Hi, trying to pass through a intel x520 VF on a Dell R620, I''m getting in domU: [ 58.162639] pci 0000:00:00.6: address space collision: [mem 0xd500c000-0xd500ffff 64bit pref] conflicts with System RAM [mem 0x00100000-0x176ffffff] [ 58.162654] pcifront pci-0: Could not claim resource 0000:00:00.6/0! Device offline. Try using e820_host=1 in the guest config. and in dom0 (XEN) mm.c:807: d2: Forcing read-only access to MFN d510c (XEN) mm.c:807: d0: Forcing read-only access to MFN d510c (XEN) mm.c:618:d0 Could not get page ref for pfn d510c (XEN) mm.c:618:d0 Could not get page ref for pfn d510c (XEN) mm.c:618:d2 Could not get page ref for pfn d510c and then in domU [ 1.703044] BUG: unable to handle kernel paging request at ffffc90000c0a00c [ 1.703061] IP: [<ffffffff8121be22>] __msix_mask_irq+0x21/0x24 [ 1.703078] PGD 1b7c43067 PUD 1b7c44067 PMD 1ac44d067 PTE 80100000d510c465 [ 1.703098] Oops: 0003 [#1] SMP [ 1.703108] Modules linked in: ixgbevf(+) This works with kernel 3.8 with iommu=soft, but 3.10 and 3.11-rc4 does not work. Dom0 memory is less than 4G and I tried with e820_host=1 but still get the error. Not sure if the following has anything to do with it: These are in dom0 on all kernels: [ 24.445555] pci 0000:01:00.1: address space collision: [mem 0xd8000000-0xd807ffff pref] conflicts with 0000:01:00.0 [mem 0xd8000000-0xd807ffff pref] [ 24.445722] pci 0000:07:00.1: address space collision: [mem 0xda000000-0xda07ffff pref] conflicts with 0000:07:00.0 [mem 0xda000000-0xda07ffff pref] Dom0 boots via PXE if that could mean something. DomU boots via pvgrub. Any hints? Thanks Kristoffer
Kristoffer Harthing Egefelt
2013-Aug-15 06:42 UTC
Re: pci passthrough don''t work with kernels > 3.8
Same problem with xen 4.2, 4.3 and 4.4. If I add pci=nomsi to domU it fails with [ 1.541856] ixgbevf: Intel(R) 10 Gigabit PCI Express Virtual Function Network Driver - version 2.7.12-k [ 1.541881] ixgbevf: Copyright (c) 2009 - 2012 Intel Corporation. [ 1.542020] ixgbevf 0000:00:00.6: enabling device (0000 -> 0002) [ 1.828215] ixgbevf 0000:00:00.6: Xen PCI mapped GSI0 to IRQ44 [ 1.840120] ixgbevf 0000:00:00.6: Unable to allocate MSI-X interrupts [ 1.840355] ixgbevf: probe of 0000:00:00.6 failed with error -22 -- View this message in context: http://xen.1045712.n5.nabble.com/pci-passthrough-don-t-work-with-kernels-3-8-tp5718172p5718174.html Sent from the Xen - User mailing list archive at Nabble.com.
Same Problem here with kernel 3.10.17 and 3.4.44. Was anyone able to get ixgbe sriov working with kernel 3.10.17 and 3.4.44 and xen 4.3? [ 4.124868] ixgbevf 0000:00:00.5: Xen PCI mapped GSI0 to IRQ25 [ 4.126071] alg: No test for __gcm-aes-aesni (__driver-gcm-aes-aesni) [ 4.410125] BUG: unable to handle kernel paging request at ffffc90000c3600c [ 4.410135] IP: [<ffffffff811df70a>] __msix_mask_irq+0x21/0x24 [ 4.410149] PGD 1ff01e067 PUD 1ff01f067 PMD 1f10a7067 PTE 80100000f7c18465 [ 4.410149] PGD 1ff01e067 PUD 1ff01f067 PMD 1f10a7067 PTE 80100000f7c18465 [ 4.410161] Oops: 0003 [#1] SMP [ 4.410166] CPU 0 [ 4.410168] Modules linked in: ixgbevf(+) evdev aesni_intel aes_x86_64 aes_generic cryptd microcode snd_pcm snd_page_alloc snd_timer snd soundcore pcspkr xen_pcifront nfs nfs_acl auth_rpcgss fscache lockd sunrpc tg3 libphy aufs xen_netfront [ 4.410196] [ 4.410200] Pid: 313, comm: modprobe Not tainted 3.4.44 #1 [ 4.410206] RIP: e030:[<ffffffff811df70a>] [<ffffffff811df70a>] __msix_mask_irq+0x21/0x24 [ 4.410215] RSP: e02b:ffff88023f01bbe0 EFLAGS: 00010286 [ 4.410219] RAX: 0000000000000001 RBX: ffff880200796000 RCX: 0000000000000001 [ 4.410225] RDX: ffffc90000c3600c RSI: 0000000000000001 RDI: ffff88023cbe5b40 [ 4.410230] RBP: ffff88023cbe5b40 R08: ffff8801fec00050 R09: ffff880200568000 [ 4.410235] R10: ffff88020049d800 R11: ffff880200568000 R12: 0000000000000000 [ 4.410240] R13: ffff880241a22ce0 R14: 0000000000000072 R15: ffffc90000c3600c [ 4.410249] FS: 00007f2ae15d3700(0000) GS:ffff8801ffc00000(0000) knlGS:0000000000000000 -- View this message in context: http://xen.1045712.n5.nabble.com/pci-passthrough-don-t-work-with-kernels-3-8-tp5718172p5719963.html Sent from the Xen - User mailing list archive at Nabble.com.
