From: Stephen Hemminger Subject: Re: [PATCH net 2/2] 8139cp: reset BQL when ring tx ring cleared Date: Mon, 20 May 2013 17:27:49 -0700 Message-ID: <20130520172749.2a4447b6@nehalam.linuxnetplumber.net> References: <20130520095443.2d18b2cb@nehalam.linuxnetplumber.net> <20130520213728.GA28252@electric-eye.fr.zoreil.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: David Miller , netdev@vger.kernel.org To: Francois Romieu Return-path: Received: from mail-pb0-f54.google.com ([209.85.160.54]:35355 "EHLO mail-pb0-f54.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751572Ab3EUA1y (ORCPT ); Mon, 20 May 2013 20:27:54 -0400 Received: by mail-pb0-f54.google.com with SMTP id ro12so39325pbb.27 for ; Mon, 20 May 2013 17:27:54 -0700 (PDT) In-Reply-To: <20130520213728.GA28252@electric-eye.fr.zoreil.com> Sender: netdev-owner@vger.kernel.org List-ID: On Mon, 20 May 2013 23:37:28 +0200 Francois Romieu wrote: > cp_stop_hw includes netdev_reset_queue. > > You have imho exhibited a start_xmit after cp_stop_hw race - not sure if > it happens in cp_tx_timeout or cp_change_mtu. Reverting the analysis above, > I have not found a place where cp_stop_hw could be called without being > followed by a cp_clean_rings. The netdev_reset_queue in cp_stop_hw, now > useless, should thus be removed. > > Does it make sense ? Your right, you could probably remove it. It doesn't solve the problem, still seeing transmit timeouts. Looks like what happens with DHCP is something else. zebra[2210]: interface eth0 index 2 changed . [ 23.960059] ------------[ cut here ]------------ [ 23.960107] WARNING: at net/sched/sch_generic.c:255 dev_watchdog+0x243/0x250() [ 23.960110] Hardware name: Bochs [ 23.960114] Modules linked in: 9p 9pnet fscache ipv6 tun mperf cpufreq_stats cpufreq_ondemand freq_table cpufreq_userspace cpufreq_conservative virtio_uio(O) cpufreq_powersave igb_uio(O) uio cirrus ttm drm_kms_helper drm sysimgblt sysfillrect virtio_balloon i2c_piix4 i2c_core hid_generic syscopyarea psmouse evdev serio_raw microcode pcspkr intel_agp intel_gtt agpgart processor thermal_sys button vfat fat usb_storage ohci_hcd squashfs loop overlayfs raid10 raid456 async_pq async_xor xor async_memcpy async_raid6_recov usbhid hid raid6_pq async_tx raid1 raid0 multipath linear md_mod pata_acpi ata_generic virtio_blk virtio_net 8139cp ata_piix 8139too mii virtio_pci virtio_ring virtio floppy [ 23.960177] Pid: 0, comm: swapper/3 Tainted: G O 3.9.3-1-amd64-vyatta #1 [ 23.960179] Call Trace: [ 23.960181] [] ? warn_slowpath_common+0x7b/0xc0 [ 23.960212] [] ? warn_slowpath_fmt+0x45/0x50 [ 23.960228] [] ? _raw_spin_lock+0x5/0x10 [ 23.960232] [] ? dev_watchdog+0x243/0x250 [ 23.960245] [] ? __queue_work+0x2b0/0x2b0 [ 23.960253] [] ? __netdev_watchdog_up+0x70/0x70 [ 23.960257] [] ? __netdev_watchdog_up+0x70/0x70 [ 23.960262] [] ? call_timer_fn+0x43/0x130 [ 23.960265] [] ? run_timer_softirq+0x237/0x280 [ 23.960269] [] ? update_rmtp+0x80/0x80 [ 23.960272] [] ? _raw_spin_lock+0x5/0x10 [ 23.960276] [] ? __do_softirq+0xe1/0x270 [ 23.960280] [] ? irq_exit+0xb5/0xc0 [ 23.960291] [] ? smp_apic_timer_interrupt+0x68/0xa0 [ 23.960302] [] ? apic_timer_interrupt+0x6d/0x80 [ 23.960304] [] ? native_safe_halt+0x2/0x10 [ 23.960318] [] ? default_idle+0x3f/0xf0 [ 23.960323] [] ? cpu_idle+0x88/0xd0 [ 23.960330] [] ? start_secondary+0x211/0x216 [ 23.960333] ---[ end trace a519f7f509aceb9a ]--- [ 23.960407] 8139cp 0000:00:03.0 eth0: Transmit timeout, status d 3b 4 80ff [ 35.960137] 8139cp 0000:00:03.0 eth0: Transmit timeout, status d 3b 5 80ff [ 47.960108] 8139cp 0000:00:03.0 eth0: Transmit timeout, status d 3b 5 80ff [ 59.960117] 8139cp 0000:00:03.0 eth0: Transmit timeout, status d 3b 5 80ff [ 71.960115] 8139cp 0000:00:03.0 eth0: Transmit timeout, status d 3b 5 80ff [ 83.960117] 8139cp 0000:00:03.0 eth0: Transmit timeout, status d 3b 5 80ff [ 95.960131] 8139cp 0000:00:03.0 eth0: Transmit timeout, status d 3b 5 80ff ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ "messages" 35 lines, 2958 characters written