# HG changeset patch
# User redestad
# Date 1497047044 -7200
#      Sat Jun 10 00:24:04 2017 +0200
# Node ID 15dae6e0d183e8b8cf128ebe0ec9b3bcab3a624a
# Parent  875fa66a13a242ec947ba309c99e2a88991dbf4f
8181147: JNI_GetStringPlatformChars should have a fast path for UTF-8
Reviewed-by: shade, chegar, erikj

diff --git a/make/test/JtregNative.gmk b/make/test/JtregNative.gmk
--- a/make/test/JtregNative.gmk
+++ b/make/test/JtregNative.gmk
@@ -44,12 +44,19 @@
 # Add more directories here when needed.
 BUILD_JDK_JTREG_NATIVE_SRC += \
     $(JDK_TOPDIR)/test/native_sanity \
+    $(JDK_TOPDIR)/test/java/lang/String/nativeEncoding \
     #
 
 BUILD_JDK_JTREG_OUTPUT_DIR := $(BUILD_OUTPUT)/support/test/jdk/jtreg/native
 
 BUILD_JDK_JTREG_IMAGE_DIR := $(TEST_IMAGE_DIR)/jdk/jtreg
 
+ifeq ($(OPENJDK_TARGET_OS), windows)
+    BUILD_JDK_JTREG_LIBRARIES_LIBS_libstringPlatformChars := java.lib
+else
+    BUILD_JDK_JTREG_LIBRARIES_LIBS_libstringPlatformChars := -ljava
+endif
+
 $(eval $(call SetupTestFilesCompilation, BUILD_JDK_JTREG_LIBRARIES, \
     TYPE := LIBRARY, \
     SOURCE_DIRS := $(BUILD_JDK_JTREG_NATIVE_SRC), \
diff --git a/src/java.base/share/native/include/jni.h b/src/java.base/share/native/include/jni.h
--- a/src/java.base/share/native/include/jni.h
+++ b/src/java.base/share/native/include/jni.h
@@ -611,6 +611,10 @@
       (JNIEnv *env, jstring str, jboolean *isCopy);
     void (JNICALL *ReleaseStringChars)
       (JNIEnv *env, jstring str, const jchar *chars);
+    const jbyte *(JNICALL *GetStringBytesCritical)
+      (JNIEnv *env, jstring str, jint *len, jbyte *encoding, jboolean *isCopy);
+    void (JNICALL *ReleaseStringBytesCritical)
+      (JNIEnv *env, jstring str, const jbyte *bytes);
 
     jstring (JNICALL *NewStringUTF)
       (JNIEnv *env, const char *utf);
@@ -1611,6 +1615,13 @@
     void ReleaseStringChars(jstring str, const jchar *chars) {
         functions->ReleaseStringChars(this,str,chars);
     }
+    const jbyte *GetStringBytesCritical(jstring str, jint *len, jbyte *encoding, jboolean *isCopy) {
+        return functions->GetStringBytesCritical(this,str,len,encoding,isCopy);
+    }
+    void ReleaseStringBytesCritical(jstring str, const jbyte *bytes) {
+        functions->ReleaseStringBytesCritical(this,str,bytes);
+    }
+
 
     jstring NewStringUTF(const char *utf) {
         return functions->NewStringUTF(this,utf);
diff --git a/src/java.base/share/native/libjava/jni_util.c b/src/java.base/share/native/libjava/jni_util.c
--- a/src/java.base/share/native/libjava/jni_util.c
+++ b/src/java.base/share/native/libjava/jni_util.c
@@ -444,9 +444,8 @@
 
 /* Optimized for char set ISO_8559_1 */
 static jstring
-newString8859_1(JNIEnv *env, const char *str)
+newSizedString8859_1(JNIEnv *env, const char *str, int len)
 {
-    int len = (int)strlen(str);
     jchar buf[512];
     jchar *str1;
     jstring result;
@@ -469,6 +468,13 @@
     return result;
 }
 
