RFR: 8357289: Break down the String constructor into smaller methods [v6]
Chen Liang
liach at openjdk.org
Mon May 26 14:58:00 UTC 2025
On Mon, 26 May 2025 13:23:07 GMT, Shaojin Wen <swen at openjdk.org> wrote:
>> Through JVM Option +PrintInlining, we found that String has a constructor codeSize of 852, which is too large. This caused failed to inline.
>>
>> The following is the output information of PrintInlining:
>>
>> @ 9 java.lang.String::<init> (12 bytes) inline (hot)
>> !m @ 1 java.nio.charset.Charset::defaultCharset (52 bytes) inline (hot)
>> ! @ 8 java.lang.String::<init> (852 bytes) failed to inline: hot method too big
>>
>>
>> In Java code, the big method that cannot be inlined is the following constructor
>>
>>
>> String(Charset charset, byte[] bytes, int offset, int length) {}
>>
>> The above String constructor is too large; break it down into smaller methods with a codeSize under 325 to allow them to be inlined by the C2.
>
> Shaojin Wen has updated the pull request incrementally with one additional commit since the last revision:
>
> from @liach
src/java.base/share/classes/java/lang/String.java line 661:
> 659: // ascii
> 660: if (ad.isASCIICompatible() && !StringCoding.hasNegatives(bytes, offset, length)) {
> 661: if (COMPACT_STRINGS) {
Replace this with `return iso88591(bytes, offset, length)`;
src/java.base/share/classes/java/lang/String.java line 673:
> 671: value = new byte[length];
> 672: ad.decodeToLatin1(bytes, offset, length, value);
> 673: coder = LATIN1;
With the other optimizations, you can make this `return new String(value, LATIN1)` and remove the `value` and `coder` local variable declarations.
src/java.base/share/classes/java/lang/String.java line 680:
> 678: char[] ca = new char[en];
> 679: int clen = ad.decode(bytes, offset, length, ca);
> 680: if (COMPACT_STRINGS) {
This can be `return new String(ca, 0, cLen, null);` too,
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/25290#discussion_r2107508429
PR Review Comment: https://git.openjdk.org/jdk/pull/25290#discussion_r2107510678
PR Review Comment: https://git.openjdk.org/jdk/pull/25290#discussion_r2107506207
More information about the core-libs-dev
mailing list