Hi, I'm experiencing heavy TX Unit hangs in a TYAN Toledo i3210W/i3200R S5211 with both onboard e1000 NICs since I bought the board in May 2008. These hangs therefore occur in several kernel versions now (2.6.31.*, 2.6.30.*, 2.6.29.*).
I replaced the unmanaged switch with a managed one without any improvement. Because of the new switch's trunking capability it runs in bonding mode now but the error also occurs without bonding on both interfaces. It doesn't occur with older 66MHz PCI based e1000 cards. I'm running a Gentoo box with a vanilla kernel. Kernel cfg can be provided just like any more information which could be helpful. Cheers, Vasco ------------[ cut here ]------------ WARNING: at net/sched/sch_generic.c:261 dev_watchdog+0x260/0x270() Hardware name: empty NETDEV WATCHDOG: eth1 (e1000e): transmit queue 0 timed out Modules linked in: md5 xts gf128mul tun nfsd lockd nfs_acl auth_rpcgss sunrpc af_packet bonding ipv6 coretemp lm85 hwmon_vid hwmon ext4 mbcache jbd2 crc16 ansi_cprng krng eseqiv rng aes_x86_64 aes_generic cbc cryptomgr aead pcompress dm_crypt cryptd crypto_hash crypto_wq crypto_blkcipher crypto_algapi fan cpufreq_ondemand acpi_cpufreq freq_table msr cpuid usbhid hid uhci_hcd ehci_hcd 8250_pnp usbcore 8250 sg rtc_cmos rtc_core rtc_lib serial_core i2c_i801 i2c_core nls_base container e1000e evdev pcspkr psmouse processor thermal button unix [last unloaded: microcode] Pid: 0, comm: swapper Not tainted 2.6.32 #1 Call Trace: <IRQ> [<ffffffff812d0790>] ? dev_watchdog+0x260/0x270 [<ffffffff8103e678>] warn_slowpath_common+0x78/0xd0 [<ffffffff8103e754>] warn_slowpath_fmt+0x64/0x70 [<ffffffff8103a590>] ? try_to_wake_up+0xa0/0x210 [<ffffffff8103a70d>] ? default_wake_function+0xd/0x10 [<ffffffff811b196a>] ? strlcpy+0x4a/0x60 [<ffffffff812bc023>] ? netdev_drivername+0x43/0x50 [<ffffffff812d0790>] dev_watchdog+0x260/0x270 [<ffffffff81050900>] ? delayed_work_timer_fn+0x0/0x40 [<ffffffff81050887>] ? __queue_work+0x77/0x90 [<ffffffff8103b875>] ? scheduler_tick+0x125/0x290 [<ffffffff812d0530>] ? dev_watchdog+0x0/0x270 [<ffffffff81048bd6>] run_timer_softirq+0x136/0x200 [<ffffffff8105fbaf>] ? clockevents_program_event+0x4f/0x90 [<ffffffff81044137>] __do_softirq+0xa7/0x130 [<ffffffff8100c33c>] call_softirq+0x1c/0x30 [<ffffffff8100e225>] do_softirq+0x55/0x90 [<ffffffff81043e15>] irq_exit+0x75/0x90 [<ffffffff8101f68d>] smp_apic_timer_interrupt+0x6d/0xa0 [<ffffffff8100bd33>] apic_timer_interrupt+0x13/0x20 <EOI> [<ffffffff81012e96>] ? mwait_idle+0x66/0x80 [<ffffffff8100a520>] ? enter_idle+0x20/0x30 [<ffffffff8100a593>] ? cpu_idle+0x63/0xb0 [<ffffffff8131f6d1>] ? rest_init+0x61/0x70 [<ffffffff81475c55>] ? start_kernel+0x300/0x3c3 [<ffffffff81475289>] ? x86_64_start_reservations+0x99/0xb9 [<ffffffff81475389>] ? x86_64_start_kernel+0xe0/0xf2 ---[ end trace 3cf2b6d30cbd8fab ]--- e1000e: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX/TX bonding: bond0: link status definitely up for interface eth0. 