Discussion on compiler differences under different build commands

[tzSharing]

We all know that there are many ways to build a v compiler. What's the difference? Which is better? So I tried to do some simple comparisons and tests.

v doctor

V full version: V 0.3.4 34163ae
OS: windows, Microsoft Windows 10 רҵ v19045 64 λ
Processor: 4 cpus, 64bit, little endian,

getwd: D:\v
vexe: D:\v\v.exe
vexe mtime: 2023-06-15 18:57:44

vroot: OK, value: D:\v
VMODULES: OK, value: C:\Users\TZ\.vmodules
VTMP: OK, value: C:\Users\TZ\AppData\Local\Temp\v_0

Git version: git version 2.36.0.windows.1
Git vroot status: weekly.2023.24-8-g34163ae4 (12 commit(s) behind V master)
.git/config present: true

CC version: Error: exec failed (CreateProcess) with code 2: 系统找不到指定的文件。
 cmd: cc --version
thirdparty/tcc status: thirdparty-windows-amd64 e90c2620

result

command	v.exe size	benchmark(quintic average)
make -tcc	9464Kb	1711.1916ms
make -gcc	6831Kb	1732.874ms
make -clang	compile error	-
v self	9429Kb	1677.5034ms
v self -prod	3706Kb	1695.5788ms

benchmark code

module main

import rand
import benchmark

struct Person {
	id   int
	name string
	age  int
mut:
	group string
}

fn main() {
	mut b := benchmark.start()

	mut arr := []Person{}
	num := 1000000 //百万
	for i in 0 .. num {
		p := Person{
			id: rand.int_in_range(1, num) or { 1 }
			name: 'tom${i}'
			age: i % 150
			group: ''
		}
		arr << p
	}

	for index, mut item in arr {
		if index % 2 == 0 {
			item.group = 'A'
		} else {
			item.group = 'B'
		}
	}

	arr.sort_with_compare(fn (a &Person, b &Person) int {
		if a.id < b.id {
			return -1
		}
		if a.id > b.id {
			return 1
		}
		return 0
	})

	// println('总数=${arr.len}')
	// println('后10=${arr[arr.len - 10..]}')
	b.measure('stop')
}

Conclusion

As you can see, the biggest difference is only the speed and size of v.exe generation, although the performance is also different but not obvious, so make-tcc or-gcc is the best choice, and the speed of generating v.exe is very fast. Of course, this is just a simple and not very rigorous test, welcome to continue research and discussion.