[Mono-bugs] [Bug 73527][Maj] New - code execution problems- nonexisting code execution

bugzilla-daemon@bugzilla.ximian.com bugzilla-daemon@bugzilla.ximian.com
Wed, 9 Mar 2005 18:13:20 -0500 (EST)


Please do not reply to this email- if you want to comment on the bug, go to the
URL shown below and enter your comments there.

Changed by plahcim@o2.pl.

http://bugzilla.ximian.com/show_bug.cgi?id=73527

--- shadow/73527	2005-03-09 18:13:20.000000000 -0500
+++ shadow/73527.tmp.29644	2005-03-09 18:13:20.000000000 -0500
@@ -0,0 +1,102 @@
+Bug#: 73527
+Product: Mono: Runtime
+Version: unspecified
+OS: 
+OS Details: amd64
+Status: NEW   
+Resolution: 
+Severity: 
+Priority: Major
+Component: JIT
+AssignedTo: mono-bugs@ximian.com                            
+ReportedBy: plahcim@o2.pl               
+QAContact: mono-bugs@ximian.com
+TargetMilestone: ---
+URL: 
+Cc: 
+Summary: code execution problems- nonexisting code execution
+
+mono from svn
+Linux on amd64 - 64 bit kernel and userland
+mcs source:  $(top_srcdir)/../mcs
+	GC:	     included
+	Globalization:     normal;  default
+	TLS:         __thread
+	SIGALTSTACK: yes
+	Engine:      Building and using the JIT
+	2.0 Alpha:   yes
+	JNI support: IKVM Native
+
+Please fill in this template when reporting a bug, unless you know what you
+are doing.
+Description of Problem: I wrote a piece od code using generics (mono
+compiled prom svn), and after succesfull compilation during program
+execution exception appears:
+Unhandled Exception: System.NullReferenceException: Object reference not
+set to an instance of an object
+in [0x00011] (at
+/root/komp/mono/mcs/class/corlib/System.Globalization/NumberFormatInfo.cs:895)
+System.Globalization.NumberFormatInfo:GetInstance (IFormatProvider provider)
+in [0x00001] (at /root/komp/mono/mcs/class/corlib/System/UInt16.cs:255)
+System.UInt16:ToString (System.String format, IFormatProvider provider)
+in [0x00092] (at
+/home/users/michalp/prog/lab/oblint/pryzaj/ann/src/intcomp/neural/Perceptron.cs:84)
+intcomp.neural.Perceptron:feed ()
+
+but, in line 84 there is only a primitive assignment (double<-double)
+
+Steps to reproduce the problem:
+No idea where the problem lies, other generics-using parts of code are
+working propely. Field that is causing a problem is testa during assignment
+in constructor and is ok, but any use of it after fails.
+Type examining gives a type mismatch. Code:
+namespace intcomp.numeric{
+	public abstract class Function<TW,TA> {
+		public abstract TW compute(TA arg);
+	}
+}
+namespace intcomp.numeric{
+	public class Sigmoid : Function<double,double> {
+		protected double beta;
+		public override double compute(double arg){
+			return 1.0/(1.0+System.Math.Exp(-beta*arg));
+		}
+		public Sigmoid(double beta){this.beta=beta;}
+	}
+}
+
+---------
+using intcomp.numeric;
+namespace intcomp.neural {
+	public class ActivationFunction <WY,WE>{
+		public Function<WY,WE> f;
+		public Function<WY,WE> df;
+		
+		public ActivationFunction(Function<WY,WE> f,Function<WY,WE> df){
+			this.f=f;this.df=df;
+		}
+	}
+}
+
+and than in for example feed (that is screaming now):
+double tmp=this.af.f.compute(activation);
+output=tmp; <-- line 84
+
+you can email my for full source code for testing,
+but the code is not finished;
+
+Code works fine in windows :(
+
+Actual Results:
+Type testing on af fields shows type mismatch.
+Not Function <t,t2> but int or uint depending on the nature of the next
+statment
+Made some changes to (class->interface->class)
+
+Expected Results:
+proper function invocation :)
+
+How often does this happen? 
+Every time with this code;
+
+Additional Information: