X-Git-Url: http://git.annexia.org/?a=blobdiff_plain;f=README;h=42c6cedcc40c304a033fdf54ec1643d7ce19c160;hb=19e979ec1896146465bdde1ea7c039d0c96e2026;hp=c78d43585199c1df63d9b421b788ad9978fefe2b;hpb=71281547e5d6756ead341e1708e5ec41d4e09452;p=supernested.git

diff --git a/README b/README
index c78d435..42c6ced 100644
--- a/README
+++ b/README
@@ -23,6 +23,10 @@ Gnu General Public License, version 2 or above.  See 'COPYING'.
 Requirements
 ----------------------------------------------------------------------
 
+ - Have a look at the 'build-supernested.sh.in' script for a variable
+   called $packages.  You need to install all of those packages on the
+   host.
+
  - supermin version 5
    https://github.com/libguestfs/supermin
    Note: older versions of supermin will not work
@@ -49,6 +53,10 @@ To build the disk image, do:
   ./configure
   make
 
+If supermin and qemu are not in the normal location on the path, you
+can set SUPERMIN and QEMU environment variables before running
+./configure.
+
 Assuming that works you will end up with 4 files:
 
  - 'supernested-<VERSION>-<DISTRO>.qcow2'
@@ -72,9 +80,18 @@ on baremetal, on a machine which is KVM-capable, a modern CPU, with
 lots of free memory, and nothing much else running.
 
 You need to have the kernel KVM module (eg. kvm_amd or kvm_intel)
-loaded, with the nested flag set.  The nested flag defaults to true in
-recent kernels.  /dev/kvm needs to be accessible.  On some distros
-you may need to add your user to a special 'kvm' group.
+loaded, with the nested=1 flag.  The nested flag defaults to true for
+AMD in recent kernels, but not for Intel.  To force this, do:
+
+  mkdir -p /etc/modprobe.d
+  echo "options kvm_intel nested=1" > /etc/modprobe.d/kvm.conf
+
+To see if nested is enabled, use:
+
+  cat /sys/modules/kvm_(intel|amd)/parameters/nested
+
+/dev/kvm needs to be accessible.  On some distros you may need to add
+your user to a special 'kvm' group.
 
 You *don't* need to run this command as root.  Running it as root may
 give different results.  Consult a qemu/KVM expert.