Search Results

Search found 330 results on 14 pages for 'unsafe'.

Page 2/14 | < Previous Page | 1 2 3 4 5 6 7 8 9 10 11 12  | Next Page >

  • Copy Small Bitmaps on to Large Bitmap with Transparency Blend: What is faster than graphics.DrawImag

    - by Glenn
    I have identified this call as a bottleneck in a high pressure function. graphics.DrawImage(smallBitmap, x , y); Is there a faster way to blend small semi transparent bitmaps into a larger semi transparent one? Example Usage: XY[] locations = GetLocs(); Bitmap[] bitmaps = GetBmps(); //small images sizes vary approx 30px x 30px using (Bitmap large = new Bitmap(500, 500, PixelFormat.Format32bppPArgb)) using (Graphics largeGraphics = Graphics.FromImage(large)) { for(var i=0; i < largeNumber; i++) { //this is the bottleneck largeGraphics.DrawImage(bitmaps[i], locations[i].x , locations[i].y); } } var done = new MemoryStream(); large.Save(done, ImageFormat.Png); done.Position = 0; return (done); The DrawImage calls take a small 32bppPArgb bitmaps and copies them into a larger bitmap at locations that vary and the small bitmaps might only partially overlap the larger bitmaps visible area. Both images have semi transparent contents that get blended by DrawImage in a way that is important to the output. I've done some testing with BitBlt but not seen significant speed improvement and the alpha blending didn't come out the same in my tests. I'm open to just about any method including a better call to bitblt or unsafe c# code.

    Read the article

  • Explain this C# code: byte* p = (byte*) (void*) Scan0;

    - by qulzam
    I found the code from the net in which i cant understand this line:- byte* p = (byte*)(void*)Scan0; There Scan0 is System.IntPtr. It is code of C#.Net. Plz Explain the above line. The complete code is given below. this is code to convert a image in grayscale. public static Image GrayScale(Bitmap b) { BitmapData bmData = b.LockBits(new Rectangle(0, 0, b.Width, b.Height), ImageLockMode.ReadWrite, PixelFormat.Format24bppRgb); int stride = bmData.Stride; System.IntPtr Scan0 = bmData.Scan0; unsafe { byte* p = (byte*)(void*)Scan0; int nOffset = stride - b.Width * 3; byte red, green, blue; for (int y = 0; y < b.Height; ++y) { for (int x = 0; x < b.Width; ++x) { blue = p[0]; green = p[1]; red = p[2]; p[0] = p[1] = p[2] = (byte)(.299 * red + .587 * green + .114 * blue); p += 3; } p += nOffset; } } b.UnlockBits(bmData); return (Image)b; } I understand all the code but only have the problem on this line. byte* p = (byte*)(void*)Scan0;

    Read the article

  • Calling AuditQuerySystemPolicy() (advapi32.dll) from C# returns "The parameter is incorrect"

    - by JCCyC
    The sequence is like follows: Open a policy handle with LsaOpenPolicy() (not shown) Call LsaQueryInformationPolicy() to get the number of categories; For each category: Call AuditLookupCategoryGuidFromCategoryId() to turn the enum value into a GUID; Call AuditEnumerateSubCategories() to get a list of the GUIDs of all subcategories; Call AuditQuerySystemPolicy() to get the audit policies for the subcategories. All of these work and return expected, sensible values except the last. Calling AuditQuerySystemPolicy() gets me a "The parameter is incorrect" error. I'm thinking there must be some subtle unmarshaling problem. I'm probably misinterpreting what exactly AuditEnumerateSubCategories() returns, but I'm stumped. You'll see (commented) I tried to dereference the return pointer from AuditEnumerateSubCategories() as a pointer. Doing or not doing that gives the same result. Code: #region LSA types public enum POLICY_INFORMATION_CLASS { PolicyAuditLogInformation = 1, PolicyAuditEventsInformation, PolicyPrimaryDomainInformation, PolicyPdAccountInformation, PolicyAccountDomainInformation, PolicyLsaServerRoleInformation, PolicyReplicaSourceInformation, PolicyDefaultQuotaInformation, PolicyModificationInformation, PolicyAuditFullSetInformation, PolicyAuditFullQueryInformation, PolicyDnsDomainInformation } public enum POLICY_AUDIT_EVENT_TYPE { AuditCategorySystem, AuditCategoryLogon, AuditCategoryObjectAccess, AuditCategoryPrivilegeUse, AuditCategoryDetailedTracking, AuditCategoryPolicyChange, AuditCategoryAccountManagement, AuditCategoryDirectoryServiceAccess, AuditCategoryAccountLogon } [StructLayout(LayoutKind.Sequential, CharSet = CharSet.Unicode)] public struct POLICY_AUDIT_EVENTS_INFO { public bool AuditingMode; public IntPtr EventAuditingOptions; public UInt32 MaximumAuditEventCount; } [StructLayout(LayoutKind.Sequential, CharSet = CharSet.Unicode)] public struct GUID { public UInt32 Data1; public UInt16 Data2; public UInt16 Data3; public Byte Data4a; public Byte Data4b; public Byte Data4c; public Byte Data4d; public Byte Data4e; public Byte Data4f; public Byte Data4g; public Byte Data4h; public override string ToString() { return Data1.ToString("x8") + "-" + Data2.ToString("x4") + "-" + Data3.ToString("x4") + "-" + Data4a.ToString("x2") + Data4b.ToString("x2") + "-" + Data4c.ToString("x2") + Data4d.ToString("x2") + Data4e.ToString("x2") + Data4f.ToString("x2") + Data4g.ToString("x2") + Data4h.ToString("x2"); } } #endregion #region LSA Imports [DllImport("kernel32.dll")] extern static int GetLastError(); [DllImport("advapi32.dll", CharSet = CharSet.Unicode, PreserveSig = true)] public static extern UInt32 LsaNtStatusToWinError( long Status); [DllImport("advapi32.dll", CharSet = CharSet.Unicode, PreserveSig = true)] public static extern long LsaOpenPolicy( ref LSA_UNICODE_STRING SystemName, ref LSA_OBJECT_ATTRIBUTES ObjectAttributes, Int32 DesiredAccess, out IntPtr PolicyHandle ); [DllImport("advapi32.dll", CharSet = CharSet.Unicode, PreserveSig = true)] public static extern long LsaClose(IntPtr PolicyHandle); [DllImport("advapi32.dll", CharSet = CharSet.Unicode, PreserveSig = true)] public static extern long LsaFreeMemory(IntPtr Buffer); [DllImport("advapi32.dll", CharSet = CharSet.Unicode, PreserveSig = true)] public static extern void AuditFree(IntPtr Buffer); [DllImport("advapi32.dll", SetLastError = true, PreserveSig = true)] public static extern long LsaQueryInformationPolicy( IntPtr PolicyHandle, POLICY_INFORMATION_CLASS InformationClass, out IntPtr Buffer); [DllImport("advapi32.dll", SetLastError = true, PreserveSig = true)] public static extern bool AuditLookupCategoryGuidFromCategoryId( POLICY_AUDIT_EVENT_TYPE AuditCategoryId, IntPtr pAuditCategoryGuid); [DllImport("advapi32.dll", SetLastError = true, PreserveSig = true)] public static extern bool AuditEnumerateSubCategories( IntPtr pAuditCategoryGuid, bool bRetrieveAllSubCategories, out IntPtr ppAuditSubCategoriesArray, out ulong pCountReturned); [DllImport("advapi32.dll", SetLastError = true, PreserveSig = true)] public static extern bool AuditQuerySystemPolicy( IntPtr pSubCategoryGuids, ulong PolicyCount, out IntPtr ppAuditPolicy); #endregion Dictionary<string, UInt32> retList = new Dictionary<string, UInt32>(); long lretVal; uint retVal; IntPtr pAuditEventsInfo; lretVal = LsaQueryInformationPolicy(policyHandle, POLICY_INFORMATION_CLASS.PolicyAuditEventsInformation, out pAuditEventsInfo); retVal = LsaNtStatusToWinError(lretVal); if (retVal != 0) { LsaClose(policyHandle); throw new System.ComponentModel.Win32Exception((int)retVal); } POLICY_AUDIT_EVENTS_INFO myAuditEventsInfo = new POLICY_AUDIT_EVENTS_INFO(); myAuditEventsInfo = (POLICY_AUDIT_EVENTS_INFO)Marshal.PtrToStructure(pAuditEventsInfo, myAuditEventsInfo.GetType()); IntPtr subCats = IntPtr.Zero; ulong nSubCats = 0; for (int audCat = 0; audCat < myAuditEventsInfo.MaximumAuditEventCount; audCat++) { GUID audCatGuid = new GUID(); if (!AuditLookupCategoryGuidFromCategoryId((POLICY_AUDIT_EVENT_TYPE)audCat, new IntPtr(&audCatGuid))) { int causingError = GetLastError(); LsaFreeMemory(pAuditEventsInfo); LsaClose(policyHandle); throw new System.ComponentModel.Win32Exception(causingError); } if (!AuditEnumerateSubCategories(new IntPtr(&audCatGuid), true, out subCats, out nSubCats)) { int causingError = GetLastError(); LsaFreeMemory(pAuditEventsInfo); LsaClose(policyHandle); throw new System.ComponentModel.Win32Exception(causingError); } // Dereference the first pointer-to-pointer to point to the first subcategory // subCats = (IntPtr)Marshal.PtrToStructure(subCats, subCats.GetType()); if (nSubCats > 0) { IntPtr audPolicies = IntPtr.Zero; if (!AuditQuerySystemPolicy(subCats, nSubCats, out audPolicies)) { int causingError = GetLastError(); if (subCats != IntPtr.Zero) AuditFree(subCats); LsaFreeMemory(pAuditEventsInfo); LsaClose(policyHandle); throw new System.ComponentModel.Win32Exception(causingError); } AUDIT_POLICY_INFORMATION myAudPol = new AUDIT_POLICY_INFORMATION(); for (ulong audSubCat = 0; audSubCat < nSubCats; audSubCat++) { // Process audPolicies[audSubCat], turn GUIDs into names, fill retList. // http://msdn.microsoft.com/en-us/library/aa373931%28VS.85%29.aspx // http://msdn.microsoft.com/en-us/library/bb648638%28VS.85%29.aspx IntPtr itemAddr = IntPtr.Zero; IntPtr itemAddrAddr = new IntPtr(audPolicies.ToInt64() + (long)(audSubCat * (ulong)Marshal.SizeOf(itemAddr))); itemAddr = (IntPtr)Marshal.PtrToStructure(itemAddrAddr, itemAddr.GetType()); myAudPol = (AUDIT_POLICY_INFORMATION)Marshal.PtrToStructure(itemAddr, myAudPol.GetType()); retList[myAudPol.AuditSubCategoryGuid.ToString()] = myAudPol.AuditingInformation; } if (audPolicies != IntPtr.Zero) AuditFree(audPolicies); } if (subCats != IntPtr.Zero) AuditFree(subCats); subCats = IntPtr.Zero; nSubCats = 0; } lretVal = LsaFreeMemory(pAuditEventsInfo); retVal = LsaNtStatusToWinError(lretVal); if (retVal != 0) throw new System.ComponentModel.Win32Exception((int)retVal); lretVal = LsaClose(policyHandle); retVal = LsaNtStatusToWinError(lretVal); if (retVal != 0) throw new System.ComponentModel.Win32Exception((int)retVal);

