From 18accade5bb062cb17705fc2816cddd08f4eefac Mon Sep 17 00:00:00 2001 From: David Ellis Date: Mon, 11 Nov 2024 14:19:21 -0600 Subject: [PATCH 1/4] Implement trigonometry functions for the CPU and GPU --- alan/src/compile/integration_tests.rs | 263 ++++++++------- alan/src/std/root.ln | 446 +++++++++++++++++++++++++- 2 files changed, 599 insertions(+), 110 deletions(-) diff --git a/alan/src/compile/integration_tests.rs b/alan/src/compile/integration_tests.rs index e086653f..285b8e69 100644 --- a/alan/src/compile/integration_tests.rs +++ b/alan/src/compile/integration_tests.rs @@ -990,6 +990,14 @@ test_full!(f32_abs => r#" export fn main = print(-3.f32.abs);"#; stdout "3\n"; ); +test_full!(f32_floor => r#" + export fn main = 2.5.f32.floor.print;"#; + stdout "2\n"; +); +test_full!(f32_ceil => r#" + export fn main = 2.5.f32.ceil.print;"#; + stdout "3\n"; +); test_full!(f64_add => r#" export fn main { @@ -1047,6 +1055,14 @@ test_full!(f64_abs => r#" export fn main = print(-3.0.abs);"#; stdout "3\n"; ); +test_full!(f64_floor => r#" + export fn main = 2.5.floor.print;"#; + stdout "2\n"; +); +test_full!(f64_ceil => r#" + export fn main = 2.5.ceil.print;"#; + stdout "3\n"; +); test_full!(grouping => r#" export fn main { @@ -3007,133 +3023,162 @@ test_ignore!(module_level_constant_from_function_call => r#" stdout "3\n5\n"; ); -// @std/trig +// Trigonometry + +test_full!(cpu_trig => r#" + export fn main { + 'Logarithms and e^x'.print; + print(exp(e).string(4)); + print(ln(e).string(4)); + print(log10(e).string(4)); + print(log2(e).string(4)); -test_ignore!(std_trig => r#" - import @std/trig - from @std/trig import e, pi, tau - // shouldn't be necessary, but compiler issue makes it so + 'Basic Trig functions'.print; + print(sin(tau / 6.0).string(4)); + print(cos(tau / 6.0).string(4)); + print(tan(tau / 6.0).string(4)); + print(sec(tau / 6.0).string(4)); + print(csc(tau / 6.0).string(4)); + print(cot(tau / 6.0).string(4)); + + 'Inverse Trig functions'.print; + asin(0.0).string(4).print; + acos(1.0).string(4).print; + atan(0.0).string(4).print; + atan2(1.0, 2.0).string(4).print; + print(asec(tau / 6.0).string(4)); + print(acsc(tau / 6.0).string(4)); + print(acot(tau / 6.0).string(4)); + 'Hyperbolic Trig functions'.print; + print(sinh(tau / 6.0).string(4)); + print(cosh(tau / 6.0).string(4)); + print(tanh(tau / 6.0).string(4)); + print(sech(tau / 6.0).string(4)); + print(csch(tau / 6.0).string(4)); + print(coth(tau / 6.0).string(4)); + + 'Inverse Hyperbolic Trig functions'.print; + print(asinh(tau / 6.0).string(4)); + print(acosh(tau / 6.0).string(4)); + print(atanh(pi / 6.0).string(4)); + print(asech(0.5).string(4)); + print(acsch(tau / 6.0).string(4)); + print(acoth(tau / 6.0).string(4)); + }"#; + stdout r#"Logarithms and e^x +15.1543 +1.0000 +0.4343 +1.4427 +Basic Trig functions +0.8660 +0.5000 +1.7321 +2.0000 +1.1547 +0.5774 +Inverse Trig functions +0.0000 +0.0000 +0.0000 +0.4636 +0.3014 +1.2694 +0.7623 +Hyperbolic Trig functions +1.2494 +1.6003 +0.7807 +0.6249 +0.8004 +1.2809 +Inverse Hyperbolic Trig functions +0.9144 +0.3060 +0.5813 +1.3170 +0.8491 +1.8849 +"#; +); + +test_gpgpu!(gpu_trig => r#" export fn main { 'Logarithms and e^x'.print; - print(trig.exp(e)); - print(trig.ln(e)); - print(trig.log(e)); + // Contrived way to get the GPU to do this work, don't follow this pattern for real GPU usage + GBuffer([e.f32]).map(fn (v: gf32) = exp(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([e.f32]).map(fn (v: gf32) = ln(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([e.f32]).map(fn (v: gf32) = log10(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([e.f32]).map(fn (v: gf32) = log2(v)).read{f32}[0].getOrExit.string(4).print; 'Basic Trig functions'.print; - print(trig.sin(tau / 6.0)); - print(trig.cos(tau / 6.0)); - print(trig.tan(tau / 6.0)); - print(trig.sec(tau / 6.0)); - print(trig.csc(tau / 6.0)); - print(trig.cot(tau / 