Hi, I received one of these 4-port NICs today but have had a thoroughly disastrous time with it :( In short, it's very easy to make any port hang. I'm running with vanilla kernel 2.4.26 and the sundance driver - if I log into another machine and issue a flood ping directed at the IP address of one of the NIC's ports, within 10 seconds the machine will see no more packets. Using vmstat during the start of the flood shows a reliable 10000-15000 interrupts per second, which then drops to the usual background 160 after the card 'stops'. I've tried to use the MMIO mode of the sundance driver but this doesn't work at all - it shows the MAC address of each of the 4 ports as something like '6f:ef:6f:ef:6f:ef' and fails to insert. Using Becker's latest driver produces another bad result : pci-scan.c:v1.12 7/30/2003 Donald Becker <becker@scyld.com> http://www.scyld.com/linux/drivers.html sundance.c:v1.11 2/4/2003 Written by Donald Becker <becker@scyld.com> http://www.scyld.com/network/sundance.html eth1: D-Link DFE-580TX (Kendin/Sundance ST201 Alta) at 0xd000, 00:0d:88:c5:95:c0, IRQ 12. invalid operand: 0000 CPU: 0 EIP: 0010:[<d0808204>] Not tainted EFLAGS: 00010246 eax: 00000020 ebx: 00000000 ecx: 00000005 edx: cf8f7800 esi: cf8f7800 edi: ffffffff ebp: cf940200 esp: cf57deb0 ds: 0018 es: 0018 ss: 0018 Process insmod (pid: 502, stackpage=cf57d000) Stack: 00000000 0000d000 00000005 0000d000 cf8f7800 c01c6f9c c137ac00 ffffffff d080631c c137ac00 00000000 0000d000 0000000c 00000000 00000000 d0808000 00000001 00000001 ffffffea 00000000 d080a9a8 c137ac00 00000000 00000000 Call Trace: [<c01c6f9c>] [<d080631c>] [<d080a9a8>] [<d0809d60>] [<d080aaa0>] [<c011234d>] [<d0808060>] [<c0106ac3>] Code: 0f 45 7d 1c 89 7c 24 1c 83 7c 24 38 07 7f 15 8b 44 24 38 83 The killer is..... the card worked fine in another PC :( I have tried to change the PCI latency, upgrade the BIOS, twiddle with settings in the BIOS to do with PCI timing, etc. but none of it has made the slightest bit of difference. Since it's a 1U machine with a small PSU, I've even tried a (brand new) 300W PSU but again, nothing changed. Does anyone have any ideas? :) I've included the output of Becker's 'alta-diag' for the first port in case it helps... pebble:~# ./alta-diag -aem alta-diag.c:v2.03 11/18/2002 Donald Becker (becker@scyld.com) http://www.scyld.com/diag/index.html Index #1: Found a Sundance Technology Alta adapter at 0xd000. Station address 00:0d:88:c5:95:c0. Receive mode is 0x15: Normal unicast and hashed multicast. MAC mode is 0020: full duplex. Tx status 00, threshold 8188. Interrupt status is 0040: Software triggered interrupt EEPROM address length 8, 256 words. EEPROM Subsystem IDs, Vendor 1186 Device 1012. EEPROM Station address is 00:0d:88:c5:95:c0. Configuration 2afc, ASIC Control c063. Checksum ceff MII PHY found at address 1, status 0x7809. MII PHY #1 transceiver registers: 3100 7809 0243 0c40 05e1 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 7000 4040 0010 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000. Cheers, Gavin. - : send the line "unsubscribe linux-net" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
