[Bug 99028] New: [KVM][GVT-d] [BDW & SKL ]Ubuntu 16.04 guest boot up with GPU Hang with drm-intel kernel
bugzilla-daemon at freedesktop.org
bugzilla-daemon at freedesktop.org
Thu Dec 8 11:20:41 UTC 2016
https://bugs.freedesktop.org/show_bug.cgi?id=99028
Bug ID: 99028
Summary: [KVM][GVT-d] [BDW & SKL ]Ubuntu 16.04 guest boot up
with GPU Hang with drm-intel kernel
Product: DRI
Version: DRI git
Hardware: x86-64 (AMD64)
OS: Linux (All)
Status: NEW
Severity: major
Priority: medium
Component: DRM/Intel
Assignee: intel-gfx-bugs at lists.freedesktop.org
Reporter: terrence.xu at intel.com
QA Contact: intel-gfx-bugs at lists.freedesktop.org
CC: gordon.jin at intel.com,
intel-gfx-bugs at lists.freedesktop.org,
jani.saarinen at intel.com, zhiyuan.lv at intel.com
i915 platform: BDW, SKL
System Environment
=======
Host kernel repo: kvm.git
Host commit: master-813ae37e
Guest repo: drm-intel.git
Guest commit: drm-intel-next-queued-312c3c46
Regression?
=======
No
Bug detailed description
=======
The guest boot up with the drm-intel 4.8.0-rc2+ kernel, with GPU hang be found.
The guest boot up with the latest drm-intel 4.9.0-rc4+ kernel with kernel panic
(as https://bugs.freedesktop.org/show_bug.cgi?id=99025), but if we try to
remove the kernel panic error handling manually, the GPU hang issue also won
existed.
This is KVM GVT-d environment issue.
Reproduce Steps
==============
Boot up Ubuntu 16.04 guest with the drm-intel kernel, the command as below:
qemu-system-x86_64 --enable-kvm -m 2048 -smp 4 -hda /root/ubuntu-16.04.img -usb
-usbdevice tablet -device virtio-net-pci,netdev=nic0,mac=00:16:3e:60:0a:50
-netdev tap,id=nic0,script=/etc/kvm/qemu-ifup -serial stdio
Expected Result
=============
Guest boot up without any issues.
Actual Result
===========
Guest boot up with GPU hang.
Analysis & Root Cause
===================
[ 6.939377] [drm] GPU HANG: ecode 8:0:0xfffffffe, reason: No progress on
rend er ring, action: reset
[ 6.939378] [drm] GPU hangs can indicate a bug anywhere in the entire gfx
stack, including userspace.
[ 6.939379] [drm] Please file a _new_ bug report on bugs.freedesktop.org
against DRI -> DRM/Intel
[ 6.939379] [drm] drm/i915 developers can then reassign to the right compone
t if it's not a kernel issue.
[ 6.939380] [drm] The gpu crash dump is required to analyze gpu hangs, so
please always attach it.
[ 6.939380] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[ 6.939429] drm/i915: Resetting chip after gpu hang
[ 7.869346] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts:
(null)
[ 7.964728] systemd[1]: systemd 229 running in system mode. (+PAM +AUDIT
+SELINUX +IMA +APPARMOR +SMACK +SYSVINIT +UTMP +LIBCRYPTSETUP +GCRYPT +GNUTLS
+ACL +
XZ -LZ4 +SECCOMP +BLKID +ELFUTILS
+KMOD -IDN)
[ 7.964756] systemd[1]: Detected virtualization qemu.
[ 7.964758] systemd[1]: Detected architecture x86-64.
[ 7.965211] systemd[1]: Set hostname to <gvt-ub16>.
[ 8.092829] systemd[1]: Listening on fsck to fsckd communication Socket.
[ 8.092875] systemd[1]: Reached target User and Group Name Lookups.
[ 8.092882] systemd[1]: Reached target Remote File Systems (Pre).
[ 8.092902] systemd[1]: Listening on udev Kernel Socket.
[ 8.092945] systemd[1]: Created slice System Slice.
[ 8.092976] systemd[1]: Listening on Syslog Socket.
[ 8.278376] EXT4-fs (sda1): re-mounted. Opts: errors=remount-ro
[ 8.349290] systemd-journald[205]: Received request to flush runtime journal
from PID 1
[ 8.481575] random: crng init done
[ 8.889146] drm/i915: Resetting chip after gpu hang
[ 9.060874] Adding 4000764k swap on /dev/sda5. Priority:-1 extents:1
across: 4000764k
[ 10.364525] e1000: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control:
RX
[ 10.873269] drm/i915: Resetting chip after gpu hang
[ 12.923317] drm/i915: Resetting chip after gpu hang
[ 14.905157] drm/i915: Resetting chip after gpu hang
[ 16.900984] drm/i915: Resetting chip after gpu hang
[ 18.885876] drm/i915: Resetting chip after gpu hang
[ 20.921190] drm/i915: Resetting chip after gpu hang
[ 22.905154] drm/i915: Resetting chip after gpu hang
--
You are receiving this mail because:
You are the assignee for the bug.
You are on the CC list for the bug.
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20161208/71b216c0/attachment.html>
More information about the intel-gfx-bugs
mailing list