Bug 21304
| Summary: | [dc] dc0 watchdog timeouts on NetGear FA310TX | ||
|---|---|---|---|
| Product: | Base System | Reporter: | david <david> |
| Component: | kern | Assignee: | Martin Blapp <mbr> |
| Status: | Closed FIXED | ||
| Severity: | Affects Only Me | ||
| Priority: | Normal | ||
| Version: | 4.1-STABLE | ||
| Hardware: | Any | ||
| OS: | Any | ||
Two things I forgot to mention: 1) I tried this with two different (known working) cables on two different (known working) hubs. You also get the same errors to console with _nothing_ plugged into the card. 2) I forgot to put the errors being logged! Sorry: dc0: watchdog timeout dc0: failed to force tx and rx to idle state alternating, about once or twice a second. --David Bushong Ho hum. Someone just mentioned on -stable, I believe that he's seen 4 bad FA310TX's in 6 months. If no one else pops up with this problem (which I only reported because it was identical to a problem that was cropping up around September 7th with dc0), I'll just pop off and return the stupid card and go hunt up an fxp1 to stick in this box. Responsible Changed From-To: freebsd-bugs->wpaul Over to Bill who has done alot of recent changes to dc(4). Bill are you the dc maintainer? I'm very sorry to have wasted anyone's time. It would appear the dc0 was trying to share irq 10 with my fxp0. (Slots 4 and 5 on my P5A motherboard share IRQs, sigh). Very sorry, please close this entry. --David Bushong The GENERIC kernel for 4.7-RELEASE also exhibits frequent dc0 watchdog timeouts. The system loaded easily enough, but the problem appeared promptly after going multi-user. The hardware is older yet similar to David's: K6/200 on an Asus T2P4; only other card besides this Kingston KNE110TX on the PCI bus is a Matrox Millenium. atapci0: <Intel PIIX3 ATA controller> port 0xe800-0xe80f at device 7.1 on pci0 ata0: at 0x1f0 irq 14 on atapci0 ata1: at 0x170 irq 15 on atapci0 pci0: <Matrox MGA Millennium 2064W graphics accelerator> at 9.0 irq 10 dc0: <82c169 PNIC 10/100BaseTX> port 0xe000-0xe0ff mem 0xe6000000-0xe60000ff irq 11 at device 11.0 on pci0 OK, another ``never mind, my bad.'' The fxp runs fine on a 41.7 MHz PCI, but the KNE110TX cannot cope. I forgot to reset things to spec on resurrecting this old system. Responsible Changed From-To: wpaul->mbr Take this PR over from wpaul. State Changed From-To: open->analyzed Hi, is your problem still persistent or fixed with newer releases of FreeBSD ? Thank you ! State Changed From-To: analyzed->closed Feedback timeout (> 1 year). |
While others with the exact same card have reported no problems, any attempt to ifconfig dc0 results in a stream of watchdog timeout messages to the console. (Additionally, no traffic can be sent) dmesg lines look like: dc0: <82c169 PNIC 10/100BaseTX> port 0xa400-0xa4ff mem 0xdd800000-0xdd8000ff irq 10 at device 13.0 on pci0 dc0: Ethernet address: 00:a0:cc:d3:54:45 miibus0: <MII bus> on dc0 ukphy0: <Generic IEEE 802.3u media interface> on miibus0 ukphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto I have two other cards in this box: an intel eepro (fxp0) and a 3c905 (xl0). The ifconfig looks like this when it's up: dc0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500 inet 196.168.1.2 netmask 0xffffff00 broadcast 196.168.1.255 ether 00:a0:cc:d3:54:45 media: autoselect (10baseT/UTP) status: active supported media: autoselect 100baseTX <full-duplex> 100baseTX 10baseT/UTP <full-duplex> 10baseT/UTP none How-To-Repeat: (My guess): Make a system out of: K6-2/400 Asus P5A motherboard put in it: Matrox Millenium II (PCI), Intel EtherExpress Pro 10/100, 3Com 3c905 (no A, B, or C), NetGear FA310TX, 1IDE drive, 1IDE cdrom. Build a kernel with dc0, xl0, miibus, and fxp0 built-in, as well as IPFIREWALL, IPFILTER (trying them both out), BRIDGE, NETGRAPH, IPSTEALTH, and other fairly standard stuff. Try to use dc0. This is with stable as of the date and time given in the Environment (~3pm PDT on Sept 15th, 2000)