[cp-patches] FYI: Small additional ClassLoader VM extensibility point
Hi,
I committed the attached patch, that allows a VM implementer to hook
Classloader.findLoadedClass.
Regards,
Jeroen
2005-07-25 Jeroen Frijters [EMAIL PROTECTED]
* java/lang/ClassLoader.java
(findLoadedClass): Call VMClassLoader.findLoadedClass.
* vm/reference/java/lang/VMClassLoader.java
(findLoadedClass): New method.
Index: java/lang/ClassLoader.java
===
RCS file: /cvsroot/classpath/classpath/java/lang/ClassLoader.java,v
retrieving revision 1.52
diff -u -r1.52 ClassLoader.java
--- java/lang/ClassLoader.java 2 Jul 2005 20:32:38 - 1.52
+++ java/lang/ClassLoader.java 25 Jul 2005 09:28:46 -
@@ -556,9 +556,7 @@
*/
protected final synchronized Class findLoadedClass(String name)
{
-// NOTE: If the VM is keeping its own cache, it may make sense to have
-// this method be native.
-return (Class) loadedClasses.get(name);
+return VMClassLoader.findLoadedClass(this, name);
}
/**
Index: vm/reference/java/lang/VMClassLoader.java
===
RCS file:
/cvsroot/classpath/classpath/vm/reference/java/lang/VMClassLoader.java,v
retrieving revision 1.26
diff -u -r1.26 VMClassLoader.java
--- vm/reference/java/lang/VMClassLoader.java 2 Jul 2005 20:33:08 -
1.26
+++ vm/reference/java/lang/VMClassLoader.java 25 Jul 2005 09:28:53 -
@@ -282,4 +282,12 @@
{
return ClassLoader.defaultGetSystemClassLoader();
}
+
+ /**
+ * If the VM wants to keep its own cache, this method can be replaced.
+ */
+ static Class findLoadedClass(ClassLoader cl, String name)
+ {
+return (Class) cl.loadedClasses.get(name);
+ }
}
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
Re: [cp-patches] [generics] Enum#valueOf implementation
On Mon, 2005-07-25 at 00:33 -0600, Tom Tromey wrote:
Ewout == Ewout Prangsma [EMAIL PROTECTED] writes:
Ewout Here's the unified diff. Again my question if someone can
Ewout commit this, or give me CVS access rights.
It still needs a ChangeLog entry.
Also the code is not formatted according to the Classpath style.
Here is an example of a correct ChangeLog entry and the patch formatted
as (I also added a little documentation, please check).
2005-07-25 Ewout Prangsma [EMAIL PROTECTED]
* java/lang/Enum.java (valueOf): implemented.
ChangeLog and Coding style is discussed int he hacking guide:
http://www.gnu.org/software/classpath/docs/hacking.html#SEC6
http://www.gnu.org/software/classpath/docs/hacking.html#SEC8
One question. Will the IllegalAccessException really never be thrown? I
can imagine that the Enum value fields are always public or package
private to java.lang in which case this is true (but I haven't studied
Enums at all). Otherwise we might need a PrivilegedAction that calls
setAccessible(true).
Andrew, could you please take a look and tell me whether it is OK to
commit (compiles, don't have ecj or gcjx installed here) and whether or
not it interferes with you current merging work?
Thanks,
Mark
Index: java/lang/Enum.java
===
RCS file: /cvsroot/classpath/classpath/java/lang/Attic/Enum.java,v
retrieving revision 1.1.2.4
diff -u -r1.1.2.4 Enum.java
--- java/lang/Enum.java 21 Mar 2005 12:10:17 - 1.1.2.4
+++ java/lang/Enum.java 25 Jul 2005 10:37:30 -
@@ -1,5 +1,5 @@
/* Enum.java - Base class for all enums
- Copyright (C) 2004 Free Software Foundation
+ Copyright (C) 2004, 2005 Free Software Foundation
This file is part of GNU Classpath.
@@ -60,11 +60,32 @@
this.ordinal = ordinal;
}
+ /**
+ * Returns an Enum for a enum class given a description string of
+ * the enum constant.
+ *
+ * @exception NullPointerException when etype or s are null.
+ * @exception IllegalArgumentException when there is no value s in
+ * the enum etype.
+ */
+ @SuppressWarnings(unchecked)
public static S extends EnumS Enum valueOf(ClassS etype, String s)
{
if (etype == null || s == null)
throw new NullPointerException();
-return null; // FIXME
+
+try
+ {
+ return (S) etype.getDeclaredField(s).get(null);
+ }
+catch (NoSuchFieldException exception)
+ {
+ throw new IllegalArgumentException(s);
+ }
+catch (IllegalAccessException exception)
+ {
+ throw new Error(Unable to access Enum class);
+ }
}
public final boolean equals(Object o)
signature.asc
Description: This is a digitally signed message part
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches] RFC: cleanup in nio native code
While merging some nio code I came upon some strange code in
gnu_java_nio_channels_FileChannelImpl.c. There is some code that
initializes the field in, out and err, which I would think would better
fit in the Java code. I propose the following patch to move this init
code there. Are there any objections? May I commit this?
2005-07-25 Roman Kennke [EMAIL PROTECTED]
* gnu/java/nio/channels/FileChannelImpl.java
(static initializer): Init out, err and in here.
* native/jni/java-nio/gnu_java_nio_channels_FileChannelImpl.c
(Java_gnu_java_nio_channels_FileChannelImpl_init): Moved init
code
for in, out and err to Java code.
/Roman
Index: gnu/java/nio/channels/FileChannelImpl.java
===
RCS file:
/cvsroot/classpath/classpath/gnu/java/nio/channels/FileChannelImpl.java,v
retrieving revision 1.15
diff -u -r1.15 FileChannelImpl.java
--- gnu/java/nio/channels/FileChannelImpl.java 11 Jul 2005 17:27:55 -
1.15
+++ gnu/java/nio/channels/FileChannelImpl.java 25 Jul 2005 10:46:41 -
@@ -73,6 +73,10 @@
public static final int SYNC = 16;
public static final int DSYNC = 32;
+ public static FileChannelImpl in;
+ public static FileChannelImpl out;
+ public static FileChannelImpl err;
+
private static native void init();
static
@@ -83,6 +87,12 @@
}
init();
+
+// Replaced native code by simpler, equivalent Java code for
+// readability and analysability.
+in = new FileChannelImpl(0,READ);
+out = new FileChannelImpl(1,WRITE);
+err = new FileChannelImpl(2,WRITE);
}
/**
@@ -130,16 +140,20 @@
}
}
- /* Used by init() (native code) */
+ /**
+ * Constructor for default channels in, out and err.
+ *
+ * Used by init() (native code).
+ *
+ * @param fd the file descriptor (0, 1, 2 for stdin, stdout, stderr).
+ *
+ * @param mode READ or WRITE
+ */
FileChannelImpl (int fd, int mode)
{
this.fd = fd;
this.mode = mode;
}
-
- public static FileChannelImpl in;
- public static FileChannelImpl out;
- public static FileChannelImpl err;
private native int open (String path, int mode) throws FileNotFoundException;
Index: native/jni/java-nio/gnu_java_nio_channels_FileChannelImpl.c
===
RCS file:
/cvsroot/classpath/classpath/native/jni/java-nio/gnu_java_nio_channels_FileChannelImpl.c,v
retrieving revision 1.19
diff -u -r1.19 gnu_java_nio_channels_FileChannelImpl.c
--- native/jni/java-nio/gnu_java_nio_channels_FileChannelImpl.c 24 Jul 2005
03:35:58 - 1.19
+++ native/jni/java-nio/gnu_java_nio_channels_FileChannelImpl.c 25 Jul 2005
10:46:41 -
@@ -136,28 +136,6 @@
}
native_fd_fieldID = field;
-
- constructor = (*env)-GetMethodID (env, clazz, init, (II)V);
- if (!constructor)
-return;
-
-#define INIT_FIELD(FIELDNAME, FDVALUE, MODE) \
- field = (*env)-GetStaticFieldID (env, clazz, FIELDNAME, \
- Lgnu/java/nio/channels/FileChannelImpl;);
\
- if (! field) \
-return;\
- obj = (*env)-NewObject (env, clazz, constructor, FDVALUE, MODE);\
- if (! obj) \
-return;\
- (*env)-SetStaticObjectField (env, clazz, field, obj); \
- if ((*env)-ExceptionOccurred (env)) \
-return;
-
- INIT_FIELD (in, 0, FILECHANNELIMPL_READ);
- INIT_FIELD (out, 1, FILECHANNELIMPL_WRITE);
- INIT_FIELD (err, 2, FILECHANNELIMPL_WRITE);
-
-#undef INIT_FIELD
}
/*
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches] FYI: Update assign faq entry
Hi, I got some questions about our contribution requirements so I asked FSF-legal for a better explanation of the assignment policy to add to the FAQ explaining that it is a contract with the FSF committing the FSF to distribute your contribution and derived works under free terms and that you can get a grant-back for any contribution you made for use in other projects (even proprietary) if you wish. Here it is. 2005-07-25 Mark Wielaard [EMAIL PROTECTED] * doc/www.gnu.org/faq.wml: Expand contribution assign answer. Committed, Mark ? doc/www.gnu.org/downloads.diff ? doc/www.gnu.org/escape.patch ? doc/www.gnu.org/gjdoc-home.patch ? doc/www.gnu.org/cp-tools/cp-tools.html Index: doc/www.gnu.org/faq/faq.wml === RCS file: /cvsroot/classpath/classpath/doc/www.gnu.org/faq/faq.wml,v retrieving revision 1.18 diff -u -r1.18 faq.wml --- doc/www.gnu.org/faq/faq.wml 4 Jul 2005 14:31:01 - 1.18 +++ doc/www.gnu.org/faq/faq.wml 25 Jul 2005 11:29:16 - @@ -304,6 +304,28 @@ not see this as a personal offence. pGiving the copyright to the FSF also gives us a clear paper trail where changes come from, which confirms our clean-room status. +/p + +p +The assignment contract commits the foundation to setting distribution terms +that permit free redistribution. +/p + +p +The assignment contract we normally use has a clause that permits you to +use your code in proprietary programs, on 30 days' notice. +(The 30 days' notice is there because, through a legal technicality, +it would improve our position in a suit against a hoarder.) +Although we believe that proprietary software is wrong, we include this +clause because it would serve no purpose to ask you to promise not to do +it. You're giving us a gift in the first place. +/p + +p +You don't need to invoke this clause in order to distribute +copies as free software under the GNU GPL, since everyone is +allowed to do that. +/p pSee also a href=http://www.gnu.org/licenses/why-assign.html;http://www.gnu.org/licenses/why-assign.html/a. /p signature.asc Description: This is a digitally signed message part ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches] RFC: RMI Class loader fix
Hi,
I'm not an RMI person, but an IKVM user ran into an RMI problem that I
think is caused by a problem in the RMI class loader that the attached
patch should fix.
