Support Java behaviour w.r.t fmin/fmax/dmin/dmax on Z #20716
Conversation
|
please review @r30shah |
| @@ -1,33 +1,42 @@ | |||
| /* | |||
| * Copyright IBM Corp. and others 2019 | |||
There was a problem hiding this comment.
Original copy right was correct, there was not issue with space here.
| public class TestJavaLangMath { | ||
|
|
||
| /** | ||
| * Tests the constant corner cases defined by the {@link Math.sqrt} method. | ||
| * Tests the constant corner cases defined by the {@link Math.sqrt} method. |
There was a problem hiding this comment.
It should be consistent for comments.
/**
*
*
*/
| @@ -0,0 +1,292 @@ | |||
| /* | |||
| * Copyright IBM Corp. and others 2024 | |||
There was a problem hiding this comment.
Please fix the space here. Check the copyright in other files.
matthewhall2
force-pushed
the
Math_Max_Min_FD
branch
from
7b70d70 to
20d8a6c
Compare
Thanks @r30shah |
|
@matthewhall2 Can you resolve the merge conflict? I see that your OMR changes have been propagated to openj9-omr so we can target to merge this today |
- Adds java_lang_Math_max/min for float/double as a recognized method - Adds a SupportsInlineMath_MaxMin_FD flag to the Z code generator - Flag is only set in Z if the TR_disableInlineMath_MaxMin_FD environment variable is not set - If the flag is set, call nodes are transformed to a functionally equivalent tree that uses fmin/fmax/dmin/dmax nodes Signed-off-by: Matthew Hall <[email protected]>
- use new omr min/max helper - +0.0 compares as strictly greater than -0.0 - returns first NaN unchanged if present Signed-off-by: Matthew Hall <[email protected]>
- tests float corner cases with +/-0 - NaN cases - confirms respective OMR changes Signed-off-by: Matthew Hall <[email protected]>
matthewhall2
force-pushed
the
Math_Max_Min_FD
branch
from
20d8a6c to
f1e4af6
Compare
done. Running internal sanity checks again |
|
jenkins test sanity zlinux jdk11,jdk21 |
|
still waiting on the internal sanity run @r30shah. will let you know when it finishes (the only failures right now are for x86) |
|
All the tests have been on jenkins passes. Merging this based on the the sanity run build from #20716 (comment) and #20716 (comment). @matthewhall2 We should keep monitor OpenJ9 Jenkins build and internal builds in case something pop-up. |
Uses SIMD instructions for floats on z14+, and for doubles on z13+ (fixes errors on z13 machines that were seen after #20530).
https://github.ibm.com/runtimes/openj9-jit-z/issues/935#issuecomment-98667018
depends on eclipse-omr/omr#7572
vmfarm
vmfarm w recognizedmethodtests