Crash - vmx_deliver_posted_interrupt

Issues related to applications and software problems
Post Reply
nokiacephusers
Posts: 9
Joined: 2017/09/15 12:02:31

Crash - vmx_deliver_posted_interrupt

Post by nokiacephusers » 2021/10/04 11:36:08

hi ,

I noticed this crash in Centos 7.6 version used on one of our servers. There was no specific activity ongoing on the server that could explain the crash.
https://bugzilla.redhat.com/show_bug.cgi?id=1717212
I could find a redhat link for the same crash which is fixed in 7.9 but it was in kernel-rt component whereas I am using kernel 3.10.0-957.27.2 so I am not sure if the similar issue applies to this distribution as well and could be fixed in a higher version like maybe 7.9.
Has anyone seen similar issue in their server?

reason: WARNING: CPU: 32 PID: 27908 at arch/x86/kvm/vmx.c:5188 vmx_deliver_posted_interrupt+0x156/0x170 [kvm_intel]
time: Saturday 28 August 2021 11:57:36 AM UTC
cmdline: BOOT_IMAGE=/vmlinuz-3.10.0-957.27.2.el7.x86_64 root=/dev/mapper/vg_root-root ro console=tty0 crashkernel=auto rd.lvm.lv=vg_root/root rhgb quiet LANG=en_US.UTF-8 biosdevname=1 consoleblank=0 intel_idle.max_cstate=0 intel_iommu=on intremap=no_x2apic_optout iommu=pt mce=ignore_ce net.ifnames=0 nosoftlockup
package: kernel
uid: 0 (root)
count: 1
Directory: /var/spool/abrt/oops-2021-08-28-11:57:36-2081982-0

[11518613.658724] Call Trace:
[11518613.658730] [<ffffffff9a564147>] dump_stack+0x19/0x1b
[11518613.658734] [<ffffffff99e98848>] __warn+0xd8/0x100
[11518613.658736] [<ffffffff99e9898d>] warn_slowpath_null+0x1d/0x20
[11518613.658740] [<ffffffffc1eb9fe6>] vmx_deliver_posted_interrupt+0x156/0x170 [kvm_intel]
[11518613.658760] [<ffffffffc089f324>] __apic_accept_irq+0x2a4/0x3b0 [kvm]
[11518613.658769] [<ffffffffc089f52a>] kvm_apic_set_irq+0x2a/0x30 [kvm]
[11518613.658777] [<ffffffffc089f6db>] kvm_irq_delivery_to_apic_fast+0x14b/0x3f0 [kvm]
[11518613.658785] [<ffffffffc08a4345>] kvm_irq_delivery_to_apic+0x65/0x2f0 [kvm]
[11518613.658792] [<ffffffffc089ea1f>] kvm_lapic_reg_write+0x11f/0x630 [kvm]
[11518613.658800] [<ffffffffc08a12cd>] kvm_x2apic_msr_write+0x5d/0xa0 [kvm]
[11518613.658807] [<ffffffffc08774fd>] kvm_set_msr_common+0x2dd/0xc10 [kvm]
[11518613.658811] [<ffffffffc1ebe073>] ? vmx_vcpu_run+0x363/0xb10 [kvm_intel]
[11518613.658814] [<ffffffffc1ebe067>] ? vmx_vcpu_run+0x357/0xb10 [kvm_intel]
[11518613.658817] [<ffffffffc1ebe073>] ? vmx_vcpu_run+0x363/0xb10 [kvm_intel]
[11518613.658820] [<ffffffffc1ec309a>] vmx_set_msr+0xda/0x590 [kvm_intel]
[11518613.658827] [<ffffffffc0868067>] kvm_set_msr+0x47/0x80 [kvm]
[11518613.658830] [<ffffffffc1eb8ae8>] handle_wrmsr+0x58/0x120 [kvm_intel]
[11518613.658833] [<ffffffffc1ec25c4>] vmx_handle_exit+0x294/0xc90 [kvm_intel]
[11518613.658836] [<ffffffffc1ebe067>] ? vmx_vcpu_run+0x357/0xb10 [kvm_intel]
[11518613.658839] [<ffffffffc1ebe620>] ? vmx_vcpu_run+0x910/0xb10 [kvm_intel]
[11518613.658842] [<ffffffffc1ebdc88>] ? vmx_hwapic_irr_update+0xb8/0xc0 [kvm_intel]
[11518613.658848] [<ffffffffc0878598>] vcpu_enter_guest+0x768/0x1470 [kvm]
[11518613.658856] [<ffffffffc0880228>] kvm_arch_vcpu_ioctl_run+0x358/0x480 [kvm]
[11518613.658862] [<ffffffffc0862801>] kvm_vcpu_ioctl+0x2c1/0x6e0 [kvm]
[11518613.658865] [<ffffffff9a057840>] do_vfs_ioctl+0x3a0/0x5a0
[11518613.658867] [<ffffffff9a057ae1>] SyS_ioctl+0xa1/0xc0
[11518613.658870] [<ffffffff9a576ddb>] system_call_fastpath+0x22/0x27
[11518613.658872] ---[ end trace 1bdd7051a99b6e9f ]---

Thanks in advance.

User avatar
TrevorH
Site Admin
Posts: 33202
Joined: 2009/09/24 10:40:56
Location: Brighton, UK

Re: Crash - vmx_deliver_posted_interrupt

Post by TrevorH » 2021/10/04 12:23:58

The bugzilla entry for the non-rt kernel is listed in the one you linked: 1719525 and the rpm -q --changelog output for the current kernel has

Code: Select all

- [x86] kvm: vmx: remove WARN_ON_ONCE in kvm_vcpu_trigger_posted_interrupt (Peter Xu) [1719525]
listed as one of the fixes. They were both released on the same date: 2019-08-06 along with a few hundred other packages so that was part of a point release kernel in the 3.10.0-1066.el7 series which has been superceded by the 7.9 kernel series 3.10.0-1160, currently 3.10.0-1160.42.2.el7.
The future appears to be RHEL or Debian. I think I'm going Debian.
Info for USB installs on http://wiki.centos.org/HowTos/InstallFromUSBkey
CentOS 5 and 6 are deadest, do not use them.
Use the FAQ Luke

nokiacephusers
Posts: 9
Joined: 2017/09/15 12:02:31

Re: Crash - vmx_deliver_posted_interrupt

Post by nokiacephusers » 2021/10/05 05:48:19

Thank you TrevorH

Post Reply