Merge branch 'main' into ml-allow-null-function-name

Author: jonathan-buttner

docs/changelog/125479.yaml

@@ -0,0 +1,6 @@
+pr: 125479
+summary: ES|QL - Allow full text functions to be used in STATS
+area: ES|QL
+type: enhancement
+issues:
+ - 125481

docs/changelog/127355.yaml

@@ -0,0 +1,5 @@
+pr: 127355
+summary: '`text ==` and `text !=` pushdown'
+area: ES|QL
+type: enhancement
+issues: []

docs/changelog/127877.yaml

@@ -0,0 +1,5 @@
+pr: 127877
+summary: Check hidden frames in entitlements
+area: Infra/Core
+type: bug
+issues: []

docs/changelog/127921.yaml

@@ -0,0 +1,5 @@
+pr: 127921
+summary: "[9.x] Revert \"Enable madvise by default for all builds\""
+area: Vector Search
+type: bug
+issues: []

docs/changelog/127949.yaml

@@ -0,0 +1,5 @@
+pr: 127949
+summary: Ensure ordinal builder emit ordinal blocks
+area: ES|QL
+type: bug
+issues: []

libs/entitlement/bridge/src/main/java/org/elasticsearch/entitlement/bridge/Util.java

@@ -10,8 +10,10 @@
 package org.elasticsearch.entitlement.bridge;
 
 import java.util.Optional;
+import java.util.Set;
 
 import static java.lang.StackWalker.Option.RETAIN_CLASS_REFERENCE;
+import static java.lang.StackWalker.Option.SHOW_HIDDEN_FRAMES;
 
 public class Util {
     /**
@@ -23,6 +25,8 @@ public class Util {
     public static final Class<?> NO_CLASS = new Object() {
     }.getClass();
 
+    private static final Set<String> skipInternalPackages = Set.of("java.lang.invoke", "java.lang.reflect", "jdk.internal.reflect");
+
     /**
      * Why would we write this instead of using {@link StackWalker#getCallerClass()}?
      * Because that method throws {@link IllegalCallerException} if called from the "outermost frame",
@@ -32,9 +36,10 @@ public class Util {
      */
     @SuppressWarnings("unused") // Called reflectively from InstrumenterImpl
     public static Class<?> getCallerClass() {
-        Optional<Class<?>> callerClassIfAny = StackWalker.getInstance(RETAIN_CLASS_REFERENCE)
+        Optional<Class<?>> callerClassIfAny = StackWalker.getInstance(Set.of(RETAIN_CLASS_REFERENCE, SHOW_HIDDEN_FRAMES))
             .walk(
                 frames -> frames.skip(2) // Skip this method and its caller
+                    .filter(frame -> skipInternalPackages.contains(frame.getDeclaringClass().getPackageName()) == false)
                     .findFirst()
                     .map(StackWalker.StackFrame::getDeclaringClass)
             );

libs/entitlement/src/main/java/org/elasticsearch/entitlement/runtime/policy/entitlements/FilesEntitlement.java

@@ -182,7 +182,9 @@ private static BaseDir parseBaseDir(String baseDir) {
             case "config" -> BaseDir.CONFIG;
             case "data" -> BaseDir.DATA;
             case "home" -> BaseDir.USER_HOME;
-            // NOTE: shared_repo is _not_ accessible to policy files, only internally
+            // it would be nice to limit this to just ES modules, but we don't have a way to plumb that through to here
+            // however, we still don't document in the error case below that shared_repo is valid
+            case "shared_repo" -> BaseDir.SHARED_REPO;
             default -> throw new PolicyValidationException(
                 "invalid relative directory: " + baseDir + ", valid values: [config, data, home]"
             );

server/src/main/java/org/elasticsearch/common/text/Text.java renamed to libs/x-content/src/main/java/org/elasticsearch/xcontent/Text.java

@@ -6,22 +6,17 @@
  * your election, the "Elastic License 2.0", the "GNU Affero General Public
  * License v3.0 only", or the "Server Side Public License, v 1".
  */
-package org.elasticsearch.common.text;
-
-import org.apache.lucene.util.BytesRef;
-import org.elasticsearch.common.bytes.BytesArray;
-import org.elasticsearch.common.bytes.BytesReference;
-import org.elasticsearch.xcontent.ToXContentFragment;
-import org.elasticsearch.xcontent.XContentBuilder;
+package org.elasticsearch.xcontent;
 
 import java.io.IOException;
+import java.nio.ByteBuffer;
 import java.nio.charset.StandardCharsets;
 