Please comment.
Regards,
Jeroen
* java/rmi/server/RMIClassLoader.java
(loadClass(String, String)): Use Class.forName() instead of
directly calling ClassLoader.loadClass(), to handle array
types correctly.
Index: java/rmi/server/RMIClassLoader.java
===
RCS file: /cvsroot/classpath/classpath/java/rmi/server/RMIClassLoader.java,v
retrieving revision 1.16
diff -u -r1.16 RMIClassLoader.java
--- java/rmi/server/RMIClassLoader.java 2 Jul 2005 20:32:40 - 1.16
+++ java/rmi/server/RMIClassLoader.java 25 Jul 2005 12:48:48 -
@@ -214,7 +214,7 @@
//try context class loader first
try
{
-return loader.loadClass (name);
+return Class.forName(name, false, loader);
}
catch (ClassNotFoundException e)
{
@@ -237,7 +237,7 @@
) at codebase ( + codebases + ));
}
-return loader.loadClass (name);
+return Class.forName(name, false, loader);
}
/**
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
Re: [cp-patches] RFC: cleanup in nio native code
Ok, as I was told on IRC, I have committed the attached slightly (only
comments) changed patch.
2005-07-25 Roman Kennke [EMAIL PROTECTED]
* gnu/java/nio/channels/FileChannelImpl.java
(static initializer): Init out, err and in here.
* native/jni/java-nio/gnu_java_nio_channels_FileChannelImpl.c
(Java_gnu_java_nio_channels_FileChannelImpl_init): Moved init
code
for in, out and err to Java code.
/Roman
Index: gnu/java/nio/channels/FileChannelImpl.java
===
RCS file:
/cvsroot/classpath/classpath/gnu/java/nio/channels/FileChannelImpl.java,v
retrieving revision 1.15
diff -u -r1.15 FileChannelImpl.java
--- gnu/java/nio/channels/FileChannelImpl.java 11 Jul 2005 17:27:55 -
1.15
+++ gnu/java/nio/channels/FileChannelImpl.java 25 Jul 2005 13:52:11 -
@@ -73,6 +73,10 @@
public static final int SYNC = 16;
public static final int DSYNC = 32;
+ public static FileChannelImpl in;
+ public static FileChannelImpl out;
+ public static FileChannelImpl err;
+
private static native void init();
static
@@ -83,6 +87,10 @@
}
init();
+
+in = new FileChannelImpl(0,READ);
+out = new FileChannelImpl(1,WRITE);
+err = new FileChannelImpl(2,WRITE);
}
/**
@@ -130,16 +138,20 @@
}
}
- /* Used by init() (native code) */
+ /**
+ * Constructor for default channels in, out and err.
+ *
+ * Used by init() (native code).
+ *
+ * @param fd the file descriptor (0, 1, 2 for stdin, stdout, stderr).
+ *
+ * @param mode READ or WRITE
+ */
FileChannelImpl (int fd, int mode)
{
this.fd = fd;
this.mode = mode;
}
-
- public static FileChannelImpl in;
- public static FileChannelImpl out;
- public static FileChannelImpl err;
private native int open (String path, int mode) throws FileNotFoundException;
Index: native/jni/java-nio/gnu_java_nio_channels_FileChannelImpl.c
===
RCS file:
/cvsroot/classpath/classpath/native/jni/java-nio/gnu_java_nio_channels_FileChannelImpl.c,v
retrieving revision 1.19
diff -u -r1.19 gnu_java_nio_channels_FileChannelImpl.c
--- native/jni/java-nio/gnu_java_nio_channels_FileChannelImpl.c 24 Jul 2005
03:35:58 - 1.19
+++ native/jni/java-nio/gnu_java_nio_channels_FileChannelImpl.c 25 Jul 2005
13:52:11 -
@@ -136,28 +136,6 @@
}
native_fd_fieldID = field;
-
- constructor = (*env)-GetMethodID (env, clazz, init, (II)V);
- if (!constructor)
-return;
-
-#define INIT_FIELD(FIELDNAME, FDVALUE, MODE) \
- field = (*env)-GetStaticFieldID (env, clazz, FIELDNAME, \
- Lgnu/java/nio/channels/FileChannelImpl;);
\
- if (! field) \
-return;\
- obj = (*env)-NewObject (env, clazz, constructor, FDVALUE, MODE);\
- if (! obj) \
-return;\
- (*env)-SetStaticObjectField (env, clazz, field, obj); \
- if ((*env)-ExceptionOccurred (env)) \
-return;
-
- INIT_FIELD (in, 0, FILECHANNELIMPL_READ);
- INIT_FIELD (out, 1, FILECHANNELIMPL_WRITE);
- INIT_FIELD (err, 2, FILECHANNELIMPL_WRITE);
-
-#undef INIT_FIELD
}
/*
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches] FYI: Some fixes and optimizations for Container
Hi,
I checked in some fixes and optimizations that I came upon while merging
with Jamaica's AWT. Most importantly the Lightweightdispatcher has been
reworked so that it is reentrant, and preferredSize is now working
optimized, so that it only recalculates the preferredSize if the
container is invalid.
2005-07-25 Roman Kennke [EMAIL PROTECTED]
* gnu/java/awt/AWTUtilities.java:
Added methods from SwingUtilities so that AWT does not have to
depend on Swing.
* java/awt/Component.java:
Reverted my DEFAULT_FONT patch from yesterday. This does not
seem to work with the Gtk peers.
* java/awt/Container.java
(addImpl): Call addNotify() on the added child. Invalidate not
only the container but also the added child. Repaint the
container.
(remove): Repaint the container.
(invalidate): Also invalidate the LayoutManager.
(invalidateTree): Call super.invalidate to invalidate the
container itself. Also invalidate the LayoutManager.
(setFont): Only set the font if the specified argument actually
differs from the current font.
(preferredSize): Optimized this method so the LayoutManager is
only called if the layout is invalid. Otherwise we return the
preferred size that has been stored during last
validation/layout.
(getAlignmentX): Despite common belief, this method does _not_
call the LayoutManagers getAlignmentX in the JDK. So we also
don't.
(getAlignmentY): Despite common belief, this method does _not_
call the LayoutManagers getAlignmentX in the JDK. So we also
don't.
(dispatchEventImpl): Let the dispatcher decide if it is enabled
for the incoming event type.
(eventTypeEnabled): Enables only container events for
containers.
(addNotifyContainerChildren): Coalesced two if statements into
one.
Enable events on the dispatcher for this container.
(LightweightDispatcher): Made this class reentrant. Handle
events enabling/disabling here.
/Roman
Index: gnu/java/awt/AWTUtilities.java
===
RCS file: /cvsroot/classpath/classpath/gnu/java/awt/AWTUtilities.java,v
retrieving revision 1.3
diff -u -r1.3 AWTUtilities.java
--- gnu/java/awt/AWTUtilities.java 2 Jul 2005 20:32:10 - 1.3
+++ gnu/java/awt/AWTUtilities.java 25 Jul 2005 13:45:05 -
@@ -1,5 +1,4 @@
-/* AWTUtilities.java -- Common utility methods for AWT and Swing.
- Copyright (C) 2005 Free Software Foundation, Inc.
+/* Copyright (C) 2004 Free Software Foundation
This file is part of GNU Classpath.
@@ -15,8 +14,8 @@
You should have received a copy of the GNU General Public License
along with GNU Classpath; see the file COPYING. If not, write to the
-Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
-02110-1301 USA.
+Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+02111-1307 USA.
Linking this library statically or dynamically with other modules is
making a combined work based on this library. Thus, the terms and
@@ -34,20 +33,30 @@
this exception to your version of the library, but you are not
obligated to do so. If you do not wish to do so, delete this
exception statement from your version. */
-
package gnu.java.awt;
+import java.applet.Applet;
import java.awt.Component;
import java.awt.Container;
+import java.awt.Font;
+import java.awt.FontMetrics;
+import java.awt.Insets;
+import java.awt.Point;
+import java.awt.Rectangle;
+import java.awt.Toolkit;
+import java.awt.Window;
+import java.awt.event.MouseEvent;
import java.util.AbstractSequentialList;
import java.util.List;
import java.util.ListIterator;
import java.util.NoSuchElementException;
import java.util.WeakHashMap;
+import java.lang.reflect.InvocationTargetException;
/**
- * This class provides utility methods that are commonly used in AWT
- * (and Swing).
+ * This class mirrors the javax.swing.SwingUtilities class. It
+ * provides commonly needed functionalities for AWT classes without
+ * the need to reference classes in the javax.swing package.
*/
public class AWTUtilities
{
@@ -317,5 +326,366 @@
visibleChildren = (VisibleComponentList) o;
return visibleChildren;
+ }
+
+ /**
+ * Calculates the portion of the base rectangle which is inside the
+ * insets.
+ *
+ * @param base The rectangle to apply the insets to
+ * @param insets The insets to apply to the base rectangle
+ * @param ret A rectangle to use for storing the return value, or
+ * codenull/code
+ *
+ * @return The calculated area inside the base rectangle and its insets,
+ * either stored in ret or a new Rectangle if ret is codenull/code
+ *
+ * @see #calculateInnerArea
+ */
+ public static Rectangle calculateInsetArea(Rectangle base, Insets insets,
+
[cp-patches] FYI: JTable selectAll fix
This patch ensures selectAll doesn't affect the lead selection indices
for the row or column ListSelectionModel. Since this is fixed, I also
replaced some direct code in BasicListUI with a call to selectAll. Also
a small touch-up to make the ENTER key press compatible with the JDK.
Patch attached.
2005-07-25 Anthony Balkissoon [EMAIL PROTECTED]
* javax/swing/JTable.java:
(selectAll): Store the lead selection indices and restore them after
selecting all cells.
* javax/swing/plaf/basic/BasicTableUI.java:
(KeyHandler.keyPressed): Changed the criteria for only one selection
when the ENTER key is pressed to match the behavior of the JDK. Also
replaced direct code for CTRL-A with call to JTable.selectAll().
- Tony
Index: javax/swing/JTable.java
===
RCS file: /cvsroot/classpath/classpath/javax/swing/JTable.java,v
retrieving revision 1.35
diff -u -r1.35 JTable.java
--- javax/swing/JTable.java 20 Jul 2005 20:25:08 - 1.35
+++ javax/swing/JTable.java 25 Jul 2005 14:46:07 -
@@ -2070,8 +2070,17 @@
public void selectAll()
{
+// rowLead and colLead store the current lead selection indices
+int rowLead = selectionModel.getLeadSelectionIndex();
+int colLead = getColumnModel().getSelectionModel().getLeadSelectionIndex();
+// the following calls to setSelectionInterval change the lead selection
+// indices
setColumnSelectionInterval(0, getColumnCount() - 1);
setRowSelectionInterval(0, getRowCount() - 1);
+// the following addSelectionInterval calls restore the lead selection
+// indices to their previous values
+addColumnSelectionInterval(colLead,colLead);
+addRowSelectionInterval(rowLead, rowLead);
}
public Object getValueAt(int row, int column)
Index: javax/swing/plaf/basic/BasicTableUI.java
===
RCS file: /cvsroot/classpath/classpath/javax/swing/plaf/basic/BasicTableUI.java,v
retrieving revision 1.14
diff -u -r1.14 BasicTableUI.java
--- javax/swing/plaf/basic/BasicTableUI.java 22 Jul 2005 19:44:35 - 1.14
+++ javax/swing/plaf/basic/BasicTableUI.java 25 Jul 2005 14:46:07 -
@@ -265,12 +265,17 @@
return;
}
- // If there is just one cell selected, select the next row, and wrap
+ // If there is just one selection, select the next row, and wrap
// when you get to the edges of the table.