    Read the article

  • Bitwise operation on void* in C#

    - by code poet
    So I am Reflector-ing some framework 2.0 code and end up with the following deconstruction fixed (void* voidRef3 = ((void*) & _someMember)) { ... } This won't compile due to 'The right hand side of a fixed statement assignment may not be a cast expression' I understand that Reflector can only approximate and generally I can see a clear path but this is a bit outside my experience. Question: what is Reflector trying to describe to me? Update: Am also seeing the following fixed (IntPtr* ptrRef3 = ((IntPtr*) & this._someMember))

    Read the article

  • Converting C# void* to byte[]

    - by Yurik
    In C#, I need to write T[] to a stream, ideally without any additional buffers. I have a dynamic code that converts T[] (where T is a no-objects struct) to a void* and fixes it in memory, and that works great. When the stream was a file, I could use native Windows API to pass the void * directly, but now I need to write to a generic Stream object that takes byte[]. Can anyone suggest a hack way to create a dummy array object which does not actually have any heap allocations, but rather points to an already existing (and fixed) heap location. This is the pseudo-code that I need: void Write(Stream stream, T[] buffer) { fixed( void* ptr = &buffer ) // done with dynamic code generation { int typeSize = sizeof(T); // done as well byte[] dummy = (byte[]) ptr; // <-- how do I create this fake array? stream.Write( dummy, 0, buffer.Length*typeSize ); } }

    Read the article

  • Help translating Reflector deconstruction into compilable code

    - by code poet
    So I am Reflector-ing some framework 2.0 code and end up with the following deconstruction fixed (void* voidRef3 = ((void*) &_someMember)) { ... } This won't compile due to 'The right hand side of a fixed statement assignment may not be a cast expression' I understand that Reflector can only approximate and generally I can see a clear path but this is a bit outside my experience. Question: what is Reflector trying to describe to me? Update: Am also seeing the following fixed (IntPtr* ptrRef3 = ((IntPtr*) &this._someMember)) Update: So, as Mitch says, it is not a bitwise operator, but an addressOf operator. Question is now: fixed (IntPtr* ptrRef3 = &_someMember) fails with an 'Cannot implicitly convert type 'xxx*' to 'System.IntPtr*'. An explicit conversion exists (are you missing a cast?)' compilation error. So I seemed to be damned if I do and damned if I dont. Any ideas?

    Read the article

  • How can I handle "NTFS partition is in unsafe state"?

    - by user211040
    Error mounting /dev/sda3 at /media/franklcohen/OS: Command-line `mount -t "ntfs" -o "uhelper=udisks2,nodev,nosuid,uid=1000,gid=1000,dmask=0077,fmask=0177" "/dev/sda3" "/media/franklcohen/OS"' exited with non-zero exit status 14: Windows is hibernated, refused to mount. Failed to mount '/dev/sda3': Operation not permitted The NTFS partition is in an unsafe state. Please resume and shutdown Windows fully (no hibernation or fast restarting), or mount the volume read-only with the 'ro' mount option. i get this error i have disabled fast start up in windows 8. what can i do i shut down my computer 4 times in windows and disabled fast start in windows 8. i'm using Ubuntu 13.10. please help thanks.

