On 01/29/2011 10:43 PM, Osier Yang wrote:
* tools/virsh.pod
---
tools/virsh.pod | 6 +++++-
1 files changed, 5 insertions(+), 1 deletions(-)
diff --git a/tools/virsh.pod b/tools/virsh.pod
index 441ca8a..d3262b6 100644
--- a/tools/virsh.pod
+++ b/tools/virsh.pod
@@ -566,7 +566,11 @@ Xen (credit scheduler): weight, cap
ESX (allocation scheduler): reservation, limit, shares
-B<Note>: The cpu_shares parameter has a valid value range of 0-262144.
+B<Note>: The cpu_shares parameter has a valid value range of 0-262144, however,
+negative number and positive number greater than 262144 are allowed, but
+negative number will be converted into positive number, and positive number
+greater than 262144 will be reduced to 262144 by cgroup in the end, e.g. -1
+can be used as useful shorthand for 262144.
The range of 262144 is specific to cgroups (qemu and lxc); it is
possible that some other hypervisor may add similar support for this
parameter, but have a different range. Then again, we already
documented it is specific to qemu and lxc, so we could update a change
in range at the time of introducing the cpu_shares parameter to another
hypervisor.
However, your attempt is still wordy. Can we go with something more
compact, as in:
B<Note>: The cpu_shares parameter has a valid value range of 0-262144;
negative values are wrapped to positive, and larger values are capped at
the maximum. Therefore, -1 is a useful shorthand for 262144.
--
Eric Blake eblake(a)redhat.com +1-801-349-2682
Libvirt virtualization library
http://libvirt.org