-
Bug
-
Resolution: Done
-
Undefined
-
None
-
CentOS Stream 9
-
None
-
No
-
None
-
rhel-sst-virtualization
-
ssg_virtualization
-
None
-
False
-
-
None
-
None
-
None
-
None
-
-
x86_64
-
None
What were you trying to do that didn't work?
In centos-ci (aws, xen virtualization, TCG), sometimes (maybe once in 20 retries) we see the following crash when executing qemu-kvm VM:
# coredumpctl dump PID: 69976 (qemu-kvm) UID: 0 (root) GID: 0 (root) Signal: 11 (SEGV) Timestamp: Thu 2024-09-05 10:41:33 UTC (1h 14min ago) Command Line: /usr/libexec/qemu-kvm -cpu max -machine q35 -smp 6 -m 1200 -nodefaults -vga none -display none -no-reboot -device virtio-rng -rtc base=localtime -pidfile /var/dracut_test/qemu.pid -drive format=raw,index=0,media=disk,file=/var/dracut_test/client_log.img -drive format=raw,index=1,media=disk,file=/var/dracut_test/client_check.img -drive format=raw,index=2,media=disk,file=/var/dracut_test/client_dumps.img -device virtio-net,netdev=nfs,mac=52:54:00:12:34:00 -netdev tap,id=nfs,script=/root/NetworkManager-ci/contrib/dracut/qemu-ifup/nfs -append $'panic=1 systemd.crash_reboot rd.shell=0 biosdevname=0 net.ifnames=0 noapic loglevel=7 enforcing=0 console=ttyS0,115200n81, intel_iommu=on root=dhcp ro nm.debug' -initrd /var/dracut_test/initramfs.client.NM -serial stdio -kernel /lib/modules/5.14.0-503.el9.x86_64/vmlinuz Executable: /usr/libexec/qemu-kvm Control Group: /user.slice/user-0.slice/session-3.scope Unit: session-3.scope Slice: user-0.slice Session: 3 Owner UID: 0 (root) Boot ID: 36431b96c20c43c2a12509ddafa5d687 Machine ID: c13ef1c4f89d456ba87137edc1b0ed80 Hostname: n27-17-146.pool.ci.centos.org Storage: /var/lib/systemd/coredump/core.qemu-kvm.0.36431b96c20c43c2a12509ddafa5d687.69976.1725532893000000.zst (truncated) Size on Disk: 59.4M Message: Process 69976 (qemu-kvm) of user 0 dumped core. Stack trace of thread 69986: #0 0x000056464af352b3 address_space_translate_for_iotlb (/usr/libexec/qemu-kvm + 0x7712b3) #1 0x000056464af834cc tlb_set_page_full (/usr/libexec/qemu-kvm + 0x7bf4cc) #2 0x000056464ade5059 x86_cpu_tlb_fill (/usr/libexec/qemu-kvm + 0x621059) #3 0x000056464ade6111 mmu_translate (/usr/libexec/qemu-kvm + 0x622111) #4 0x000056464ade4f86 x86_cpu_tlb_fill (/usr/libexec/qemu-kvm + 0x620f86) #5 0x000056464af8e470 mmu_lookup1 (/usr/libexec/qemu-kvm + 0x7ca470) #6 0x000056464af8e1e0 mmu_lookup.llvm.10739711994244380661 (/usr/libexec/qemu-kvm + 0x7ca1e0) #7 0x000056464af852f7 do_ld4_mmu.llvm.10739711994244380661 (/usr/libexec/qemu-kvm + 0x7c12f7) #8 0x000056464ae792c2 do_interrupt_all (/usr/libexec/qemu-kvm + 0x6b52c2) #9 0x000056464adebf61 x86_cpu_exec_interrupt (/usr/libexec/qemu-kvm + 0x627f61) #10 0x000056464af6f189 cpu_exec_loop (/usr/libexec/qemu-kvm + 0x7ab189) #11 0x000056464af6e451 cpu_exec_setjmp (/usr/libexec/qemu-kvm + 0x7aa451) #12 0x000056464af6e27e cpu_exec (/usr/libexec/qemu-kvm + 0x7aa27e) #13 0x00007efe98d7019c n/a (n/a + 0x0) ELF object binary architecture: AMD x86-64
The corefile is saved here:
http://tools.lab.eng.brq2.redhat.com/~fpokryvk/reports/qemu-kvm.core
Seems very similar to upstream issue:
https://gitlab.com/qemu-project/qemu/-/issues/2220
Please provide the package NVR for which bug is seen:
qemu-kvm-9.0.0-9.el9.x86_64
How reproducible:
sometimes (maybe 1:20)
Steps to reproduce
- run the following test from NetworkManager-ci in the loop:
- ./test_run.sh dracut_NM_NFS_root_dhcp_nm_debug
Expected results
the test should pass reliably
Actual results
test sometimes fails due to qemu-kvm crash
