JDK-8217561 : x86: Add floating-point Math.min/max intrinsics
  • Type: Enhancement
  • Component: hotspot
  • Sub-Component: compiler
  • Affected Version: 11,13
  • Priority: P4
  • Status: Resolved
  • Resolution: Fixed
  • CPU: x86
  • Submitted: 2019-01-22
  • Updated: 2022-05-20
  • Resolved: 2019-03-07
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.
JDK 11 JDK 13
11.0.12Fixed 13 b12Fixed
Related Reports
Relates :  
JDK-8218550 - Add test omitted from JDK-8212043
Relates :  
JDK-8220407 - compiler/intrinsics/math/TestFpMinMaxIntrinsics.java timedout
Relates :  
JDK-8212043 - AArch64: Add floating-point Math.min/max intrinsics
Relates :  
JDK-8227222 - vmTestbase/jit/FloatingPoint/gen_math/Loops04/Loops04.java hits assert(((!attributes->uses_vl()) || (attributes->get_vector_len() == AVX_512bit) || (!_legacy_mode_vl) || (attributes->is_legacy_mode()))) failed: XMM register should be 0-15
Relates :  
JDK-8287087 - C2: perform SLP reduction analysis on-demand
Relates :  
JDK-8226627 - assert(t->singleton()) failed: must be a constant
Description
Intrinsic implementation taking care of the non-conformance with the Java API documentation of x86 min/max instructions for +/-0.0 and NaN.
Comments
I believe JDK-8226627 should be backported to 11u, too.
28-03-2021
Fix Request [11u]. This is a follow up to the ARM64 11u backport of JDK-8212043, see https://mail.openjdk.java.net/pipermail/jdk-updates-dev/2021-January/004734.html. Applies cleanly except that the second hunk of x86.ad required a context modification, and the first hunk of library_call.cpp was removed due to a later copyright date update. This is the first of three backports that will be pushed together. The other two apply cleanly after 8217561. JDK-8227222: vmTestbase/jit/FloatingPoint/gen_math/Loops04/Loops04.java hits assert(((!attributes->uses_vl()) || (attributes->get_vector_len() == AVX_512bit) || (!_legacy_mode_vl) || (attributes->is_legacy_mode()))) failed: XMM register should be 0-15 JDK-8220407: compiler/intrinsics/math/TestFpMinMaxIntrinsics.java timedout Review request: https://mail.openjdk.java.net/pipermail/jdk-updates-dev/2021-March/005343.html, reviewed by simonis.
17-03-2021
URL: http://hg.openjdk.java.net/jdk/jdk/rev/ff399127078a User: bsrbnd Date: 2019-03-07 14:37:48 +0000
07-03-2019
compiler/intrinsics/math/TestFpMinMaxIntrinsics.java failed on SPARC for last new @run commands with -XX:CompileCommand=dontinline,TestFpMinMaxIntrinsics.min* TestFpMinMaxIntrinsics sortedSearchTree 1
07-03-2019
Latest webrev: http://cr.openjdk.java.net/~bsrbnd/jdk8217561/webrev.06/
07-03-2019
I'll disable the printing of the tree when pushing which might cause stack overflows on some systems, see: http://mail.openjdk.java.net/pipermail/hotspot-compiler-dev/2019-March/033055.html
07-03-2019
Review thread of the patch provided by Jatin Bhateja: http://mail.openjdk.java.net/pipermail/hotspot-compiler-dev/2019-January/032521.html http://mail.openjdk.java.net/pipermail/hotspot-compiler-dev/2019-February/032626.html
07-02-2019