[PATCH] 8217561 : X86: Add floating-point Math.min/max intrinsics, approval request

[B. Blaser]

Thanks for your approval, Vladimir, I'll push it.

I had timeouts on jdk/submit with the initial search tree but reducing
the loop number made the test pass [1].
It seems you have a stack overflow only when printing the tree at the
end of the test on your SPARC system which isn't concerned by this
fix.
I'll comment out the recursive printing of the tree when pushing, the
insertion being iterative.
However, if we still have timeout/overflow reports on some systems,
I'll comment out the search tree example as I added it only to try a
realistic use case.

Bernard

[1] http://hg.openjdk.java.net/jdk/submit/rev/d164e0b595e6#l2.7


On Thu, 7 Mar 2019 at 07:11, Vladimir Kozlov <vladimir.kozlov at oracle.com> wrote:
>
> I run tier1-3 testing with these changes.
> All passed except compiler/intrinsics/math/TestFpMinMaxIntrinsics.java failed on SPARC:
>
> java.lang.reflect.InvocationTargetException
>         at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
>         at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
> ...
> Caused by: java.lang.StackOverflowError
>         at compiler.intrinsics.math.TestFpMinMaxIntrinsics$Node.toString(TestFpMinMaxIntrinsics.java:262)
>         at java.base/java.lang.invoke.StringConcatFactory$Stringifiers$ObjectStringifier.valueOf(StringConcatFactory.java:1702)
>         at compiler.intrinsics.math.TestFpMinMaxIntrinsics$Node.toString(TestFpMinMaxIntrinsics.java:264)
>         at java.base/java.lang.invoke.StringConcatFactory$Stringifiers$ObjectStringifier.valueOf(StringConcatFactory.java:1702)
>         at compiler.intrinsics.math.TestFpMinMaxIntrinsics$Node.toString(TestFpMinMaxIntrinsics.java:264)
>
> Looks like deep recursion.
>
> It failed for last new @run commands with -XX:CompileCommand=dontinline,TestFpMinMaxIntrinsics.min*
> TestFpMinMaxIntrinsics sortedSearchTree 1
>
> Vladimir
>
> On 3/6/19 4:55 PM, Vladimir Kozlov wrote:
> > Okay. Lets push this version. Do you need sponsor to push?
> >
> > Thanks,
> > Vladimir
> >
> > On 3/6/19 2:25 PM, B. Blaser wrote:
> >> Here it is:
> >>
> >> http://cr.openjdk.java.net/~bsrbnd/jdk8217561/webrev.06/
> >>
> >> Any feedback is welcome (jdk/submit report is good),
> >> Bernard