Hi, On Wed, Nov 20, rishi wrote:> Same Problem here with kernel 3.10.17 and 3.4.44. Was anyone able to get > ixgbe sriov working with kernel 3.10.17 and 3.4.44 and xen 4.3?I had also problems with xen 4.3.0 (4.2.2 worked fine) and PV domU. After upgrade to 4.3.1 all works fine. So please try xen 4.3.1 -- Regards Dieter -- I do not get viruses because I do not use MS software. If you use Outlook then please do not put my email address in your address-book so that WHEN you get a virus it won''t use my address in the From field.
Not sure exactly what device you are trying to pass through (what is ixgbevf?) but mine''s working just fine with 3.9.9 at the moment (USB, audio, GPU). On 11/20/2013 04:46 PM, rishi wrote:> Same Problem here with kernel 3.10.17 and 3.4.44. Was anyone able to get > ixgbe sriov working with kernel 3.10.17 and 3.4.44 and xen 4.3? > > [ 4.124868] ixgbevf 0000:00:00.5: Xen PCI mapped GSI0 to IRQ25 > [ 4.126071] alg: No test for __gcm-aes-aesni (__driver-gcm-aes-aesni) > [ 4.410125] BUG: unable to handle kernel paging request at > ffffc90000c3600c > [ 4.410135] IP: [<ffffffff811df70a>] __msix_mask_irq+0x21/0x24 > [ 4.410149] PGD 1ff01e067 PUD 1ff01f067 PMD 1f10a7067 PTE > 80100000f7c18465 > [ 4.410149] PGD 1ff01e067 PUD 1ff01f067 PMD 1f10a7067 PTE > 80100000f7c18465 > [ 4.410161] Oops: 0003 [#1] SMP > [ 4.410166] CPU 0 > [ 4.410168] Modules linked in: ixgbevf(+) evdev aesni_intel aes_x86_64 > aes_generic cryptd microcode snd_pcm snd_page_alloc snd_timer snd soundcore > pcspkr xen_pcifront nfs nfs_acl auth_rpcgss fscache lockd sunrpc tg3 libphy > aufs xen_netfront > [ 4.410196] > [ 4.410200] Pid: 313, comm: modprobe Not tainted 3.4.44 #1 > [ 4.410206] RIP: e030:[<ffffffff811df70a>] [<ffffffff811df70a>] > __msix_mask_irq+0x21/0x24 > [ 4.410215] RSP: e02b:ffff88023f01bbe0 EFLAGS: 00010286 > [ 4.410219] RAX: 0000000000000001 RBX: ffff880200796000 RCX: > 0000000000000001 > [ 4.410225] RDX: ffffc90000c3600c RSI: 0000000000000001 RDI: > ffff88023cbe5b40 > [ 4.410230] RBP: ffff88023cbe5b40 R08: ffff8801fec00050 R09: > ffff880200568000 > [ 4.410235] R10: ffff88020049d800 R11: ffff880200568000 R12: > 0000000000000000 > [ 4.410240] R13: ffff880241a22ce0 R14: 0000000000000072 R15: > ffffc90000c3600c > [ 4.410249] FS: 00007f2ae15d3700(0000) GS:ffff8801ffc00000(0000) > knlGS:0000000000000000 > > > > > > -- > View this message in context: http://xen.1045712.n5.nabble.com/pci-passthrough-don-t-work-with-kernels-3-8-tp5718172p5719963.html > Sent from the Xen - User mailing list archive at Nabble.com. > > _______________________________________________ > Xen-users mailing list > Xen-users@lists.xen.org > http://lists.xen.org/xen-users >
I am trying to provide passing through 10G ethernet intel 82599 card for domU (ixgbevf driver is for intel card). I tried with xen 4.3.1, the same issue occurs. -- View this message in context: http://xen.1045712.n5.nabble.com/pci-passthrough-don-t-work-with-kernels-3-8-tp5718172p5719972.html Sent from the Xen - User mailing list archive at Nabble.com.
Maybe Matching Threads
- issue using SRIOV "Unable to start - perhaps the PF driver is not up yet", while PF driver is actually up
- issue using SRIOV "Unable to start - perhaps the PF driver is not up yet", while PF driver is actually up
- [BUG] domU kernel crash at igbvf module loading / __msix_mask_irq
- automatically detaching PCI host devices from guest for SRIOV usage.
- [RFC] virtio-net: help live migrate SR-IOV devices