The arguments to a log call are always evaluated, even if the log event is discarded.
If possible, defer computation so that it happens only if the value is actually logged.
For example, consider the call

	slog.Info("starting request", "url", r.URL.String())  // may compute String unnecessarily

The URL.String method will be called even if the logger discards Info-level events.
Instead, pass the URL directly:

    TFE_Executor*, TF_Status* status);

// When an error happens, any pending operations are discarded, and newly issued
// ops return an error. This call clears the error state and re-enables
// execution of newly issued ops.
//
// Note that outputs of discarded ops remain in a corrupt state and should not
// be used for future calls.

//
// It calls the getPackage function for the package path string of
// each dependency (perhaps indirect) that it encounters in the
// encoding. If the function returns nil, the fact is discarded.
//
// This function is preferred over [NewDecoder] when the client is
// capable of efficient look-up of packages by package path.
func NewDecoderFunc(pkg *types.Package, getPackage GetPackageFunc) *Decoder {

[[sec:partial-rewrite]]
==== Partial rewrite of existing content

A "complete rewrite" causes all existing content to be discarded, thereby losing any changes made directly in the IDE. The `beforeMerged` hook makes it possible to overwrite just certain parts of the existing content. The following example removes all existing dependencies from the `Module` domain object:

 * the configuration cache, so that its outputs in turn become fixed. The node can then be discarded from the graph and replaced with its (now fixed) outputs.</p>
 */
public interface ProviderInternal<T> extends Provider<T>, ValueSupplier, TaskDependencyContainer, EvaluationContext.EvaluationOwner {
    /**

    }

    /**
     * Is this node ready to execute or discard (eg because a dependency has failed)?
     */
    public boolean allDependenciesComplete() {
        return state == ExecutionState.SHOULD_RUN && dependenciesState != DependenciesState.NOT_COMPLETE;
    }

    /**
     * Can this node execute or should it be discarded? Should only be called when {@link #allDependenciesComplete()} returns true.
     */

The property getter method must be `public` or `protected`. The method can be `abstract` or, in cases where this isn't possible, can have a dummy method body.
The method body is discarded.

Here is an example that shows a task type that receives a two services via property getter methods:

====
[source.multi-language-sample,java]
.Download.java
----

	// at least one buffer of work and reduces contention on the
	// global work lists.
	//
	// wbuf1 is always the buffer we're currently pushing to and
	// popping from and wbuf2 is the buffer that will be discarded
	// next.
	//
	// Invariant: Both wbuf1 and wbuf2 are nil or neither are.
	wbuf1, wbuf2 *workbuf

	// Bytes marked (blackened) on this gcWork. This is aggregated
	// into work.bytesMarked by dispose.

[[sec:partial-overwrite]]
==== Partial overwrite of existing content

// additional pages after the segment end.
// The function returns a slice of pointers to the headers in the input
// slice, which are valid only while phdrs is not modified or discarded.
func ProgramHeadersForMapping(phdrs []elf.ProgHeader, mapOff, mapSz uint64) []*elf.ProgHeader {
	const (
		// pageSize defines the virtual memory page size used by the loader. This