+++ This bug was initially created as a clone of Bug #1901004 +++ Description of problem: Failed to create 1G huge pages on rhel guest. mount -t hugetlbfs -o pagesize=1G none /dev/hugepages1G Version-Release number of selected component (if applicable): Red Hat OpenStack Platform release 16.1.3 GA (Train) RHOS-16.1-RHEL-8-20201110.n.1 How reproducible: Always, with regression of NFV perf ci, Steps to Reproduce: 1. Deploy, Director w/ ovs+dpdk, not a must, with the following templates. https://code.engineering.redhat.com/gerrit/gitweb?p=nfv-qe.git;a=tree;f=ospd-16.1-vxlan-dpdk-sriov-ctlplane-dataplane-bonding-hybrid;h=247e16dd031230866fa91e9121305b6409b77d0c;hb=refs/heads/ci 2. Deploy vm rhel7.6 image and the follwoing extra-specs aggregate_instance_extra_specs:flavor='dut_ag', hw:cpu_policy='dedicated', hw:emulator_threads_policy='share', hw:mem_page_size='1GB', hw:numa_cpus.1='0,1,2,3,4,5,6,7', hw:numa_mem.1='8192', hw:numa_nodes='1' 3. ssh to guest, Update kernel with the following: Add the following to /etc/default/grub default_hugepagesz=1G hugepagesz=1G hugepages=2 4. grub2-mkconfig -o /boot/grub2/grub.cfg This is causing to kernel crash delete from /etc/fstab: nodev /mnt/huge hugetlbfs pagesize=1GB 0 0 5. try the following: mount -t hugetlbfs nodev /mnt/huge mkdir /dev/hugepages1G mount -t hugetlbfs -o pagesize=1G none /dev/hugepages1G Actual results: Receive the following err: [ 402.080533] hugetlbfs: Unsupported page size 1024 MB Expected results: mount should complete success Additional info: sos report will be added as first comment with link --- Additional comment from Miguel Angel Nieto on 2020-11-24 15:01:31 UTC --- I have checked that with RHOS-16.1-RHEL-8-20201021.n.0 we had in the configuration: <cpu mode='custom' match='exact' check='full'> <model fallback='forbid'>Broadwell-IBRS</model> <vendor>Intel</vendor> <topology sockets='8' dies='1' cores='1' threads='2'/> <feature policy='require' name='vme'/> <feature policy='require' name='ss'/> <feature policy='require' name='vmx'/> <feature policy='require' name='f16c'/> <feature policy='require' name='rdrand'/> <feature policy='require' name='hypervisor'/> <feature policy='require' name='arat'/> <feature policy='require' name='tsc_adjust'/> <feature policy='require' name='umip'/> <feature policy='require' name='stibp'/> <feature policy='require' name='arch-capabilities'/> <feature policy='require' name='xsaveopt'/> <feature policy='require' name='pdpe1gb'/> <feature policy='require' name='abm'/> <feature policy='require' name='ibpb'/> <feature policy='require' name='skip-l1dfl-vmentry'/> <feature policy='require' name='pschange-mc-no'/> <numa> <cell id='0' cpus='0-15' memory='8388608' unit='KiB' memAccess='shared'/> </numa> </cpu> while in RHOS-16.1-RHEL-8-20201110.n.1 we have the configuration: <cpu mode='custom' match='exact' check='full'> <model fallback='forbid'>qemu64</model> <topology sockets='8' dies='1' cores='1' threads='2'/> <feature policy='require' name='x2apic'/> <feature policy='require' name='hypervisor'/> <feature policy='require' name='lahf_lm'/> <feature policy='disable' name='svm'/> <numa> <cell id='0' cpus='0-15' memory='8388608' unit='KiB' memAccess='shared'/> </numa> </cpu> Feature pdpe1gb is missing in the guest, so I think it is not possible to create 1G hugepages --- Additional comment from Miguel Angel Nieto on 2020-11-24 15:20:22 UTC --- this is from /proc/cpuinfo in the guest for both versions: RHOS-16.1-RHEL-8-20201021.n.0: processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 61 model name : Intel Core Processor (Broadwell, IBRS) stepping : 2 microcode : 0x1 cpu MHz : 2199.998 cache size : 16384 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 1 apicid : 0 initial apicid : 0 fpu : yes fpu_exception : yes cpuid level : 13 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology eagerfpu pni pclmulqdq vmx ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm rdseed adx smap xsaveopt arat umip spec_ctrl intel_stibp arch_capabilities bogomips : 4399.99 clflush size : 64 cache_alignment : 64 address sizes : 46 bits physical, 48 bits virtual power management: RHOS-16.1-RHEL-8-20201110.n.1: processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 13 model name : QEMU Virtual CPU version 2.5+ stepping : 3 microcode : 0x1 cpu MHz : 2199.998 cache size : 16384 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 1 apicid : 0 initial apicid : 0 fpu : yes fpu_exception : yes cpuid level : 13 wp : yes flags : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse336 clflush mmx fxsr sse sse2 ht syscall nx lm rep_good nopl xtopology eagerfpu pnni cx16 x2apic hypervisor lahf_lm bogomips : 4399.99 clflush size : 64 cache_alignment : 64 address sizes : 46 bits physical, 48 bits virtual power management: --- Additional comment from Martin Schuppert on 2020-11-25 10:22:39 UTC --- https://github.com/openstack/tripleo-heat-templates/commit/772b7398a7222e8b286848ba00c06006d6b68785 introduced THT parameters to set libvirt/cpu_mode. The patch sets the NovaLibvirtCPUMode wrong to 'none' string which results in puppet-nova not to handle the default cases correct and sets libvirt/cpu_mode to none which results in qemu64 CPU model be used. A workaround should be to set NovaLibvirtCPUMode to host-model
1901004 was already verified by DFG NFV
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Red Hat OpenStack Platform (RHOSP) 16.2 enhancement advisory), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHEA-2021:3483