    Read the article

  • value types in the vm

    - by john.rose
    value types in the vm p.p1 {margin: 0.0px 0.0px 0.0px 0.0px; font: 14.0px Times} p.p2 {margin: 0.0px 0.0px 14.0px 0.0px; font: 14.0px Times} p.p3 {margin: 0.0px 0.0px 12.0px 0.0px; font: 14.0px Times} p.p4 {margin: 0.0px 0.0px 15.0px 0.0px; font: 14.0px Times} p.p5 {margin: 0.0px 0.0px 0.0px 0.0px; font: 14.0px Courier} p.p6 {margin: 0.0px 0.0px 0.0px 0.0px; font: 14.0px Courier; min-height: 17.0px} p.p7 {margin: 0.0px 0.0px 0.0px 0.0px; font: 14.0px Times; min-height: 18.0px} p.p8 {margin: 0.0px 0.0px 0.0px 36.0px; text-indent: -36.0px; font: 14.0px Times; min-height: 18.0px} p.p9 {margin: 0.0px 0.0px 12.0px 0.0px; font: 14.0px Times; min-height: 18.0px} p.p10 {margin: 0.0px 0.0px 12.0px 0.0px; font: 14.0px Times; color: #000000} li.li1 {margin: 0.0px 0.0px 0.0px 0.0px; font: 14.0px Times} li.li7 {margin: 0.0px 0.0px 0.0px 0.0px; font: 14.0px Times; min-height: 18.0px} span.s1 {font: 14.0px Courier} span.s2 {color: #000000} span.s3 {font: 14.0px Courier; color: #000000} ol.ol1 {list-style-type: decimal} Or, enduring values for a changing world. Introduction A value type is a data type which, generally speaking, is designed for being passed by value in and out of methods, and stored by value in data structures. The only value types which the Java language directly supports are the eight primitive types. Java indirectly and approximately supports value types, if they are implemented in terms of classes. For example, both Integer and String may be viewed as value types, especially if their usage is restricted to avoid operations appropriate to Object. In this note, we propose a definition of value types in terms of a design pattern for Java classes, accompanied by a set of usage restrictions. We also sketch the relation of such value types to tuple types (which are a JVM-level notion), and point out JVM optimizations that can apply to value types. This note is a thought experiment to extend the JVM’s performance model in support of value types. The demonstration has two phases.  Initially the extension can simply use design patterns, within the current bytecode architecture, and in today’s Java language. But if the performance model is to be realized in practice, it will probably require new JVM bytecode features, changes to the Java language, or both.  We will look at a few possibilities for these new features. An Axiom of Value In the context of the JVM, a value type is a data type equipped with construction, assignment, and equality operations, and a set of typed components, such that, whenever two variables of the value type produce equal corresponding values for their components, the values of the two variables cannot be distinguished by any JVM operation. Here are some corollaries: A value type is immutable, since otherwise a copy could be constructed and the original could be modified in one of its components, allowing the copies to be distinguished. Changing the component of a value type requires construction of a new value. The equals and hashCode operations are strictly component-wise. If a value type is represented by a JVM reference, that reference cannot be successfully synchronized on, and cannot be usefully compared for reference equality. A value type can be viewed in terms of what it doesn’t do. We can say that a value type omits all value-unsafe operations, which could violate the constraints on value types.  These operations, which are ordinarily allowed for Java object types, are pointer equality comparison (the acmp instruction), synchronization (the monitor instructions), all the wait and notify methods of class Object, and non-trivial finalize methods. The clone method is also value-unsafe, although for value types it could be treated as the identity function. Finally, and most importantly, any side effect on an object (however visible) also counts as an value-unsafe operation. A value type may have methods, but such methods must not change the components of the value. It is reasonable and useful to define methods like toString, equals, and hashCode on value types, and also methods which are specifically valuable to users of the value type. Representations of Value Value types have two natural representations in the JVM, unboxed and boxed. An unboxed value consists of the components, as simple variables. For example, the complex number x=(1+2i), in rectangular coordinate form, may be represented in unboxed form by the following pair of variables: /*Complex x = Complex.valueOf(1.0, 2.0):*/ double x_re = 1.0, x_im = 2.0; These variables might be locals, parameters, or fields. Their association as components of a single value is not defined to the JVM. Here is a sample computation which computes the norm of the difference between two complex numbers: double distance(/*Complex x:*/ double x_re, double x_im,         /*Complex y:*/ double y_re, double y_im) {     /*Complex z = x.minus(y):*/     double z_re = x_re - y_re, z_im = x_im - y_im;     /*return z.abs():*/     return Math.sqrt(z_re*z_re + z_im*z_im); } A boxed representation groups component values under a single object reference. The reference is to a ‘wrapper class’ that carries the component values in its fields. (A primitive type can naturally be equated with a trivial value type with just one component of that type. In that view, the wrapper class Integer can serve as a boxed representation of value type int.) The unboxed representation of complex numbers is practical for many uses, but it fails to cover several major use cases: return values, array elements, and generic APIs. The two components of a complex number cannot be directly returned from a Java function, since Java does not support multiple return values. The same story applies to array elements: Java has no ’array of structs’ feature. (Double-length arrays are a possible workaround for complex numbers, but not for value types with heterogeneous components.) By generic APIs I mean both those which use generic types, like Arrays.asList and those which have special case support for primitive types, like String.valueOf and PrintStream.println. Those APIs do not support unboxed values, and offer some problems to boxed values. Any ’real’ JVM type should have a story for returns, arrays, and API interoperability. The basic problem here is that value types fall between primitive types and object types. Value types are clearly more complex than primitive types, and object types are slightly too complicated. Objects are a little bit dangerous to use as value carriers, since object references can be compared for pointer equality, and can be synchronized on. Also, as many Java programmers have observed, there is often a performance cost to using wrapper objects, even on modern JVMs. Even so, wrapper classes are a good starting point for talking about value types. If there were a set of structural rules and restrictions which would prevent value-unsafe operations on value types, wrapper classes would provide a good notation for defining value types. This note attempts to define such rules and restrictions. Let’s Start Coding Now it is time to look at some real code. Here is a definition, written in Java, of a complex number value type. @ValueSafe public final class Complex implements java.io.Serializable {     // immutable component structure:     public final double re, im;     private Complex(double re, double im) {         this.re = re; this.im = im;     }     // interoperability methods:     public String toString() { return "Complex("+re+","+im+")"; }     public List<Double> asList() { return Arrays.asList(re, im); }     public boolean equals(Complex c) {         return re == c.re && im == c.im;     }     public boolean equals(@ValueSafe Object x) {         return x instanceof Complex && equals((Complex) x);     }     public int hashCode() {         return 31*Double.valueOf(re).hashCode()                 + Double.valueOf(im).hashCode();     }     // factory methods:     public static Complex valueOf(double re, double im) {         return new Complex(re, im);     }     public Complex changeRe(double re2) { return valueOf(re2, im); }     public Complex changeIm(double im2) { return valueOf(re, im2); }     public static Complex cast(@ValueSafe Object x) {         return x == null ? ZERO : (Complex) x;     }     // utility methods and constants:     public Complex plus(Complex c)  { return new Complex(re+c.re, im+c.im); }     public Complex minus(Complex c) { return new Complex(re-c.re, im-c.im); }     public double abs() { return Math.sqrt(re*re + im*im); }     public static final Complex PI = valueOf(Math.PI, 0.0);     public static final Complex ZERO = valueOf(0.0, 0.0); } This is not a minimal definition, because it includes some utility methods and other optional parts.  The essential elements are as follows: The class is marked as a value type with an annotation. The class is final, because it does not make sense to create subclasses of value types. The fields of the class are all non-private and final.  (I.e., the type is immutable and structurally transparent.) From the supertype Object, all public non-final methods are overridden. The constructor is private. Beyond these bare essentials, we can observe the following features in this example, which are likely to be typical of all value types: One or more factory methods are responsible for value creation, including a component-wise valueOf method. There are utility methods for complex arithmetic and instance creation, such as plus and changeIm. There are static utility constants, such as PI. The type is serializable, using the default mechanisms. There are methods for converting to and from dynamically typed references, such as asList and cast. The Rules In order to use value types properly, the programmer must avoid value-unsafe operations.  A helpful Java compiler should issue errors (or at least warnings) for code which provably applies value-unsafe operations, and should issue warnings for code which might be correct but does not provably avoid value-unsafe operations.  No such compilers exist today, but to simplify our account here, we will pretend that they do exist. A value-safe type is any class, interface, or type parameter marked with the @ValueSafe annotation, or any subtype of a value-safe type.  If a value-safe class is marked final, it is in fact a value type.  All other value-safe classes must be abstract.  The non-static fields of a value class must be non-public and final, and all its constructors must be private. Under the above rules, a standard interface could be helpful to define value types like Complex.  Here is an example: @ValueSafe public interface ValueType extends java.io.Serializable {     // All methods listed here must get redefined.     // Definitions must be value-safe, which means     // they may depend on component values only.     List<? extends Object> asList();     int hashCode();     boolean equals(@ValueSafe Object c);     String toString(); } //@ValueSafe inherited from supertype: public final class Complex implements ValueType { … The main advantage of such a conventional interface is that (unlike an annotation) it is reified in the runtime type system.  It could appear as an element type or parameter bound, for facilities which are designed to work on value types only.  More broadly, it might assist the JVM to perform dynamic enforcement of the rules for value types. Besides types, the annotation @ValueSafe can mark fields, parameters, local variables, and methods.  (This is redundant when the type is also value-safe, but may be useful when the type is Object or another supertype of a value type.)  Working forward from these annotations, an expression E is defined as value-safe if it satisfies one or more of the following: The type of E is a value-safe type. E names a field, parameter, or local variable whose declaration is marked @ValueSafe. E is a call to a method whose declaration is marked @ValueSafe. E is an assignment to a value-safe variable, field reference, or array reference. E is a cast to a value-safe type from a value-safe expression. E is a conditional expression E0 ? E1 : E2, and both E1 and E2 are value-safe. Assignments to value-safe expressions and initializations of value-safe names must take their values from value-safe expressions. A value-safe expression may not be the subject of a value-unsafe operation.  In particular, it cannot be synchronized on, nor can it be compared with the “==” operator, not even with a null or with another value-safe type. In a program where all of these rules are followed, no value-type value will be subject to a value-unsafe operation.  Thus, the prime axiom of value types will be satisfied, that no two value type will be distinguishable as long as their component values are equal. More Code To illustrate these rules, here are some usage examples for Complex: Complex pi = Complex.valueOf(Math.PI, 0); Complex zero = pi.changeRe(0);  //zero = pi; zero.re = 0; ValueType vtype = pi; @SuppressWarnings("value-unsafe")   Object obj = pi; @ValueSafe Object obj2 = pi; obj2 = new Object();  // ok List<Complex> clist = new ArrayList<Complex>(); clist.add(pi);  // (ok assuming List.add param is @ValueSafe) List<ValueType> vlist = new ArrayList<ValueType>(); vlist.add(pi);  // (ok) List<Object> olist = new ArrayList<Object>(); olist.add(pi);  // warning: "value-unsafe" boolean z = pi.equals(zero); boolean z1 = (pi == zero);  // error: reference comparison on value type boolean z2 = (pi == null);  // error: reference comparison on value type boolean z3 = (pi == obj2);  // error: reference comparison on value type synchronized (pi) { }  // error: synch of value, unpredictable result synchronized (obj2) { }  // unpredictable result Complex qq = pi; qq = null;  // possible NPE; warning: “null-unsafe" qq = (Complex) obj;  // warning: “null-unsafe" qq = Complex.cast(obj);  // OK @SuppressWarnings("null-unsafe")   Complex empty = null;  // possible NPE qq = empty;  // possible NPE (null pollution) The Payoffs It follows from this that either the JVM or the java compiler can replace boxed value-type values with unboxed ones, without affecting normal computations.  Fields and variables of value types can be split into their unboxed components.  Non-static methods on value types can be transformed into static methods which take the components as value parameters. Some common questions arise around this point in any discussion of value types. Why burden the programmer with all these extra rules?  Why not detect programs automagically and perform unboxing transparently?  The answer is that it is easy to break the rules accidently unless they are agreed to by the programmer and enforced.  Automatic unboxing optimizations are tantalizing but (so far) unreachable ideal.  In the current state of the art, it is possible exhibit benchmarks in which automatic unboxing provides the desired effects, but it is not possible to provide a JVM with a performance model that assures the programmer when unboxing will occur.  This is why I’m writing this note, to enlist help from, and provide assurances to, the programmer.  Basically, I’m shooting for a good set of user-supplied “pragmas” to frame the desired optimization. Again, the important thing is that the unboxing must be done reliably, or else programmers will have no reason to work with the extra complexity of the value-safety rules.  There must be a reasonably stable performance model, wherein using a value type has approximately the same performance characteristics as writing the unboxed components as separate Java variables. There are some rough corners to the present scheme.  Since Java fields and array elements are initialized to null, value-type computations which incorporate uninitialized variables can produce null pointer exceptions.  One workaround for this is to require such variables to be null-tested, and the result replaced with a suitable all-zero value of the value type.  That is what the “cast” method does above. Generically typed APIs like List<T> will continue to manipulate boxed values always, at least until we figure out how to do reification of generic type instances.  Use of such APIs will elicit warnings until their type parameters (and/or relevant members) are annotated or typed as value-safe.  Retrofitting List<T> is likely to expose flaws in the present scheme, which we will need to engineer around.  Here are a couple of first approaches: public interface java.util.List<@ValueSafe T> extends Collection<T> { … public interface java.util.List<T extends Object|ValueType> extends Collection<T> { … (The second approach would require disjunctive types, in which value-safety is “contagious” from the constituent types.) With more transformations, the return value types of methods can also be unboxed.  