RFR: 8320448: Accelerate IndexOf using AVX2 [v20]
Sandhya Viswanathan
sviswanathan at openjdk.org
Fri May 24 22:33:13 UTC 2024
On Fri, 17 May 2024 23:47:45 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:
>
> Addressing lots of comments. Interim commit.
src/hotspot/cpu/x86/c2_MacroAssembler_x86.cpp line 4737:
> 4735: bind(COMPARE_BYTE);
> 4736: } else {
> 4737: lea(ary1, Address(ary1, expand_ary2 ? 4 : 2));
This change is not required. expand_ary2 code doesn't come here.
src/hotspot/cpu/x86/stubGenerator_x86_64_string.cpp line 1233:
> 1231: __ andq(eq_mask, rTmp);
> 1232:
> 1233: __ testl(eq_mask, eq_mask);
Mismatch of operation size q vs l: andq and testl.
src/hotspot/cpu/x86/stubGenerator_x86_64_string.cpp line 1623:
> 1621: ////////////////////////////////////////////////////////////////////////////////////////
> 1622: //
> 1623: // Small haystack (<32 bytes) switch
This should be <= 32 bytes.
src/hotspot/cpu/x86/stubGenerator_x86_64_string.cpp line 1709:
> 1707: // XMM_BYTE_K - last element of needle, broadcast
> 1708: //
> 1709: // The haystack is >= 32 bytes
Should this be > 32 bytes?
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/16753#discussion_r1609023624
PR Review Comment: https://git.openjdk.org/jdk/pull/16753#discussion_r1609043720
PR Review Comment: https://git.openjdk.org/jdk/pull/16753#discussion_r1609160143
PR Review Comment: https://git.openjdk.org/jdk/pull/16753#discussion_r1609163535
More information about the core-libs-dev
mailing list