- if ((table.getSelectedRowCount() = 1
- table.getSelectedColumnCount() = 1)
- || (table.getRowSelectionAllowed() == false
- table.getColumnSelectionAllowed() == false))
+ boolean multRowsSelected, multColsSelected;
+ multRowsSelected = (table.getSelectedRowCount() 1) ||
+(!table.getRowSelectionAllowed()
+ table.getSelectedColumnCount() 0);
+ multColsSelected = (table.getSelectedColumnCount() 1) ||
+(!table.getColumnSelectionAllowed()
+ table.getSelectedRowCount() 0);
+
+ if (!multColsSelected || !multRowsSelected)
{
rowModel.setSelectionInterval((rowLead + 1)%(rowMax + 1),
(rowLead + 1)%(rowMax + 1));
@@ -350,13 +355,7 @@
else if ((evt.getKeyCode() == KeyEvent.VK_A || evt.getKeyCode()
== KeyEvent.VK_SLASH) evt.isControlDown())
{
- rowModel.setSelectionInterval(0, rowMax);
- colModel.setSelectionInterval(0, colMax);
- // the next two lines are to restore the lead selection indices to
- // their previous values, because select-all operations shouldn't
- // change them
- rowModel.addSelectionInterval(rowLead, rowLead);
- colModel.addSelectionInterval(colLead, colLead);
+ table.selectAll();
}
else if (evt.getKeyCode() == KeyEvent.VK_BACK_SLASH
evt.isControlDown())
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
RE: [cp-patches] FYI: ClassLoader caching Part 2
Mark Wielaard wrote: Yes, but unfortunately that doesn't change the byte-code of the ClassLoader class. static final primitive (and String) field values are actually inlined into other classes. So you can replace VMClassLoader all you want with a version the has VM_USE_CACHE set to another value, that won't actually change the ClassLoader code paths unless ClassLoader is also recompiled against this new value. Really, try it out. Trust me, I know all this ;-) It is an annoying byte code optimization. It's Java's version of #define. So for things that can be different at runtime like the VMClasses you unfortunately cannot use static final fields of primitive types. As long as you build everything from source (with the replaced VM* classes) it works, but I must admit that I have no clue about how everyone builds the class library (personally, I don't use any of the build infrastructure that comes with GNU Classpath). Regards, Jeroen ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
RE: [cp-patches] RFC: RMI Class loader fix
Mark Wielaard wrote:
On Mon, 2005-07-25 at 14:52 +0200, Jeroen Frijters wrote:
I'm not an RMI person, but an IKVM user ran into an RMI
problem that I
think is caused by a problem in the RMI class loader that
the attached
patch should fix.
Please comment.
I assume this is because the code tries to load an array class and
forName() allows to specify arrays as type strings, but
loadClass() only allows for fully qualified class names? Our
documentation isn't very clear here. But if that is the case then
this patch looks good.
Unfortunately the Sun documentation isn't very clear either, but you are
correct that I believe this to be the case. BTW, clearly Class.forName()
allows array types, but ClassLoader.loadClass() isn't very clear and the
Sun implementation does some weird things. Here's an example I tried:
public class test
{
public static void main(String[] args) throws Exception
{
System.out.println(test.class.getClassLoader().loadClass(test));
Class.forName([Ltest;);
System.out.println(test.class.getClassLoader().loadClass([Ltest;));
}
}
If you run this on JDK 1.5, the second loadClass succeeds, but if you
comment out the Class.forName(), the second load class fails. On JDK
1.4, the Class.forName() isn't required, but if you comment out the
first loadClass, the second one will fail...
I will commit the RMI fix.
Regards,
Jeroen
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
Re: [cp-patches] RFC: RMI Class loader fix
Hi, On Mon, 2005-07-25 at 14:52 +0200, Jeroen Frijters wrote: I'm not an RMI person, but an IKVM user ran into an RMI problem that I think is caused by a problem in the RMI class loader that the attached patch should fix. Please comment. I assume this is because the code tries to load an array class and forName() allows to specify arrays as type strings, but loadClass() only allows for fully qualified class names? Our documentation isn't very clear here. But if that is the case then this patch looks good. Cheers, Mark signature.asc Description: This is a digitally signed message part ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
RE: [cp-patches] FYI: ClassLoader caching Part 2
Jeroen Frijters wrote: Mark Wielaard wrote: On Mon, 2005-07-25 at 16:29 +0200, Jeroen Frijters wrote: I committed the attached patch to complete to ability of the VM to bypass the class caching in ClassLoader. 2005-07-25 Jeroen Frijters [EMAIL PROTECTED] * java/lang/ClassLoader.java (loadedClasses): Set based on VMClassLoader.VM_USE_CACHE. (defineClass): Modified to respect VMClassLoader.VM_USE_CACHE. * vm/reference/java/lang/VMClassLoader.java (VM_USE_CACHE): New field. This won't work if you make the VM_USE_CACHE field static final. Then the constant will be compiled into ClassLoader making it impossible to override for the runtime vm-classes later. It has to be a static method for that to work. Like Archie says, your comment doesn't make sense ;-) If a VM decides not to use the cache in ClassLoader, it replaces VMClassLoader and sets the constant to true. Oh, I'm sorry, I think see what you mean. If a VM would take a binary GNU Classpath jar, it wouldn't be able to override the value of the flag. Is this an important scenario? I thought you long haired types didn't like binary distributions ;-) Regards, Jeroen ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches] FYI: Container fixlet
This patch fixes a typo in Container.java that was causing an array out of bounds exception with calls to setFocusedTraversalKeys. Patch attached. 2005-07-25 Anthony Balkissoon [EMAIL PROTECTED] * java/awt/Container.java: (setFocusTraversalKeys): Instantiate focusTraversalKeys to an array of size 4, not 3. This must have been a typo. - Tony Index: java/awt/Container.java === RCS file: /cvsroot/classpath/classpath/java/awt/Container.java,v retrieving revision 1.55 diff -u -r1.55 Container.java --- java/awt/Container.java 25 Jul 2005 14:18:02 - 1.55 +++ java/awt/Container.java 25 Jul 2005 16:31:35 - @@ -1225,7 +1225,7 @@ } if (focusTraversalKeys == null) - focusTraversalKeys = new Set[3]; + focusTraversalKeys = new Set[4]; keystrokes = Collections.unmodifiableSet (new HashSet (keystrokes)); firePropertyChange (name, focusTraversalKeys[id], keystrokes); ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches] FYI: RMI Class loader fix
Hi,
I applied the attached fix.
Regards,
Jeroen
* java/rmi/server/RMIClassLoader.java
(loadClass(String, String)): Use Class.forName() instead of
directly calling ClassLoader.loadClass(), to handle array
types correctly.
Index: java/rmi/server/RMIClassLoader.java
===
RCS file: /cvsroot/classpath/classpath/java/rmi/server/RMIClassLoader.java,v
retrieving revision 1.16
diff -u -r1.16 RMIClassLoader.java
--- java/rmi/server/RMIClassLoader.java 2 Jul 2005 20:32:40 - 1.16
+++ java/rmi/server/RMIClassLoader.java 25 Jul 2005 12:48:48 -
@@ -214,7 +214,7 @@
//try context class loader first
try
{
-return loader.loadClass (name);
+return Class.forName(name, false, loader);
}
catch (ClassNotFoundException e)
{
@@ -237,7 +237,7 @@
) at codebase ( + codebases + ));
}
-return loader.loadClass (name);
+return Class.forName(name, false, loader);
}
/**
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
Re: [cp-patches] FYI: ClassLoader caching Part 2
Jeroen Frijters wrote: Oh, I'm sorry, I think see what you mean. If a VM would take a binary GNU Classpath jar, it wouldn't be able to override the value of the flag. Is this an important scenario? I thought you long haired types didn't like binary distributions ;-) OK now I get it too... I call this idea of imposing VM-specific versions of a class in front of the stock Classpath versions overlaying instead of overriding to avoid confusion :-) Yes, I think not breaking overlaying is important.. so a static method (with comment why being used) would be better. -Archie __ Archie Cobbs *CTO, Awarix* http://www.awarix.com ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
Re: [cp-patches] RFC: New JCL helper functions for RawData handling
Guilhem Lavaux wrote: Ok. Then I will make a set of patches to first use the new common JCL interface. And then I will rename RawData* to Pointer*. Is it ok ? Absolutely. -- Chris Burdess ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches]: Patch: JTree painting
A fix for the case where there are no icons.
2005-07-25 Lillian Angel [EMAIL PROTECTED]
* javax/swing/plaf/basic/BasicTreeUI.java
(paint): moved code to paintNode
(getCellBounds): width increased, slightly short before
(paintNode): changed to paint node depending if icons exist
? examples/gnu/classpath/examples/swing/Demo$1.class
? examples/gnu/classpath/examples/swing/Demo$10.class
? examples/gnu/classpath/examples/swing/Demo$11.class
? examples/gnu/classpath/examples/swing/Demo$2.class
? examples/gnu/classpath/examples/swing/Demo$3.class
? examples/gnu/classpath/examples/swing/Demo$4.class
? examples/gnu/classpath/examples/swing/Demo$5.class
? examples/gnu/classpath/examples/swing/Demo$6.class
? examples/gnu/classpath/examples/swing/Demo$7.class
? examples/gnu/classpath/examples/swing/Demo$8.class
? examples/gnu/classpath/examples/swing/Demo$9.class
? examples/gnu/classpath/examples/swing/Demo$CheckCellRenderer.class
? examples/gnu/classpath/examples/swing/Demo$LabelCellRenderer.class
? examples/gnu/classpath/examples/swing/Demo$LaterMain.class
? examples/gnu/classpath/examples/swing/Demo$PopUpAction.class
? examples/gnu/classpath/examples/swing/Demo.class
? examples/gnu/classpath/examples/swing/GNULookAndFeel.class
Index: examples/gnu/classpath/examples/swing/Demo.java
===
RCS file: /cvsroot/classpath/classpath/examples/gnu/classpath/examples/swing/Demo.java,v
retrieving revision 1.16
diff -u -r1.16 Demo.java
--- examples/gnu/classpath/examples/swing/Demo.java 21 Jul 2005 14:13:36 - 1.16
+++ examples/gnu/classpath/examples/swing/Demo.java 25 Jul 2005 17:35:03 -
@@ -70,7 +70,7 @@
Look and Feel notice,
JOptionPane.INFORMATION_MESSAGE);
-UIManager.setLookAndFeel(new GNULookAndFeel());
+UIManager.setLookAndFeel(new MetalLookAndFeel());
}
}
catch (UnsupportedLookAndFeelException e)
Index: javax/swing/plaf/basic/BasicIconFactory.java
===
RCS file: /cvsroot/classpath/classpath/javax/swing/plaf/basic/BasicIconFactory.java,v
retrieving revision 1.10
diff -u -r1.10 BasicIconFactory.java
--- javax/swing/plaf/basic/BasicIconFactory.java 21 Jul 2005 12:07:06 - 1.10
+++ javax/swing/plaf/basic/BasicIconFactory.java 25 Jul 2005 17:35:04 -
@@ -113,7 +113,7 @@
// The icon is empty and needs no painting.