This may require significant bytecode-level transformations, and would work best in the presence of a bytecode representation for multiple value groups, which I have proposed elsewhere under the title “Tuples in the VM”. But for starters, the JVM can apply this transformation under the covers, to internally compiled methods.  This would give a way to express multiple return values and structured return values, which is a significant pain-point for Java programmers, especially those who work with low-level structure types favored by modern vector and graphics processors.  The lack of multiple return values has a strong distorting effect on many Java APIs. Even if the JVM fails to unbox a value, there is still potential benefit to the value type.  Clustered computing systems something have copy operations (serialization or something similar) which apply implicitly to command operands.  When copying JVM objects, it is extremely helpful to know when an object’s identity is important or not.  If an object reference is a copied operand, the system may have to create a proxy handle which points back to the original object, so that side effects are visible.  Proxies must be managed carefully, and this can be expensive.  On the other hand, value types are exactly those types which a JVM can “copy and forget” with no downside. Array types are crucial to bulk data interfaces.  (As data sizes and rates increase, bulk data becomes more important than scalar data, so arrays are definitely accompanying us into the future of computing.)  Value types are very helpful for adding structure to bulk data, so a successful value type mechanism will make it easier for us to express richer forms of bulk data. Unboxing arrays (i.e., arrays containing unboxed values) will provide better cache and memory density, and more direct data movement within clustered or heterogeneous computing systems.  They require the deepest transformations, relative to today’s JVM.  There is an impedance mismatch between value-type arrays and Java’s covariant array typing, so compromises will need to be struck with existing Java semantics.  It is probably worth the effort, since arrays of unboxed value types are inherently more memory-efficient than standard Java arrays, which rely on dependent pointer chains. It may be sufficient to extend the “value-safe” concept to array declarations, and allow low-level transformations to change value-safe array declarations from the standard boxed form into an unboxed tuple-based form.  Such value-safe arrays would not be convertible to Object[] arrays.  Certain connection points, such as Arrays.copyOf and System.arraycopy might need additional input/output combinations, to allow smooth conversion between arrays with boxed and unboxed elements. Alternatively, the correct solution may have to wait until we have enough reification of generic types, and enough operator overloading, to enable an overhaul of Java arrays. Implicit Method Definitions The example of class Complex above may be unattractively complex.  I believe most or all of the elements of the example class are required by the logic of value types. If this is true, a programmer who writes a value type will have to write lots of error-prone boilerplate code.  On the other hand, I think nearly all of the code (except for the domain-specific parts like plus and minus) can be implicitly generated. Java has a rule for implicitly defining a class’s constructor, if no it defines no constructors explicitly.  Likewise, there are rules for providing default access modifiers for interface members.  Because of the highly regular structure of value types, it might be reasonable to perform similar implicit transformations on value types.  Here’s an example of a “highly implicit” definition of a complex number type: public class Complex implements ValueType {  // implicitly final     public double re, im;  // implicitly public final     //implicit methods are defined elementwise from te fields:     //  toString, asList, equals(2), hashCode, valueOf, cast     //optionally, explicit methods (plus, abs, etc.) would go here } In other words, with the right defaults, a simple value type definition can be a one-liner.  The observant reader will have noticed the similarities (and suitable differences) between the explicit methods above and the corresponding methods for List<T>. Another way to abbreviate such a class would be to make an annotation the primary trigger of the functionality, and to add the interface(s) implicitly: public @ValueType class Complex { … // implicitly final, implements ValueType (But to me it seems better to communicate the “magic” via an interface, even if it is rooted in an annotation.) Implicitly Defined Value Types So far we have been working with nominal value types, which is to say that the sequence of typed components is associated with a name and additional methods that convey the intention of the programmer.  A simple ordered pair of floating point numbers can be variously interpreted as (to name a few possibilities) a rectangular or polar complex number or Cartesian point.  The name and the methods convey the intended meaning. But what if we need a truly simple ordered pair of floating point numbers, without any further conceptual baggage?  Perhaps we are writing a method (like “divideAndRemainder”) which naturally returns a pair of numbers instead of a single number.  Wrapping the pair of numbers in a nominal type (like “QuotientAndRemainder”) makes as little sense as wrapping a single return value in a nominal type (like “Quotient”).  What we need here are structural value types commonly known as tuples. For the present discussion, let us assign a conventional, JVM-friendly name to tuples, roughly as follows: public class java.lang.tuple.$DD extends java.lang.tuple.Tuple {      double $1, $2; } Here the component names are fixed and all the required methods are defined implicitly.  The supertype is an abstract class which has suitable shared declarations.  The name itself mentions a JVM-style method parameter descriptor, which may be “cracked” to determine the number and types of the component fields. The odd thing about such a tuple type (and structural types in general) is it must be instantiated lazily, in response to linkage requests from one or more classes that need it.  The JVM and/or its class loaders must be prepared to spin a tuple type on demand, given a simple name reference, $xyz, where the xyz is cracked into a series of component types.  (Specifics of naming and name mangling need some tasteful engineering.) Tuples also seem to demand, even more than nominal types, some support from the language.  (This is probably because notations for non-nominal types work best as combinations of punctuation and type names, rather than named constructors like Function3 or Tuple2.)  At a minimum, languages with tuples usually (I think) have some sort of simple bracket notation for creating tuples, and a corresponding pattern-matching syntax (or “destructuring bind”) for taking tuples apart, at least when they are parameter lists.  Designing such a syntax is no simple thing, because it ought to play well with nominal value types, and also with pre-existing Java features, such as method parameter lists, implicit conversions, generic types, and reflection.  That is a task for another day. Other Use Cases Besides complex numbers and simple tuples there are many use cases for value types.  Many tuple-like types have natural value-type representations. These include rational numbers, point locations and pixel colors, and various kinds of dates and addresses. Other types have a variable-length ‘tail’ of internal values. The most common example of this is String, which is (mathematically) a sequence of UTF-16 character values. Similarly, bit vectors, multiple-precision numbers, and polynomials are composed of sequences of values. Such types include, in their representation, a reference to a variable-sized data structure (often an array) which (somehow) represents the sequence of values. The value type may also include ’header’ information. Variable-sized values often have a length distribution which favors short lengths. In that case, the design of the value type can make the first few values in the sequence be direct ’header’ fields of the value type. In the common case where the header is enough to represent the whole value, the tail can be a shared null value, or even just a null reference. Note that the tail need not be an immutable object, as long as the header type encapsulates it well enough. This is the case with String, where the tail is a mutable (but never mutated) character array. Field types and their order must be a globally visible part of the API.  The structure of the value type must be transparent enough to have a globally consistent unboxed representation, so that all callers and callees agree about the type and order of components  that appear as parameters, return types, and array elements.  This is a trade-off between efficiency and encapsulation, which is forced on us when we remove an indirection enjoyed by boxed representations.  A JVM-only transformation would not care about such visibility, but a bytecode transformation would need to take care that (say) the components of complex numbers would not get swapped after a redefinition of Complex and a partial recompile.  Perhaps constant pool references to value types need to declare the field order as assumed by each API user. This brings up the delicate status of private fields in a value type.  It must always be possible to load, store, and copy value types as coordinated groups, and the JVM performs those movements by moving individual scalar values between locals and stack.  If a component field is not public, what is to prevent hostile code from plucking it out of the tuple using a rogue aload or astore instruction?  Nothing but the verifier, so we may need to give it more smarts, so that it treats value types as inseparable groups of stack slots or locals (something like long or double). My initial thought was to make the fields always public, which would make the security problem moot.  But public is not always the right answer; consider the case of String, where the underlying mutable character array must be encapsulated to prevent security holes.  I believe we can win back both sides of the tradeoff, by training the verifier never to split up the components in an unboxed value.  Just as the verifier encapsulates the two halves of a 64-bit primitive, it can encapsulate the the header and body of an unboxed String, so that no code other than that of class String itself can take apart the values. Similar to String, we could build an efficient multi-precision decimal type along these lines: public final class DecimalValue extends ValueType {     protected final long header;     protected private final BigInteger digits;     public DecimalValue valueOf(int value, int scale) {         assert(scale >= 0);         return new DecimalValue(((long)value << 32) + scale, null);     }     public DecimalValue valueOf(long value, int scale) {         if (value == (int) value)             return valueOf((int)value, scale);         return new DecimalValue(-scale, new BigInteger(value));     } } Values of this type would be passed between methods as two machine words. Small values (those with a significand which fits into 32 bits) would be represented without any heap data at all, unless the DecimalValue itself were boxed. (Note the tension between encapsulation and unboxing in this case.  It would be better if the header and digits fields were private, but depending on where the unboxing information must “leak”, it is probably safer to make a public revelation of the internal structure.) Note that, although an array of Complex can be faked with a double-length array of double, there is no easy way to fake an array of unboxed DecimalValues.  (Either an array of boxed values or a transposed pair of homogeneous arrays would be reasonable fallbacks, in a current JVM.)  Getting the full benefit of unboxing and arrays will require some new JVM magic. Although the JVM emphasizes portability, system dependent code will benefit from using machine-level types larger than 64 bits.  For example, the back end of a linear algebra package might benefit from value types like Float4 which map to stock vector types.  This is probably only worthwhile if the unboxing arrays can be packed with such values. More Daydreams A more finely-divided design for dynamic enforcement of value safety could feature separate marker interfaces for each invariant.  An empty marker interface Unsynchronizable could cause suitable exceptions for monitor instructions on objects in marked classes.  More radically, a Interchangeable marker interface could cause JVM primitives that are sensitive to object identity to raise exceptions; the strangest result would be that the acmp instruction would have to be specified as raising an exception. @ValueSafe public interface ValueType extends java.io.Serializable,         Unsynchronizable, Interchangeable { … public class Complex implements ValueType {     // inherits Serializable, Unsynchronizable, Interchangeable, @ValueSafe     … It seems possible that Integer and the other wrapper types could be retro-fitted as value-safe types.  This is a major change, since wrapper objects would be unsynchronizable and their references interchangeable.  It is likely that code which violates value-safety for wrapper types exists but is uncommon.  It is less plausible to retro-fit String, since the prominent operation String.intern is often used with value-unsafe code. We should also reconsider the distinction between boxed and unboxed values in code.  The design presented above obscures that distinction.  As another thought experiment, we could imagine making a first class distinction in the type system between boxed and unboxed representations.  Since only primitive types are named with a lower-case initial letter, we could define that the capitalized version of a value type name always refers to the boxed representation, while the initial lower-case variant always refers to boxed.  For example: complex pi = complex.valueOf(Math.PI, 0); Complex boxPi = pi;  // convert to boxed myList.add(boxPi); complex z = myList.get(0);  // unbox Such a convention could perhaps absorb the current difference between int and Integer, double and Double. It might also allow the programmer to express a helpful distinction among array types. As said above, array types are crucial to bulk data interfaces, but are limited in the JVM.  Extending arrays beyond the present limitations is worth thinking about; for example, the Maxine JVM implementation has a hybrid object/array type.  Something like this which can also accommodate value type components seems worthwhile.  On the other hand, does it make sense for value types to contain short arrays?  And why should random-access arrays be the end of our design process, when bulk data is often sequentially accessed, and it might make sense to have heterogeneous streams of data as the natural “jumbo” data structure.  These considerations must wait for another day and another note. More Work It seems to me that a good sequence for introducing such value types would be as follows: Add the value-safety restrictions to an experimental version of javac. Code some sample applications with value types, including Complex and DecimalValue. Create an experimental JVM which internally unboxes value types but does not require new bytecodes to do so.  Ensure the feasibility of the performance model for the sample applications. Add tuple-like bytecodes (with or without generic type reification) to a major revision of the JVM, and teach the Java compiler to switch in the new bytecodes without code changes. A staggered roll-out like this would decouple language changes from bytecode changes, which is always a convenient thing. A similar investigation should be applied (concurrently) to array types.  In this case, it seems to me that the starting point is in the JVM: Add an experimental unboxing array data structure to a production JVM, perhaps along the lines of Maxine hybrids.  No bytecode or language support is required at first; everything can be done with encapsulated unsafe operations and/or method handles. Create an experimental JVM which internally unboxes value types but does not require new bytecodes to do so.  Ensure the feasibility of the performance model for the sample applications. Add tuple-like bytecodes (with or without generic type reification) to a major revision of the JVM, and teach the Java compiler to switch in the new bytecodes without code changes. That’s enough musing me for now.  Back to work!

