uint256: optimize div-related functions by removing unnecessary computation

[AaronChen0]

The functions Div, SDiv and MulDivOverflow don't need the remainder, so we change the api of udivrem to pass nil to skip the last part of the computatoin of rem. This PR doesn't affect the computation of quotient quot.

Test

go test ./...

ok  	github.com/holiman/uint256	1.366s

Benchmark

goos: linux
goarch: amd64
pkg: github.com/holiman/uint256
cpu: AMD Ryzen 7 7735H with Radeon Graphics         
                                 │     old      │                 new                 │
                                 │    sec/op    │   sec/op     vs base                │
Div/small/uint256-16                3.105n ± 0%   3.105n ± 1%        ~ (p=0.988 n=10)
Div/mod64/uint256-16                21.57n ± 2%   21.20n ± 2%   -1.72% (p=0.017 n=10)
Div/mod128/uint256-16               38.65n ± 2%   37.28n ± 1%   -3.53% (p=0.000 n=10)
Div/mod192/uint256-16               34.94n ± 1%   32.80n ± 2%   -6.14% (p=0.000 n=10)
Div/mod256/uint256-16               27.29n ± 1%   23.99n ± 1%  -12.08% (p=0.000 n=10)
Mod/small/uint256-16                4.065n ± 2%   4.082n ± 1%        ~ (p=0.617 n=10)
Mod/mod64/uint256-16                27.16n ± 2%   22.75n ± 1%  -16.24% (p=0.000 n=10)
Mod/mod128/uint256-16               40.44n ± 2%   39.97n ± 2%        ~ (p=0.060 n=10)
Mod/mod192/uint256-16               36.69n ± 1%   36.53n ± 0%        ~ (p=0.101 n=10)
Mod/mod256/uint256-16               28.71n ± 1%   28.48n ± 2%   -0.78% (p=0.001 n=10)
AddMod/small/uint256-16             6.372n ± 2%   6.205n ± 2%   -2.63% (p=0.000 n=10)
AddMod/mod64/uint256-16            10.080n ± 1%   9.048n ± 0%  -10.23% (p=0.000 n=10)
AddMod/mod128/uint256-16            18.88n ± 2%   18.15n ± 1%   -3.84% (p=0.000 n=10)
AddMod/mod192/uint256-16            20.69n ± 1%   19.75n ± 0%   -4.54% (p=0.000 n=10)
AddMod/mod256/uint256-16            6.434n ± 1%   6.457n ± 1%   +0.36% (p=0.020 n=10)
MulMod/small/uint256-16             17.61n ± 1%   15.95n ± 1%   -9.40% (p=0.000 n=10)
MulMod/mod64/uint256-16             36.81n ± 2%   29.49n ± 1%  -19.88% (p=0.000 n=10)
MulMod/mod128/uint256-16            56.83n ± 2%   52.33n ± 1%   -7.91% (p=0.000 n=10)
MulMod/mod192/uint256-16            71.45n ± 2%   66.25n ± 1%   -7.28% (p=0.000 n=10)
MulMod/mod256/uint256-16            87.00n ± 1%   86.02n ± 1%        ~ (p=0.190 n=10)
MulMod/mod256/uint256r-16           37.83n ± 2%   37.84n ± 2%        ~ (p=0.839 n=10)
SDiv/large/uint256-16               35.52n ± 1%   33.66n ± 1%   -5.25% (p=0.000 n=10)
MulDivOverflow/small/uint256-16     22.30n ± 2%   21.40n ± 1%   -4.06% (p=0.000 n=10)
MulDivOverflow/div64/uint256-16     27.43n ± 1%   24.33n ± 1%  -11.29% (p=0.000 n=10)
MulDivOverflow/div128/uint256-16    41.56n ± 0%   39.32n ± 2%   -5.40% (p=0.000 n=10)
MulDivOverflow/div192/uint256-16    53.98n ± 1%   50.98n ± 1%   -5.57% (p=0.000 n=10)
MulDivOverflow/div256/uint256-16    69.49n ± 1%   65.30n ± 1%   -6.04% (p=0.000 n=10)
geomean                             24.99n        23.61n        -5.54%

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 100.00%. Comparing base (11a325c) to head (21940a9).

Additional details and impacted files

@@            Coverage Diff            @@
##            master      #169   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            5         5           
  Lines         1628      1632    +4     
=========================================
+ Hits          1628      1632    +4     

[holiman]

LGTM