}
}
-
+
/**
* The icon used for RadioButtons in the BasicLookAndFeel. This is an empty
* icon with a size of 13x13 pixels.
@@ -163,7 +163,7 @@
/** The cached CheckBoxIcon instance. */
private static RadioButtonIcon radioButtonIcon;
-
+
public static Icon getMenuItemCheckIcon()
{
return new DummyIcon();
@@ -217,7 +217,7 @@
checkBoxIcon = new CheckBoxIcon();
return checkBoxIcon;
}
-
+
/**
* Returns an icon for RadioButtons in the BasicLookAndFeel. RadioButton
* icons in the Basic Lamp;F are empty and have a size of 13x13 pixels.
Index: javax/swing/plaf/basic/BasicLookAndFeel.java
===
RCS file: /cvsroot/classpath/classpath/javax/swing/plaf/basic/BasicLookAndFeel.java,v
retrieving revision 1.40
diff -u -r1.40 BasicLookAndFeel.java
--- javax/swing/plaf/basic/BasicLookAndFeel.java 22 Jul 2005 16:45:07 - 1.40
+++ javax/swing/plaf/basic/BasicLookAndFeel.java 25 Jul 2005 17:35:04 -
@@ -1013,8 +1013,6 @@
}),
Tree.background, new ColorUIResource(light),
Tree.changeSelectionWithFocus, Boolean.TRUE,
-// Tree.closedIcon, new IconUIResource(new ImageIcon(icons/TreeClosed.png)),
-// Tree.collapsedIcon, new IconUIResource(new ImageIcon(icons/TreeCollapsed.png)),
Tree.drawsFocusBorderAroundIcon, Boolean.FALSE,
Tree.editorBorder, new BorderUIResource.LineBorderUIResource(Color.lightGray),
Tree.focusInputMap, new UIDefaults.LazyInputMap(new Object[] {
Index: javax/swing/plaf/basic/BasicTreeUI.java
===
RCS file: /cvsroot/classpath/classpath/javax/swing/plaf/basic/BasicTreeUI.java,v
retrieving revision 1.47
diff -u -r1.47 BasicTreeUI.java
--- javax/swing/plaf/basic/BasicTreeUI.java 22 Jul 2005 08:36:52 - 1.47
+++ javax/swing/plaf/basic/BasicTreeUI.java 25 Jul 2005 17:35:04 -
@@ -1240,16 +1240,6 @@
if (hasControlIcons())
paintControlIcons(g, 0, 0, 0, 0, tree, mod, root);
-
- TreePath lead = tree.getLeadSelectionPath();
- if (lead != null tree.isPathSelected(lead))
- {
- Rectangle cell = getPathBounds(tree, lead);
- g.setColor(UIManager.getLookAndFeelDefaults().getColor(
- Tree.selectionBorderColor));
- g.drawRect(cell.x + rightChildIndent - 4, cell.y,
-
[cp-patches] Patch: FYI: Properties fixlet
I'm checking this in. This fixes a bug in Properties.load that was
found with FindBugs.
Tom
Index: ChangeLog
from Tom Tromey [EMAIL PROTECTED]
* java/util/Properties.java (load): Handle case where backslash
appears at EOF when reading the key. PR classpath/22994.
Index: java/util/Properties.java
===
RCS file: /cvsroot/classpath/classpath/java/util/Properties.java,v
retrieving revision 1.29
diff -u -r1.29 Properties.java
--- java/util/Properties.java 5 Jul 2005 10:28:03 - 1.29
+++ java/util/Properties.java 25 Jul 2005 17:27:22 -
@@ -209,8 +209,12 @@
{
if (pos == line.length())
{
-// The line continues on the next line.
+// The line continues on the next line. If there
+// is no next line, just treat it as a key with an
+// empty value.
line = reader.readLine();
+ if (line == null)
+ line = ;
pos = 0;
while (pos line.length()
Character.isWhitespace(c = line.charAt(pos)))
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
Re: [cp-patches]: Patch: JTree painting
Done it again!
heres the correct patch.
On Mon, 2005-07-25 at 13:39 -0400, Lillian Angel wrote:
A fix for the case where there are no icons.
2005-07-25 Lillian Angel [EMAIL PROTECTED]
* javax/swing/plaf/basic/BasicTreeUI.java
(paint): moved code to paintNode
(getCellBounds): width increased, slightly short before
(paintNode): changed to paint node depending if icons exist
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
Index: javax/swing/plaf/basic/BasicTreeUI.java
===
RCS file: /cvsroot/classpath/classpath/javax/swing/plaf/basic/BasicTreeUI.java,v
retrieving revision 1.47
diff -u -r1.47 BasicTreeUI.java
--- javax/swing/plaf/basic/BasicTreeUI.java 22 Jul 2005 08:36:52 - 1.47
+++ javax/swing/plaf/basic/BasicTreeUI.java 25 Jul 2005 17:35:04 -
@@ -1240,16 +1240,6 @@
if (hasControlIcons())
paintControlIcons(g, 0, 0, 0, 0, tree, mod, root);
-
- TreePath lead = tree.getLeadSelectionPath();
- if (lead != null tree.isPathSelected(lead))
- {
- Rectangle cell = getPathBounds(tree, lead);
- g.setColor(UIManager.getLookAndFeelDefaults().getColor(
- Tree.selectionBorderColor));
- g.drawRect(cell.x + rightChildIndent - 4, cell.y,
- cell.width + 4, cell.height);
- }
}
/**
@@ -1928,7 +1918,7 @@
boolean cntlClick = false;
Rectangle bounds = BasicTreeUI.this.getPathBounds(
BasicTreeUI.this.tree, path);
-// include icon
+
bounds.x -= rightChildIndent - 4;
bounds.width += rightChildIndent + 4;
@@ -2517,7 +2507,7 @@
Font f = tree.getFont();
FontMetrics fm = tree.getToolkit().getFontMetrics(tree.getFont());
- return new Rectangle(x, y, SwingUtilities.computeStringWidth(fm, s),
+ return new Rectangle(x, y, SwingUtilities.computeStringWidth(fm, s) + 4,
fm.getHeight());
}
return null;
@@ -2581,7 +2571,8 @@
TreePath curr = new TreePath(((DefaultMutableTreeNode) node).getPath());
boolean selected = tree.isPathSelected(curr);
boolean expanded = false;
-
+ boolean hasIcons = false;
+
if (tree.isVisible(curr))
{
DefaultTreeCellRenderer dtcr = (DefaultTreeCellRenderer) tree
@@ -2589,21 +2580,46 @@
if (!isLeaf)
expanded = tree.isExpanded(curr);
-
+
+ Icon icon = null;
+ if (!isLeaf expanded)
+icon = dtcr.getOpenIcon();
+ else if (!isLeaf !expanded)
+icon = dtcr.getClosedIcon();
+ else
+icon = dtcr.getLeafIcon();
+
+ if (icon.getIconHeight() -1 icon.getIconWidth() -1)
+hasIcons = true;
+
Component c = dtcr.getTreeCellRendererComponent(tree, node, selected,
expanded, isLeaf, 0, false);
- if (selected)
+ if (hasIcons)
{
-Rectangle cell = getPathBounds(tree, curr);
-g.setColor(dtcr.getBackgroundSelectionColor());
-g.fillRect(cell.x + rightChildIndent - 4, cell.y, cell.width + 4,
- cell.height);
+if (selected)
+{
+ Rectangle cell = getPathBounds(tree, curr);
+ g.setColor(dtcr.getBackgroundSelectionColor());
+ g.fillRect(cell.x + icon.getIconWidth()/2, cell.y, cell.width,
+ cell.height);
+
+ if (curr.equals(tree.getLeadSelectionPath()))
+ {
+ g.setColor(UIManager.getLookAndFeelDefaults().getColor(
+Tree.selectionBorderColor));
+ g.drawRect(cell.x + icon.getIconWidth()/2, cell.y,
+cell.width, cell.height);
+ }
+}
+
+g.translate(x, y);
+c.paint(g);
+g.translate(-x, -y);
}
-
- g.translate(x, y);
- c.paint(g);
- g.translate(-x, -y);
+ else
+rendererPane.paintComponent(g, c, c.getParent(),
+ getCellBounds(x, y, node));
}
}
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
Re: [cp-patches] RFC: RMI Class loader fix
Jeroen == Jeroen Frijters [EMAIL PROTECTED] writes: Jeroen * java/rmi/server/RMIClassLoader.java Jeroen (loadClass(String, String)): Use Class.forName() instead of Jeroen directly calling ClassLoader.loadClass(), to handle array Jeroen types correctly. Andrew, didn't we run into code that expected ClassLoader.loadClass to handle array signatures? Tom ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
Re: [cp-patches] FYI: Some fixes and optimizations for Container
Roman == Roman Kennke [EMAIL PROTECTED] writes: Roman (getAlignmentX): Despite common belief, this method does _not_ Roman call the LayoutManagers getAlignmentX in the JDK. So we also Roman don't. Information like this should be in comments in the code, not in the ChangeLog entry. Generally speaking the ChangeLog should describe what was done and the code should explain why. Tom ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches] rmic error handling
Attached is a patch to the cp-tools rmic that centralizes calls to
System.exit and makes handling of exceptions and error conditions more
uniform.
With the patch, rmic (built with libgcj-3.3.3-7 on FC2) passes the mauve
rmic tests.
-- Archit Shah
Index: src/gnu/classpath/tools/rmi/rmic/RMIC.java
===
RCS file: /cvsroot/classpath/cp-tools/src/gnu/classpath/tools/rmi/rmic/RMIC.java,v
retrieving revision 1.2
diff -u -r1.2 RMIC.java
--- src/gnu/classpath/tools/rmi/rmic/RMIC.java 5 Jul 2005 18:11:36 - 1.2
+++ src/gnu/classpath/tools/rmi/rmic/RMIC.java 25 Jul 2005 18:21:50 -
@@ -1,4 +1,4 @@
-/* ASMRMIC.java --
+/* RMIC.java --
Copyright (c) 1996, 1997, 1998, 1999, 2001, 2002, 2003, 2004, 2005
Free Software Foundation, Inc.