    Read the article

  • "Unsafe JavaScript attempt to access frame with URL..." error when developing for Facebook Open Grap

    - by Neil Sarkar
    Chrome (or any other webkit browser) throws a ton of these "Unsafe JavaScript attempt to access frame with URL..." when working with the Facebook Open Graph API. It doesn't interfere with actual operation, but it does make the javascript console basically unusable. I'd like to know if there is a way to suppress warnings in the console? Or if there are other solutions you guys can think of, I would really appreciate it. Thanks.

    Read the article

  • How do I fix "java.lang.OutOfMemoryError at sun.misc.Unsafe.allocateMemory(Native Method)"?

    - by Jephir
    I'm making a Java application that uses the Slick library to load images. However, on some computers, I get this error when trying to run the program: Exception in thread "main" java.lang.OutOfMemoryError at sun.misc.Unsafe.allocateMemory(Native Method) at java.nio.DirectByteBuffer.<init>(DirectByteBuffer.java:99) at java.nio.ByteBuffer.allocateDirect(ByteBuffer.java:288) at org.lwjgl.BufferUtils.createByteBuffer(BufferUtils.java:60) at org.newdawn.slick.opengl.PNGImageData.loadImage(PNGImageData.java:692) at org.newdawn.slick.opengl.CompositeImageData.loadImage(CompositeImageData.java:62) at org.newdawn.slick.opengl.CompositeImageData.loadImage(CompositeImageData.java:43) My VM options are: -Djava.library.path=lib -Xms1024M -Xmx1024M -XX:PermSize=256M -XX:MaxPermSize=256M The program loads a few large images (1024 x 768 resolution) at the beginning. Any help to solve this problem would be greatly appreciated.

    Read the article

  • I can't find the cause of an "unchecked or unsafe operations" warning in Java.

    - by Peter
    Hello, as per the title I am struggling to find the cause of an "unchecked or unsafe operations" warning in some code. If I have the following code, it compiles without any warnings: public void test() { Set<String> mySet = new HashSet<String>(); Set<String> myNewSet = mySet; //do stuff } Now, if I change where mySet comes from, specifically as the result of a method call, I get the "unchecked yadda yadda" warning: public void test() { Set<String> myNewSet = this.getSet(); //do stuff } public Set getSet() { Set<String> set = new HashSet<String>(); return set; } I have tried and tried to work out what the problem is and I am completely stumped. The issue is present whether I use Sets or Lists. Why would the Set returned by the getSet method be any different to the Set in the first example? Any help would be greatly appreciated as while the warning isn't the end of the world, it is bugging the hell out of me! :( Regards

    Read the article

  • What is the fastest (possibly unsafe) way to read a byte[]?

    - by Aidiakapi
    I'm working on a server project in C#, and after a TCP message is received, it is parsed, and stored in a byte[] of exact size. (Not a buffer of fixed length, but a byte[] of an absolute length in which all data is stored.) Now for reading this byte[] I'll be creating some wrapper functions (also for compatibility), these are the signatures of all functions I need: public byte ReadByte(); public sbyte ReadSByte(); public short ReadShort(); public ushort ReadUShort(); public int ReadInt(); public uint ReadUInt(); public float ReadFloat(); public double ReadDouble(); public string ReadChars(int length); public string ReadString(); The string is a \0 terminated string, and is probably encoded in ASCII or UTF-8, but I cannot tell that for sure, since I'm not writing the client. The data exists of: byte[] _data; int _offset; Now I can write all those functions manually, like this: public byte ReadByte() { return _data[_offset++]; } public sbyte ReadSByte() { byte r = _data[_offset++]; if (r >= 128) return (sbyte)(r - 256); else return (sbyte)r; } public short ReadShort() { byte b1 = _data[_offset++]; byte b2 = _data[_offset++]; if (b1 >= 128) return (short)(b1 * 256 + b2 - 65536); else return (short)(b1 * 256 + b2); } public short ReadUShort() { byte b1 = _data[_offset++]; return (short)(b1 * 256 + _data[_offset++]); } But I wonder if there's a faster way, not excluding the use of unsafe code, since this seems a little bit too much work for simple processing.

    Read the article

  • Java map / nio / NFS issue causing a VM fault: "a fault occurred in a recent unsafe memory access op

    - by Matthew Bloch
    I have written a parser class for a particular binary format (nfdump if anyone is interested) which uses java.nio's MappedByteBuffer to read through files of a few GB each. The binary format is just a series of headers and mostly fixed-size binary records, which are fed out to the called by calling nextRecord(), which pushes on the state machine, returning null when it's done. It performs well. It works on a development machine. On my production host, it can run for a few minutes or hours, but always seems to throw "java.lang.InternalError: a fault occurred in a recent unsafe memory access operation in compiled Java code", fingering one of the Map.getInt, getShort methods, i.e. a read operation in the map. The uncontroversial (?) code that sets up the map is this: /** Set up the map from the given filename and position */ protected void open() throws IOException { // Set up buffer, is this all the flexibility we'll need? channel = new FileInputStream(file).getChannel(); MappedByteBuffer map1 = channel.map(FileChannel.MapMode.READ_ONLY, 0, channel.size()); map1.load(); // we want the whole thing, plus seems to reduce frequency of crashes? map = map1; // assumes the host writing the files is little-endian (x86), ought to be configurable map.order(java.nio.ByteOrder.LITTLE_ENDIAN); map.position(position); } and then I use the various map.get* methods to read shorts, ints, longs and other sequences of bytes, before hitting the end of the file and closing the map. I've never seen the exception thrown on my development host. But the significant point of difference between my production host and development is that on the former, I am reading sequences of these files over NFS (probably 6-8TB eventually, still growing). On my dev machine, I have a smaller selection of these files locally (60GB), but when it blows up on the production host it's usually well before it gets to 60GB of data. Both machines are running java 1.6.0_20-b02, though the production host is running Debian/lenny, the dev host is Ubuntu/karmic. I'm not convinced that will make any difference. Both machines have 16GB RAM, and are running with the same java heap settings. I take the view that if there is a bug in my code, there is enough of a bug in the JVM not to throw me a proper exception! But I think it is just a particular JVM implementation bug due to interactions between NFS and mmap, possibly a recurrence of 6244515 which is officially fixed. I already tried adding in a "load" call to force the MappedByteBuffer to load its contents into RAM - this seemed to delay the error in the one test run I've done, but not prevent it. Or it could be coincidence that was the longest it had gone before crashing! If you've read this far and have done this kind of thing with java.nio before, what would your instinct be? Right now mine is to rewrite it without nio :)

    Read the article

  • BizTalk and SQL: Alternatives to the SQL receive adapter. Using Msmq to receive SQL data

