Unsafe.{get,put}-X-Unaligned; Efficient array comparison intrinsics

[Andrew Haley]

On 02/24/2015 02:47 PM, Paul Sandoz wrote:
> On Feb 24, 2015, at 2:48 PM, Andrew Haley <aph at redhat.com> wrote:

>>> With that in mind is there any need to intrinsify the new methods
>>> at all given those new Java methods can defer to the older ones
>>> based on a constant check? Also should that anyway be done for the
>>> interpreter?
>>>
>>> public void putIntUnaligned(Object o, long offset, int x) { if (IS_UNALIGNED || (offset & 3) == 0) { putInt(o, offset, x); } else if (byteOrder == BIG_ENDIAN) { putIntB(o, offset, x); } else { putIntL(o, offset, x); } }
>>
>> Yes.  It certainly could be done like this but I think C1 doesn't do
>> the optimization to remove the IS_UNALIGNED test, so we'd still want
>> the C1 builtins.
> 
> Hmm.... if i run the following code on my Mac:
> 
> 
>     public int x(int i) {
>         if (X || (i & 3) == 0) {
>             return i + i;
>         }
>         else {
>             return Integer.sum(i, i);
>         }
>     }

> With the options:
> 
>   -XX:TieredStopAtLevel=1 -XX:+UnlockDiagnosticVMOptions -XX:+PrintAssembly
> 
> Then Y.x is compiled to:
> 
>   0x000000011155ea80: mov    %eax,-0x14000(%rsp)
>   0x000000011155ea87: push   %rbp
>   0x000000011155ea88: sub    $0x30,%rsp         ;*getstatic X
>                                                 ; - Y::x at 0 (line 43)
> 
>   0x000000011155ea8c: mov    %rdx,%rax
>   0x000000011155ea8f: add    %edx,%eax
>   0x000000011155ea91: add    $0x30,%rsp
>   0x000000011155ea95: pop    %rbp
>   0x000000011155ea96: test   %eax,-0x62ca9c(%rip)        # 0x0000000110f32000
>                                                 ;   {poll_return}
>   0x000000011155ea9c: retq   

Oh, right, better than I thought.  I should know by now to check
before posting.  But even so, we don't know that the method will
be inlined.

>>> If we expose the endianness query via a new method in unsafe we
>>> should reuse that in java.nio.Bits and get rid of the associated
>>> static code block.
>>
>> Sure, I already did that.
>>
> 
> Locally i guess? (just in case i missed something in the current webrev).

Ah.  I used the query but I forgot to get rid of the static code
block:

http://cr.openjdk.java.net/~aph/unaligned.jdk.1/src/java.base/share/classes/java/nio/Bits.java.cdiff.html

Andrew.