6.0)); + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sin(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cos(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = tan(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sec(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = csc(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cot(v)).read{f32}[0].getOrExit.string(4).print; 'Inverse Trig functions'.print; - print(trig.arcsine(0.0)); - print(trig.arccosine(1.0)); - print(trig.arctangent(0.0)); - print(trig.arcsecant(tau / 6.0)); - print(trig.arccosecant(tau / 6.0)); - print(trig.arccotangent(tau / 6.0)); - - 'Historic Trig functions (useful for navigation and as a teaching aid: https://en.wikipedia.org/wiki/File:Circle-trig6.svg )'.print; - print(trig.versine(pi / 3.0)); - print(trig.vercosine(pi / 3.0)); - print(trig.coversine(pi / 3.0)); - print(trig.covercosine(pi / 3.0)); - print(trig.haversine(pi / 3.0)); - print(trig.havercosine(pi / 3.0)); - print(trig.hacoversine(pi / 3.0)); - print(trig.hacovercosine(pi / 3.0)); - print(trig.exsecant(pi / 3.0)); - print(trig.excosecant(pi / 3.0)); - print(trig.chord(pi / 3.0)); - - 'Historic Inverse Trig functions'.print; - print(trig.aver(0.0)); - print(trig.avcs(0.5)); - print(trig.acvs(1.0)); - print(trig.acvc(1.0)); - print(trig.ahav(0.5)); - print(trig.ahvc(0.5)); - print(trig.ahcv(0.5)); - print(trig.ahcc(0.5)); - print(trig.aexs(0.5)); - print(trig.aexc(0.5)); - print(trig.acrd(0.5)); + GBuffer([0.0.f32]).map(fn (v: gf32) = asin(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([1.0.f32]).map(fn (v: gf32) = acos(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([0.0.f32]).map(fn (v: gf32) = atan(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([1.0.f32]).map(fn (v: gf32) = atan2(v, 2.0)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = asec(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acsc(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acot(v)).read{f32}[0].getOrExit.string(4).print; 'Hyperbolic Trig functions'.print; - print(trig.sinh(tau / 6.0)); - print(trig.cosh(tau / 6.0)); - print(trig.tanh(tau / 6.0)); - print(trig.sech(tau / 6.0)); - print(trig.csch(tau / 6.0)); - print(trig.coth(tau / 6.0)); + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sinh(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cosh(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = tanh(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sech(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = csch(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = coth(v)).read{f32}[0].getOrExit.string(4).print; 'Inverse Hyperbolic Trig functions'.print; - print(trig.hyperbolicArcsine(tau / 6.0)); - print(trig.hyperbolicArccosine(tau / 6.0)); - print(trig.hyperbolicArctangent(tau / 6.0)); - print(trig.hyperbolicArcsecant(0.5)); - print(trig.hyperbolicArccosecant(tau / 6.0)); - print(trig.hyperbolicArccotangent(tau / 6.0)); + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = asinh(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acosh(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([pi.f32 / 6.0.f32]).map(fn (v: gf32) = atanh(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([0.5.f32]).map(fn (v: gf32) = asech(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acsch(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acoth(v)).read{f32}[0].getOrExit.string(4).print; }"#; stdout r#"Logarithms and e^x -15.154262241479259 -1 -0.4342944819032518 +15.1543 +1.0000 +0.4343 +1.4427 Basic Trig functions -0.8660254037844386 -0.5000000000000001 -1.7320508075688767 -1.9999999999999996 -1.1547005383792517 -0.577350269189626 +0.8660 +0.5000 +1.7321 +2.0000 +1.1547 +0.5774 Inverse Trig functions -0 -0 -0 -0.3013736097452911 -1.2694227170496055 -0.7623475341648746 -Historic Trig functions (useful for navigation and as a teaching aid: https://en.wikipedia.org/wiki/File:Circle-trig6.svg ) -0.4999999999999999 -1.5 -0.1339745962155614 -1.8660254037844386 -0.24999999999999994 -0.75 -0.0669872981077807 -0.9330127018922193 -0.9999999999999996 -0.15470053837925168 -0.9999999999999999 -Historic Inverse Trig functions -0 -2.0943951023931957 -0 -0 -1.5707963267948966 -1.5707963267948966 -0 -0 -0.8410686705679303 -0.7297276562269663 -0.5053605102841573 +0.0000 +0.0000 +0.0000 +0.4636 +0.3013 +1.2698 +0.7623 Hyperbolic Trig functions -1.2493670505239751 -1.600286857702386 -0.7807144353592677 -0.6248879662960872 -0.8004052928885931 -1.2808780710450447 +1.2494 +1.6003 +0.7807 +0.6249 +0.8004 +1.2809 Inverse Hyperbolic Trig functions -0.9143566553928857 -0.3060421086132653 -1.8849425394276085 -1.3169578969248166 -0.8491423010640059 -1.8849425394276085 +0.9144 +0.3060 +0.5813 +1.3170 +0.8491 +1.8849 "#; ); diff --git a/alan/src/std/root.ln b/alan/src/std/root.ln index a3de33e9..240b4d7a 100644 --- a/alan/src/std/root.ln +++ b/alan/src/std/root.ln @@ -179,6 +179,11 @@ export type{Js} Dict{K, V} = Binds{"Map", K, V}; export type{Rs} Set{V} = Binds{"std::collections::HashSet", V}; export type{Js} Set{V} = Binds{"Set", V}; +// Basic trig constants (this language is meant for GPGPU, this makes sense in the root scope) +export const e = 2.718281828459045; +export const pi = 3.141592653589793; +export const tau = 6.283185307179586; + /// Functions for (potentially) every type export fn{Rs} clone{T} (v: T) -> T = {Method{"clone"} :: T -> T}(v); // TODO: This needs to be turned into a JS function that's bound @@ -534,6 +539,7 @@ export fn{Rs} gte Infix{">="} :: (Deref{i8}, Deref{i8}) -> bool; export fn{Js} gte "((a, b) => new alan_std.Bool(a.val >= b.val))" <- RootBacking :: (i8, i8) -> bool; export fn min (a: i8, b: i8) = if(a.lte(b), a, b); export fn max (a: i8, b: i8) = if(a.gte(b), a, b); +export fn clamp(v: i8, l: i8, h: i8) = if(v.lte(l), l, if(v.gte(h), h, v)); export fn{Rs} shl (a: i8, b: i8) = {Method{"wrapping_shl"} :: (i8, Deref{u32}) -> i8}(a, b.u32); export fn{Js} shl Method{"wrappingShl"} :: (i8, i8) -> i8; export fn{Rs} shr (a: i8, b: i8) = {Method{"wrapping_shr"} :: (i8, Deref{u32}) -> i8}(a, b.u32); @@ -584,6 +590,7 @@ export fn{Rs} gte Infix{">="} :: (Deref{i16}, Deref{i16}) -> bool; export fn{Js} gte "((a, b) => new alan_std.Bool(a.val >= b.val))" <- RootBacking :: (i16, i16) -> bool; export fn min (a: i16, b: i16) = if(a.lte(b), a, b); export fn max (a: i16, b: i16) = if(a.gte(b), a, b); +export fn clamp(v: i16, l: i16, h: i16) = if(v.lte(l), l, if(v.gte(h), h, v)); export fn{Rs} shl (a: i16, b: i16) = {Method{"wrapping_shl"} :: (i16, Deref{u32}) -> i16}(a, b.u32); export fn{Js} shl Method{"wrappingShl"} :: (i16, i16) -> i16; export fn{Rs} shr (a: i16, b: i16) = {Method{"wrapping_shr"} :: (i16, Deref{u32}) -> i16}(a, b.u32); @@ -634,6 +641,7 @@ export fn{Rs} gte Infix{">="} :: (Deref{i32}, Deref{i32}) -> bool; export fn{Js} gte "((a, b) => new alan_std.Bool(a.val >= b.val))" :: (i32, i32) -> bool; export fn min (a: i32, b: i32) = if(a.lte(b), a, b); export fn max (a: i32, b: i32) = if(a.gte(b), a, b); +export fn clamp(v: i32, l: i32, h: i32) = if(v.lte(l), l, if(v.gte(h), h, v)); export fn{Rs} shl (a: i32, b: i32) = {Method{"wrapping_shl"} :: (i32, Deref{u32}) -> i32}(a, b.u32); export fn{Js} shl Method{"wrappingShl"} :: (i32, i32) -> i32; export fn{Rs} shr (a: i32, b: i32) = {Method{"wrapping_shr"} :: (i32, Deref{u32}) -> i32}(a, b.u32); @@ -684,6 +692,7 @@ export fn{Rs} gte Infix{">="} :: (Deref{i64}, Deref{i64}) -> bool; export fn{Js} gte "((a, b) => new alan_std.Bool(a.val >= b.val))" <- RootBacking :: (i64, i64) -> bool; export fn min (a: i64, b: i64) = if(a.lte(b), a, b); export fn max (a: i64, b: i64) = if(a.gte(b), a, b); +export fn clamp(v: i64, l: i64, h: i64) = if(v.lte(l), l, if(v.gte(h), h, v)); export fn{Rs} shl (a: i64, b: i64) = {Method{"wrapping_shl"} :: (i64, Deref{u32}) -> i64}(a, b.u32); export fn{Js} shl