    - by Leonid Ganeline
    If we have to get data from the SQL database, the standard way is to use a receive port with SQL adapter. SQL receive adapter is a solicit-response adapter. It periodically polls the SQL database with queries. That’s only way it can work. Sometimes it is undesirable. With new WCF-SQL adapter we can use the lightweight approach but still with the same principle, the WCF-SQL adapter periodically solicits the database with queries to check for the new records. Imagine the situation when the new records can appear in very broad time limits, some - in a second interval, others - in the several minutes interval. Our requirement is to process the new records ASAP. That means the polling interval should be near the shortest interval between the new records, a second interval. As a result the most of the poll queries would return nothing and would load the database without good reason. If the database is working under heavy payload, it is very undesirable. Do we have other choices? Sure. We can change the polling to the “eventing”. The good news is the SQL server could issue the event in case of new records with triggers. Got a new record –the trigger event is fired. No new records – no the trigger events – no excessive load to the database. The bad news is the SQL Server doesn’t have intrinsic methods to send the event data outside. For example, we would rather use the adapters that do listen for the data and do not solicit. There are several such adapters-listeners as File, Ftp, SOAP, WCF, and Msmq. But the SQL Server doesn’t have methods to create and save files, to consume the Web-services, to create and send messages in the queue, does it? Can we use the File, FTP, Msmq, WCF adapters to get data from SQL code? Yes, we can. The SQL Server 2005 and 2008 have the possibility to use .NET code inside SQL code. See the SQL Integration. How it works for the Msmq, for example: ·         New record is created, trigger is fired ·         Trigger calls the CLR stored procedure and passes the message parameters to it ·         The CLR stored procedure creates message and sends it to the outgoing queue in the SQL Server computer. ·         Msmq service transfers message to the queue in the BizTalk Server computer. ·         WCF-NetMsmq adapter receives the message from this queue. For the File adapter the idea is the same, the CLR stored procedure creates and stores the file with message, and then the File adapter picks up this file. Using WCF-NetMsmq adapter to get data from SQL I am describing the full set of the deployment and development steps for the case with the WCF-NetMsmq adapter. Development: 1.       Create the .NET code: project, class and method to create and send the message to the MSMQ queue. 2.       Create the SQL code in triggers to call the .NET code. Installation and Deployment: 1.       SQL Server: a.       Register the CLR assembly with .NET (CLR) code b.      Install the MSMQ Services 2.       BizTalk Server: a.       Install the MSMQ Services b.      Create the MSMQ queue c.       Create the WCF-NetMsmq receive port. The detailed description is below. Code .NET code … using System.Xml; using System.Xml.Linq; using System.Xml.Serialization;   //namespace MyCompany.MySolution.MyProject – doesn’t work. The assembly name is MyCompany.MySolution.MyProject // I gave up with the compound namespace. Seems the CLR Integration cannot work with it L. Maybe I’m wrong.     public class Event     {         static public XElement CreateMsg(int par1, int par2, int par3)         {             XNamespace ns = "http://schemas.microsoft.com/Sql/2008/05/TypedPolling/my_storedProc";             XElement xdoc =                 new XElement(ns + "TypedPolling",                     new XElement(ns + "TypedPollingResultSet0",                         new XElement(ns + "TypedPollingResultSet0",                             new XElement(ns + "par1", par1),                             new XElement(ns + "par2", par2),                             new XElement(ns + "par3", par3),                         )                     )                 );             return xdoc;         }     }   //////////////////////////////////////////////////////////////////////// … using System.ServiceModel; using System.ServiceModel.Channels; using System.Transactions; using System.Data; using System.Data.Sql; using System.Data.SqlTypes;   public class MsmqHelper {     [Microsoft.SqlServer.Server.SqlProcedure]     // msmqAddress as "net.msmq://localhost/private/myapp.myqueue";     public static void SendMsg(string msmqAddress, string action, int par1, int par2, int par3)     {         using (TransactionScope scope = new TransactionScope(TransactionScopeOption.Suppress))         {             NetMsmqBinding binding = new NetMsmqBinding(NetMsmqSecurityMode.None);             binding.ExactlyOnce = true;             EndpointAddress address = new EndpointAddress(msmqAddress);               using (ChannelFactory<IOutputChannel> factory = new ChannelFactory<IOutputChannel>(binding, address))             {                 IOutputChannel channel = factory.CreateChannel();                 try                 {                     XElement xe = Event.CreateMsg(par1, par2, par3);                     XmlReader xr = xe.CreateReader();                     Message msg = Message.CreateMessage(MessageVersion.Default, action, xr);                     channel.Send(msg);                     //SqlContext.Pipe.Send(…); // to test                 }                 catch (Exception ex)                 { …                 }             }             scope.Complete();         }     }   SQL code in triggers   -- sp_SendMsg was registered as a name of the MsmqHelper.SendMsg() EXEC sp_SendMsg'net.msmq://biztalk_server_name/private/myapp.myqueue', 'Create', @par1, @par2, @par3   Installation and Deployment On the SQL Server Registering the CLR assembly 1.       Prerequisites: .NET 3.5 SP1 Framework. It could be the issue for the production SQL Server! 2.       For more information, please, see the link http://nielsb.wordpress.com/sqlclrwcf/ 3.       Copy files: >copy “\Windows\Microsoft.net\Framework\v3.0\Windows Communication Foundation\Microsoft.Transactions.Bridge.dll” “\Program Files\Reference Assemblies\Microsoft\Framework\v3.0 \Microsoft.Transactions.Bridge.dll” If your machine is a 64-bit, run two commands: >copy “\Windows\Microsoft.net\Framework\v3.0\Windows Communication Foundation\Microsoft.Transactions.Bridge.dll” “\Program Files (x86)\Reference Assemblies\Microsoft\Framework\v3.0 \Microsoft.Transactions.Bridge.dll” >copy “\Windows\Microsoft.net\Framework64\v3.0\Windows Communication Foundation\Microsoft.Transactions.Bridge.dll” “\Program Files\Reference Assemblies\Microsoft\Framework\v3.0 \Microsoft.Transactions.Bridge.dll” 4.       Execute the SQL code to register the .NET assemblies: -- For x64 OS: CREATE ASSEMBLY SMdiagnostics AUTHORIZATION dbo FROM 'C:\Windows\Microsoft.NET\Framework\v3.0\Windows Communication Foundation\SMdiagnostics.dll' WITH permission_set = unsafe CREATE ASSEMBLY [System.Web] AUTHORIZATION dbo FROM 'C:\Windows\Microsoft.NET\Framework64\v2.0.50727\System.Web.dll' WITH permission_set = unsafe CREATE ASSEMBLY [System.Messaging] AUTHORIZATION dbo FROM 'C:\Windows\Microsoft.NET\Framework\v2.0.50727\System.Messaging.dll' WITH permission_set = unsafe CREATE ASSEMBLY [System.ServiceModel] AUTHORIZATION dbo FROM 'C:\Program Files (x86)\Reference Assemblies\Microsoft\Framework\v3.0\System.ServiceModel.dll' WITH permission_set = unsafe CREATE ASSEMBLY [System.Xml.Linq] AUTHORIZATION dbo FROM 'C:\Program Files\Reference Assemblies\Microsoft\Framework\v3.5\System.Xml.Linq.dll' WITH permission_set = unsafe   -- For x32 OS: --CREATE ASSEMBLY SMdiagnostics AUTHORIZATION dbo FROM 'C:\Windows\Microsoft.NET\Framework\v3.0\Windows Communication Foundation\SMdiagnostics.dll' WITH permission_set = unsafe --CREATE ASSEMBLY [System.Web] AUTHORIZATION dbo FROM 'C:\Windows\Microsoft.NET\Framework\v2.0.50727\System.Web.dll' WITH permission_set = unsafe --CREATE ASSEMBLY [System.Messaging] AUTHORIZATION dbo FROM 'C:\Windows\Microsoft.NET\Framework\v2.0.50727\System.Messaging.dll' WITH permission_set = unsafe --CREATE ASSEMBLY [System.ServiceModel] AUTHORIZATION dbo FROM 'C:\Program Files\Reference Assemblies\Microsoft\Framework\v3.0\System.ServiceModel.dll' WITH permission_set = unsafe 5.       Register the assembly with the external stored procedure: CREATE ASSEMBLY [HelperClass] AUTHORIZATION dbo FROM ’<FilePath>MyCompany.MySolution.MyProject.dll' WITH permission_set = unsafe where the <FilePath> - the path of the file on this machine! 6. Create the external stored procedure CREATE PROCEDURE sp_SendMsg (        @msmqAddress nvarchar(100),        @Action NVARCHAR(50),        @par1 int,        @par2 int,        @par3 int ) AS EXTERNAL NAME HelperClear.MsmqHelper.SendMsg   Installing the MSMQ Services 1.       Check if the MSMQ service is NOT installed. To check:  Start / Administrative Tools / Computer Management, on the left pane open the “Services and Applications”, search to the “Message Queuing”. If you cannot see it, follow next steps. 2.       Start / Control Panel / Programs and Features 3.       Click “Turn Windows Features on or off” 4.       Click Features, click “Add Features” 5.       Scroll down the feature list; open the “Message Queuing” / “Message Queuing Services”; and check the “Message Queuing Server” option  6.       Click Next; Click Install; wait to the successful finish of the installation Creating the MSMQ queue We don’t need to create the queue on the “sender” side. On the BizTalk Server Installing the MSMQ Services The same is as for the SQL Server. Creating the MSMQ queue 1.       Start / Administrative Tools / Computer Management, on the left pane open the “Services and Applications”, open the “Message Queuing”, and open the “Private Queues”. 2.       Right-click the “Private Queues”; choose New; choose “Private Queue”. 3.       Type the Queue name as ’myapp.myqueue'; check the “Transactional” option. Creating the WCF-NetMsmq receive port I will not go through this step in all details. It is straightforward. URI for this receive location should be 'net.msmq://localhost/private/myapp.myqueue'. Notes ·         The biggest problem is usually on the step the “Registering the CLR assembly”. It is hard to predict where are the assemblies from the assembly list, what version should be used, x86 or x64. It is pity of such “rude” integration of the SQL with .NET. ·         In couple cases the new WCF-NetMsmq port was not able to work with the queue. Try to replace the WCF- NetMsmq port with the WCF-Custom port with netMsmqBinding. It was working fine for me. ·         To test how messages go through the queue you can turn on the Journal /Enabled option for the queue. I used the QueueExplorer utility to look to the messages in Journal. The Computer Management can also show the messages but it shows only small part of the message body and in the weird format. The QueueExplorer can do the better job; it shows the whole body and Xml messages are in good color format.

