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

[Vladimir Kozlov]

Thank you, Bernard

I looked on our regular testing results after your push and there are no failures I observed before.

Vladimir

On 3/7/19 7:21 AM, B. Blaser wrote:
> Pushed:
> http://hg.openjdk.java.net/jdk/jdk/rev/ff399127078a
> 
> Search tree printing commented here:
> http://hg.openjdk.java.net/jdk/jdk/rev/ff399127078a#l8.221
> 
> Thanks to Jatin Bhateja for his contribution!
> Bernard
> 
> On Thu, 7 Mar 2019 at 12:42, B. Blaser <bsrbnd at gmail.com> wrote:
>>
>> 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