Method{"wrappingShl"} :: (i64, i64) -> i64; export fn{Rs} shr (a: i64, b: i64) = {Method{"wrapping_shr"} :: (i64, Deref{u32}) -> i64}(a, b.u32); @@ -731,6 +740,7 @@ export fn{Rs} gte Infix{">="} :: (Deref{u8}, Deref{u8}) -> bool; export fn{Js} gte "((a, b) => new alan_std.Bool(a.val >= b.val))" <- RootBacking :: (u8, u8) -> bool; export fn min (a: u8, b: u8) = if(a.lte(b), a, b); export fn max (a: u8, b: u8) = if(a.gte(b), a, b); +export fn clamp(v: u8, l: u8, h: u8) = if(v.lte(l), l, if(v.gte(h), h, v)); export fn{Rs} shl (a: u8, b: u8) = {Method{"wrapping_shl"} :: (u8, Deref{u32}) -> u8}(a, b.u32); export fn{Js} shl Method{"wrappingShl"} :: (u8, u8) -> u8; export fn{Rs} shr (a: u8, b: u8) = {Method{"wrapping_shr"} :: (u8, Deref{u32}) -> u8}(a, b.u32); @@ -777,6 +787,7 @@ export fn{Rs} gte Infix{">="} :: (Deref{u16}, Deref{u16}) -> bool; export fn{Js} gte "((a, b) => new alan_std.Bool(a.val >= b.val))" <- RootBacking :: (u16, u16) -> bool; export fn min (a: u16, b: u16) = if(a.lte(b), a, b); export fn max (a: u16, b: u16) = if(a.gte(b), a, b); +export fn clamp(v: u16, l: u16, h: u16) = if(v.lte(l), l, if(v.gte(h), h, v)); export fn{Rs} shl (a: u16, b: u16) = {Method{"wrapping_shl"} :: (u16, Deref{u32}) -> u16}(a, b.u32); export fn{Js} shl Method{"wrappingShl"} :: (u16, u16) -> u16; export fn{Rs} shr (a: u16, b: u16) = {Method{"wrapping_shr"} :: (u16, Deref{u32}) -> u16}(a, b.u32); @@ -823,6 +834,7 @@ export fn{Rs} gte Infix{">="} :: (Deref{u32}, Deref{u32}) -> bool; export fn{Js} gte "((a, b) => new alan_std.Bool(a.val >= b.val))" <- RootBacking :: (u32, u32) -> bool; export fn min (a: u32, b: u32) = if(a.lte(b), a, b); export fn max (a: u32, b: u32) = if(a.gte(b), a, b); +export fn clamp(v: u32, l: u32, h: u32) = if(v.lte(l), l, if(v.gte(h), h, v)); export fn{Rs} shl (a: u32, b: u32) = {Method{"wrapping_shl"} :: (u32, Deref{u32}) -> u32}(a, b.u32); export fn{Js} shl Method{"wrappingShl"} :: (u32, u32) -> u32; export fn{Rs} shr (a: u32, b: u32) = {Method{"wrapping_shr"} :: (u32, Deref{u32}) -> u32}(a, b.u32); @@ -869,6 +881,7 @@ export fn{Rs} gte Infix{">="} :: (Deref{u64}, Deref{u64}) -> bool; export fn{Js} gte "((a, b) => new alan_std.Bool(a.val >= b.val))" <- RootBacking :: (u64, u64) -> bool; export fn min (a: u64, b: u64) = if(a.lte(b), a, b); export fn max (a: u64, b: u64) = if(a.gte(b), a, b); +export fn clamp(v: u64, l: u64, h: u64) = if(v.lte(l), l, if(v.gte(h), h, v)); export fn{Rs} shl (a: u64, b: u64) = {Method{"wrapping_shl"} :: (u64, Deref{u32}) -> u64}(a, b.u32); export fn{Js} shl Method{"wrappingShl"} :: (u64, u64) -> u64; export fn{Rs} shr (a: u64, b: u64) = {Method{"wrapping_shr"} :: (u64, Deref{u32}) -> u64}(a, b.u32); @@ -909,6 +922,57 @@ export fn{Rs} gte Infix{">="} :: (Deref{f32}, Deref{f32}) -> bool; export fn{Js} gte "((a, b) => new alan_std.Bool(a.val >= b.val))" <- RootBacking :: (f32, f32) -> bool; export fn min (a: f32, b: f32) = if(a.lte(b), a, b); export fn max (a: f32, b: f32) = if(a.gte(b), a, b); +export fn clamp(v: f32, l: f32, h: f32) = if(v.lte(l), l, if(v.gte(h), h, v)); +export fn{Rs} floor Method{"floor"} :: f32 -> f32; +export fn{Js} floor "((a) => new alan_std.F32(Math.floor(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} ceil Method{"ceil"} :: f32 -> f32; +export fn{Js} ceil "((a) => new alan_std.F32(Math.ceil(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} acos Method{"acos"} :: f32 -> f32; +export fn{Js} acos "((a) => new alan_std.F32(Math.acos(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} acosh Method{"acosh"} :: f32 -> f32; +export fn{Js} acosh "((a) => new alan_std.F32(Math.acosh(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} asin Method{"asin"} :: f32 -> f32; +export fn{Js} asin "((a) => new alan_std.F32(Math.asin(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} asinh Method{"asinh"} :: f32 -> f32; +export fn{Js} asinh "((a) => new