[11u] RFR 8245512: CRC32 optimization using AVX512 instructions

[Langer, Christoph]

Hi,

as we should be cautious about backports to an LTS release if something is not just a bugfix, I'm not sure whether this backport for a performance enhancement is good to go. Generally, if it can be proven that it's really just a potential performance improvement without side effects to other scenarios, I would be fine with it. But as I'm not an expert in this area, I'd prefer if it there was some further endorsement (or disapproval)...

@Andrew Haley; @Lindenmaier, Goetz; @Doerr, Martin, what's your assessment?

Thanks
Christoph

> -----Original Message-----
> From: jdk-updates-dev <jdk-updates-dev-retn at openjdk.java.net> On
> Behalf Of Viswanathan, Sandhya
> Sent: Freitag, 11. Dezember 2020 00:47
> To: Vladimir Kozlov <vladimir.kozlov at oracle.com>; jdk-updates-
> dev at openjdk.java.net; hotspot-compiler-dev at openjdk.java.net
> Subject: RE: [11u] RFR 8245512: CRC32 optimization using AVX512 instructions
> 
> Thanks a lot Vladimir for review and link to the process. I requested the
> backport in JBS.
> 
> Best Regards,
> Sandhya
> 
> -----Original Message-----
> From: Vladimir Kozlov <vladimir.kozlov at oracle.com>
> Sent: Thursday, December 10, 2020 3:07 PM
> To: Viswanathan, Sandhya <sandhya.viswanathan at intel.com>; jdk-updates-
> dev at openjdk.java.net; hotspot-compiler-dev at openjdk.java.net
> Subject: Re: [11u] RFR 8245512: CRC32 optimization using AVX512 instructions
> 
> Changes are fine but you need to follow process:
> 
> http://openjdk.java.net/projects/jdk-updates/approval.html
> 
> Also it is performance improvement which may not be accepted.
> 
> Regards,
> Vladimi
> 
> On 12/10/20 1:00 PM, Viswanathan, Sandhya wrote:
> > I would like to backport the CRC32 optimization using AVX512 instructions
> to JDK 11u.
> > This optimization was introduced in JDK 15.
> >
> > JBS: https://bugs.openjdk.java.net/browse/JDK-8245512
> > Webrev: http://cr.openjdk.java.net/~sviswanathan/8245512/webrev.00/
> >
> > Only one change was needed to the original patch in
> src/hotspot/cpu/x86/assembler_x86.cpp.
> >
> > The following in Assembler::blendvpb(), line 7908:
> >
> >   emit_int24(0x4C, (0xC0 | encode), (0xF0 & src2_enc << 4)); was
> > replaced by equivalent:
> >    emit_int8((unsigned char)0x4C);
> > emit_int8((unsigned char)(0xC0 | encode)); emit_int8((unsigned
> > char)(0xF0 & src2_enc<<4));
> >
> > The patch applies cleanly otherwise.
> >
> > Please review this backport to 11u.
> >
> > Best Regards,
> > Sandhya
> >