RFR: 8302358: Behavior of adler32 changes after JDK-8300208

[Jatin Bhateja]

On Thu, 16 Feb 2023 01:29:44 GMT, Sandhya Viswanathan <sviswanathan at openjdk.org> wrote:

> This PR fixes the issue in Adler computation on x86_64. The problem was  due to a missing register clear. Also the test is updated to capture the failed test case.
> 
> Please review.
> 
> Best Regards,
> Sandhya

src/hotspot/cpu/x86/stubGenerator_x86_64_adler.cpp line 285:

> 283:   // continue loop
> 284:   __ movdl(xa, a_d);
> 285:   __ vpxor(yb, yb, yb, Assembler::AVX_256bit);

Thanks @sviswa7 , can you please move this to start of SLOOP1, we can then remove two [vpxors]() before the start for AVX2 and AVX3 loops. Also there was a minor typo in the comment over algorithm, attaching the change file.
[suggestion.txt](https://github.com/openjdk/jdk/files/10752579/suggestion.txt)

-------------

PR: https://git.openjdk.org/jdk/pull/12581