multicast join flood messes up sk0

[Sean McNeil]

My sk0 is rendered useless when flooded with multicast join requests.

Here is my setup:

FreeBSD server.mcneil.com 6.0-STABLE FreeBSD 6.0-STABLE #94: Mon Nov  7
23:51:05 PST 2005     root at server.mcneil.com:/usr/obj/usr/src/sys/AMD64
amd64

CPU: AMD Athlon(tm) 64 X2 Dual Core Processor 3800+ (2009.79-MHz
K8-class CPU)
  Origin = "AuthenticAMD"  Id = 0x20f32  Stepping = 2

Features=0x178bfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,MMX,FXSR,SSE,SSE2,HTT>
  Features2=0x1<SSE3>
  AMD Features=0xe2500800<SYSCALL,NX,MMX+,<b25>,LM,3DNow+,3DNow>
real memory  = 2147418112 (2047 MB)
avail memory = 2064441344 (1968 MB)
FreeBSD/SMP: Multiprocessor System Detected: 2 CPUs
 cpu0 (BSP): APIC ID:  0
 cpu1 (AP): APIC ID:  1
skc0: <Marvell Gigabit Ethernet> port 0xa800-0xa8ff mem
0xf5000000-0xf5003fff irq 19 at device 11.0 on pci2
skc0: Marvell Yukon Lite Gigabit Ethernet rev. (0x9)
sk0: <Marvell Semiconductor, Inc. Yukon> on skc0
sk0: Ethernet address: 00:14:85:85:27:b3
miibus1: <MII bus> on sk0
e1000phy0: <Marvell 88E1000 Gigabit PHY> on miibus1
e1000phy0:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX,
1000baseTX-FDX, auto

I'm streaming MPEG with vls to a target device that is running Linux.
Something goes wrong (not related) and I stop the streaming, but the
Linux target appears to start flooding multicast join requests.  When
this happens, my sk0 nic suddenly becomes useless.  I get messages like:

sk0: watchdog timeout
sk0: link state changed to DOWN
sk0: watchdog timeout

dhcpd: send_packet: No buffer space available

If I kill my Linux target, the interface recovers just fine.

The sk0 is attached to a gigE linksys switch which is attached to a
100BT switch which is attached to the Linux target.  There should be no
way that it could use up all the resources on my machine.

Sean