f.Modified = modified.In(timeZone(msdosModified.Sub(modified)))
		}
	}

	// Assume that uncompressed size 2³²-1 could plausibly happen in
	// an old zip32 file that was sharding inputs into the largest chunks
	// possible (or is just malicious; search the web for 42.zip).
	// If needUSize is true still, it means we didn't see a zip64 extension.
	// As long as the compressed size is not also 2³²-1 (implausible)

to disable optimizations that are invalid for Go.
</p>

<p>
Note that the prohibition on introducing data races
does not apply if the compiler can prove that the races
do not affect correct execution on the target platform.
For example, on essentially all CPUs, it is valid to rewrite
</p>

<pre>
n := 0
for i := 0; i < m; i++ {
	n += *shared
}
</pre>

into:

<pre>
n := 0

  @Override
  public abstract UnmodifiableIterator<E> descendingIterator();

  /** Returns the position of an element within the set, or -1 if not present. */
  abstract int indexOf(@CheckForNull Object target);

  /*
   * This class is used to serialize all ImmutableSortedSet instances,
   * regardless of implementation type. It captures their "logical contents"

  // for a GetCapacityRequest with topology and parameters that match the
  // previous fields.
  //
  // This is defined since CSI spec 1.4.0 as the largest size
  // that may be used in a
  // CreateVolumeRequest.capacity_range.required_bytes field to
  // create a volume with the same parameters as those in
  // GetCapacityRequest. The corresponding value in the Kubernetes

### Other notable changes

* Test x509 intermediates correctly ([#34524](https://github.com/kubernetes/kubernetes/pull/34524), [@liggitt](https://github.com/liggitt))
* Remove headers that are unnecessary for proxy target ([#34076](https://github.com/kubernetes/kubernetes/pull/34076), [@mbohlool](https://github.com/mbohlool))



# v1.3.8

     *
     * Two other points:
     *
     * 1. We'll want to use something like @PolyNull once we can make that work for the various
     * platforms we target.
     *
     * 2. Kotlin's Map type has a getOrDefault method that accepts and returns a "plain V," in
     * contrast to the "V?" type that we're using. As a result, Kotlin sees a conflict between the

	pendingCount    uint64
	failedCount     uint64
	replTargetStats map[string]replTargetSizeSummary
	tiers           map[string]tierStats
}

// replTargetSizeSummary holds summary of replication stats by target
type replTargetSizeSummary struct {
	replicatedSize  int64
	replicatedCount int64
	pendingSize     int64
	failedSize      int64
	pendingCount    uint64
	failedCount     uint64
}

* Updated Go to 1.7 ([#28742](https://github.com/kubernetes/kubernetes/pull/28742), [@jessfraz](https://github.com/jessfraz))
* HPA: fixed wrong count for target replicas calculations ([#34821](https://github.com/kubernetes/kubernetes/pull/34821)). ([#34955](https://github.com/kubernetes/kubernetes/pull/34955), [@jszczepkowski](https://github.com/jszczepkowski))

	}

	// In case of SSE-C, we have to decrypt the OEK using the client-provided key.
	// In case of a SSE-C server-side copy, the client might provide two keys,
	// one for the source and one for the target. This reader is the source.
	var ssecClientKey []byte
	if crypto.SSEC.IsEncrypted(oi.UserDefined) {
		if copySource && crypto.SSECopy.IsRequested(h) {
			key, err := crypto.SSECopy.ParseHTTP(h)
			if err != nil {

     *
     * Two other points:
     *
     * 1. We'll want to use something like @PolyNull once we can make that work for the various
     * platforms we target.
     *
     * 2. Kotlin's Map type has a getOrDefault method that accepts and returns a "plain V," in
     * contrast to the "V?" type that we're using. As a result, Kotlin sees a conflict between the