alan_std.F32(Math.asinh(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} atan Method{"atan"} :: f32 -> f32; +export fn{Js} atan "((a) => new alan_std.F32(Math.atan(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} atanh Method{"atanh"} :: f32 -> f32; +export fn{Js} atanh "((a) => new alan_std.F32(Math.atanh(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} atan2 Method{"atan2"} :: (f32, Deref{f32}) -> f32; +export fn{Js} atan2 "((a, b) => new alan_std.F32(Math.atan2(a.val, b.val)))" <- RootBacking :: (f32, f32) -> f32; +export fn{Rs} cos Method{"cos"} :: f32 -> f32; +export fn{Js} cos "((a) => new alan_std.F32(Math.cos(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} cosh Method{"cosh"} :: f32 -> f32; +export fn{Js} cosh "((a) => new alan_std.F32(Math.cosh(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} sin Method{"sin"} :: f32 -> f32; +export fn{Js} sin "((a) => new alan_std.F32(Math.sin(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} sinh Method{"sinh"} :: f32 -> f32; +export fn{Js} sinh "((a) => new alan_std.F32(Math.sinh(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} tan Method{"tan"} :: f32 -> f32; +export fn{Js} tan "((a) => new alan_std.F32(Math.tan(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} tanh Method{"tanh"} :: f32 -> f32; +export fn{Js} tanh "((a) => new alan_std.F32(Math.tanh(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} exp Method{"exp"} :: f32 -> f32; +export fn{Js} exp "((a) => new alan_std.F32(Math.exp(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} ln Method{"ln"} :: f32 -> f32; +export fn{Js} ln "((a) => new alan_std.F32(Math.log(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} log2 Method{"log2"} :: f32 -> f32; +export fn{Js} log2 "((a) => new alan_std.F32(Math.log2(a.val)))" <- RootBacking :: f32 -> f32; +export fn{Rs} log10 Method{"log10"} :: f32 -> f32; +export fn{Js} log10 "((a) => new alan_std.F32(Math.log10(a.val)))" <- RootBacking :: f32 -> f32; +export fn sec(x: f32) = 1.0.f32.div(cos(x)); +export fn csc(x: f32) = 1.0.f32.div(sin(x)); +export fn cot(x: f32) = 1.0.f32.div(tan(x)); +export fn asec(x: f32) = acos(1.0.f32.div(x)); +export fn acsc(x: f32) = asin(1.0.f32.div(x)); +export fn acot(x: f32) = pi.f32.div(2.0.f32).sub(atan(x)); +export fn sech(x: f32) = 1.0.f32.div(cosh(x)); +export fn csch(x: f32) = 1.0.f32.div(sinh(x)); +export fn coth(x: f32) = 1.0.f32.div(tanh(x)); +export fn asech(x: f32) = ln(1.0.f32.add(sqrt(1.0.f32.sub(x.pow(2.0.f32)))).div(x)); +export fn acsch(x: f32) = ln(1.0.f32.div(x).add(sqrt(1.0.f32.div(x.pow(2.0.f32)).add(1.0.f32)))); +export fn acoth(x: f32) = ln(x.add(1.0.f32).div(x.sub(1.0.f32))).div(2.0.f32); export fn{Rs} add Infix{"+"} :: (f64, f64) -> f64; export fn{Js} add "((a, b) => new alan_std.F64(a.val + b.val))" <- RootBacking :: (f64, f64) -> f64; @@ -940,6 +1004,57 @@ export fn{Rs} gte Infix{">="} :: (Deref{f64}, Deref{f64}) -> bool; export fn{Js} gte "((a, b) => new alan_std.F64(a.val >= b.val))" <- RootBacking :: (f64, f64) -> bool; export fn min (a: f64, b: f64) = if(a.lte(b), a, b); export fn max (a: f64, b: f64) = if(a.gte(b), a, b); +export fn clamp(v: f64, l: f64, h: f64) = if(v.lte(l), l, if(v.gte(h), h, v)); +export fn{Rs} floor Method{"floor"} :: f64 -> f64; +export fn{Js} floor "((a) => new alan_std.F64(Math.floor(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} ceil Method{"ceil"} :: f64 -> f64; +export fn{Js} ceil "((a) => new alan_std.F64(Math.ceil(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} acos Method{"acos"} :: f64 -> f64; +export fn{Js} acos "((a) => new alan_std.F64(Math.acos(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} acosh Method{"acosh"} :: f64 -> f64; +export fn{Js} acosh "((a) => new alan_std.F64(Math.acosh(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} asin Method{"asin"} :: f64 -> f64; +export fn{Js} asin "((a) => new alan_std.F64(Math.asin(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} asinh