0000:0d:00.0: eth1: Detected Tx Unit Hang: TDH <b9> TDT <a6> next_to_use <a6> next_to_clean <b7> buffer_info[next_to_clean]: time_stamp <1011a215d> next_to_watch <b9> jiffies <1011a2262> next_to_watch.status <0> 0000:0d:00.0: eth1: Detected Tx Unit Hang: TDH <b9> TDT <a6> next_to_use <a6> next_to_clean <b7> buffer_info[next_to_clean]: time_stamp <1011a215d> next_to_watch <b9> jiffies <1011a2456> next_to_watch.status <0> bonding: bond0: link status definitely down for interface eth1, disabling it 0000:0f:00.0: eth0: Detected Tx Unit Hang: TDH <20> TDT <9> next_to_use <9> next_to_clean <1d> buffer_info[next_to_clean]: time_stamp <1023c04d4> next_to_watch <20> jiffies <1023c0649> next_to_watch.status <0> 0000:0f:00.0: eth0: Detected Tx Unit Hang: TDH <20> TDT <9> next_to_use <9> next_to_clean <1d> buffer_info[next_to_clean]: time_stamp <1023c04d4> next_to_watch <20> jiffies <1023c083d> next_to_watch.status <0> 0000:0d:00.0: eth0: (PCI Express:2.5GB/s:Width x1) 00:e0:81:b1:83:2e 0000:0d:00.0: eth0: Intel(R) PRO/1000 Network Connection 0000:0d:00.0: eth0: MAC: 2, PHY: 2, PBA No: ffffff-0ff e1000e 0000:0f:00.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17 e1000e 0000:0f:00.0: setting latency timer to 64 e1000e 0000:0f:00.0: irq 29 for MSI/MSI-X 0000:0f:00.0: eth1_rename: (PCI Express:2.5GB/s:Width x1) 00:e0:81:b1:83:2f 0000:0f:00.0: eth1_rename: Intel(R) PRO/1000 Network Connection 0000:0f:00.0: eth1_rename: MAC: 2, PHY: 2, PBA No: ffffff-0ff +-1c.4-[0000:0d]----00.0 Intel Corporation 82573V Gigabit Ethernet Controller (Copper) +-1c.5-[0000:0f]----00.0 Intel Corporation 82573L Gigabit Ethernet Controller continuity ~ # ethtool -i eth0 driver: e1000e version: 1.0.2-k2 firmware-version: 1.0-2 bus-info: 0000:0f:00.0 continuity ~ # ethtool -i eth1 driver: e1000e version: 1.0.2-k2 firmware-version: 1.0-2 bus-info: 0000:0d:00.0 continuity ~ # ethtool -g eth0 Ring parameters for eth0: Pre-set maximums: RX: 4096 RX Mini: 0 RX Jumbo: 0 TX: 4096 Current hardware settings: RX: 256 RX Mini: 0 RX Jumbo: 0 TX: 256 continuity ~ # ethtool -g eth1 Ring parameters for eth1: Pre-set maximums: RX: 4096 RX Mini: 0 RX Jumbo: 0 TX: 4096 Current hardware settings: RX: 256 RX Mini: 0 RX Jumbo: 0 TX: 256 continuity ~ # ethtool -a eth0 Pause parameters for eth0: Autonegotiate: on RX: on TX: on continuity ~ # ethtool -a eth1 Pause parameters for eth1: Autonegotiate: on RX: on TX: on ------------------------------------------------------------------------------ This SF.Net email is sponsored by the Verizon Developer Community Take advantage of Verizon's best-in-class app development support A streamlined, 14 day to market process makes app distribution fast and easy Join now and get one step closer to millions of Verizon customers http://p.sf.net/sfu/verizon-dev2dev _______________________________________________ E1000-devel mailing list E1000-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/e1000-devel