RFR: 8289552: Make intrinsic conversions between bit representations of half precision values and floats [v8]

[Vladimir Kozlov]

On Thu, 29 Sep 2022 17:45:49 GMT, Smita Kamath <svkamath at openjdk.org> wrote:

>> tier2 and 3 passed. I will wait update with test and JVMCI fix before doing more testing.
>
> @vnkozlov I have addressed all review comments. Could you please run the patch through your testing? Thanks a lot for all the help.

@smita-kamath I have builds failures. Please, build and test yourself to verify changes.

src/hotspot/cpu/x86/assembler_x86.cpp: In member function 'void Assembler::evcvtps2ph(Address, KRegister, XMMRegister, int, int)':
src/hotspot/cpu/x86/assembler_x86.cpp:1950:15: error: no matching function for call to 'Assembler::emit_operand(XMMRegister&, Address&)'
 emit_operand(src, dst);

-------------

PR: https://git.openjdk.org/jdk/pull/9781