I recently created a test program that simply receives data as fast as possible and sends some data also via TCP using the Zedboard ethernet port running at a giga-bit.
Occasional I get the following error on the console. The only way to get networking to work again is to reboot? Anyone else seen this? Any solution?
The only other thing to note is CPU#0 was 100% utilized mostly with system and interrupt time to handle this.
Kernel: 3.6.0-xilinx #1 SMP PREEMPT
WARNING: at net/sched/sch_generic.c:255 dev_watchdog+0x15c/0x26c()
NETDEV WATCHDOG: eth0 (xemacps): transmit queue 0 timed out
[<c0013fdc>] (unwind_backtrace+0x0/0x11c) from [<c0020e6c>] (warn_slowpath_common+0x4c/0x64)
[<c0020e6c>] (warn_slowpath_common+0x4c/0x64) from [<c0020f04>] (warn_slowpath_fmt+0x2c/0x3c)
[<c0020f04>] (warn_slowpath_fmt+0x2c/0x3c) from [<c0300de0>] (dev_watchdog+0x15c/0x26c)
[<c0300de0>] (dev_watchdog+0x15c/0x26c) from [<c002cbc0>] (run_timer_softirq+0x190/0x244)
[<c002cbc0>] (run_timer_softirq+0x190/0x244) from [<c0027350>] (__do_softirq+0xb8/0x174)
[<c0027350>] (__do_softirq+0xb8/0x174) from [<c0027860>] (irq_exit+0x48/0xa0)
[<c0027860>] (irq_exit+0x48/0xa0) from [<c000e920>] (handle_IRQ+0x7c/0xb8)
[<c000e920>] (handle_IRQ+0x7c/0xb8) from [<c0008484>] (gic_handle_irq+0x38/0x5c)
[<c0008484>] (gic_handle_irq+0x38/0x5c) from [<c000dc80>] (__irq_svc+0x40/0x70)
Exception stack(0xde06bfa0 to 0xde06bfe8)
bfa0: de06a030 c000ebac de06bff0 00000000 1e05c06a 00000015 10c0387d c0554bf4
bfc0: 0000406a 413fc090 00000000 00000000 00000000 de06bfe8 c000ebcc c000ebd0
bfe0: 60000013 ffffffff
[<c000dc80>] (__irq_svc+0x40/0x70) from [<c000ebd0>] (default_idle+0x24/0x2c)
[<c000ebd0>] (default_idle+0x24/0x2c) from [<c000edd0>] (cpu_idle+0xa0/0xfc)
[<c000edd0>] (cpu_idle+0xa0/0xfc) from [<00383594>] (0x383594)
---[ end trace 97652d78abb4988e]---
xemacps e000b000.ps7-ethernet: transmit timeout 3000 ms, reseting...
BUG: scheduling while atomic: swapper/1/0/0x00000105
