[Xen-bugs] [Bug 1742] Passthrough VF does not work after DomU reboot with non-distro kernel (stock 2.6.35)

[bugzilla-daemon@xxxxxxxxxxxxxxxxxxx]

http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=1742





------- Comment #10 from jason.kwon@xxxxxxxxxxxx  2011-02-16 14:50 -------
I cloned the repository you specified and built/booted a kernel, but ran into
two issues.  One, the bridged ethernet device for my guest stopped working; the
tail of dmesg in DomU showed:

[   76.020088] ------------[ cut here ]------------
[   76.020098] WARNING: at net/sched/sch_generic.c:256
dev_watchdog+0x293/0x2a0()
[   76.020101] Hardware name: HVM domU
[   76.020103] NETDEV WATCHDOG: eth0 (8139cp): transmit queue 0 timed out
[   76.020105] Modules linked in: lp ppdev parport_pc parport psmouse serio_raw
8139too floppy 8139cp mii
[   76.020116] Pid: 0, comm: kworker/0:0 Not tainted 2.6.38-rc5+ #2
[   76.020118] Call Trace:
[   76.020121]  <IRQ>  [<ffffffff810613cf>] ? warn_slowpath_common+0x7f/0xc0
[   76.020132]  [<ffffffff810614c6>] ? warn_slowpath_fmt+0x46/0x50
[   76.020135]  [<ffffffff814d5993>] ? dev_watchdog+0x293/0x2a0
[   76.020138]  [<ffffffff81091bbf>] ? tick_program_event+0x1f/0x30
[   76.020142]  [<ffffffff81086d70>] ? hrtimer_interrupt+0x130/0x220
[   76.020145]  [<ffffffff81070c4a>] ? run_timer_softirq+0x16a/0x360
[   76.020147]  [<ffffffff814d5700>] ? dev_watchdog+0x0/0x2a0
[   76.020150]  [<ffffffff8106808b>] ? __do_softirq+0xab/0x200
[   76.020153]  [<ffffffff8100ce9c>] ? call_softirq+0x1c/0x30
[   76.020156]  [<ffffffff8100e8a5>] ? do_softirq+0x65/0xa0
[   76.020158]  [<ffffffff81067f55>] ? irq_exit+0x85/0x90
[   76.020161]  [<ffffffff81351e65>] ? xen_evtchn_do_upcall+0x35/0x50
[   76.020164]  [<ffffffff8100cfc3>] ? xen_hvm_callback_vector+0x13/0x20
[   76.020165]  <EOI>  [<ffffffff81035ecb>] ? native_safe_halt+0xb/0x10
[   76.020172]  [<ffffffff81014ccf>] ? default_idle+0x5f/0x170
[   76.020176]  [<ffffffff8100b067>] ? cpu_idle+0xb7/0x110
[   76.020179]  [<ffffffff815a3b81>] ? start_secondary+0x1c0/0x1c7
[   76.020181] ---[ end trace 5abf079aca6bd7f2 ]---
[   76.020286] 8139cp 0000:00:04.0: eth0: Transmit timeout, status  d   3b   15
80ff

The second issue is that I could not load the VF driver in the guest:

