mirror of
https://codeberg.org/ziglang/zig.git
synced 2025-12-09 07:08:59 +00:00
b5d5685a4e
This change:
- Adds generic implementation of the float -> integer conversion
functions floatXiYf, including support for f80
- Updates the existing implementation of integer -> float conversion
fixXiYf to support f16 and f80
- Fixes the handling of the explicit integer bit in `__trunctfxf2`
- Combines the test cases for fixXfYi/floatXiYf into a single file
- Renames `fmodl` to `fmodq`, since it operates on 128-bit floats
The new implementation for floatXiYf has been benchmarked, and generally
provides equal or better performance versus the current implementations:
Throughput (MiB/s) - Before
| u32 | i32 | u64 | i64 | u128 | i128 |
-----|----------|----------|----------|----------|----------|----------|
f16 | none | none | none | none | none | none |
f32 | 2231.67 | 2001.19 | 1745.66 | 1405.77 | 2173.99 | 1874.63 |
f64 | 1407.17 | 1055.83 | 2911.68 | 2437.21 | 1676.05 | 1476.67 |
f80 | none | none | none | none | none | none |
f128 | 327.56 | 321.25 | 645.92 | 654.52 | 1153.56 | 1096.27 |
Throughput (MiB/s) - After
| u32 | i32 | u64 | i64 | u128 | i128 |
-----|----------|----------|----------|----------|----------|----------|
f16 | 1407.61 | 1637.25 | 3555.03 | 2594.56 | 3680.60 | 3063.34 |
f32 | 2101.36 | 2122.62 | 3225.46 | 3123.86 | 2860.05 | 1985.21 |
f64 | 1395.57 | 1314.87 | 2409.24 | 2196.30 | 2384.95 | 1908.15 |
f80 | 475.53 | 457.92 | 884.50 | 812.12 | 1475.27 | 1382.16 |
f128 | 359.60 | 350.91 | 723.08 | 706.80 | 1296.42 | 1198.87 |
114 lines
3 KiB
Zig
114 lines
3 KiB
Zig
//
|
|
// SPARC uses a different naming scheme for its support routines so we map it here to the x86 name.
|
|
|
|
const std = @import("std");
|
|
const builtin = @import("builtin");
|
|
|
|
// The SPARC Architecture Manual, Version 9:
|
|
// A.13 Floating-Point Compare
|
|
const FCMP = enum(i32) {
|
|
Equal = 0,
|
|
Less = 1,
|
|
Greater = 2,
|
|
Unordered = 3,
|
|
};
|
|
|
|
// Basic arithmetic
|
|
|
|
pub fn _Qp_add(c: *f128, a: *f128, b: *f128) callconv(.C) void {
|
|
c.* = @import("addXf3.zig").__addtf3(a.*, b.*);
|
|
}
|
|
|
|
pub fn _Qp_div(c: *f128, a: *f128, b: *f128) callconv(.C) void {
|
|
c.* = @import("divtf3.zig").__divtf3(a.*, b.*);
|
|
}
|
|
|
|
pub fn _Qp_mul(c: *f128, a: *f128, b: *f128) callconv(.C) void {
|
|
c.* = @import("mulXf3.zig").__multf3(a.*, b.*);
|
|
}
|
|
|
|
pub fn _Qp_sub(c: *f128, a: *f128, b: *f128) callconv(.C) void {
|
|
c.* = @import("addXf3.zig").__subtf3(a.*, b.*);
|
|
}
|
|
|
|
// Comparison
|
|
|
|
pub fn _Qp_cmp(a: *f128, b: *f128) callconv(.C) i32 {
|
|
return @enumToInt(@import("compareXf2.zig").cmp(f128, FCMP, a.*, b.*));
|
|
}
|
|
|
|
pub fn _Qp_feq(a: *f128, b: *f128) callconv(.C) bool {
|
|
return _Qp_cmp(a, b) == @enumToInt(FCMP.Equal);
|
|
}
|
|
|
|
pub fn _Qp_fne(a: *f128, b: *f128) callconv(.C) bool {
|
|
return _Qp_cmp(a, b) != @enumToInt(FCMP.Equal);
|
|
}
|
|
|
|
pub fn _Qp_flt(a: *f128, b: *f128) callconv(.C) bool {
|
|
return _Qp_cmp(a, b) == @enumToInt(FCMP.Less);
|
|
}
|
|
|
|
pub fn _Qp_fle(a: *f128, b: *f128) callconv(.C) bool {
|
|
const cmp = _Qp_cmp(a, b);
|
|
return cmp == @enumToInt(FCMP.Less) or cmp == @enumToInt(FCMP.Equal);
|
|
}
|
|
|
|
pub fn _Qp_fgt(a: *f128, b: *f128) callconv(.C) bool {
|
|
return _Qp_cmp(a, b) == @enumToInt(FCMP.Greater);
|
|
}
|
|
|
|
pub fn _Qp_fge(a: *f128, b: *f128) callconv(.C) bool {
|
|
const cmp = _Qp_cmp(a, b);
|
|
return cmp == @enumToInt(FCMP.Greater) or cmp == @enumToInt(FCMP.Equal);
|
|
}
|
|
|
|
// Conversion
|
|
|
|
pub fn _Qp_itoq(c: *f128, a: i32) callconv(.C) void {
|
|
c.* = @import("floatXiYf.zig").__floatsitf(a);
|
|
}
|
|
|
|
pub fn _Qp_uitoq(c: *f128, a: u32) callconv(.C) void {
|
|
c.* = @import("floatXiYf.zig").__floatunsitf(a);
|
|
}
|
|
|
|
pub fn _Qp_xtoq(c: *f128, a: i64) callconv(.C) void {
|
|
c.* = @import("floatXiYf.zig").__floatditf(a);
|
|
}
|
|
|
|
pub fn _Qp_uxtoq(c: *f128, a: u64) callconv(.C) void {
|
|
c.* = @import("floatXiYf.zig").__floatunditf(a);
|
|
}
|
|
|
|
pub fn _Qp_stoq(c: *f128, a: f32) callconv(.C) void {
|
|
c.* = @import("extendXfYf2.zig").__extendsftf2(a);
|
|
}
|
|
|
|
pub fn _Qp_dtoq(c: *f128, a: f64) callconv(.C) void {
|
|
c.* = @import("extendXfYf2.zig").__extenddftf2(a);
|
|
}
|
|
|
|
pub fn _Qp_qtoi(a: *f128) callconv(.C) i32 {
|
|
return @import("fixXfYi.zig").__fixtfsi(a.*);
|
|
}
|
|
|
|
pub fn _Qp_qtoui(a: *f128) callconv(.C) u32 {
|
|
return @import("fixXfYi.zig").__fixunstfsi(a.*);
|
|
}
|
|
|
|
pub fn _Qp_qtox(a: *f128) callconv(.C) i64 {
|
|
return @import("fixXfYi.zig").__fixtfdi(a.*);
|
|
}
|
|
|
|
pub fn _Qp_qtoux(a: *f128) callconv(.C) u64 {
|
|
return @import("fixXfYi.zig").__fixunstfdi(a.*);
|
|
}
|
|
|
|
pub fn _Qp_qtos(a: *f128) callconv(.C) f32 {
|
|
return @import("truncXfYf2.zig").__trunctfsf2(a.*);
|
|
}
|
|
|
|
pub fn _Qp_qtod(a: *f128) callconv(.C) f64 {
|
|
return @import("truncXfYf2.zig").__trunctfdf2(a.*);
|
|
}
|