Problem
I am using the Intel eeE driver on an Intel chipset 82559ER and during heavy TCP/IP load my server panics with the following example panic stack traces:
debug0:1> stack
E0000C78 scodb_call(? 4, 0, F019B350, E) <- cmn_err+203
E0000C9C cmn_err(3, F0164358, E, E0000CDC) <- k_trap+2D8
E0000CD0 k_trap(? E0000CDC, 0, 0, 160) <- kern_trap+B
kern_trap() <- INT [E0000CDC]
E0000D2C INTERRUPT <- freemsg+1C
E0000D2C freemsg(1201010A) <- in_freemsg+C
E0000D38 in_freemsg(1201010A, F105DDC0) <- ip_freef+26
E0000D54 ip_freef(F105DDC0, 11, F019B060) <- ip_slowtimo+6B
E0000D84 ip_slowtimo(0) <- in_slowtimo+19
E0000D98 in_slowtimo(? 0, E351, F0C02F40, F0161134) <- timein+82
E0000DC8 timein(? 0, 0, 160, 160) <- cmnintret+23
cmnintret() <- INT [E0000D4C]
E0000E24 INTERRUPT <- hlted
E0000E24 hlted(? F00E4971, 0, F0160C3C, F0058AC4) <- swtch+5There may be other reasons why a panic may be generated by freemsg other than the above.
Cause
The eeE driver supplied by SCO does not support the Intel 82559ER chipset (0x80861209) nor it is supported by Intel in the server environment.
This is because the chipset is designed for desktop machines, not server machines.
Solution
Install a supported NIC or supported chipset.
A complete list of certified hardware can be found at: http://www.sco.com/chwp
NOTE: This chipset is not supported for the UnixWare or OpenUnix 8 operating systems.
See also
ta101596, "Using Intel Pro/100B NIC: "WARNING: eeE: Allocb failure during initialization". ta109718, "Panic "Watchdog ReInitMemory for 0 Unexpected trap in Kernel Mode" with Intel Pro100b".