@@ -65,7 +65,7 @@
{
private String[] args;
private int next;
- private Exception exception;
+ private List errors = new ArrayList();
private boolean keep = false;
private boolean need11Stubs = true;
private boolean need12Stubs = true;
@@ -97,40 +97,66 @@
public static void main(String[] args)
{
+if (rmic(args))
+ System.exit(0);
+else
+ System.exit(1);
+ }
+
+ /**
+ * @return true if compilation was successful
+ */
+ public static boolean rmic(String[] args)
+ {
RMIC r = new RMIC(args);
-if (r.run() == false)
- {
- Exception e = r.getException();
- if (e != null)
- e.printStackTrace();
- else
- System.exit(1);
- }
+return r.run();
}
- public boolean run()
+ /**
+ * @return true if run was successful
+ */
+ private boolean run()
{
-parseOptions();
+boolean done = parseOptions();
+if (done)
+ return errorCount == 0;
+
if (next = args.length)
- error(no class names found);
+ {
+usage();
+return false;
+ }
+
for (int i = next; i args.length; i++)
{
try
{
- if (verbose)
+if (verbose)
System.out.println([Processing class + args[i] + .class]);
processClass(args[i].replace(File.separatorChar, '.'));
}
- catch (Exception e)
- {
- exception = e;
- return (false);
- }
+catch (IOException e)
+ {
+errors.add(e);
+ }
+catch (RMICException e)
+ {
+errors.add(e);
+ }
}
-return (true);
+if (errors.size() 0)
+ {
+for (Iterator it = errors.iterator(); it.hasNext(); )
+ {
+Exception ex = (Exception) it.next();
+logError(ex);
+ }
+ }
+
+return errorCount == 0;
}
- private boolean processClass(String cls) throws Exception
+ private void processClass(String cls) throws IOException, RMICException
{
// reset class specific vars
clazz = null;
@@ -142,18 +168,14 @@
skelname = null;
mRemoteInterfaces = new ArrayList();
-errorCount = 0;
-
analyzeClass(cls);
-if (errorCount 0)
- System.exit(1);
generateStub();
if (need11Stubs)
generateSkel();
-return (true);
}
- private void analyzeClass(String cname) throws Exception
+ private void analyzeClass(String cname)
+throws RMICException
{
if (verbose)
System.out.println([analyze class + cname + ]);
@@ -168,35 +190,35 @@
findRemoteMethods();
}
+ /**
+ * @deprecated
+ */
public Exception getException()
{
-return (exception);
+return errors.size() == 0 ? null : (Exception) errors.get(0);
}
private void findClass()
+throws RMICException
{
+ClassLoader cl = (loader == null
+ ? ClassLoader.getSystemClassLoader()
+ : loader);
try
{
-ClassLoader cl = (loader == null
- ? ClassLoader.getSystemClassLoader()
- : loader);
clazz = Class.forName(fullclassname, false, cl);
}
catch (ClassNotFoundException cnfe)
{
-System.err.println(fullclassname + not found in + classpath);
-throw new RuntimeException(cnfe);
+throw new RMICException
+ (Class + fullclassname + not found in classpath, cnfe);
}
if (! Remote.class.isAssignableFrom(clazz))
{
-logError(Class + clazz.getName() + is not a remote object.
- + It does not implement an interface that is a
- + java.rmi.Remote-interface.);
-throw new RuntimeException
- (Class + clazz.getName() + is not a remote object.
- + It does not implement an interface that is a
- + java.rmi.Remote-interface.);
+throw new RMICException
+ (Class + clazz.getName()
+ + does not implement a remote interface.);
}
}
@@ -412,7 +434,8 @@
}
}
- private void generateStub() throws IOException
+ private void generateStub()
+throws IOException
{
RE: [cp-patches] FYI: ClassLoader caching Part 2
On Mon, 2005-07-25 at 18:20 +0200, Mark Wielaard wrote:
On Mon, 2005-07-25 at 17:55 +0200, Jeroen Frijters wrote:
This won't work if you make the VM_USE_CACHE field static final. Then
the constant will be compiled into ClassLoader making it impossible to
override for the runtime vm-classes later. It has to be a
static method for that to work.
Like Archie says, your comment doesn't make sense ;-) If a VM decides
not to use the cache in ClassLoader, it replaces VMClassLoader and sets
the constant to true.
Yes, but unfortunately that doesn't change the byte-code of the
ClassLoader class. static final primitive (and String) field values are
actually inlined into other classes. So you can replace VMClassLoader
all you want with a version the has VM_USE_CACHE set to another value,
that won't actually change the ClassLoader code paths unless ClassLoader
is also recompiled against this new value. Really, try it out. It is an
annoying byte code optimization. So for things that can be different at
runtime like the VMClasses you unfortunately cannot use static final
fields of primitive types.
I haven't looked at jdwp yet, but hopefully jdwp will provide its own
VM interface instead of messing around with the internals of other
classes.
Yeah that would be the clean way to do it indeed.
Well if it were decided to have some minor changes to allow VM re-use
this is pretty much how I would alter ClassLoader.
We're not ready to use it anyways but I'll put it out there so people
know what would be required.
Aaron
Index: java/lang/ClassLoader.java
===
RCS file: /cvsroot/classpath/classpath/java/lang/ClassLoader.java,v
retrieving revision 1.54
diff -u -p -r1.54 ClassLoader.java
--- java/lang/ClassLoader.java 25 Jul 2005 14:28:42 - 1.54
+++ java/lang/ClassLoader.java 25 Jul 2005 18:37:38 -
@@ -40,6 +40,7 @@ package java.lang;
import gnu.classpath.SystemProperties;
import gnu.classpath.VMStackWalker;
+import gnu.classpath.jdwp.Jdwp;
import gnu.java.util.DoubleEnumeration;
import gnu.java.util.EmptyEnumeration;
@@ -152,6 +153,14 @@ public abstract class ClassLoader
*/
private final boolean initialized;
+ /**
+ * This list is kept around for JDWP. We need a way to know every class
+ * that this ClassLoader has been asked to load, thus they will be added
+ * here. There's no need to do anything else, JDWP is able to sniff out
+ * private fields so we'll just sniff this field as well!
+ */
+ final ArrayList loadRequests = new ArrayList();
+
static class StaticData
{
/**
@@ -333,23 +342,25 @@ public abstract class ClassLoader
{
if (parent == null)
{
- c = VMClassLoader.loadClass(name, resolve);
- if (c != null)
- return c;
- }
- else
- {
- return parent.loadClass(name, resolve);
- }
- }
- catch (ClassNotFoundException e)
- {
- }
- // Still not found, we have to do it ourself.
- c = findClass(name);
+ c = VMClassLoader.loadClass(name, resolve);
+ }
+else
+ {
+ c = parent.loadClass(name, resolve);
+ }
+ }
+ catch (ClassNotFoundException e)
+ {
+ }
+ // Still not found, we have to do it ourself.
+ if (c == null)
+c = findClass(name);
}
+}
if (resolve)
resolveClass(c);
+if (Jdwp.isDebugging)
+ loadRequests.add(c);
return c;
}
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches] Patch: FYI: fix PR classpath/22986
I'm checking this in. FindBugs pointed out that we were using equals() to compare arrays. In this case, this is incorrect, and we want Arrays.equals(). Tom Index: ChangeLog from Tom Tromey [EMAIL PROTECTED] * java/text/CollationKey.java (equals): Use Arrays.equals. PR classpath/22986. Index: java/text/CollationKey.java === RCS file: /cvsroot/classpath/classpath/java/text/CollationKey.java,v retrieving revision 1.15 diff -u -r1.15 CollationKey.java --- java/text/CollationKey.java 23 Jul 2005 20:25:15 - 1.15 +++ java/text/CollationKey.java 25 Jul 2005 18:26:23 - @@ -38,6 +38,8 @@ package java.text; +import java.util.Arrays; + /* Written using Java Class Libraries, 2nd edition, plus online * API docs for JDK 1.2 from http://www.javasoft.com. * Status: Believed complete and correct. @@ -154,7 +156,7 @@ if (!ck.getSourceString ().equals (getSourceString ())) return false; -if (!ck.toByteArray ().equals (toByteArray ())) +if (! Arrays.equals (ck.toByteArray (), toByteArray ())) return false; return true; ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
Re: [cp-patches] FYI: Some fixes and optimizations for Container
Hi, -/* AWTUtilities.java -- Common utility methods for AWT and Swing. - Copyright (C) 2005 Free Software Foundation, Inc. +/* Copyright (C) 2004 Free Software Foundation Don't do that. sorry, I simply copied over the file (from our repo). I will take more care in the future. I'll fix this tomorrow. -Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA -02110-1301 USA. +Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA +02111-1307 USA. Please keep the new address in place everywhere. See above. + public static Rectangle calculateInsetArea(Rectangle base, Insets insets, + Rectangle ret) This looks strange, did you forget a return here? There are a couple of similar things like this in the patch. Now I was confused. I thought you meant a missing return statement which would have been rejected by the compiler. I see you mean the bad formatting. I will go over this file and fix everything tomorrow. Sorry for this crap. /Roman ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
RE: [cp-patches] FYI: ClassLoader caching Part 2
Hi Jeroen, On Mon, 2005-07-25 at 18:17 +0200, Jeroen Frijters wrote: Is this an important scenario? I thought you long haired types didn't like binary distributions ;-) He he. Sure we don't. But if you follow the harmony effort of the cleanly shaved apache types you will have noticed that they might actually be paranoid enough to not look at or use something that comes with source code :) But I do care about binary installations! I have here two runtimes (jamvm and kissme) who both use the same GNU Classpath installation, but both overlay the VM classes differently. I would actually like that to be true for some of the others I have installed here too. Cheers, Mark signature.asc Description: This is a digitally signed message part ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
Re: [cp-patches] RFC: New JCL helper functions for RawData handling
Hi, On Mon, 2005-07-25 at 19:21 +0200, Guilhem Lavaux wrote: Ok. Then I will make a set of patches to first use the new common JCL interface. And then I will rename RawData* to Pointer*. Is it ok ? Yes please. And for extra bonus points you could add a little thing to the Hacker Guide about using these new JCL functions! Cheers, Mark signature.asc Description: This is a digitally signed message part ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches] FYI: BasicTableUI TAB, SHIFT-TAB, and SHIFT-ENTER
This patch implements TAB, SHIFT-TAB, and SHIFT-ENTER key actions and
merges them with the code for the ENTER action because of the
similarities.
Note that you won't be able to test the TAB or SHIFT-TAB unless you also
apply the Hack.diff patch because JTable currently uses TABs to change
focus. This is incorrect and will be fixed separately, but applying the
small Hack.diff will allow you to see the implemented key actions.
Patch attached.