    Read the article

  • From Binary to Data Structures

    - by Cédric Menzi
    Table of Contents Introduction PE file format and COFF header COFF file header BaseCoffReader Byte4ByteCoffReader UnsafeCoffReader ManagedCoffReader Conclusion History This article is also available on CodeProject Introduction Sometimes, you want to parse well-formed binary data and bring it into your objects to do some dirty stuff with it. In the Windows world most data structures are stored in special binary format. Either we call a WinApi function or we want to read from special files like images, spool files, executables or may be the previously announced Outlook Personal Folders File. Most specifications for these files can be found on the MSDN Libarary: Open Specification In my example, we are going to get the COFF (Common Object File Format) file header from a PE (Portable Executable). The exact specification can be found here: PECOFF PE file format and COFF header Before we start we need to know how this file is formatted. The following figure shows an overview of the Microsoft PE executable format. Source: Microsoft Our goal is to get the PE header. As we can see, the image starts with a MS-DOS 2.0 header with is not important for us. From the documentation we can read "...After the MS DOS stub, at the file offset specified at offset 0x3c, is a 4-byte...". With this information we know our reader has to jump to location 0x3c and read the offset to the signature. The signature is always 4 bytes that ensures that the image is a PE file. The signature is: PE\0\0. To prove this we first seek to the offset 0x3c, read if the file consist the signature. So we need to declare some constants, because we do not want magic numbers.   private const int PeSignatureOffsetLocation = 0x3c; private const int PeSignatureSize = 4; private const string PeSignatureContent = "PE";   Then a method for moving the reader to the correct location to read the offset of signature. With this method we always move the underlining Stream of the BinaryReader to the start location of the PE signature.   private void SeekToPeSignature(BinaryReader br) { // seek to the offset for the PE signagure br.BaseStream.Seek(PeSignatureOffsetLocation, SeekOrigin.Begin); // read the offset int offsetToPeSig = br.ReadInt32(); // seek to the start of the PE signature br.BaseStream.Seek(offsetToPeSig, SeekOrigin.Begin); }   Now, we can check if it is a valid PE image by reading of the next 4 byte contains the content PE.   private bool IsValidPeSignature(BinaryReader br) { // read 4 bytes to get the PE signature byte[] peSigBytes = br.ReadBytes(PeSignatureSize); // convert it to a string and trim \0 at the end of the content string peContent = Encoding.Default.GetString(peSigBytes).TrimEnd('\0'); // check if PE is in the content return peContent.Equals(PeSignatureContent); }   With this basic functionality we have a good base reader class to try the different methods of parsing the COFF file header. COFF file header The COFF header has the following structure: Offset Size Field 0 2 Machine 2 2 NumberOfSections 4 4 TimeDateStamp 8 4 PointerToSymbolTable 12 4 NumberOfSymbols 16 2 SizeOfOptionalHeader 18 2 Characteristics If we translate this table to code, we get something like this:   [StructLayout(LayoutKind.Sequential, CharSet = CharSet.Unicode)] public struct CoffHeader { public MachineType Machine; public ushort NumberOfSections; public uint TimeDateStamp; public uint PointerToSymbolTable; public uint NumberOfSymbols; public ushort SizeOfOptionalHeader; public Characteristic Characteristics; } BaseCoffReader All readers do the same thing, so we go to the patterns library in our head and see that Strategy pattern or Template method pattern is sticked out in the bookshelf. I have decided to take the template method pattern in this case, because the Parse() should handle the IO for all implementations and the concrete parsing should done in its derived classes.   public CoffHeader Parse() { using (var br = new BinaryReader(File.Open(_fileName, FileMode.Open, FileAccess.Read, FileShare.Read))) { SeekToPeSignature(br); if (!IsValidPeSignature(br)) { throw new BadImageFormatException(); } return ParseInternal(br); } } protected abstract CoffHeader ParseInternal(BinaryReader br);   First we open the BinaryReader, seek to the PE signature then we check if it contains a valid PE signature and rest is done by the derived implementations. Byte4ByteCoffReader The first solution is using the BinaryReader. It is the general way to get the data. We only need to know which order, which data-type and its size. If we read byte for byte we could comment out the first line in the CoffHeader structure, because we have control about the order of the member assignment.   protected override CoffHeader ParseInternal(BinaryReader br) { CoffHeader coff = new CoffHeader(); coff.Machine = (MachineType)br.ReadInt16(); coff.NumberOfSections = (ushort)br.ReadInt16(); coff.TimeDateStamp = br.ReadUInt32(); coff.PointerToSymbolTable = br.ReadUInt32(); coff.NumberOfSymbols = br.ReadUInt32(); coff.SizeOfOptionalHeader = (ushort)br.ReadInt16(); coff.Characteristics = (Characteristic)br.ReadInt16(); return coff; }   If the structure is as short as the COFF header here and the specification will never changed, there is probably no reason to change the strategy. But if a data-type will be changed, a new member will be added or ordering of member will be changed the maintenance costs of this method are very high. UnsafeCoffReader Another way to bring the data into this structure is using a "magically" unsafe trick. As above, we know the layout and order of the data structure. Now, we need the StructLayout attribute, because we have to ensure that the .NET Runtime allocates the structure in the same order as it is specified in the source code. We also need to enable "Allow unsafe code (/unsafe)" in the project's build properties. Then we need to add the following constructor to the CoffHeader structure.   [StructLayout(LayoutKind.Sequential, CharSet = CharSet.Unicode)] public struct CoffHeader { public CoffHeader(byte[] data) { unsafe { fixed (byte* packet = &data[0]) { this = *(CoffHeader*)packet; } } } }   The "magic" trick is in the statement: this = *(CoffHeader*)packet;. What happens here? We have a fixed size of data somewhere in the memory and because a struct in C# is a value-type, the assignment operator = copies the whole data of the structure and not only the reference. To fill the structure with data, we need to pass the data as bytes into the CoffHeader structure. This can be achieved by reading the exact size of the structure from the PE file.   protected override CoffHeader ParseInternal(BinaryReader br) { return new CoffHeader(br.ReadBytes(Marshal.SizeOf(typeof(CoffHeader)))); }   This solution is the fastest way to parse the data and bring it into the structure, but it is unsafe and it could introduce some security and stability risks. ManagedCoffReader In this solution we are using the same approach of the structure assignment as above. But we need to replace the unsafe part in the constructor with the following managed part:   [StructLayout(LayoutKind.Sequential, CharSet = CharSet.Unicode)] public struct CoffHeader { public CoffHeader(byte[] data) { IntPtr coffPtr = IntPtr.Zero; try { int size = Marshal.SizeOf(typeof(CoffHeader)); coffPtr = Marshal.AllocHGlobal(size); Marshal.Copy(data, 0, coffPtr, size); this = (CoffHeader)Marshal.PtrToStructure(coffPtr, typeof(CoffHeader)); } finally { Marshal.FreeHGlobal(coffPtr); } } }     Conclusion We saw that we can parse well-formed binary data to our data structures using different approaches. The first is probably the clearest way, because we know each member and its size and ordering and we have control about the reading the data for each member. But if add member or the structure is going change by some reason, we need to change the reader. The two other solutions use the approach of the structure assignment. In the unsafe implementation we need to compile the project with the /unsafe option. We increase the performance, but we get some security risks.

    Read the article

  • Java ThreadPoolExecutor getting stuck while using ArrayBlockingQueue

    - by Ravi Rao
    Hi, I'm working on some application and using ThreadPoolExecutor for handling various tasks. ThreadPoolExecutor is getting stuck after some duration. To simulate this in a simpler environment, I've written a simple code where I'm able to simulate the issue. import java.util.concurrent.ArrayBlockingQueue; import java.util.concurrent.RejectedExecutionHandler; import java.util.concurrent.ThreadPoolExecutor; import java.util.concurrent.TimeUnit; public class MyThreadPoolExecutor { private int poolSize = 10; private int maxPoolSize = 50; private long keepAliveTime = 10; private ThreadPoolExecutor threadPool = null; private final ArrayBlockingQueue&lt;Runnable&gt; queue = new ArrayBlockingQueue&lt;Runnable&gt;( 100000); public MyThreadPoolExecutor() { threadPool = new ThreadPoolExecutor(poolSize, maxPoolSize, keepAliveTime, TimeUnit.SECONDS, queue); threadPool.setRejectedExecutionHandler(new RejectedExecutionHandler() { @Override public void rejectedExecution(Runnable runnable, ThreadPoolExecutor threadPoolExecutor) { System.out .println(&quot;Execution rejected. Please try restarting the application.&quot;); } }); } public void runTask(Runnable task) { threadPool.execute(task); } public void shutDown() { threadPool.shutdownNow(); } public ThreadPoolExecutor getThreadPool() { return threadPool; } public void setThreadPool(ThreadPoolExecutor threadPool) { this.threadPool = threadPool; } public static void main(String[] args) { MyThreadPoolExecutor mtpe = new MyThreadPoolExecutor(); for (int i = 0; i &lt; 1000; i++) { final int j = i; mtpe.runTask(new Runnable() { @Override public void run() { System.out.println(j); } }); } } } Try executing this code a few times. It normally print outs the number on console and when all threads end, it exists. But at times, it finished all task and then is not getting terminated. The thread dump is as follows: MyThreadPoolExecutor [Java Application] MyThreadPoolExecutor at localhost:2619 (Suspended) Daemon System Thread [Attach Listener] (Suspended) Daemon System Thread [Signal Dispatcher] (Suspended) Daemon System Thread [Finalizer] (Suspended) Object.wait(long) line: not available [native method] ReferenceQueue&lt;T&gt;.remove(long) line: not available ReferenceQueue&lt;T&gt;.remove() line: not available Finalizer$FinalizerThread.run() line: not available Daemon System Thread [Reference Handler] (Suspended) Object.wait(long) line: not available [native method] Reference$Lock(Object).wait() line: 485 Reference$ReferenceHandler.run() line: not available Thread [pool-1-thread-1] (Suspended) Unsafe.park(boolean, long) line: not available [native method] LockSupport.park(Object) line: not available AbstractQueuedSynchronizer$ConditionObject.await() line: not available ArrayBlockingQueue&lt;E&gt;.take() line: not available ThreadPoolExecutor.getTask() line: not available ThreadPoolExecutor$Worker.run() line: not available Thread.run() line: not available Thread [pool-1-thread-2] (Suspended) Unsafe.park(boolean, long) line: not available [native method] LockSupport.park(Object) line: not available AbstractQueuedSynchronizer$ConditionObject.await() line: not available ArrayBlockingQueue&lt;E&gt;.take() line: not available ThreadPoolExecutor.getTask() line: not available ThreadPoolExecutor$Worker.run() line: not available Thread.run() line: not available Thread [pool-1-thread-3] (Suspended) Unsafe.park(boolean, long) line: not available [native method] LockSupport.park(Object) line: not available AbstractQueuedSynchronizer$ConditionObject.await() line: not available ArrayBlockingQueue&lt;E&gt;.take() line: not available ThreadPoolExecutor.getTask() line: not available ThreadPoolExecutor$Worker.run() line: not available Thread.run() line: not available Thread [pool-1-thread-4] (Suspended) Unsafe.park(boolean, long) line: not available [native method] LockSupport.park(Object) line: not available AbstractQueuedSynchronizer$ConditionObject.await() line: not available ArrayBlockingQueue&lt;E&gt;.take() line: not available ThreadPoolExecutor.getTask() line: not available ThreadPoolExecutor$Worker.run() line: not available Thread.run() line: not available Thread [pool-1-thread-6] (Suspended) Unsafe.park(boolean, long) line: not available [native method] LockSupport.park(Object) line: not available AbstractQueuedSynchronizer$ConditionObject.await() line: not available ArrayBlockingQueue&lt;E&gt;.take() line: not available ThreadPoolExecutor.getTask() line: not available ThreadPoolExecutor$Worker.run() line: not available Thread.run() line: not available Thread [pool-1-thread-8] (Suspended) Unsafe.park(boolean, long) line: not available [native method] LockSupport.park(Object) line: not available AbstractQueuedSynchronizer$ConditionObject.await() line: not available ArrayBlockingQueue&lt;E&gt;.take() line: not available ThreadPoolExecutor.getTask() line: not available ThreadPoolExecutor$Worker.run() line: not available Thread.run() line: not available Thread [pool-1-thread-5] (Suspended) Unsafe.park(boolean, long) line: not available [native method] LockSupport.park(Object) line: not available AbstractQueuedSynchronizer$ConditionObject.await() line: not available ArrayBlockingQueue&lt;E&gt;.take() line: not available ThreadPoolExecutor.getTask() line: not available ThreadPoolExecutor$Worker.run() line: not available Thread.run() line: not available Thread [pool-1-thread-10] (Suspended) Unsafe.park(boolean, long) line: not available [native method] LockSupport.park(Object) line: not available AbstractQueuedSynchronizer$ConditionObject.await() line: not available ArrayBlockingQueue&lt;E&gt;.take() line: not available ThreadPoolExecutor.getTask() line: not available ThreadPoolExecutor$Worker.run() line: not available Thread.run() line: not available Thread [pool-1-thread-9] (Suspended) Unsafe.park(boolean, long) line: not available [native method] LockSupport.park(Object) line: not available AbstractQueuedSynchronizer$ConditionObject.await() line: not available ArrayBlockingQueue&lt;E&gt;.take() line: not available ThreadPoolExecutor.getTask() line: not available ThreadPoolExecutor$Worker.run() line: not available Thread.run() line: not available Thread [pool-1-thread-7] (Suspended) Unsafe.park(boolean, long) line: not available [native method] LockSupport.park(Object) line: not available AbstractQueuedSynchronizer$ConditionObject.await() line: not available ArrayBlockingQueue&lt;E&gt;.take() line: not available ThreadPoolExecutor.getTask() line: not available ThreadPoolExecutor$Worker.run() line: not available Thread.run() line: not available Thread [DestroyJavaVM] (Suspended) C:\Program Files\Java\jre1.6.0_07\bin\javaw.exe (Jun 17, 2010 10:42:33 AM) In my actual application,ThreadPoolExecutor threads go in this state and then it stops responding. Regards, Ravi Rao

