RFR: 8320448: Accelerate IndexOf using AVX2 [v19]

[Sandhya Viswanathan]

On Sat, 4 May 2024 19:35:21 GMT, Scott Gibbons <sgibbons at openjdk.org> wrote:

>> Re-write the IndexOf code without the use of the pcmpestri instruction, only using AVX2 instructions.  This change accelerates String.IndexOf on average 1.3x for AVX2.  The benchmark numbers:
>> 
>> 
>> Benchmark	                                               Score		Latest		
>> StringIndexOf.advancedWithMediumSub   343.573		317.934		0.925375393x
>> StringIndexOf.advancedWithShortSub1	  1039.081		1053.96		1.014319384x
>> StringIndexOf.advancedWithShortSub2	      55.828		110.541		1.980027943x
>> StringIndexOf.constantPattern	                9.361		11.906		1.271872663x
>> StringIndexOf.searchCharLongSuccess	        4.216		4.218		1.000474383x
>> StringIndexOf.searchCharMediumSuccess	3.133		3.216		1.02649218x
>> StringIndexOf.searchCharShortSuccess	3.76		        3.761		1.000265957x
>> StringIndexOf.success	                                9.186		9.713		1.057369911x
>> StringIndexOf.successBig	                      14.341		46.343		3.231504079x
>> StringIndexOfChar.latin1_AVX2_String	  6220.918		12154.52		1.953814533x
>> StringIndexOfChar.latin1_AVX2_char	  5503.556		5540.044		1.006629895x
>> StringIndexOfChar.latin1_SSE4_String	  6978.854		6818.689		0.977049957x
>> StringIndexOfChar.latin1_SSE4_char	  5657.499		5474.624		0.967675646x
>> StringIndexOfChar.latin1_Short_String	  7132.541		6863.359		0.962260014x
>> StringIndexOfChar.latin1_Short_char	16013.389	      16162.437		1.009307711x
>> StringIndexOfChar.latin1_mixed_String	  7386.123	      14771.622		1.999915517x
>> StringIndexOfChar.latin1_mixed_char	  9901.671		9782.245		0.987938803
>
> Scott Gibbons has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Rearrange; add lambdas for clarity

src/hotspot/cpu/x86/stubGenerator_x86_64_string.cpp line 1476:

> 1474:                                _masm);
> 1475: 
> 1476:   __ movq(r11, -1);

There doesn't seem to be a use of r11 below in this function.

src/hotspot/cpu/x86/stubGenerator_x86_64_string.cpp line 1493:

> 1491:   // Assume r10 is n - k
> 1492:   __ leaq(last, Address(haystack, r10, Address::times_1, isU ? -30 : -31));
> 1493:   __ jmpb(temp);

Need to pass r10 as parameter. Also temp label could be given a better name.

src/hotspot/cpu/x86/stubGenerator_x86_64_string.cpp line 1502:

> 1500: 
> 1501:   __ cmpq(hsPtrRet, last);
> 1502:   __ cmovq(Assembler::aboveEqual, hsPtrRet, last);

cmovq is expensive, better sequence would be:

__ cmpq(hsPtrRet, last);
__ jb_b(temp);
__ movq(hsPtrRet, last);

src/hotspot/cpu/x86/stubGenerator_x86_64_string.cpp line 1510:

> 1508:   compare_big_haystack_to_needle(sizeKnown, size, NUMBER_OF_NEEDLE_BYTES_TO_COMPARE, loop_top, hsPtrRet, hsLength,
> 1509:                                  needleLen, isU, DO_EARLY_BAILOUT, eq_mask, temp2, r10, _masm);
> 1510: 

At this point hsLength is not the remaining length from hsPtrRet, would that cause a problem? If not, all the special paths in compare_big_haystack_to_needle need not be generated on this call.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/16753#discussion_r1602016421
PR Review Comment: https://git.openjdk.org/jdk/pull/16753#discussion_r1601943761
PR Review Comment: https://git.openjdk.org/jdk/pull/16753#discussion_r1602251994
PR Review Comment: https://git.openjdk.org/jdk/pull/16753#discussion_r1602010926