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

[Vladimir Kozlov]

Okay. I will wait next version.

Thanks,
Vladimir

On 1/29/19 11:18 AM, Viswanathan, Sandhya wrote:
> Hi Vladimir,
> 
>  From what I understand the test is correct and the implementation needs to be fixed (the effect on registers a and b should be USE_KILL).
> Jatin plans to send an updated webrev after fixing the issue.
> 
> Best Regards,
> Sandhya
> 
> 
> -----Original Message-----
> From: hotspot-compiler-dev [mailto:hotspot-compiler-dev-bounces at openjdk.java.net] On Behalf Of Vladimir Kozlov
> Sent: Tuesday, January 29, 2019 10:26 AM
> To: hotspot-compiler-dev at openjdk.java.net
> Subject: Re: [PATCH] 8217561 : X86: Add floating-point Math.min/max intrinsics
> 
> Code change seems fine.
> 
> For -XX:TieredStopAtLevel=1 test command add explicit -XX:+TieredCompilation
> Also both JIT (not -Xint) @run command add -XX:+IgnoreUnrecognizedVMOptions flag to make sure test works with all VM
> build variants.
> 
> Was the test fixed to solve issue found by Nils?
> 
> Thanks,
> Vladimir
> 
> On 1/29/19 7:45 AM, Viswanathan, Sandhya wrote:
>> Please use the following updated link to webrev for review:
>>
>> http://cr.openjdk.java.net/~sviswanathan/Jatin/8217561/webrev.00/
>>
>> Best Regards,
>> Sandhya
>>
>>
>> -----Original Message-----
>> From: Bhateja, Jatin
>> Sent: Tuesday, January 29, 2019 2:20 AM
>> To: Andrew Dinn <adinn at redhat.com>; Viswanathan, Sandhya <sandhya.viswanathan at intel.com>
>> Cc: hotspot-compiler-dev at openjdk.java.net
>> Subject: RE: [PATCH] 8217561 : X86: Add floating-point Math.min/max intrinsics
>>
>> Hi Andrew,
>>
>> Some permission issue seem to crop up in webrev, which is why sent the patch diff along.
>>
>> Hi Sandhya,
>>
>> Kindly help in resolving this.
>>
>> Regards,
>> Jatin
>>
>> -----Original Message-----
>> From: Andrew Dinn [mailto:adinn at redhat.com]
>> Sent: Tuesday, January 29, 2019 3:25 PM
>> To: Bhateja, Jatin <jatin.bhateja at intel.com>; hotspot-compiler-dev at openjdk.java.net
>> Subject: Re: [PATCH] 8217561 : X86: Add floating-point Math.min/max intrinsics
>>
>> Hi Jatin,
>>
>> On 29/01/2019 04:18, Bhateja, Jatin wrote:
>>
>>> Please find attached a patch for intrinsification for java library
>>> methods Math.max and Math.min  for scalar floating point types (float
>>> and double).
>>>
>>> New intrinsics match the semantics of java library function and takes
>>> care of NaN and signed zero operands.
>>>
>>> Webrev : http://cr.openjdk.java.net/~sviswanathan/8217561/webrev.01/
>> The patch file attachment came through all right. However, I am getting
>> "403 - Forbidden" when I try to read the files linked in the webrev.
>> Does Sandya need to change the permissions on the files?
>>
>> regards,
>>
>>
>> Andrew Dinn
>> -----------
>> Senior Principal Software Engineer
>> Red Hat UK Ltd
>> Registered in England and Wales under Company Registration No. 03798903
>> Directors: Michael Cunningham, Michael ("Mike") O'Neill, Eric Shander
>>