    Read the article

  • Can't upload project to PPA using Quickly

    - by RobinJ
    I can't get Quickly to upload my project into my PPA. I've set up my PGP key and used it so sign the code of conduct, and the PPA exists. I don't know what other usefull information I can supply. robin@RobinJ:~/Ubuntu One/Python/gtkreddit$ quickly share --ppa robinj/gtkredditGet Launchpad Settings Launchpad connection is ok gpg: WARNING: unsafe permissions on configuration file `/home/robin/.gnupg/gpg.conf' gpg: WARNING: unsafe enclosing directory permissions on configuration file `/home/robin/.gnupg/gpg.conf' gpg: WARNING: unsafe permissions on configuration file `/home/robin/.gnupg/gpg.conf' gpg: WARNING: unsafe enclosing directory permissions on configuration file `/home/robin/.gnupg/gpg.conf' Traceback (most recent call last): File "/usr/share/quickly/templates/ubuntu-application/share.py", line 138, in <module> license.licensing() File "/usr/share/quickly/templates/ubuntu-application/license.py", line 284, in licensing {'translatable': 'yes'}) File "/usr/share/quickly/templates/ubuntu-application/internal/quicklyutils.py", line 166, in change_xml_elem xml_tree.find(parent_node).insert(0, new_node) AttributeError: 'NoneType' object has no attribute 'insert' ERROR: share command failed Aborting I reported this as a bug on Launchpad, because I assume that it is a bug. If you know a quick workaround, please let me know. https://bugs.launchpad.net/ubuntu/+source/quickly/+bug/1018138

    Read the article

  • How can I get the palette of an 8-bit surface in SDL.NET/Tao.SDL?

    - by lolmaster
    I'm looking to get the palette of an 8-bit surface in SDL.NET if possible, or (more than likely) using Tao.SDL. This is because I want to do palette swapping with the palette directly, instead of blitting surfaces together to replace colours like how you would do it with a 32-bit surface. I've gotten the SDL_Surface and the SDL_PixelFormat, however when I go to get the palette in the same way, I get a System.ExecutionEngineException: private Tao.Sdl.Sdl.SDL_Palette GetPalette(Surface surf) { // Get surface. Tao.Sdl.Sdl.SDL_Surface sdlSurface = (Tao.Sdl.Sdl.SDL_Surface)System.Runtime.InteropServices.Marshal.PtrToStructure(surf.Handle, typeof(Tao.Sdl.Sdl.SDL_Surface)); // Get pixel format. Tao.Sdl.Sdl.SDL_PixelFormat pixelFormat = (Tao.Sdl.Sdl.SDL_PixelFormat)System.Runtime.InteropServices.Marshal.PtrToStructure(sdlSurface.format, typeof(Tao.Sdl.Sdl.SDL_PixelFormat)); // Execution exception here. Tao.Sdl.Sdl.SDL_Palette palette = (Tao.Sdl.Sdl.SDL_Palette)System.Runtime.InteropServices.Marshal.PtrToStructure(pixelFormat.palette, typeof(Tao.Sdl.Sdl.SDL_Palette)); return palette; } When I used unsafe code to get the palette, I got a compile time error: "Cannot take the address of, get the size of, or declare a pointer to a managed type ('Tao.Sdl.Sdl.SDL_Palette')". My unsafe code to get the palette was this: unsafe { Tao.Sdl.Sdl.SDL_Palette* pal = (Tao.Sdl.Sdl.SDL_Palette*)pixelFormat.palette; } From what I've read, a managed type in this case is when a structure has some sort of reference inside it as a field. The SDL_Palette structure happens to have an array of SDL_Color's, so I'm assuming that's the reference type that is causing issues. However I'm still not sure how to work around that to get the underlying palette. So if anyone knows how to get the palette from an 8-bit surface, whether it's through safe or unsafe code, the help would be greatly appreciated.

    Read the article

  • How to marshal int* in C#?

    - by MartyIX
    Hi, I would like to call this method in unmanaged library: void __stdcall GetConstraints( unsigned int* puiMaxWidth, unsigned int* puiMaxHeight, unsigned int* puiMaxBoxes ); My solution: Delegate definition: [UnmanagedFunctionPointer(CallingConvention.StdCall)] private delegate void GetConstraintsDel(UIntPtr puiMaxWidth, UIntPtr puiMaxHeight, UIntPtr puiMaxBoxes); The call of the method: // PLUGIN NAME GetConstraintsDel getConstraints = (GetConstraintsDel)Marshal.GetDelegateForFunctionPointer(pAddressOfFunctionToCall, typeof(GetConstraintsDel)); uint maxWidth, maxHeight, maxBoxes; unsafe { UIntPtr a = new UIntPtr(&maxWidth); UIntPtr b = new UIntPtr(&maxHeight); UIntPtr c = new UIntPtr(&maxBoxes); getConstraints(a, b, c); } It works but I have to allow "unsafe" flag. Is there a solution without unsafe code? Or is this solution ok? I don't quite understand the implications of setting the project with unsafe flag. Thanks for help!

    Read the article

  • Emacs - Error when calling (server-start)

    - by Jonas Gorauskas
    I am currently using GNU Emacs 23.0.93.1 in Windows Vista SP1. In my .emacs file I make a call to (server-start) and that is causing an error with the message The directory ~/.emacs.d/server is unsafe. Has anyone seen this and know a fix or workaround? ... other than leaving server turned off ;) Here is the stack trace: Debugger entered--Lisp error: (error "The directory ~/.emacs.d/server is unsafe") signal(error ("The directory ~/.emacs.d/server is unsafe")) error("The directory %s is unsafe" "~/.emacs.d/server") server-ensure-safe-dir("~\\.emacs.d\\server\\") server-start(nil) call-interactively(server-start t nil) execute-extended-command(nil) call-interactively(execute-extended-command nil nil)

    Read the article

  • What is the best strategy for transforming unicode strings into filenames?

    - by David Cowden
    I have a bunch (thousands) of resources in an RDF/XML file. I am writing a certain subset of the resources to files -- one file for each, and I'm using the resource's title property as the file name. However, the titles are every day article, website, and blog post titles, so they contain characters unsafe for a URI (the necessary step for constructing a valid file path). I know of the Jersey UriBuilder but I can't quite get it to work for my needs as I detailed in a different question on SO. Some possibilities I have considered are: Since each resource should also have an associated URL, I could try to use the name of the file on the server. The down side of this is sometimes people don't name their content logically and I think the title of an article better reflects the content that will be in each text file. Construct a white list of valid characters and parse the string myself defining substitutions for unsafe characters. The downside of this is the result could be just as unreadable as the former solution because presumably the content creators went through a similar process when placing the files on their server. Choose a more generic naming scheme, place the title in the text file along with the other attributes, and tell my boss to live with it. So my question here is, what methods work well for dealing with a scenario where you need to construct file names out of strings with potentially unsafe characters? Is there a solution that better fills out my constraints?

    Read the article

  • What if I dismount main volume, where the Windows is installed

    - by ST3
    I'm writing permanent file deletion tool and accessing raw disk clusters. Since Windows Vista writing into raw disk is a bit more complicated. I have tried on my external data device first and worked fine, however one of the steps was dismounting of the volume, not sure if it is a good idea to dismount main volume where the Windows are. Want to ask that is possible consequences and if it safe/unsafe/very unsafe.

    Read the article

  • How to stop auto encoding of <%= strings %> with Rails 3.0beta

    - by christophercotton
    I'm using rails 3.0beta3. In my index.html.erb and in my index.js.erb, if I have: <%= "string with unsafe characters' like <" %> It will automatically be encoded to: string with unsafe characters&quot; like &amp; just the same as if I had used: <%=h "string with unsafe characters' like <" %> How do I get it to stop? I have stored some short bits of JavaScript that I need to insert into the template without it automatically encoding the string?

    Read the article

< Previous Page | 1 2 3 4 5 6 7 8 9 10 11 12  | Next Page >