Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines
Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.
REPEAT: Do NOT report bugs for outdated packages!
https://wiki.archlinux.org/title/Bug_reporting_guidelines
Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.
REPEAT: Do NOT report bugs for outdated packages!
FS#58463 - [linux] kvm wmwrite error and virtual machine slow down with old CPUs (e.g. Intel Core 2
Attached to Project:
Arch Linux
Opened by Andreas Baumann (andreas_baumann) - Saturday, 05 May 2018, 06:45 GMT
Last edited by Doug Newgard (Scimmia) - Friday, 01 June 2018, 15:41 GMT
Opened by Andreas Baumann (andreas_baumann) - Saturday, 05 May 2018, 06:45 GMT
Last edited by Doug Newgard (Scimmia) - Friday, 01 June 2018, 15:41 GMT
|
DetailsDescription:
starting a virtual machine with kvm acceleration almost kills the machine (with high CPU load of systemd-journald). The root cause is in KVM logging the following error over and over again: vmwrite error: reg 401e [ 3368.691017] CPU: 0 PID: 1491 Comm: CPU 0/KVM Tainted: G O 4.16.6-1-ARCH #1 [ 3368.691018] Hardware name: /DP965LT, BIOS MQ96510J.86A.1577.2006.1115.2351 11/15/2006 [ 3368.691019] Call Trace: [ 3368.691023] dump_stack+0x5c/0x85 [ 3368.691027] vmx_set_cr4+0x1d2/0x210 [kvm_intel] [ 3368.691045] kvm_set_cr4+0x18e/0x2c0 [kvm] [ 3368.691051] handle_cr+0x601/0x6a0 [kvm_intel] [ 3368.691055] ? vmx_vcpu_run+0x2f5/0x5b0 [kvm_intel] [ 3368.691060] vmx_handle_exit+0xaf/0x14e0 [kvm_intel] [ 3368.691065] ? vmx_vcpu_run+0x2f5/0x5b0 [kvm_intel] [ 3368.691069] ? vmx_vcpu_run+0x301/0x5b0 [kvm_intel] [ 3368.691074] ? vmx_vcpu_run+0x2f5/0x5b0 [kvm_intel] [ 3368.691078] ? vmx_vcpu_run+0x301/0x5b0 [kvm_intel] [ 3368.691083] ? vmx_vcpu_run+0x3bd/0x5b0 [kvm_intel] [ 3368.691102] kvm_arch_vcpu_ioctl_run+0xc39/0x1820 [kvm] The following patch seemed to solve the issue: https://bugzilla.redhat.com/show_bug.cgi?id=1566258 https://bodhi.fedoraproject.org/updates/FEDORA-2018-b6be90818 https://patchwork.kernel.org/patch/10353377/ I attached a modified PKGBUILD and the patch. Note: I didn't check yet, if 4.17 is also affected. Additional info: * package version(s): 4.16.6-1 * config and/or log files etc. Steps to reproduce: virsh start <name of vm> check output of systemd journal |
This task depends upon
Closed by Doug Newgard (Scimmia)
Friday, 01 June 2018, 15:41 GMT
Reason for closing: Fixed
Additional comments about closing: linux 4.16.11-1
Friday, 01 June 2018, 15:41 GMT
Reason for closing: Fixed
Additional comments about closing: linux 4.16.11-1
PKGBUILD
queued for 4.16.11
Edit:
4.16.11 not 4.16.10