Method{"asinh"} :: f64 -> f64; +export fn{Js} asinh "((a) => new alan_std.F64(Math.asinh(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} atan Method{"atan"} :: f64 -> f64; +export fn{Js} atan "((a) => new alan_std.F64(Math.atan(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} atanh Method{"atanh"} :: f64 -> f64; +export fn{Js} atanh "((a) => new alan_std.F64(Math.atanh(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} atan2 Method{"atan2"} :: (f64, Deref{f64}) -> f64; +export fn{Js} atan2 "((a, b) => new alan_std.F64(Math.atan2(a.val, b.val)))" <- RootBacking :: (f64, f64) -> f64; +export fn{Rs} cos Method{"cos"} :: f64 -> f64; +export fn{Js} cos "((a) => new alan_std.F64(Math.cos(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} cosh Method{"cosh"} :: f64 -> f64; +export fn{Js} cosh "((a) => new alan_std.F64(Math.cosh(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} sin Method{"sin"} :: f64 -> f64; +export fn{Js} sin "((a) => new alan_std.F64(Math.sin(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} sinh Method{"sinh"} :: f64 -> f64; +export fn{Js} sinh "((a) => new alan_std.F64(Math.sinh(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} tan Method{"tan"} :: f64 -> f64; +export fn{Js} tan "((a) => new alan_std.F64(Math.tan(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} tanh Method{"tanh"} :: f64 -> f64; +export fn{Js} tanh "((a) => new alan_std.F64(Math.tanh(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} exp Method{"exp"} :: f64 -> f64; +export fn{Js} exp "((a) => new alan_std.F64(Math.exp(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} ln Method{"ln"} :: f64 -> f64; +export fn{Js} ln "((a) => new alan_std.F64(Math.log(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} log2 Method{"log2"} :: f64 -> f64; +export fn{Js} log2 "((a) => new alan_std.F64(Math.log2(a.val)))" <- RootBacking :: f64 -> f64; +export fn{Rs} log10 Method{"log10"} :: f64 -> f64; +export fn{Js} log10 "((a) => new alan_std.F64(Math.log10(a.val)))" <- RootBacking :: f64 -> f64; +export fn sec(x: f64) = 1.0.div(cos(x)); +export fn csc(x: f64) = 1.0.div(sin(x)); +export fn cot(x: f64) = 1.0.div(tan(x)); +export fn asec(x: f64) = acos(1.0.div(x)); +export fn acsc(x: f64) = asin(1.0.div(x)); +export fn acot(x: f64) = pi.div(2.0).sub(atan(x)); +export fn sech(x: f64) = 1.0.div(cosh(x)); +export fn csch(x: f64) = 1.0.div(sinh(x)); +export fn coth(x: f64) = 1.0.div(tanh(x)); +export fn asech(x: f64) = ln(1.0.add(sqrt(1.0.sub(x.pow(2.0)))).div(x)); +export fn acsch(x: f64) = ln(1.0.div(x).add(sqrt(1.0.div(x.pow(2.0)).add(1.0)))); +export fn acoth(x: f64) = ln(x.add(1.0).div(x.sub(1.0))).div(2.0); /// String related bindings export fn{Rs} string "format!" :: ("{}", i8) -> string; @@ -962,6 +1077,15 @@ export fn{Rs} string "format!" :: ("{}", f32) -> string; export fn{Js} string "new alan_std.Str" <- RootBacking :: f32 -> string; export fn{Rs} string "format!" :: ("{}", f64) -> string; export fn{Js} string "new alan_std.Str" <- RootBacking :: f64 -> string; +// TODO: Until better printing is ready? Or perhaps this is a better fit for Alan +export fn{Rs} string (f: f32, p: i64) = {"format!" :: ("{0:.1$}", f32, Deref{Binds{"usize"}}) -> string}( + f, + {Cast{"usize"} :: Deref{i64} -> Binds{"usize"}}(p)); +export fn{Js} string "((v, p) => new alan_std.Str(v.valueOf().toFixed(Number(p))))" <- RootBacking :: (f32, i64) -> string; +export fn{Rs} string (f: f64, p: i64) = {"format!" :: ("{0:.1$}", f64, Deref{Binds{"usize"}}) -> string}( + f, + {Cast{"usize"} :: Deref{i64} -> Binds{"usize"}}(p)); +export fn{Js} string "((v, p) => new alan_std.Str(v.valueOf().toFixed(Number(p))))" <- RootBacking :: (f64, i64) -> string; export fn string(b: bool) = if(b, "true", "false"); export fn string(s: string) = s; export fn{Rs} concat "format!" :: ("{}{}", string, string) -> string; @@ -1510,7 +1634,7 @@ export