JDK-8241484 : Optimizing vector logic not operation
  • Type: Enhancement
  • Component: hotspot
  • Sub-Component: compiler
  • Affected Version: 15
  • Priority: P4
  • Status: Open
  • Resolution: Unresolved
  • CPU: x86_64
  • Submitted: 2020-03-24
  • Updated: 2020-06-03
The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.

To download the current JDK release, click here.
Other
tbdUnresolved
Related Reports
Relates :  
Relates :  
Description
- Current  Implementation of vector logic not operation performs a Xor operation between the input vector and Broadcasted -1 value which is read from an externally initialized memory. 
- The broadcast operation can be made efficient by replacing a read from external memory (which may cause a cache miss) over non-AVX3 targets.
-  Over AVX3 a single ternary logic instruction is sufficient to replace complete pattern involving Xor and broadcast operation. 
Comments
webrev: http://cr.openjdk.java.net/~jbhateja/8241484/webrev.01_notV/
24-03-2020