On Tue, Feb 04, 2014 at 12:18:34PM -0500, Marcelo Tosatti wrote:
On Tue, Feb 04, 2014 at 05:10:13PM +0000, Daniel P. Berrange wrote:
Because there is no guarantee with -mem-prealloc. For instance, if the hugepage path is not actually hugetlbfs backed, QEMU falls back to malloc().
Well if you can't fix -mem-prealloc to properly report errors for reasons of back compat, then it is certainly possible to add a further 'strict=yes|no' option to the CLI arg request that it verify this. I don't see any reason why this checking code should be in libvirt rather than QEMU.
Regards, Daniel
OK, do you want the StrictHugepage element to exist, and default libvirt behaviour to remain as it is today (fallback to malloc),
or
do you want libvirt to pass strict mode to -mem-prealloc as default when using hugepages element? (which would change behaviour of libvirt for existing guests)
I'd say the latter - if the user is requesting hugepages they should be guaranteed to get them. Libvirt doesn't ever use -mem-path with a non-hugetlbfs filesystem path AFAIR, so the fallback to malloc doesn't really make any sense for our needs. Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|