RE: [Xen-users] Re: HELP DomU (Centos 5) with dedicated e1000 (i

I got this error lately. It turns out that I was passing a device through PCI passthrough and this device was given an IRQ (from domU) which conflicted with a IRQ number already used by dom0.

Take a look at your current attributed IRQs in dom0 and your domUs and try to see if there might be any conflict.

De : xen-users-bounces@xxxxxxxxxxxxxxxxxxx [mailto:xen-users-bounces@xxxxxxxxxxxxxxxxxxx] De la part de Julien Reveillet
Envoyé : 12 décembre 2007 10:26
À : xen-users@xxxxxxxxxxxxxxxxxxx
Objet : [Xen-users] Re: HELP DomU (Centos 5) with dedicated e1000 (intel) device dropping packets

Hi everyone,

Here is what my Xen Dom0 syslog daemon reports after a freeze of my CentOs DomU network interface :

xen kernel: Disabling IRQ #17
Dec 12 16:16:18 xen kernel: [<c013af97>] __report_bad_irq+0x2b/0x69
Dec 12 16:16:18 xen kernel: [<c013b188>] note_interrupt+0x1b3/0x1eb
Dec 12 16:16:18 xen kernel: [<c013a7ed>] handle_IRQ_event+0x49/0x51
Dec 12 16:16:18 xen kernel: [<c013a8a8>] __do_IRQ+0xb3/0xe8
Dec 12 16:16:18 xen kernel: [<c010612c>] do_IRQ+0x43/0x53
Dec 12 16:16:18 xen kernel: [<c020c1cc>] evtchn_do_upcall+0x64/0x9b
Dec 12 16:16:18 xen kernel: [<c0104a51>] hypervisor_callback+0x3d/0x48
Dec 12 16:16:18 xen kernel: [<c020bbb1>] notify_remote_via_irq+0x27/0x5f
Dec 12 16:16:18 xen kernel: [<c021359b>] blkif_notify_work+0x1b/0x1d
Dec 12 16:16:18 xen kernel: [<c02137b3>] make_response+0xeb/0x12f
Dec 12 16:16:18 xen kernel: [<c021383c>] end_block_io_op+0x45/0x7e
Dec 12 16:16:18 xen kernel: [<c02137f7>] end_block_io_op+0x0/0x7e
Dec 12 16:16:18 xen kernel: [<c015d715>] bio_endio+0x50/0x55
Dec 12 16:16:18 xen kernel: [<ee1a62da>] dec_pending+0x9b/0xb3 [dm_mod]
Dec 12 16:16:18 xen kernel: [<ee1a63de>] clone_endio+0x7c/0x94 [dm_mod]
Dec 12 16:16:18 xen kernel: [<ee1a6362>] clone_endio+0x0/0x94 [dm_mod]
Dec 12 16:16:18 xen kernel: [<c015d715>] bio_endio+0x50/0x55
Dec 12 16:16:18 xen kernel: [<c01ad07b>] __end_that_request_first+0x11b/0x422
Dec 12 16:16:18 xen kernel: [<c0123662>] del_timer+0x41/0x47
Dec 12 16:16:18 xen kernel: [<ee0e2055>] scsi_end_request+0x1a/0xa9 [scsi_mod]
Dec 12 16:16:18 xen kernel: [<c013ebec>] mempool_free+0x5f/0x63
Dec 12 16:16:18 xen kernel: [<ee0e2227>] scsi_io_completion+0x143/0x2ed [scsi_mod]
Dec 12 16:16:18 xen kernel: [<c020bcf2>] unmask_evtchn+0x26/0xba
Dec 12 16:16:18 xen kernel: [<ee16f3b2>] sd_rw_intr+0x1eb/0x215 [sd_mod]
Dec 12 16:16:18 xen kernel: [<c0289aa0>] _spin_lock_irqsave+0x8/0x28
Dec 12 16:16:18 xen kernel: [<ee0de3bd>] scsi_finish_command+0x73/0x77 [scsi_mod]
Dec 12 16:16:18 xen kernel: [<c01af180>] blk_done_softirq+0x55/0x60
Dec 12 16:16:18 xen kernel: [<c011f41e>] __do_softirq+0x5e/0xc3
Dec 12 16:16:18 xen kernel: [<c011f4bd>] do_softirq+0x3a/0x4a
Dec 12 16:16:18 xen kernel: [<c0106131>] do_IRQ+0x48/0x53
Dec 12 16:16:18 xen kernel: [<c020c1cc>] evtchn_do_upcall+0x64/0x9b
Dec 12 16:16:18 xen kernel: [<c0104a51>] hypervisor_callback+0x3d/0x48
Dec 12 16:16:18 xen kernel: [<c0107342>] raw_safe_halt+0x8c/0xaf
Dec 12 16:16:18 xen kernel: [<c0102c5f>] xen_idle+0x22/0x2e
Dec 12 16:16:18 xen kernel: [<c0102d7e>] cpu_idle+0x91/0xab
Dec 12 16:16:18 xen kernel: [<c03236fc>] start_kernel+0x378/0x37f

It's apparently an Irq conflict but i don't know why?

Do someone have an idea?

Thanks.

Julien.

Julien Reveillet a écrit :

Hello everybody,

I've finished with pci export from DomU to Dom0 (Debian Etch) but now i have a new problem, and a big one.

My ethernet card is dropping packets but after some time (i can't tell how)
It can work for a day (not in production so not hard tested) and then all packets are dropped.

Look at the ifconfig output :

eth0      Link encap:Ethernet HWaddr XX:XX:XX:XX:XX:XX
          inet addr:192.168.9.52 Bcast:192.168.9.255 Mask:255.255.255.0
          inet6 addr: fe80::213:72ff:fe0f:1c6a/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
          RX packets:10505 errors:0 dropped: 352 overruns:0 frame:0
          TX packets:2921 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:2223312 (2.1 MiB) TX bytes:555445 (542.4 KiB)
          Base address:0xcce0 Memory:fe3e0000-fe400000

Today, nothing was done on it but, a few minutes before, i realised that it was down.

Dmesg output :

...
Intel(R) PRO/1000 Network Driver - version 7.1.9-k4-NAPI
Copyright (c) 1999-2006 Intel Corporation.
PCI: Enabling device 0000:00:01.0 (0000 -> 0003)
PCI: Setting latency timer of device 0000:00:01.0 to 64
e1000: 0000:00:01.0: e1000_probe: (PCI Express:2.5Gb/s:Width x1) xx:xx:xx:xx:xx:xx
e1000: eth0: e1000_probe: Intel(R) PRO/1000 Network Connection
e1000: eth0: e1000_watchdog: NIC Link is Up 1000 Mbps Full Duplex
NET: Registered protocol family 10
lo: Disabled Privacy Extensions
IPv6 over IPv4 tunneling driver
NET: Registered protocol family 5
eth0: no IPv6 routers present

The interface is still up.

I saw while googling that it could be an hardware problem but this was not in a xen environnement so...

Anybody have seen this before?

Thanks.

Best Regards.
Julien.

_______________________________________________

Xen-users mailing list

Xen-users@xxxxxxxxxxxxxxxxxxx

http://lists.xensource.com/xen-users

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users

WARNING - OLD ARCHIVES

xen-users

RE: [Xen-users] Re: HELP DomU (Centos 5) with dedicated e1000 (intel) de