fn gi32(gf: gf32) = gPrimitiveConvert{gf32, gi32}(gf); export fn gi32(gb: gbool) = gPrimitiveConvert{gbool, gi32}(gb); export fn gi32{T}(i: T) = gi32(i.i32); -export fn gf32(f: f32) = gf32(f.string, Dict{string, string}(), Set{GBuffer}()); +export fn gf32(f: f32) = gf32(if(f.string.eq(f.i32.string), f.string(1), f.string), Dict{string, string}(), Set{GBuffer}()); export fn gf32(gu: gu32) = gPrimitiveConvert{gu32, gf32}(gu); export fn gf32(gi: gi32) = gPrimitiveConvert{gi32, gf32}(gi); export fn gf32(gb: gbool) = gPrimitiveConvert{gbool, gf32}(gb); @@ -1552,18 +1676,22 @@ fn gvec2Primitive{I, O}(a: I, b: I) { export fn gvec2u() = gvec2u("vec2u()", Dict{string, string}(), Set{GBuffer}()); export fn gvec2u(a: gu32, b: gu32) = gvec2Primitive{gu32, gvec2u}(a, b); export fn gvec2u{T}(a: T, b: T) = gvec2u(a.gu32, b.gu32); +export fn gvec2u{T}(a: T) = gvec2u(a, a); export fn gvec2i() = gvec2i("vec2i()", Dict{string, string}(), Set{GBuffer}()); export fn gvec2i(a: gi32, b: gi32) = gvec2Primitive{gi32, gvec2i}(a, b); export fn gvec2i{T}(a: T, b: T) = gvec2i(a.gi32, b.gi32); +export fn gvec2i{T}(a: T) = gvec2i(a, a); export fn gvec2f() = gvec2f("vec2f()", Dict{string, string}(), Set{GBuffer}()); export fn gvec2f(a: gf32, b: gf32) = gvec2Primitive{gf32, gvec2f}(a, b); export fn gvec2f{T}(a: T, b: T) = gvec2f(a.gf32, b.gf32); +export fn gvec2f{T}(a: T) = gvec2f(a, a); export fn gvec2b() = gvec2b("vec2()", Dict{string, string}(), Set{GBuffer}()); export fn gvec2b(a: gbool, b: gbool) = gvec2Primitive{gbool, gvec2b}(a, b); export fn gvec2b{T}(a: T, b: T) = gvec2b(a.gbool, b.gbool); +export fn gvec2b{T}(a: T) = gvec2b(a, a); fn gvec3Primitive{I, O}(a: I, b: I, c: I) { let typename = {O.typeName}(); @@ -1583,18 +1711,22 @@ fn gvec3Primitive{I, O}(a: I, b: I, c: I) { export fn gvec3u() = gvec3u("vec3u()", Dict{string, string}(), Set{GBuffer}()); export fn gvec3u(a: gu32, b: gu32, c: gu32) = gvec3Primitive{gu32, gvec3u}(a, b, c); export fn gvec3u{T}(a: T, b: T, c: T) = gvec3u(a.gu32, b.gu32, c.gu32); +export fn gvec3u{T}(a: T) = gvec3u(a, a, a); export fn gvec3i() = gvec3i("vec3i()", Dict{string, string}(), Set{GBuffer}()); export fn gvec3i(a: gi32, b: gi32, c: gi32) = gvec3Primitive{gi32, gvec3i}(a, b, c); export fn gvec3i{T}(a: T, b: T, c: T) = gvec3i(a.gi32, b.gi32, c.gi32); +export fn gvec3i{T}(a: T) = gvec3i(a, a, a); export fn gvec3f() = gvec3f("vec3f()", Dict{string, string}(), Set{GBuffer}()); export fn gvec3f(a: gf32, b: gf32, c: gf32) = gvec3Primitive{gf32, gvec3f}(a, b, c); export fn gvec3f{T}(a: T, b: T, c: T) = gvec3f(a.gf32, b.gf32, c.gf32); +export fn gvec3f{T}(a: T) = gvec3f(a, a, a); export fn gvec3b() = gvec3b("vec3()", Dict{string, string}(), Set{GBuffer}()); export fn gvec3b(a: gbool, b: gbool, c: gbool) = gvec3Primitive{gbool, gvec3b}(a, b, c); export fn gvec3b{T}(a: T, b: T, c: T) = gvec3b(a.gbool, b.gbool, c.gbool); +export fn gvec3b{T}(a: T) = gvec3b(a, a, a); fn gvec4Primitive{I, O}(a: I, b: I, c: I, d: I) { let typename = {O.typeName}(); @@ -1618,24 +1750,28 @@ export fn gvec4u(a: gu32, b: gu32, c: gu32, d: gu32) = gvec4Primitive{gu32, gvec a, b, c, d ); export fn gvec4u{T}(a: T, b: T, c: T, d: T) = gvec4u(a.gu32, b.gu32, c.gu32, d.gu32); +export fn gvec4u{T}(a: T) = gvec4u(a, a, a, a); export fn gvec4i() = gvec4i("vec4i()", Dict{string, string}(), Set{GBuffer}()); export fn gvec4i(a: gi32, b: gi32, c: gi32, d: gi32) = gvec4Primitive{gi32, gvec4i}( a, b, c, d ); export fn gvec4i{T}(a: T, b: T, c: T, d: T) = gvec4i(a.gi32, b.gi32, c.gi32, d.gi32); +export fn gvec4i{T}(a: T) = gvec4i(a, a, a, a); export fn gvec4f() = gvec4f("vec4f()", Dict{string, string}(), Set{GBuffer}()); export fn gvec4f(a: gf32, b: gf32, c: gf32, d: gf32) = gvec4Primitive{gf32, gvec4f}( a, b, c, d ); export fn gvec4f{T}(a: T, b: T, c: T, d: T) = gvec4f(a.gf32, b.gf32, c.gf32, d.gf32); +export fn gvec4f{T}(a: T) = gvec4f(a, a, a, a); export fn