2005-07-25 Anthony Balkissoon [EMAIL PROTECTED]
* javax/swing/plaf/basic/BasicTableUI.java:
(KeyHandler.advanceSingleSelection): New method. Helper method for
ENTER, SHIFT-ENTER, TAB, and SHIFT-TAB key events.
(KeyHandler.advanceMultipleSelection): Likewise, used when the table
has multiple selections at the time that ENTER or TAB was pressed.
(KeyHandler.keyPressed): Implemented TAB, SHIFT-TAB, and SHIFT-ENTER
and merged these with existing code for ENTER event, because of the
similarites.
Index: javax/swing/plaf/basic/BasicTableUI.java
===
RCS file: /cvsroot/classpath/classpath/javax/swing/plaf/basic/BasicTableUI.java,v
retrieving revision 1.15
diff -u -r1.15 BasicTableUI.java
--- javax/swing/plaf/basic/BasicTableUI.java 25 Jul 2005 14:48:53 - 1.15
+++ javax/swing/plaf/basic/BasicTableUI.java 25 Jul 2005 19:50:44 -
@@ -97,6 +97,171 @@
class KeyHandler implements KeyListener
{
+
+/**
+ * A helper method for the keyPressed event. Used because the actions
+ * for TAB, SHIFT-TAB, ENTER, and SHIFT-ENTER are very similar.
+ *
+ * Selects the next (previous if SHIFT pressed) column for TAB, or row for
+ * ENTER from within the currently selected cells.
+ *
+ * @param firstModel the ListSelectionModel for columns (TAB) or
+ * rows (ENTER)
+ * @param firstMin the first selected index in firstModel
+ * @param firstMax the last selected index in firstModel
+ * @param secondModel the ListSelectionModel for rows (TAB) or
+ * columns (ENTER)
+ * @param secondMin the first selected index in secondModel
+ * @param secondMax the last selected index in secondModel
+ * @param reverse true if shift was held for the event
+ * @param eventIsTab true if TAB was pressed, false if ENTER pressed
+ */
+void advanceMultipleSelection (ListSelectionModel firstModel, int firstMin,
+ int firstMax, ListSelectionModel secondModel,
+ int secondMin, int secondMax, boolean reverse,
+ boolean eventIsTab)
+{
+ // If eventIsTab, all the firsts correspond to columns, otherwise, to rows
+ // seconds correspond to the opposite
+ int firstLead = firstModel.getLeadSelectionIndex();
+ int secondLead = secondModel.getLeadSelectionIndex();
+ int numFirsts = eventIsTab ?
+table.getModel().getColumnCount() : table.getModel().getRowCount();
+ int numSeconds = eventIsTab ?
+table.getModel().getRowCount() : table.getModel().getColumnCount();
+
+ // check if we have to wrap the firsts around, going to the other side
+ if ((firstLead == firstMax !reverse) ||
+ (reverse firstLead == firstMin))
+{
+ firstModel.addSelectionInterval(reverse ? firstMax : firstMin,
+ reverse ? firstMax : firstMin);
+
+ // check if we have to wrap the seconds
+ if ((secondLead == secondMax !reverse) ||
+ (reverse secondLead == secondMin))
+secondModel.addSelectionInterval(reverse ? secondMax : secondMin,
+ reverse ? secondMax : secondMin);
+
+ // if we're not wrapping the seconds, we have to find out where we
+ // are within the secondModel and advance to the next cell (or
+ // go back to the previous cell if reverse == true)
+ else
+{
+ int[] secondsSelected;
+ if (eventIsTab table.getRowSelectionAllowed() ||
+ !eventIsTab table.getColumnSelectionAllowed())
+secondsSelected = eventIsTab ?
+ table.getSelectedRows() : table.getSelectedColumns();
+ else
+{
+ // if row selection is not allowed, then the entire column gets
+ // selected when you click on it, so consider ALL rows selected
+ secondsSelected = new int[numSeconds];
+ for (int i = 0; i numSeconds; i++)
+ secondsSelected[i] = i;
+}
+
+ // and now find the next index within the model
+ int secondIndex = reverse ? secondsSelected.length - 1 : 0;
+ if (!reverse)
+while (secondsSelected[secondIndex] = secondLead)
+ secondIndex++;
+ else
+
[cp-patches] [patch] examples README
Hi all, a little patchlet for the README in the examples directory. Ok? Andreas 2005-07-25 Andreas Tobler [EMAIL PROTECTED] * examples/README: Add GNULookAndFeel.java to the compile command. Index: README === RCS file: /cvsroot/classpath/classpath/examples/README,v retrieving revision 1.3 diff -u -r1.3 README --- README 2 Jul 2005 20:32:08 - 1.3 +++ README 25 Jul 2005 20:12:26 - @@ -7,7 +7,8 @@ The examples can be compiled and run with gcj as follows: gcj -o swingdemo --main=gnu.classpath.examples.swing.Demo \ - gnu/classpath/examples/swing/Demo.java + gnu/classpath/examples/swing/Demo.java \ + gnu/classpath/examples/swing/GNULookAndFeel.java ./swingdemo Or with a traditional byte code interpreter like: ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches] [patch] fix memory image source handling in GTK peers
Hi,
This patch does three things:
- implements GtkComponentPeer.repaint timed repaints properly
- implements GtkComponentPeer.updateCursorImmediately
- fixes handling of memory image sources in GtkImageConsumer
I committed this to mainline. This patch gets the Cortado applet closer
to working out-of-the-box on GNU Classpath.
Tom
2005-07-25 Thomas Fitzsimmons [EMAIL PROTECTED]
* gnu/java/awt/peer/gtk/GtkComponentPeer.java (repaint): Implement
timed repaint.
(updateCursorImmediately): Implement.
(RepaintTimerTask): New class.
* gnu/java/awt/peer/gtk/GtkImageConsumer.java (imageComplete):
Don't remove consumer if source is a MemoryImageSource.
Index: gnu/java/awt/peer/gtk/GtkComponentPeer.java
===
RCS file: /cvsroot/classpath/classpath/gnu/java/awt/peer/gtk/GtkComponentPeer.java,v
retrieving revision 1.88
diff -u -r1.88 GtkComponentPeer.java
--- gnu/java/awt/peer/gtk/GtkComponentPeer.java 14 Jul 2005 22:07:02 - 1.88
+++ gnu/java/awt/peer/gtk/GtkComponentPeer.java 26 Jul 2005 01:50:52 -
@@ -70,6 +70,8 @@
import java.awt.image.ImageProducer;
import java.awt.image.VolatileImage;
import java.awt.peer.ComponentPeer;
+import java.util.Timer;
+import java.util.TimerTask;
public class GtkComponentPeer extends GtkGenericPeer
implements ComponentPeer
@@ -372,8 +374,28 @@
if (x == 0 y == 0 width == 0 height == 0)
return;
-q().postEvent (new PaintEvent (awtComponent, PaintEvent.UPDATE,
- new Rectangle (x, y, width, height)));
+Timer t = new Timer();
+
+t.schedule(new RepaintTimerTask(x, y, width, height), tm);
+ }
+
+ private class RepaintTimerTask extends TimerTask
+ {
+private int x, y, width, height;
+
+RepaintTimerTask(int x, int y, int width, int height)
+{
+ this.x = x;
+ this.y = y;
+ this.width = width;
+ this.height = height;
+}
+
+public void run()
+{
+ q().postEvent (new PaintEvent (awtComponent, PaintEvent.UPDATE,
+ new Rectangle (x, y, width, height)));
+}
}
public void requestFocus ()
@@ -586,7 +608,8 @@
public void updateCursorImmediately ()
{
-
+if (awtComponent.getCursor() != null)
+ setCursor(awtComponent.getCursor());
}
public boolean handlesWheelScrolling ()
Index: gnu/java/awt/peer/gtk/GtkImageConsumer.java
===
RCS file: /cvsroot/classpath/classpath/gnu/java/awt/peer/gtk/GtkImageConsumer.java,v
retrieving revision 1.3
diff -u -r1.3 GtkImageConsumer.java
--- gnu/java/awt/peer/gtk/GtkImageConsumer.java 2 Jul 2005 20:32:12 - 1.3
+++ gnu/java/awt/peer/gtk/GtkImageConsumer.java 26 Jul 2005 01:50:52 -
@@ -45,6 +45,7 @@
import java.awt.image.ImageConsumer;
import java.awt.image.ImageObserver;
import java.awt.image.ImageProducer;
+import java.awt.image.MemoryImageSource;
import java.util.Hashtable;
import java.util.Vector;
@@ -70,7 +71,10 @@
public synchronized void imageComplete (int status)
{
-source.removeConsumer(this);
+// we need to reuse the pixel cache for memory image sources since
+// a memory image's backing array can be updated live.
+if (!(source instanceof MemoryImageSource))
+ source.removeConsumer(this);
target.setImage(width, height, pixelCache, properties);
}
___
Classpath-patches mailing list
Classpath-patches@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath-patches
RE: ClasspathToolkit redesign.
Sven de Marothy wrote: Ok, I've been looking into the issue of our AWT peer interface now that I've been running my peers on Classpath in earnest. The goal here is to try to bring our peer interface as close to Sun's as possible. And simplify things for potential peer-authors. [...] Summary: So, my opinion: Keep most of the font stuff and the robot stuff. Get rid of the rest, and we'll have an interface which is both closer to the JDKs, and simpler. Comments? I very much agree. My peers are only toys at the moment, but I ran into many of the same issues. Especially the fact that the GTK event model is integrated in EventQueue is very bad. Method: createEmbeddedWindow () Comment: What's an EmbeddedWindow and why do we need this as part of the peer interface? I'm not sure, but I think this is used for Applets that have to live inside a non-Java window. Regards, Jeroen ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
RE: Implementation details of VMStackWalker
Ingo Prötel wrote: I just implemented VMStackWalker for our VM and have some questions. The reference implementation of 'getCallingClass()' and 'getCallingClassLoader()' just look at the third entry in the class context. Would it not be better to return the first class that is not assignable to the class in context[0] ? That way we could cope with classes that first call some private or protected classes an then ends up calling the stack walker. I'm *very* strongly opposed to this. It makes auditing the code for correctness of VMStackWalker *much* harder and it adds no value. The next thing I would like to have is a method to get the calling method name. This would be good for logging. Yeah, we should probably add a method to get the calling method and maybe another one to get the entire stack trace (as a java.lang.reflect.Method[]). For the security part: Is it enough to check if the class loader of context[0] is the boot classloader? You don't need to enforce that, the documentation is just intended to point out that this is a privileged call that is not available to untrusted code. The security model will be used to enforce that untrusted code cannot call code in the gnu.classpath.* package. Regards, Jeroen ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
RE: Implementation details of VMStackWalker
Andrew Haley wrote:
In gcj, we have a method called GetCallingClass(Class c). It searches
firstly for a method declared in class c, then for a method not
declared in c. We have found, after a certain amount of trouble, that
this is the right way to do things; it means that an arbitrary number
of stack frames can be between the direct caller of GetCallingClass
and the user code, and it also means that you don't have to check for
assignability, but for an exact match.