 /**
- * Both {@link String} and {@link BytesReference} representation of the text. Starts with one of those, and if
+ * Both {@link String} and {@link ByteBuffer} representation of the text. Starts with one of those, and if
  * the other is requests, caches the other one in a local reference so no additional conversion will be needed.
  */
-public final class Text implements Comparable<Text>, ToXContentFragment {
+public final class Text implements XContentString, Comparable<Text>, ToXContentFragment {
 
     public static final Text[] EMPTY_ARRAY = new Text[0];
 
@@ -36,31 +31,43 @@ public static Text[] convertFromStringArray(String[] strings) {
         return texts;
     }
 
-    private BytesReference bytes;
+    private ByteBuffer bytes;
     private String text;
     private int hash;
+    private int stringLength = -1;
+
+    /**
+     * Construct a Text from a UTF-8 encoded ByteBuffer. Since no string length is specified, {@link #stringLength()}
+     * will perform a string conversion to measure the string length.
+     */
+    public Text(ByteBuffer bytes) {
+        this.bytes = bytes;
+    }
 
-    public Text(BytesReference bytes) {
+    /**
+     * Construct a Text from a UTF-8 encoded ByteBuffer and an explicit string length. Used to avoid string conversion
+     * in {@link #stringLength()}.
+     */
+    public Text(ByteBuffer bytes, int stringLength) {
         this.bytes = bytes;
+        this.stringLength = stringLength;
     }
 
     public Text(String text) {
         this.text = text;
     }
 
     /**
-     * Whether a {@link BytesReference} view of the data is already materialized.
+     * Whether a {@link ByteBuffer} view of the data is already materialized.
      */
     public boolean hasBytes() {
         return bytes != null;
     }
 
-    /**
-     * Returns a {@link BytesReference} view of the data.
-     */
-    public BytesReference bytes() {
+    @Override
+    public ByteBuffer bytes() {
         if (bytes == null) {
-            bytes = new BytesArray(text.getBytes(StandardCharsets.UTF_8));
+            bytes = StandardCharsets.UTF_8.encode(text);
         }
         return bytes;
     }
@@ -72,11 +79,20 @@ public boolean hasString() {
         return text != null;
     }
 
-    /**
-     * Returns a {@link String} view of the data.
-     */
+    @Override
     public String string() {
-        return text == null ? bytes.utf8ToString() : text;
+        if (text == null) {
+            text = StandardCharsets.UTF_8.decode(bytes).toString();
+        }
+        return text;
+    }
+
+    @Override
+    public int stringLength() {
+        if (stringLength < 0) {
+            stringLength = string().length();
+        }
+        return stringLength;
     }
 
     @Override
@@ -115,8 +131,8 @@ public XContentBuilder toXContent(XContentBuilder builder, Params params) throws
         } else {
             // TODO: TextBytesOptimization we can use a buffer here to convert it? maybe add a
             // request to jackson to support InputStream as well?
-            BytesRef br = this.bytes().toBytesRef();
-            return builder.utf8Value(br.bytes, br.offset, br.length);
+            assert bytes.hasArray();
+            return builder.utf8Value(bytes.array(), bytes.arrayOffset() + bytes.position(), bytes.remaining());
         }
     }
 }

libs/x-content/src/main/java/org/elasticsearch/xcontent/XContentString.java

@@ -0,0 +1,29 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the "Elastic License
+ * 2.0", the "GNU Affero General Public License v3.0 only", and the "Server Side
+ * Public License v 1"; you may not use this file except in compliance with, at
+ * your election, the "Elastic License 2.0", the "GNU Affero General Public
+ * License v3.0 only", or the "Server Side Public License, v 1".
+ */
+
+package org.elasticsearch.xcontent;
+
+import java.nio.ByteBuffer;
+
+public interface XContentString {
+    /**
+     * Returns a {@link String} view of the data.
+     */
+    String string();
+
+    /**
+     * Returns a UTF8-encoded {@link ByteBuffer} view of the data.
+     */
+    ByteBuffer bytes();
+
+    /**
+     * Returns the number of characters in the represented string.
+     */
+    int stringLength();
+}

modules/repository-url/src/main/plugin-metadata/entitlement-policy.yaml

@@ -1,2 +1,8 @@
+org.elasticsearch.repository.url:
+  - outbound_network
+  - files:
+      - relative_path: .
+        relative_to: shared_repo
+        mode: read
 org.apache.httpcomponents.httpclient:
   - outbound_network # for URLHttpClient