1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
|
semantic: 0.923
other: 0.920
graphic: 0.920
vnc: 0.911
KVM: 0.900
assembly: 0.889
instruction: 0.884
mistranslation: 0.857
boot: 0.849
device: 0.849
network: 0.787
socket: 0.734
VFIO passthrough causes assertation failure
Since commit 5e95494380ec I am no longer able to passthrough my Nvidia GTX 770 using VFIO. Qemu terminates with:
qemu-system-x86_64: hw/pci/pcie.c:240: pcie_cap_slot_hotplug_common: Assertion `((pci_dev->devfn) & 0x07) == 0' failed.
Above output was generated using commit f55ea6297cc0.
Lspci of the vga card:
01:00.0 VGA compatible controller: NVIDIA Corporation GK104 [GeForce GTX 770] (rev a1)
Subsystem: Gigabyte Technology Co., Ltd Device 360c
Kernel driver in use: vfio-pci
01:00.1 Audio device: NVIDIA Corporation GK104 HDMI Audio Controller (rev a1)
Subsystem: Gigabyte Technology Co., Ltd Device 360c
Kernel driver in use: vfio-pci
Commandline used to start qemu:
qemu-system-x86_64 -machine accel=kvm \
-nodefaults \
-name VFIO-Test \
-machine q35 \
-cpu host \
-smp 1 \
-enable-kvm \
-m 1024 \
-k de \
-vga none \
-device ioh3420,bus=pcie.0,addr=1c.0,multifunction=on,port=1,chassis=1,id=root.1 \
-device vfio-pci,host=01:00.0,bus=root.1,addr=00.0,multifunction=on,x-vga=on \
-device vfio-pci,host=01:00.1,bus=root.1,addr=00.1 \
-rtc base=utc \
-boot order=d \
-device ide-cd,drive=drive-cd-disk1,id=cd-disk1,unit=0 \
-drive file=/home/bluebird/Downloads/systemrescuecd-x86-4.0.0.iso,if=none,id=drive-cd-disk1,media=cdrom \
-nographic
Full output of git bisect:
5e95494380ecf83c97d28f72134ab45e0cace8f9 is the first bad commit
commit 5e95494380ecf83c97d28f72134ab45e0cace8f9
Author: Igor Mammedov <email address hidden>
Date: Wed Feb 5 16:36:52 2014 +0100
hw/pci: switch to a generic hotplug handling for PCIDevice
make qdev_unplug()/device_set_realized() to call hotplug handler's
plug/unplug methods if available and remove not needed anymore
hot(un)plug handling from PCIDevice.
In case if hotplug handler is not available, revert to the legacy
hotplug method for compatibility with not yet converted buses.
Signed-off-by: Igor Mammedov <email address hidden>
Reviewed-by: Michael S. Tsirkin <email address hidden>
Signed-off-by: Michael S. Tsirkin <email address hidden>
:040000 040000 9bdab0d75fbc9be4fe2e4274e58e0cdcd347ac7e d6d6294ea9c06e80a0fc8fcabd6345dfae5137ad M hw
:040000 040000 d064d75ca8b8f169c41eee2683082e8f9104e968 f2abbf9bee754ada0f49135968455fd1a69b2186 M include
:040000 040000 c515daff6c77f9bd2cc32873be4c5c3a1c20cbb9 c506f5587afe8f7ee129a7ca6e3ae2e5118254f9 M tests
commit 6e1f0a55a14bad1d0c8b9d29626ef4e4b2617c74
Author: Igor Mammedov <email address hidden>
Date: Mon Feb 17 15:00:06 2014 +0100
PCIE: fix regression with coldplugged multifunction device
PCIE is causing asserts each time a multifunction device is added
on command line (coldplug).
This is caused by
commit a66e657e18cd9b70e9f57ae5512c07faf2bc508f
pci/pcie: convert PCIE hotplug to use hotplug-handler API
QEMU abort is caused by misplaced assertion, which should
be checked only when device is hotplugged.
Reference to regression report:
http://<email address hidden>/msg216226.html
Fixes: a66e657e18cd9b70e9f57ae5512c07faf2bc508f
Reported-By: Nigel Kukard <email address hidden>
Signed-off-by: Igor Mammedov <email address hidden>
Reviewed-by: Michael S. Tsirkin <email address hidden>
Signed-off-by: Michael S. Tsirkin <email address hidden>
|