root@xc5-pc-vm1:/home/test# modprobe ixgbevf
[  213.130657] ------------[ cut here ]------------
[  213.133878] kernel BUG at drivers/pci/msi.c:209!
[  213.137074] invalid opcode: 0000 [#1] SMP 
[  213.140071] last sysfs file:
/sys/devices/system/cpu/cpu1/cache/index2/shared_cpu_map
[  213.140071] CPU 0 
[  213.140071] Modules linked in: ixgbevf(+) lp ppdev parport_pc parport
psmouse serio_raw 8139too floppy 8139cp mii
[  213.140071] 
[  213.140071] Pid: 772, comm: modprobe Tainted: G        W   2.6.38-rc5+ #2
Xen HVM domU
[  213.140071] RIP: 0010:[<ffffffff812f9ead>]  [<ffffffff812f9ead>]
__read_msi_msg+0xed/0x100
[  213.140071] RSP: 0018:ffff88003148fb98  EFLAGS: 00010206
[  213.140071] RAX: ffff880030097e50 RBX: ffff880030097e40 RCX:
ffff88003cf958f0
[  213.140071] RDX: 0000000000000011 RSI: ffff88003148fbe8 RDI:
ffff880030097e40
[  213.140071] RBP: ffff88003148fbc8 R08: ffff88003195e598 R09:
ffff880030097cc0
[  213.140071] R10: 0000000000000068 R11: 0000000000000005 R12:
ffff88003148fbe8
[  213.140071] R13: ffff88003cf95000 R14: ffff88003cf958f0 R15:
ffff88003148fbe8
[  213.140071] FS:  00007fc5bd3c7700(0000) GS:ffff88003fc00000(0000)
knlGS:0000000000000000
[  213.140071] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[  213.140071] CR2: 00007fff2a612ef8 CR3: 000000002a17d000 CR4:
00000000000006f0
[  213.140071] DR0: 0000000000000000 DR1: 0000000000000000 DR2:
0000000000000000
[  213.140071] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7:
0000000000000400
[  213.140071] Process modprobe (pid: 772, threadinfo ffff88003148e000, task
ffff88001f8f0000)
[  213.140071] Stack:
[  213.140071]  0000000000000030 00000000f3004000 ffff880030097e40
ffff88003cf95000
[  213.140071]  0000000000000000 ffff88003cf958f0 ffff88003148fc38
ffffffff8149f09f
[  213.140071]  00000000f3005000 000000110018e000 ffff88003148fc38
ffffffff8114a735
[  213.140071] Call Trace:
[  213.140071]  [<ffffffff8149f09f>] xen_hvm_setup_msi_irqs+0x5f/0x1d0
[  213.140071]  [<ffffffff8114a735>] ? kmem_cache_alloc_trace+0xb5/0x110
[  213.140071]  [<ffffffff812fa141>] pci_enable_msix+0x281/0x410
[  213.140071]  [<ffffffffa008f9d0>] ixgbevf_probe+0x533/0x8cd [ixgbevf]
[  213.140071]  [<ffffffff812e72bf>] local_pci_probe+0x5f/0xd0
[  213.140071]  [<ffffffff812e84e1>] pci_device_probe+0x101/0x120
[  213.140071]  [<ffffffff813968da>] ? driver_sysfs_add+0x7a/0xb0
[  213.140071]  [<ffffffff81396a36>] driver_probe_device+0x96/0x1c0
[  213.140071]  [<ffffffff81396b60>] ? __driver_attach+0x0/0xa0
[  213.140071]  [<ffffffff81396bfb>] __driver_attach+0x9b/0xa0
[  213.140071]  [<ffffffff81396b60>] ? __driver_attach+0x0/0xa0
[  213.140071]  [<ffffffff81395e58>] bus_for_each_dev+0x68/0x90
[  213.140071]  [<ffffffff8139685e>] driver_attach+0x1e/0x20
[  213.140071]  [<ffffffff8139613d>] bus_add_driver+0xcd/0x270
[  213.140071]  [<ffffffffa0095000>] ? ixgbevf_init_module+0x0/0x51 [ixgbevf]
[  213.140071]  [<ffffffff81396f40>] driver_register+0x80/0x150
[  213.140071]  [<ffffffff815a9b64>] ? printk+0x68/0x6c
[  213.140071]  [<ffffffffa0095000>] ? ixgbevf_init_module+0x0/0x51 [ixgbevf]
[  213.140071]  [<ffffffff812e8766>] __pci_register_driver+0x56/0xd0
[  213.140071]  [<ffffffffa009504f>] ixgbevf_init_module+0x4f/0x51 [ixgbevf]
[  213.140071]  [<ffffffff81002053>] do_one_initcall+0x43/0x190
[  213.140071]  [<ffffffff8109e28b>] sys_init_module+0xbb/0x200
[  213.140071]  [<ffffffff8100bf82>] system_call_fastpath+0x16/0x1b
[  213.140071] Code: 75 38 49 8b 7d 10 49 8d 4c 24 04 41 8d 56 08 e8 2a 6f fe
ff 41 8b 75 38 49 8b 7d 10 48 8d 4d de 41 8d 56 0c e8 a5 6f fe ff eb bf <0f> 0b
eb fe eb 0d 90 90 90 90 90 90 90 90 90 90 90 90 90 55 48 
[  213.140071] RIP  [<ffffffff812f9ead>] __read_msi_msg+0xed/0x100
[  213.140071]  RSP <ffff88003148fb98>
[  213.389349] ---[ end trace 5abf079aca6bd7f3 ]---
Segmentation fault

Regarding the DomU kernel config, can you clarify whether the paravirtops
config options still apply?  Since this is an HVM guest, I was using the
default Ubuntu config for all my built kernels.  In any case, I did configure
the kernel as described in the wiki.

May I ask what was introduced with 2.6.38-rc5 that might produce different VF
behavior compared with the other kernels I tried?  I didn't think the issue was
tied to a particular kernel version, since both 2.6.18 and 2.6.35 distro
kernels worked.  And more confusingly, since the distro 2.6.35-22-server
worked, but recompiling the very same kernel source failed.


-- 
Configure bugmail: 
http://bugzilla.xensource.com/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

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