This is a dumb idea...
For example, you may have
class Foo
{
bar (Object O)
{
return baz(O);
}
baz (Object O)
{
... GetCallingClass (Foo) ...
}
}
class Frob extends Foo
{
baz (Object o)
{
super.baz(o);
}
}
... and you will get class of the caller of bar, not Foo.class.
No, you'll get Frob.
Regards,
Jeroen
___
Classpath mailing list
Classpath@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath
RE: Implementation details of VMStackWalker
Andrew Haley wrote:
However, as for overhead -- I don't believe it. I doubt that not
having this parameter saves anything much on any VM.
That's just your lack of imagination ;-) I was concerned with two
aspects wrt performance:
1) Class literals are inefficient -- this is no longer an issue to me,
since I now use ecj (in 1.5 mode) to compile GNU Classpath
2) I have to do multiple (partial) stack walks to retrieve multiple
frames
Remember, IKVM is built on top of another VM, so I have to use another
stack walking abstraction to walk the stack. Having to fetch multiple
frames is obviously more expensive than having to fetch a single frame.
*BUT* I have much stronger objections against this model:
1) Like Archie points out, there is no need for this additional
functionality (at least not in the common case)
2) It is very easy to introduce subtle bugs (like the one with the
override I showed)
3) It is tempting to do (either on purpose or accidentally):
class Foo
{
somemethod()
{
// we know we're always being called by Bar and we need its
caller
GetCaller(Bar.class);
}
}
4) It makes advanced optimizations by the JIT harder (the current
interface is very easy for the JIT to special case)
Regards,
Jeroen
___
Classpath mailing list
Classpath@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath
RE: Implementation details of VMStackWalker
Jeroen Frijters writes:
Andrew Haley wrote:
In gcj, we have a method called GetCallingClass(Class c). It searches
firstly for a method declared in class c, then for a method not
declared in c. We have found, after a certain amount of trouble, that
this is the right way to do things; it means that an arbitrary number
of stack frames can be between the direct caller of GetCallingClass
and the user code, and it also means that you don't have to check for
assignability, but for an exact match.
This is a dumb idea...
For example, you may have
class Foo
{
bar (Object O)
{
return baz(O);
}
baz (Object O)
{
... GetCallingClass (Foo) ...
}
}
class Frob extends Foo
{
baz (Object o)
{
super.baz(o);
}
}
... and you will get class of the caller of bar, not Foo.class.
No, you'll get Frob.
Which is the Right Answer: the caller of baz *is* Frob.
Andrew.
___
Classpath mailing list
Classpath@gnu.org
http://lists.gnu.org/mailman/listinfo/classpath
GJava_gnu_java_awt_peer_gtk_GdkGraphics_setFGColor possible crash
Hi! We have some problems with the spec jvm applet and i think i've found a problem. Java_gnu_java_awt_peer_gtk_GdkGraphics_setFGColor uses NSA_GET_PTR to get the graphics pointer, which calls cp_gtk_get_state. This function may return NULL in exception case, which lets this code crash: g = (struct graphics *) NSA_GET_PTR (env, obj); gdk_color_alloc (g-cm, color); May this be a problem? In fact, this crashes currently for me, but there is no exception, it just returns NULL. TWISTI ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
RE: Implementation details of VMStackWalker
Jeroen Frijters writes: Andrew Haley wrote: However, as for overhead -- I don't believe it. I doubt that not having this parameter saves anything much on any VM. That's just your lack of imagination ;-) I was concerned with two aspects wrt performance: 1) Class literals are inefficient -- this is no longer an issue to me, since I now use ecj (in 1.5 mode) to compile GNU Classpath 2) I have to do multiple (partial) stack walks to retrieve multiple frames Remember, IKVM is built on top of another VM, so I have to use another stack walking abstraction to walk the stack. Having to fetch multiple frames is obviously more expensive than having to fetch a single frame. As I said: IMO trying to unify low-level stack walker code is unnecessary and leads to too many compromises; it's a merge too far. Andrew. ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
RE: Implementation details of VMStackWalker
Jeroen Frijters writes: Andrew Haley wrote: Which is the Right Answer: the caller of baz *is* Frob. It *may* be the right answer, but how can you be certain that the coder of class Foo intended this? Ah, well that is a totally different matter. IMNSHO, code like this is completely unauditable (remember, this isn't just a correctness issue, access to class loaders is a potential security issue). Of course, yes. But it's security issues that I'm concerned about here: what we want to know is the first caller of Foo.method() that is not Foo. Andrew. ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
RE: Implementation details of VMStackWalker
Andrew Haley wrote: Which is the Right Answer: the caller of baz *is* Frob. It *may* be the right answer, but how can you be certain that the coder of class Foo intended this? IMNSHO, code like this is completely unauditable (remember, this isn't just a correctness issue, access to class loaders is a potential security issue). Regards, Jeroen ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
RE: Implementation details of VMStackWalker
Ingo Prötel wrote: I really like the idea of having a stack walker so I would like to use it wherever possible. For example in java.util.logging.Logger. If a class calls 'logger.warning(some warning)' this call will be forwarded through a couple of calls until it will call 'Logger.log(Level level, String message, Throwable thrown)' there the stack will be inspected to get the calling class and calling method. I think that a (potentially) more optimized version of the stack walker API can be made for this. I have no problem with that since accuracy in the logging isn't that critical (getting it wrong will be annoying but won't create a security hole). Does this mean the VM must check for every call if it is a call into the gnu.classpath.* package and if so check the classloader of the calling class? No, this is enforced by the system class loader. Untrusted code cannot even load classes from the gnu.classpath package (this isn't yet fully implemented yet, but most of the infrastructure is in place). Only if you replace the GNU Classpath system class loader do you need to worry about this. Regards, Jeroen ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
RE: Implementation details of VMStackWalker
Andrew Haley wrote: Of course, yes. But it's security issues that I'm concerned about here: what we want to know is the first caller of Foo.method() that is not Foo. Not necessarily. Typically what's important is the supplier of the arguments to the method. In the subclassing scenario, the subclass may be the one providing the arguments (i.e. passing different values then it was passed), but it may also be passing along the original values. If the subclasser is trusted but the original caller isn't, you have a problem. Now granted, this is a coding error, but I think it is facilitated by this too flexible model of walking the stack. BTW, I'm not ruling out the need for this more flexible way of getting the caller, I just want to make sure that this isn't the default and is used only very cautiously. Regards, Jeroen ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
RE: Implementation details of VMStackWalker
Am Montag, den 25.07.2005, 10:16 +0200 schrieb Jeroen Frijters: Ingo Prötel wrote: I just implemented VMStackWalker for our VM and have some questions. The reference implementation of 'getCallingClass()' and 'getCallingClassLoader()' just look at the third entry in the class context. Would it not be better to return the first class that is not assignable to the class in context[0] ? That way we could cope with classes that first call some private or protected classes an then ends up calling the stack walker. I'm *very* strongly opposed to this. It makes auditing the code for correctness of VMStackWalker *much* harder and it adds no value. I really like the idea of having a stack walker so I would like to use it wherever possible. For example in java.util.logging.Logger. If a class calls 'logger.warning(some warning)' this call will be forwarded through a couple of calls until it will call 'Logger.log(Level level, String message, Throwable thrown)' there the stack will be inspected to get the calling class and calling method. The current implementation of VMStackWalker would require the generation of the full class context to get the class that called 'Logger.warning(String)' since 'VMStackWalker.getCallingClass()' would always return 'Logger'. But creating the whole class context is expensive. And yes our AOT compiler can simply inline the methods but not the simple interpreter that we use to run dynamically loaded classes. No I'm not sure if checking for assignability really is useful. But if somebody wants to create their own Logger subclass it would just work as I would expect when a call is forwarded to the Logger. But I admit I have not yet checked what happens on a Sun VM in such a case. The next thing I would like to have is a method to get the calling method name. This would be good for logging. Yeah, we should probably add a method to get the calling method and maybe another one to get the entire stack trace (as a java.lang.reflect.Method[]). For the security part: Is it enough to check if the class loader of context[0] is the boot classloader? You don't need to enforce that, the documentation is just intended to point out that this is a privileged call that is not available to untrusted code. The security model will be used to enforce that untrusted code cannot call code in the gnu.classpath.* package. Does this mean the VM must check for every call if it is a call into the gnu.classpath.* package and if so check the classloader of the calling class? Ingo -- Ingo Prötel [EMAIL PROTECTED] aicas GmbHhttp://www.aicas.com Haid-und-Neu-Str. 18phone +49 721 663 968-32 76131 Karlsruhe fax +49 721 663 968-93 Germany ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
Using GNU Classpath while changing his hierarchy
Oups..., I completely forget to answer to the mailing list *too*... So, here's what I answered to Tom... -- Forwarded message -- From: Jean-Philippe Deblonde [EMAIL PROTECTED] Date: 21 juil. 2005 19:21 Subject: Re: Using GNU Classpath while changing his hierarchy To: [EMAIL PROTECTED] In fact, it's a lot more than a simple refactoring, because I need, by example, to rewrite the String.java to use an array of bytes (coded with UTF-8) instead of an array of char, to save some space for the smartcards... It's only one of the substitutions I need to do, and I need to do them automatically, since the developpers getting hands on the whole thing will just have to download the new version of the Classpath archive, and run a script to merge it with the rest of the project 21 Jul 2005 10:22:30 -0600, Tom Tromey [EMAIL PROTECTED]: == Jean-Philippe Deblonde [EMAIL PROTECTED] writes: The GNU Classpath project evolves while I'm doing this. And some substitution rules available for today won't be available in future releases of the Classpath project. Some new classes/substitutions will be introduced with each new version... So, I post this message to ask if I'm not missing something in the process, if there is a existing tool or some guidelines to help me keep track of the diff in a streamlined way. Do you have any comments or suggestions ? Use Eclipse to do the refactoring for you? Or, do the refactoring *after* the build using a tool like ASM? Tom ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
Re: ClasspathToolkit redesign.