+static jstring
+newString8859_1(JNIEnv *env, const char *str)
+{
+    int len = (int)strlen(str);
+    return newSizedString8859_1(env, str, len);
+}
+
 static const char*
 getString8859_1Chars(JNIEnv *env, jstring jstr)
 {
@@ -500,7 +506,6 @@
     return result;
 }
 
-
 /* Optimized for char set ISO646-US (us-ascii) */
 static jstring
 newString646_US(JNIEnv *env, const char *str)
@@ -671,8 +676,36 @@
 static jmethodID String_init_ID;        /* String(byte[], enc) */
 static jmethodID String_getBytes_ID;    /* String.getBytes(enc) */
 
-int getFastEncoding() {
-    return fastEncoding;
+/* Optimized for char set UTF-8 */
+static jstring
+newStringUTF8(JNIEnv *env, const char *str)
+{
+    jboolean isAscii = JNI_TRUE;
+    jstring result;
+    jbyteArray hab = NULL;
+    int len = 0;
+    char b;
+    for (b = str[len]; b != '\0'; len++, b = str[len]) {
+        if (isAscii && b & 0x80) {
+            isAscii = JNI_FALSE;
+        }
+    }
+
+    if (isAscii) {
+        return newSizedString8859_1(env, str, len);
+    }
+
+    hab = (*env)->NewByteArray(env, len);
+    if (hab != 0) {
+        jclass strClazz = JNU_ClassString(env);
+        CHECK_NULL_RETURN(strClazz, 0);
+        (*env)->SetByteArrayRegion(env, hab, 0, len, (jbyte *)str);
+        result = (*env)->NewObject(env, strClazz,
+                                       String_init_ID, hab, jnuEncoding);
+        (*env)->DeleteLocalRef(env, hab);
+        return result;
+    }
+    return NULL;
 }
 
 /* Initialize the fast encoding.  If the "sun.jnu.encoding" property
@@ -718,17 +751,20 @@
                     if ((strcmp(encname, "8859_1") == 0) ||
                         (strcmp(encname, "ISO8859-1") == 0) ||
                         (strcmp(encname, "ISO8859_1") == 0) ||
-                        (strcmp(encname, "ISO-8859-1") == 0))
+                        (strcmp(encname, "ISO-8859-1") == 0)) {
                         fastEncoding = FAST_8859_1;
-                    else if (strcmp(encname, "ISO646-US") == 0)
+                    } else if (strcmp(encname, "UTF-8") == 0) {
+                        fastEncoding = FAST_UTF_8;
+                        jnuEncoding = (jstring)(*env)->NewGlobalRef(env, enc);
+                    } else if (strcmp(encname, "ISO646-US") == 0) {
                         fastEncoding = FAST_646_US;
-                    else if (strcmp(encname, "Cp1252") == 0 ||
+                    } else if (strcmp(encname, "Cp1252") == 0 ||
                              /* This is a temporary fix until we move */
                              /* to wide character versions of all Windows */
                              /* calls. */
-                             strcmp(encname, "utf-16le") == 0)
+                             strcmp(encname, "utf-16le") == 0) {
                         fastEncoding = FAST_CP1252;
-                    else {
+                    } else {
                         fastEncoding = NO_FAST_ENCODING;
                         jnuEncoding = (jstring)(*env)->NewGlobalRef(env, enc);
                     }
@@ -792,6 +828,8 @@
         return newString646_US(env, str);
     if (fastEncoding == FAST_CP1252)
         return newStringCp1252(env, str);
+    if (fastEncoding == FAST_UTF_8)
+        return newStringUTF8(env, str);
 
     if ((*env)->EnsureLocalCapacity(env, 2) < 0)
         return NULL;
@@ -830,27 +868,10 @@
     return JNU_GetStringPlatformChars(env, jstr, isCopy);
 }
 
-JNIEXPORT const char * JNICALL
-JNU_GetStringPlatformChars(JNIEnv *env, jstring jstr, jboolean *isCopy)
-{
+static const char* getStringBytes(JNIEnv *env, jstring jstr) {
     char *result = NULL;
     jbyteArray hab = 0;
 
