qemu_init: increase NOFILE soft limit on POSIX (!226) · Merge requests · Red Hat / centos-stream / src / qemu-kvm

Cornelia Huck requested to merge cohuck/qemu-kvm-c9s:bump-nofile-soft-limit into c9s Feb 27, 2024

Jira: https://issues.redhat.com/browse/RHEL-26049

We're running into the NOFILE soft limit when trying to run guests with more than 246 vCPUs on aarch64, as a bunch of eventfds are used per vCPU (and the code needs to fallback to a slower, untested userspace implementation). Follow upstream in bumping the soft limit, which is the recommended way to handle that.

qemu_init: increase NOFILE soft limit on POSIX

Merge request reports