extended.openjdk jdk_tools windows jdk8 sun/tools/native2ascii/Native2AsciiTests.sh fails

extended.openjdk jdk_tools test sun/tools/native2ascii/Native2AsciiTests.sh fails on win32 and win64 jdk8 with the output message:
```
[2021-06-19T15:24:58.496Z] STDOUT:
[2021-06-19T15:24:58.496Z] *** x.out.expected Sat Jun 19 10:24:59 2021
[2021-06-19T15:24:58.496Z] --- x.out.1 Sat Jun 19 10:24:59 2021
[2021-06-19T15:24:58.496Z] ***************
[2021-06-19T15:24:58.496Z] *** 1 ****
[2021-06-19T15:24:58.496Z] ! €?‚ƒ„…†‡ˆ‰Š‹Œ?Ž??‘’“”•–—˜™š›œ?žŸ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ
[2021-06-19T15:24:58.496Z] --- 1 ----
[2021-06-19T15:24:58.496Z] ! €?‚ƒ„…†‡ˆ‰Š‹Œ?Ž??‘’“”•–—˜™š›œ?žŸ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ
[2021-06-19T15:24:58.496Z] 4630463 failed
[2021-06-19T15:24:58.496Z] STDERR:
[2021-06-19T15:24:58.496Z] + diff -c x.out.expected x.out.1
```
Downloading the test files and viewing them in a hex editor shows that:
x.out.expected file, length 129 bytes, ends in hex '0A'
x.out.1 file, length 130 bytes, ends in hex '0D0A'

So it looks like the test writes an output file with Windows CRLF but the test only expects LF.

To rerun: https://ci.adoptopenjdk.net/job/Grinder/parambuild/?JDK_VERSION=8&JDK_IMPL=hotspot&JDK_VENDOR=Temurin&BUILD_LIST=openjdk&PLATFORM=x86-64_windows_xl&TARGET=jdk_tools

Run the linked Grinder configuration for the extended.openjdk jdk_tools test sun/tools/native2ascii/Native2AsciiTests.sh on Windows JDK 8. Compare x.out.expected with x.out.1, focusing on the LF versus CRLF difference; done means the test passes on win32 and win64 without the reported diff failure.

Written by the indexing model from the issue text.