Hi Sven, Ok, I've been looking into the issue of our AWT peer interface now that I've been running my peers on Classpath in earnest. The goal here is to try to bring our peer interface as close to Sun's as possible. And simplify things for potential peer-authors. ... Summary: So, my opinion: Keep most of the font stuff and the robot stuff. Get rid of the rest, and we'll have an interface which is both closer to the JDKs, and simpler. Comments? A cleanup is very welcome. Go ahead. /Roman ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
Re: Using GNU Classpath while changing his hierarchy
Jean-Philippe Deblonde wrote: In fact, it's a lot more than a simple refactoring, because I need, by example, to rewrite the String.java to use an array of bytes (coded with UTF-8) instead of an array of char, to save some space for the smartcards.. Maybe I do not understand properly, but why on the world are you doing so strange? GNU Classpath Reader and Writer supports not just UTF-8, but also many other encodings, see release notes for details. You only write in UTF-16 if you want to. Are you using smart card as the main memory for the application? ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
Re: Using GNU Classpath while changing his hierarchy
The idea was to save memory space execution time, so the String.java use *internally* UTF format. 2005/7/25, Meskauskas Audrius [EMAIL PROTECTED]: Jean-Philippe Deblonde wrote: In fact, it's a lot more than a simple refactoring, because I need, by example, to rewrite the String.java to use an array of bytes (coded with UTF-8) instead of an array of char, to save some space for the smartcards.. Maybe I do not understand properly, but why on the world are you doing so strange? GNU Classpath Reader and Writer supports not just UTF-8, but also many other encodings, see release notes for details. You only write in UTF-16 if you want to. Are you using smart card as the main memory for the application? ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
Re: Using GNU Classpath while changing his hierarchy
Hi, On Mon, 2005-07-25 at 14:46 +0200, Jean-Philippe Deblonde wrote: In fact, it's a lot more than a simple refactoring, because I need, by example, to rewrite the String.java to use an array of bytes (coded with UTF-8) instead of an array of char, to save some space for the smartcards... This is what wonka does for internal string representations btw. It's only one of the substitutions I need to do, and I need to do them automatically, since the developpers getting hands on the whole thing will just have to download the new version of the Classpath archive, and run a script to merge it with the rest of the project It is a pain to maintain a forked version. And the best way to make sure you won't have to is to provide patches to make these things possible for the runtime to tweak. I am sure others will love to help with that if you come up with some nice optimization points for the core classes. Just supply some patches and a reference implementation and we can see if we can incorporate it in the default build. We make regular releases (once every 2 months at least). Thanks, Mark signature.asc Description: This is a digitally signed message part ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
Re: ClasspathToolkit redesign.
Hi, I'll let Thomas reply with more details since he knows a lot more about this. But here are some quick references that might help. On Sun, 2005-07-24 at 00:19 +0200, Sven de Marothy wrote: Method: getClasspathFontPeer(), getClasspathTextLayoutPeer(), getFont() Comment: We need these. No way around that. But they could perhaps be put in a seperate ClasspathFontInterface or something. Method: createFont() Comment: Um, we don't even implement this ourselves. Sorry Sacha, but there's no point. Right now anyway. Note that Sascha did make a design and a start of an implementation: http://www.brawer.ch/software/fonts/ Method: createEmbeddedWindow () Comment: What's an EmbeddedWindow and why do we need this as part of the peer interface? It is an extension for things like gcjwebplugin. We also briefly discussed this with the OpenOffice hackers who would also like some way to create embedded windows (inside documents). See http://savannah.nongnu.org/cgi-bin/viewcvs/gcjwebplugin/gcjwebplugin/src/gnu/gcjwebplugin/PluginAppletWindow.java?rev=HEADcontent-type=text/vnd.viewcvs-markup Method: registerImageIOSpis() Comment: Unnecessary. The peers can register their SPIs (if they have any) themselves without this. Agreed. Or use the META-INF/services hack. Cheers, Mark signature.asc Description: This is a digitally signed message part ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
Re: ClasspathToolkit redesign.
Hi, On Sun, 2005-07-24 at 00:19 +0200, Sven de Marothy wrote: [...] Method: getClasspathFontPeer(), getClasspathTextLayoutPeer(), getFont() Comment: We need these. No way around that. But they could perhaps be put in a seperate ClasspathFontInterface or something. OK. Method: createFont() Comment: Um, we don't even implement this ourselves. Sorry Sacha, but there's no point. Right now anyway. Method: createRobot() Comment: This we need. But it's a rather singular item. Replace with a property setting? Certain peers may not (or cannot) have Robots anyway. How could this be replaced by a property setting? Method: createEmbeddedWindow () Comment: What's an EmbeddedWindow and why do we need this as part of the peer interface? An EmbeddedWindowPeer is a wrapper around the native toolkit's implementation of the XEMBED protocol. gnu.java.awt.EmbeddedWindow should use the current toolkit's XEMBED implementation, e.g. for GTK, GtkPlug. We'll need some way to create the peer dependent on the toolkit. Whatever we do for Robot can be done here too. Method: registerImageIOSpis() Comment: Unnecessary. The peers can register their SPIs (if they have any) themselves without this. Method: nativeQueueEmpty(), wakeNativeQueue(), iterateNativeQueue() Comment: These are specific to how the GTK peers handles threads and events, and should certainly not be part of the interface. The versions of EventQueue using these methods should be reverted too. The GTK peers should extend and overload EventQueue for their custom approach instead. Summary: So, my opinion: Keep most of the font stuff and the robot stuff. Get rid of the rest, and we'll have an interface which is both closer to the JDKs, and simpler. Comments? Yeah, this sounds good. Tom ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
Re: JNI assertion failure
I checked in the aforementioned assertion patch. 2005-07-25 Archie Cobbs [EMAIL PROTECTED] * native/jni/classpath/native_state.c: add assertion for object type -Archie __ Archie Cobbs *CTO, Awarix* http://www.awarix.com ___ Classpath mailing list Classpath@gnu.org http://lists.gnu.org/mailman/listinfo/classpath
[commit-cp] classpath ./ChangeLog gnu/java/nio/channels/Fil...
CVSROOT:/cvsroot/classpath Module name:classpath Branch: Changes by: Roman Kennke [EMAIL PROTECTED]05/07/25 13:57:40 Modified files: . : ChangeLog gnu/java/nio/channels: FileChannelImpl.java native/jni/java-nio: gnu_java_nio_channels_FileChannelImpl.c Log message: 2005-07-25 Roman Kennke [EMAIL PROTECTED] * gnu/java/nio/channels/FileChannelImpl.java (static initializer): Init out, err and in here. * native/jni/java-nio/gnu_java_nio_channels_FileChannelImpl.c (Java_gnu_java_nio_channels_FileChannelImpl_init): Moved init code for in, out and err to Java code. CVSWeb URLs: http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/ChangeLog.diff?tr1=1.4213tr2=1.4214r1=textr2=text http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/gnu/java/nio/channels/FileChannelImpl.java.diff?tr1=1.15tr2=1.16r1=textr2=text http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/native/jni/java-nio/gnu_java_nio_channels_FileChannelImpl.c.diff?tr1=1.19tr2=1.20r1=textr2=text ___ Commit-classpath mailing list Commit-classpath@gnu.org http://lists.gnu.org/mailman/listinfo/commit-classpath
[commit-cp] classpath ./ChangeLog java/lang/ClassLoader.jav...
CVSROOT:/cvsroot/classpath Module name:classpath Branch: Changes by: Jeroen Frijters [EMAIL PROTECTED] 05/07/25 14:28:42 Modified files: . : ChangeLog java/lang : ClassLoader.java vm/reference/java/lang: VMClassLoader.java Log message: 2005-07-25 Jeroen Frijters [EMAIL PROTECTED] * java/lang/ClassLoader.java (loadedClasses): Set based on VMClassLoader.VM_USE_CACHE. (defineClass): Modified to respect VMClassLoader.VM_USE_CACHE. * vm/reference/java/lang/VMClassLoader.java (VM_USE_CACHE): New field. CVSWeb URLs: http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/ChangeLog.diff?tr1=1.4215tr2=1.4216r1=textr2=text http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/java/lang/ClassLoader.java.diff?tr1=1.53tr2=1.54r1=textr2=text http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/vm/reference/java/lang/VMClassLoader.java.diff?tr1=1.27tr2=1.28r1=textr2=text ___ Commit-classpath mailing list Commit-classpath@gnu.org http://lists.gnu.org/mailman/listinfo/commit-classpath
[commit-cp] classpath ./ChangeLog java/rmi/server/RMIClassL...
CVSROOT:/cvsroot/classpath Module name:classpath Branch: Changes by: Jeroen Frijters [EMAIL PROTECTED] 05/07/25 16:31:53 Modified files: . : ChangeLog java/rmi/server: RMIClassLoader.java Log message: 2005-07-25 Jeroen Frijters [EMAIL PROTECTED] * java/rmi/server/RMIClassLoader.java (loadClass(String, String)): Use Class.forName() instead of directly calling ClassLoader.loadClass(), to handle array types correctly. CVSWeb URLs: http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/ChangeLog.diff?tr1=1.4217tr2=1.4218r1=textr2=text http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/java/rmi/server/RMIClassLoader.java.diff?tr1=1.16tr2=1.17r1=textr2=text ___ Commit-classpath mailing list Commit-classpath@gnu.org http://lists.gnu.org/mailman/listinfo/commit-classpath
[commit-cp] classpath ./ChangeLog java/awt/Container.java
CVSROOT:/cvsroot/classpath Module name:classpath Branch: Changes by: Anthony Balkissoon [EMAIL PROTECTED] 05/07/25 16:33:15 Modified files: . : ChangeLog java/awt : Container.java Log message: 2005-07-25 Anthony Balkissoon [EMAIL PROTECTED] * java/awt/Container.java: (setFocusTraversalKeys): Instantiate focusTraversalKeys to an array of size 4, not 3. This must have been a typo. CVSWeb URLs: http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/ChangeLog.diff?tr1=1.4218tr2=1.4219r1=textr2=text http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/java/awt/Container.java.diff?tr1=1.55tr2=1.56r1=textr2=text ___ Commit-classpath mailing list Commit-classpath@gnu.org http://lists.gnu.org/mailman/listinfo/commit-classpath
[commit-cp] classpath ./ChangeLog java/text/CollationKey.java
CVSROOT:/cvsroot/classpath Module name:classpath Branch: Changes by: Tom Tromey [EMAIL PROTECTED] 05/07/25 18:36:35 Modified files: . : ChangeLog java/text : CollationKey.java Log message: * java/text/CollationKey.java (equals): Use Arrays.equals. PR classpath/22986. CVSWeb URLs: http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/ChangeLog.diff?tr1=1.4222tr2=1.4223r1=textr2=text http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/java/text/CollationKey.java.diff?tr1=1.15tr2=1.16r1=textr2=text ___ Commit-classpath mailing list Commit-classpath@gnu.org http://lists.gnu.org/mailman/listinfo/commit-classpath
[commit-cp] classpath ./ChangeLog native/jni/classpath/nati...
CVSROOT:/cvsroot/classpath Module name:classpath Branch: Changes by: Archie Cobbs [EMAIL PROTECTED]05/07/25 22:32:51 Modified files: . : ChangeLog native/jni/classpath: native_state.c Log message: * native/jni/classpath/native_state.c: add assertion for object type CVSWeb URLs: http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/ChangeLog.diff?tr1=1.4224tr2=1.4225r1=textr2=text http://savannah.gnu.org/cgi-bin/viewcvs/classpath/classpath/native/jni/classpath/native_state.c.diff?tr1=1.11tr2=1.12r1=textr2=text ___ Commit-classpath mailing list Commit-classpath@gnu.org http://lists.gnu.org/mailman/listinfo/commit-classpath
