``MinIO`` community welcomes your contribution. To make the process as seamless as possible, we recommend you read this contribution guide.

## Development Workflow

# Tensorflow SavedModel to StableHLO (tf-to-stablehlo-translate)

Converts TensorFlow models (SavedModel or MLIR module) to StableHLO MLIR
modules, preserving model structure and signatures. It enables seamless
integration of TensorFlow models into MLIR-based compiler frameworks for further
optimization and deployment.

## C++ APIs

```bash
tf-to-stablehlo-translate \
    --input-path=/path/to/model \

# to a subdirectory of $WORK/share. This mimics the directory structure reported
# in https://go.dev/issue/57754.
#
# Symlink everything else to the original $GOROOT to avoid needless copying work.

mkdir $WORK/lib/goroot
mkdir $WORK/share/goroot
symlink $WORK/share/goroot/src -> $GOROOT${/}src
symlink $WORK/lib/goroot/src -> ../../share/goroot/src
symlink $WORK/lib/goroot/pkg -> $GOROOT${/}pkg

func (*T4) M() {}

type T5 struct {
	T4
}

// ----------------------------------------------------------------------------
// Recursive type declarations must not lead to endless recursion.

type U1 struct {
	*U1
}

// U1.M should appear as method of U1.
func (*U1) M() {}

type U2 struct {
	*U3
}

	"rows":        contentTypePlain,
	"rowspan":     contentTypePlain,
	"sandbox":     contentTypeUnsafe,
	"spellcheck":  contentTypePlain,
	"scope":       contentTypePlain,
	"scoped":      contentTypePlain,
	"seamless":    contentTypePlain,
	"selected":    contentTypePlain,
	"shape":       contentTypePlain,
	"size":        contentTypePlain,
	"sizes":       contentTypePlain,
	"span":        contentTypePlain,

| <<xctest_plugin.adoc#xctest_plugin,XCTest>>
| Adds a target representing the XCTest bundle to the project file.
| The tested target is also configured to enable a seamless workflow from the IDE.
|===

[[sec:xcode_usage]]
== Usage

.Applying the Xcode Plugin
====
include::sample[dir="snippets/ide/xcode/kotlin",files="build.gradle.kts[tags=apply-plugin]"]

)

var (
	uploadKubeadmConfigLongDesc = fmt.Sprintf(cmdutil.LongDesc(`
		Upload the kubeadm ClusterConfiguration to a ConfigMap called %s in the %s namespace.
		This enables correct configuration of system components and a seamless user experience when upgrading.

		Alternatively, you can use kubeadm config.
		`), kubeadmconstants.KubeadmConfigConfigMap, metav1.NamespaceSystem)

	uploadKubeadmConfigExample = cmdutil.Examples(`

transformations will operate on.

We make use of two different dialects to model TensorFlow graphs in MLIR: first
the `tf_executor` dialect that represents the execution model of the TensorFlow
executor (e.g. control dependencies, deadness propagation) and the `tf` dialect
which represent the regular operations in a TensorFlow graph (the ones that
don’t have special contract with the executor).

  // node in `should_revisit` denotes that the deadness flowing out from any
  // output from said node may have changed.  This is fine; only switches
  // propagate different deadness along different output edges, and since the
  // delta is solely due to the input *values* (and not input deadness), the
  // delta should not change in the second iteration.
  std::vector<bool> should_revisit;

- Handle Generic Ephemeral Volumes properly in the node limits scheduler filter and the kubelet `hostPath` check. ([#100482](https://github.com/kubernetes/kubernetes/pull/100482), [@pohly](https://github.com/pohly))