10:07 a.m.
Daniel Veillard wrote:
On Sat, Jul 25, 2009 at 08:02:11AM -0400, Bryan Kearney wrote:
> ---
> src/org/libvirt/Connect.java | 9 +
> src/org/libvirt/Domain.java | 307 +++++++++++++---------
> src/org/libvirt/DomainBlockStats.java | 13 +
> src/org/libvirt/DomainInfo.java | 14 +
> src/org/libvirt/DomainInterfaceStats.java | 17 ++
> src/org/libvirt/ErrorHandler.java | 11 +-
> src/org/libvirt/Network.java | 8 +-
> src/org/libvirt/NodeInfo.java | 4 +-
> src/org/libvirt/SchedBooleanParameter.java | 20 ++-
> src/org/libvirt/SchedDoubleParameter.java | 13 +
> src/org/libvirt/SchedIntParameter.java | 13 +
> src/org/libvirt/SchedLongParameter.java | 14 +
> src/org/libvirt/SchedParameter.java | 43 +++
> src/org/libvirt/SchedUintParameter.java | 13 +
> src/org/libvirt/SchedUlongParameter.java | 13 +
> src/org/libvirt/VcpuInfo.java | 13 +
> src/org/libvirt/jna/Libvirt.java | 42 +++-
> src/org/libvirt/jna/virDomainBlockStats.java | 12 +
> src/org/libvirt/jna/virDomainInfo.java | 13 +
> src/org/libvirt/jna/virDomainInterfaceStats.java | 16 ++
> src/org/libvirt/jna/virSchedParameter.java | 11 +
> src/org/libvirt/jna/virSchedParameterValue.java | 13 +
> src/org/libvirt/jna/virVcpuInfo.java | 12 +
> src/test.java | 154 ++++++-----
> 24 files changed, 595 insertions(+), 203 deletions(-)
> create mode 100644 src/org/libvirt/jna/virDomainBlockStats.java
> create mode 100644 src/org/libvirt/jna/virDomainInfo.java
> create mode 100644 src/org/libvirt/jna/virDomainInterfaceStats.java
> create mode 100644 src/org/libvirt/jna/virSchedParameter.java
> create mode 100644 src/org/libvirt/jna/virSchedParameterValue.java
> create mode 100644 src/org/libvirt/jna/virVcpuInfo.java
[...]
> }
> +
> + public static int[] convertUUIDBytes(byte bytes[]) {
> + int[] returnValue = new int[Libvirt.VIR_UUID_BUFLEN] ;
> + for (int x = 0 ; x < Libvirt.VIR_UUID_BUFLEN ; x++) {
> + returnValue[x] = (int) bytes[x] ;
> + }
> + return returnValue ;
> + }
> +
bahh, it's not too bad :-)
[..]
> +import com.sun.jna.ptr.IntByReference;
[..]
> public boolean getAutostart() throws LibvirtException{
> -// return _getAutostart(VDP);
> - throw new RuntimeException("Not Implemented") ;
> + IntByReference autoStart = new IntByReference() ;
> + libvirt.virDomainGetAutostart(VDP, autoStart) ;
> + processError() ;
> + return autoStart.getValue() != 0 ? true : false ;
> }
oh, it can handle int * ... cool :-)
This is the nastiest the code gets outside of the jna package. It is
pretty nice.
[...]
> public SchedParameter[] getSchedulerParameters() throws LibvirtException{
> -// return _getSchedulerParameters(VDP);
> - throw new RuntimeException("Not Implemented") ;
> + IntByReference nParams = new IntByReference() ;
> + SchedParameter[] returnValue = new SchedParameter[0] ;
> + String scheduler = libvirt.virDomainGetSchedulerType(VDP, nParams) ;
> + processError() ;
> + if (scheduler != null) {
> + virSchedParameter[] nativeParams = new
virSchedParameter[nParams.getValue()] ;
> + returnValue = new SchedParameter[nParams.getValue()] ;
> + libvirt.virDomainGetSchedulerParameters(VDP, nativeParams, nParams) ;
> + processError() ;
> + for (int x = 0 ; x < nParams.getValue() ; x++ ) {
> + returnValue[x] = SchedParameter.create(nativeParams[x]) ;
> + }
> + }
> +
> + return returnValue ;
> }
>
> -// private native SchedParameter[] _getSchedulerParameters (long VDP) throws
LibvirtException;
>
> /**
> * Changes the scheduler parameters
> @@ -220,11 +248,16 @@ public class Domain {
> * @throws LibvirtException
> */
> public void setSchedulerParameters(SchedParameter[] params) throws
LibvirtException{
> -// _setSchedulerParameters(VDP, params);
> - throw new RuntimeException("Not Implemented") ;
> + IntByReference nParams = new IntByReference() ;
> + nParams.setValue(params.length) ;
> + virSchedParameter[] input = new virSchedParameter[params.length] ;
> + for (int x = 0 ; x < params.length ; x++) {
> + input[x] = SchedParameter.toNative(params[x]) ;
> + }
> + libvirt.virDomainSetSchedulerParameters(VDP, input, nParams) ;
> + processError() ;
> }
Get and SetSchedulerParameters are the most free-form APIs in libvirt,
if that's all it takes, I would say congrats to JNA design, of course
there is a separate classes for them but they are really simple and the
the result looks natural, great !
[...]
> +++ b/src/org/libvirt/jna/virDomainBlockStats.java
> @@ -0,0 +1,12 @@
> +package org.libvirt.jna;
> +
> +import com.sun.jna.Structure;
> +
> +public class virDomainBlockStats extends Structure
> +{
> + public long rd_req ;
> + public long rd_bytes ;
> + public long wr_req ;
> + public long wr_bytes ;
> + public long errs ;
> +}
Hum ... in C we have
struct _virDomainBlockStats {
long long rd_req; /* number of read requests */
long long rd_bytes; /* number of read bytes */
long long wr_req; /* number of write requests */
long long wr_bytes; /* number of written bytes */
long long errs; /* In Xen this returns the mysterious 'oo_req'. */
};
will java long really be able to match C long long scope ?
Per the JNA mappings, long long equals java long... so this is correct.
I did add an explicit comment on each line which uses a long in java
instead of a NativeLong.
> +package org.libvirt.jna;
> +
> +import com.sun.jna.Structure;
> +
> +public class virDomainInterfaceStats extends Structure
> +{
> + public long rx_bytes;
> + public long rx_packets;
> + public long rx_errs;
> + public long rx_drop;
> + public long tx_bytes;
> + public long tx_packets;
> + public long tx_errs;
> + public long tx_drop;
> +
> +}
and same question here.
Except for this question looks fine, ACK !
Same as above.