-    if (isCopy)
-        *isCopy = JNI_TRUE;
-
-    if (fastEncoding == NO_ENCODING_YET) {
-        initializeEncoding(env);
-        JNU_CHECK_EXCEPTION_RETURN(env, 0);
-    }
-
-    if ((fastEncoding == FAST_8859_1) || (fastEncoding == NO_ENCODING_YET))
-        return getString8859_1Chars(env, jstr);
-    if (fastEncoding == FAST_646_US)
-        return getString646_USChars(env, jstr);
-    if (fastEncoding == FAST_CP1252)
-        return getStringCp1252Chars(env, jstr);
-
     if ((*env)->EnsureLocalCapacity(env, 2) < 0)
         return 0;
 
@@ -883,6 +904,74 @@
     return result;
 }
 
+static const char*
+getStringUTF8(JNIEnv *env, jstring jstr)
+{
+    int i;
+    char *result;
+    jbyte encoding = 0;
+    jint len = 0;
+    const jbyte *str = (*env)->GetStringBytesCritical(env, jstr, &len, &encoding, NULL);
+
+    if (str == NULL) {
+        return NULL;
+    }
+
+    // only support ASCII encoding natively.
+    if (encoding == 0) {
+        result = (char*)str;
+        for (i = 0; i < len; i++) {
+            if (str[i] < 0) {
+                // not ASCII, abort to slow path
+                str = NULL;
+                break;
+            }
+        }
+    }
+
+
+    if (str == NULL) {
+        (*env)->ReleaseStringBytesCritical(env, jstr, str);
+        return getStringBytes(env, jstr);
+    }
+
+    result = MALLOC_MIN4(len);
+    if (result == 0) {
+        (*env)->ReleaseStringBytesCritical(env, jstr, str);
+        JNU_ThrowOutOfMemoryError(env, 0);
+        return NULL;
+    }
+    for (i = 0; i < len; i++) {
+        result[i] = (char)str[i];
+    }
+    result[len] = '\0';
+    (*env)->ReleaseStringBytesCritical(env, jstr, str);
+    return result;
+}
+
+JNIEXPORT const char * JNICALL
+JNU_GetStringPlatformChars(JNIEnv *env, jstring jstr, jboolean *isCopy)
+{
+
+    if (isCopy)
+        *isCopy = JNI_TRUE;
+
+    if (fastEncoding == NO_ENCODING_YET) {
+        initializeEncoding(env);
+        JNU_CHECK_EXCEPTION_RETURN(env, 0);
+    }
+
+    if ((fastEncoding == FAST_8859_1) || (fastEncoding == NO_ENCODING_YET))
+        return getString8859_1Chars(env, jstr);
+    if (fastEncoding == FAST_646_US)
+        return getString646_USChars(env, jstr);
+    if (fastEncoding == FAST_CP1252)
+        return getStringCp1252Chars(env, jstr);
+    if (fastEncoding == FAST_UTF_8)
+        return getStringUTF8(env, jstr);
+    else
+        return getStringBytes(env, jstr);
+}
 JNIEXPORT void JNICALL
 JNU_ReleaseStringPlatformChars(JNIEnv *env, jstring jstr, const char *str)
 {
diff --git a/src/java.base/share/native/libjava/jni_util.h b/src/java.base/share/native/libjava/jni_util.h
--- a/src/java.base/share/native/libjava/jni_util.h
+++ b/src/java.base/share/native/libjava/jni_util.h
@@ -382,7 +382,8 @@
     NO_FAST_ENCODING,           /* Platform encoding is not fast */
     FAST_8859_1,                /* ISO-8859-1 */
     FAST_CP1252,                /* MS-DOS Cp1252 */
-    FAST_646_US                 /* US-ASCII : ISO646-US */
+    FAST_646_US,                /* US-ASCII : ISO646-US */
+    FAST_UTF_8
 };
 
 int getFastEncoding();