lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Date:	Mon, 14 Jul 2008 23:30:40 -0400
From:	"Andrew Paprocki" <andrew@...iboo.com>
To:	netdev@...r.kernel.org
Cc:	"Francois Romieu" <romieu@...zoreil.com>
Subject: r8169 device jumps from 12k/sec interrupts with traffic to over 100k/sec interrupts without traffic

I'm currently running 2.6.26-rc8 (NAPI enabled) and my r8169 device is
going a little crazy after a short time under load.

Test scenario is as follows...
r8169-system: nc -l -p 8888 > /dev/null
remote-system: cat /dev/zero | nc r8169-system 8888

While this is running, I also do the following:
r8169-system: dd if=/dev/sda of=/dev/null

While both commands are running under the r8169 system, I run the
following to see the # of interrupts/sec occuring:
while true; do cat /proc/interrupts | grep eth0; sleep 1; done

I see roughly 12k/sec interrupts on eth0 and ~ 400/sec interrupts on
sata_sil. This continues for some time at which point the system
appears to hang (I no longer see the /proc/interrupts output every
second). I switched back to the r8169 netcat and ctrl-c'd it. At this
point the system is still acting really really sluggish and finally
the cat /proc/interrupts output begins again. The only difference is
now that I see it has exploded to 110k/sec interrupts on eth0 even
though there is no data flowing to the system! The stats on ifconfig
eth0 show no rx/tx byte counters increasing yet there is a constant
100+k/sec interrupt load on the device.

It appears to be affecting the rest of the system because during this
time I also get other strange behavior.. For instance I see libata
timeouts on the disks and they soft reset occasionally. Another
example of something which happens in this state:

NETDEV WATCHDOG: eth0: transmit timed out
------------[ cut here ]------------
WARNING: at net/sched/sch_generic.c:222 dev_watchdog+0x80/0xc3()
Pid: 738, comm: dd Not tainted 2.6.26-rc8 #8
 [<c0216072>] warn_on_slowpath+0x3b/0x61
 [<c0222d35>] autoremove_wake_function+0xc/0x2b
 [<c02124f8>] __wake_up_common+0x2d/0x52
 [<c03a6879>] dev_watchdog+0x0/0xc3
 [<c0213bd0>] __wake_up+0xf/0x15
 [<c0216416>] wake_up_klogd+0x2b/0x2d
 [<c0224c16>] hrtimer_forward+0xe2/0xfe
 [<c0206bf0>] read_tsc+0x6/0x22
 [<c0226b5b>] getnstimeofday+0x32/0xaf
 [<c0206a53>] pit_next_event+0x16/0x1a
 [<c0228e78>] clockevents_program_event+0xd1/0xe0
 [<c0220de2>] queue_delayed_work_on+0x70/0x7b
 [<c03a6879>] dev_watchdog+0x0/0xc3
 [<c0220e13>] queue_delayed_work+0x16/0x18
 [<c03a6879>] dev_watchdog+0x0/0xc3
 [<c03a68f9>] dev_watchdog+0x80/0xc3
 [<c021c051>] run_timer_softirq+0xf3/0x137
 [<c02195c0>] __do_softirq+0x35/0x75
 [<c0219622>] do_softirq+0x22/0x26
 [<c021989e>] irq_exit+0x25/0x53
 [<c0204cc3>] do_IRQ+0x4d/0x5d
 [<c0203833>] common_interrupt+0x23/0x28
 [<c0260000>] sys_futimesat+0x6/0x81
 [<c0247eeb>] rw_verify_area+0x5d/0x78
 [<c0248402>] vfs_write+0x62/0xec
 [<c0248863>] sys_write+0x3c/0x63
 [<c02036aa>] syscall_call+0x7/0xb
 =======================
---[ end trace f4d1fed7dd1e3699 ]---
r8169: eth0: link up

What can I do to further debug this problem?

Thanks,
-Andrew
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Powered by blists - more mailing lists