[jdk17u-dev] RFR: 8273454: C2: Transform (-a)*(-b) into a*b [v3]

[Aleksey Shipilev]

On Mon, 11 Dec 2023 14:36:56 GMT, Goetz Lindenmaier <goetz at openjdk.org> wrote:

>> I backport this for parity with 17.0.11-oracle.
>> 
>> This contains follow up fix 8274060: C2: Incorrect computation after JDK-8273454 and a part of JDK-8274130) C2: MulNode::Ideal chained transformations may act on wrong nodes which is already backported, but omitted the part to be applied to this patch.
>> This is also the reason why a clean backport failed.
>
> Goetz Lindenmaier has refreshed the contents of this pull request, and previous commits have been removed. The incremental views will show differences compared to the previous content of the PR. The pull request contains two new commits since the last revision:
> 
>  - Backport c77ebe88748b0a55f1fc7a5497314a752eab1e2a
>  - Backport 7c9868c0b3c9bd3d305e71f91596190813cdccce

I never thought we would backport this. But now that we do, this hunk is missing:
https://github.com/openjdk/jdk/commit/756d22c3563ac92e74bb68d5eecb86d4fbab2c6b#diff-b1bd52f0743843e15452764f48ff43c15dd3192a28bfb684b34149f0e964996eR83-R84

When we backported [JDK-8274130](https://bugs.openjdk.org/browse/JDK-8274130), we skipped that hunk.

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

Changes requested by shade (Reviewer).

PR Review: https://git.openjdk.org/jdk17u-dev/pull/2035#pullrequestreview-1775306325