev. 3, which uses b43. The other is a BCM4303, which uses b43legacy. The output of lspci -nn for these devices is

01:09.0 Network controller [0280]: Broadcom Corporation BCM4306 802.11b/g Wireless LAN Controller [14e4:4320] (rev 03)
01:0a.0 Network controller [0280]: Broadcom Corporation BCM4303 802.11b Wireless LAN Controller [14e4:4301] (rev 02)

Upon booting, I noticed the following messages in the log:

b44: b44.c:v2.0
b44: Invalid MAC address found in EEPROM
b44 ssb1:1: Problem fetching invariants of chip, aborting

b44: probe of ssb1:1 failed with error -22

As this box does not have a b44 installed, I wondered why this was happening. When I unloaded all the drivers and used modprobe to load ssb, I found that b43, b43legacy and b44 were all loaded. The console output is

finger at pam:~> lsmod | grep b4 <== none loaded
finger at pam:~> sudo modprobe -v ssb <== load ssb
insmod /lib/modules/2.6.37-wl+/kernel/drivers/ssb/ssb.ko

The above looks normal, but look at what is now resident!

finger at pam:~> lsmod | grep b4

b43legacy             115302  0
b44                    28767  0
b43                   174321  0
ssb                    38157  3 b43legacy,b44,b43
mac80211              266240  2 b43legacy,b43
cfg80211              161930  3 b43legacy,b43,mac80211
Quoted 7 lines

Any idea why loading ssb should silently load b43legacy AND b44? Any ideas on where to look?

Thanks,

Larry

Does one of these wireless cards have a dangling ethernet core? I would not be surprised...

Signature · 1 line
Greetings Michael.
#2 lore raw
Quoted 3 lines · Michael Büsch

Does one of these wireless cards have a dangling ethernet core? I would not be surprised...

Yes. The core scan for the BCM4303 is as follows:

ssb: Core 0 found: IEEE 802.11 (cc 0x812, rev 0x02, vendor 0x4243)
ssb: Core 1 found: PCMCIA (cc 0x80D, rev 0x00, vendor 0x4243)
ssb: Core 2 found: Fast Ethernet (cc 0x806, rev 0x02, vendor 0x4243)
ssb: Core 3 found: V90 (cc 0x807, rev 0x01, vendor 0x4243)
ssb: Core 4 found: PCI (cc 0x804, rev 0x03, vendor 0x4243)
ssb: Sonics Silicon Backplane found on PCI device 0000:01:09.0

Larry

#3 lore raw
Quoted 7 lines · Larry Finger
Quoted 3 lines · Michael B?sch

Does one of these wireless cards have a dangling ethernet core? I would not be surprised...

Yes. The core scan for the BCM4303 is as follows:

Quoted 9 lines

ssb: Core 0 found: IEEE 802.11 (cc 0x812, rev 0x02, vendor 0x4243)
ssb: Core 1 found: PCMCIA (cc 0x80D, rev 0x00, vendor 0x4243)
ssb: Core 2 found: Fast Ethernet (cc 0x806, rev 0x02, vendor 0x4243)
ssb: Core 3 found: V90 (cc 0x807, rev 0x01, vendor 0x4243)
ssb: Core 4 found: PCI (cc 0x804, rev 0x03, vendor 0x4243)
ssb: Sonics Silicon Backplane found on PCI device 0000:01:09.0

Larry

Can you please try this patch?

drivers/ssb/scan.c +10 -0
Index: linux-2.6.37/drivers/ssb/scan.c
===================================================================
@@ -420,6 +420,16 @@
 			bus->pcicore.dev = dev;
 #endif /* CONFIG_SSB_DRIVER_PCICORE */
 			break;
+		case SSB_DEV_ETHERNET:
+			if (bus->bustype == SSB_BUSTYPE_PCI) {
+				if (bus->host_pci->vendor == PCI_VENDOR_ID_BROADCOM &&
+				    (bus->host_pci->device & 0xFF00) == 0x4300) {
+					/* This is a dangling ethernet core on a
+					 * wireless device. Ignore it. */
+					continue;
+				}
+			}
+			break;
 		default:
 			break;
 		}
Signature · 1 line
Greetings Michael.
#4 lore raw
Quoted 34 lines · Michael Büsch
Quoted 7 lines · Larry Finger
Quoted 3 lines · Michael Büsch

Does one of these wireless cards have a dangling ethernet core? I would not be surprised...

Yes. The core scan for the BCM4303 is as follows:

Quoted 9 lines

ssb: Core 0 found: IEEE 802.11 (cc 0x812, rev 0x02, vendor 0x4243)
ssb: Core 1 found: PCMCIA (cc 0x80D, rev 0x00, vendor 0x4243)
ssb: Core 2 found: Fast Ethernet (cc 0x806, rev 0x02, vendor 0x4243)
ssb: Core 3 found: V90 (cc 0x807, rev 0x01, vendor 0x4243)
ssb: Core 4 found: PCI (cc 0x804, rev 0x03, vendor 0x4243)
ssb: Sonics Silicon Backplane found on PCI device 0000:01:09.0

Larry

Can you please try this patch?

Index: linux-2.6.37/drivers/ssb/scan.c
===================================================================

drivers/ssb/scan.c +3 -0
@@ -420,6 +420,16 @@
 			bus->pcicore.dev = dev;
 #endif /* CONFIG_SSB_DRIVER_PCICORE */
 			break;
+		case SSB_DEV_ETHERNET:
+			if (bus->bustype == SSB_BUSTYPE_PCI) {
+				if (bus->host_pci->vendor == PCI_VENDOR_ID_BROADCOM &&

drivers/ssb/scan.c line 435

Quoted 10 lines
drivers/ssb/scan.c +7 -0
+				    (bus->host_pci->device & 0xFF00) == 0x4300) {
+					/* This is a dangling ethernet core on a
+					 * wireless device. Ignore it. */
+					continue;
+				}
+			}
+			break;
 		default:
 			break;
 		}

That patch fixes the problem. Thanks.

Larry