JDK-8262355 : Support for AVX-512 opmask register allocation.
- Type: Enhancement
- Component: hotspot
- Sub-Component: compiler
- Affected Version: 17
- Priority: P4
- Status: Resolved
- Resolution: Fixed
- CPU: x86
- Submitted: 2021-02-25
- Updated: 2025-01-21
- Resolved: 2021-04-04
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.
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.
| JDK 17 |
|---|
17 b17Fixed  |
Related Reports
|
Relates :
|
|
|
Relates :
|
|
|
Relates :
|
|
|
Relates :
|
|
|
Relates :
|
|
|
Relates :
|
Description
- AVX-512 added 8 new 64 bit opmask register (k0-k7). These registers enable predicated vector instruction such that an operation is performed only over vector lanes for which corresponding bit is set in opmask register. - Currently in order to support predicated vector operations in macro assembly routines computed masks are stored in hard coded opmask registers. - Cross instruction mask propagation is done either using a GPR or a vector a vector register. - Register allocation support for opmask register (k1-k7) will facilitate mask propagation across instruction and thus enable emitting efficient instruction sequence over X86 targets supporting AVX-512 feature.
Comments
|