[fpc-pascal] Program efficiency

[Vojtěch Čihák]

Hi,
 
I noticed this too a few years ago when I translated xxHash from C to FPC.
Pushing the code to less lines improved speed so xxHash32 was as fast as in C.
 
do { v1 = XXH32_round(v1, XXH_get32bits(input)); input += 4; v2 = XXH32_round(v2, XXH_get32bits(input)); input += 4; v3 = XXH32_round(v3, XXH_get32bits(input)); input += 4; v4 = XXH32_round(v4, XXH_get32bits(input)); input += 4; } while (input < limit);
 
 
    repeat
        v1 := cPrime32x1 * RolDWord(v1 + cPrime32x2 * PLongWord(ABuffer)^, 13);
        v2 := cPrime32x1 * RolDWord(v2 + cPrime32x2 * PLongWord(ABuffer+4)^, 13);
        v3 := cPrime32x1 * RolDWord(v3 + cPrime32x2 * PLongWord(ABuffer+8)^, 13);
        v4 := cPrime32x1 * RolDWord(v4 + cPrime32x2 * PLongWord(ABuffer+12)^, 13);
        inc(ABuffer, 16);
      until not (ABuffer <= pLimit);       
 
______________________________________________________________
> Od: "James Richters via fpc-pascal" <fpc-pascal at lists.freepascal.org>
> Komu: "'FPC-Pascal users discussions'" <fpc-pascal at lists.freepascal.org>
> Datum: 09.11.2022 13:47
> Předmět: [fpc-pascal] Program efficiency
>
I was wondering if breaking up large complicated formulas into smaller
 sections that use more variables is less efficient than just the large
 formula.
 
 In other words.. is
 
 A:=(B+C/D)-E^F;
 G:=H*(I+J);
 K:=L+M/N;
 O:= A+G-K;
 
 Less efficient than
 O:= ((B+C/D)-E^F)+ (H*(I+J))-(L+M/N);
 

 James
 
 _______________________________________________
 fpc-pascal maillist  -  fpc-pascal at lists.freepascal.org
 https://lists.freepascal.org/cgi-bin/mailman/listinfo/fpc-pascal <https://lists.freepascal.org/cgi-bin/mailman/listinfo/fpc-pascal>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freepascal.org/pipermail/fpc-pascal/attachments/20221109/0a220612/attachment.htm>