RFR 8179602: Fix for JDK-8165437 is broken on 32-bit Linux

[Michael McMahon]

Hi Vyom,

As discussed, I think there are still some issues including on 32 bit 
Linux and since we don't
have too many of those machines left to debug with, it's probably best 
to back out the original
change, and go back to the original webrev. We can then review an 
updated version which takes
care of the 32/64 bit issues that were found. Below is a backout webrev, 
with no other changes

http://cr.openjdk.java.net/~michaelm/8179602/webrev.1/

Thanks,
Michael



On 04/05/2017, 09:50, Vyom Tewari wrote:
> Hi All,
>
> Please  review the below change.
>
> Webrev: http://cr.openjdk.java.net/~vtewari/8179602/webrev0.0/index.html
>
> Bugid: https://bugs.openjdk.java.net/browse/JDK-8179602
>
> This issue is because of side effect of "JDK-8165437" where we are 
> using "JVM_NanoTime" which returns a 64 bit jlong and return value was 
> getting assigned to long type. On 32 bit OS long is 4 byte, which 
> leads to integer overflow.
>
> Our internal test JPRT is  still running.
>
> Thanks,
>
> Vyom
>