gvec4b() = gvec4b("vec4()", Dict{string, string}(), Set{GBuffer}()); export fn gvec4b(a: gbool, b: gbool, c: gbool, d: gbool) = gvec4Primitive{gbool, gvec4b}( a, b, c, d ); export fn gvec4b{T}(a: T, b: T, c: T, d: T) = gvec4b(a.gbool, b.gbool, c.gbool, d.gbool); +export fn gvec4b{T}(a: T) = gvec4b(a, a, a, a); // The global_invocation_id; the entry value to a compute shader, and it's constructor, a // specialized version of gvec3u that initializes a bit differently. @@ -3867,6 +4003,298 @@ export fn mod(a: gvec4f, b: gvec4f) = gmod(a, b); export fn mod(a: gvec4f, b: gf32) = gmod(a, b); export fn mod(a: gf32, b: gvec4f) = gmodRev(a, b); +fn gpow{A, B}(a: A, b: B) { + let varName = 'pow('.concat(a.varName).concat(', ').concat(b.varName).concat(')'); + let statements = a.statements.concat(b.statements); + let buffers = a.buffers.union(b.buffers); + return {A}(varName, statements, buffers); +} +export fn pow(a: gf32, b: gf32) = gpow(a, b); +export fn pow{T}(a: gf32, b: T) = pow(a, b.gf32); +export fn pow{T}(a: T, b: gf32) = pow(a.gf32, b); +export fn pow(a: gvec2f, b: gvec2f) = gpow(a, b); +export fn pow{T}(a: gvec2f, b: T) = pow(a, b.gvec2f); +export fn pow{T}(a: T, b: gvec2f) = pow(a.gvec2f, b); +export fn pow(a: gvec3f, b: gvec3f) = gpow(a, b); +export fn pow{T}(a: gvec3f, b: T) = pow(a, b.gvec3f); +export fn pow{T}(a: T, b: gvec3f) = pow(a.gvec3f, b); +export fn pow(a: gvec4f, b: gvec4f) = gpow(a, b); +export fn pow{T}(a: gvec4f, b: T) = pow(a, b.gvec4f); +export fn pow{T}(a: T, b: gvec4f) = pow(a.gvec4f, b); + +fn gsqrt{I}(v: I) { + let varName = 'sqrt('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn sqrt(v: gf32) = gsqrt{gf32}(v); +export fn sqrt(v: gvec2f) = gsqrt{gvec2f}(v); +export fn sqrt(v: gvec3f) = gsqrt{gvec3f}(v); +export fn sqrt(v: gvec4f) = gsqrt{gvec4f}(v); + +fn gacos{I}(v: I) { + let varName = 'acos('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn acos(v: gf32) = gacos{gf32}(v); +export fn acos(v: gvec2f) = gacos{gvec2f}(v); +export fn acos(v: gvec3f) = gacos{gvec3f}(v); +export fn acos(v: gvec4f) = gacos{gvec4f}(v); + +fn gacosh{I}(v: I) { + let varName = 'acosh('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn acosh(v: gf32) = gacosh{gf32}(v); +export fn acosh(v: gvec2f) = gacosh{gvec2f}(v); +export fn acosh(v: gvec3f) = gacosh{gvec3f}(v); +export fn acosh(v: gvec4f) = gacosh{gvec4f}(v); + +fn gasin{I}(v: I) { + let varName = 'asin('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn asin(v: gf32) = gasin{gf32}(v); +export fn asin(v: gvec2f) = gasin{gvec2f}(v); +export fn asin(v: gvec3f) = gasin{gvec3f}(v); +export fn asin(v: gvec4f) = gasin{gvec4f}(v); + +fn gasinh{I}(v: I) { + let varName = 'asinh('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn asinh(v: gf32) = gasinh{gf32}(v); +export fn asinh(v: gvec2f) = gasinh{gvec2f}(v); +export fn asinh(v: gvec3f) = gasinh{gvec3f}(v); +export fn asinh(v: gvec4f) = gasinh{gvec4f}(v); + +fn gatan{I}(v: I) { + let varName = 'atan('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn atan(v: gf32) = gatan{gf32}(v); +export fn atan(v: gvec2f) = gatan{gvec2f}(v); +export fn atan(v: gvec3f) = gatan{gvec3f}(v); +export fn atan(v: gvec4f) = gatan{gvec4f}(v); + +fn gatanh{I}(v: I) { + let varName = 'atanh('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn atanh(v: gf32) = gatanh{gf32}(v); +export fn atanh(v: gvec2f) = gatanh{gvec2f}(v); +export fn atanh(v: gvec3f) = gatanh{gvec3f}(v); +export fn atanh(v: gvec4f) = gatanh{gvec4f}(v); + +fn gatan2{I}(a: I, b: I) { + let varName = 'atan2('.concat(a.varName).concat(', ').concat(b.varName).concat(')'); + let statements = a.statements.concat(b.statements); + let buffers = a.buffers.union(b.buffers); + return {I}(varName, statements, buffers); +} +export fn atan2(a: gf32, b: gf32) = gatan2{gf32}(a, b); +export fn atan2{T}(a: gf32, b: T) = gatan2{gf32}(a, b.gf32); +export fn atan2{T}(a: T, b: gf32) = gatan2{gf32}(a.gf32, b); +export fn atan2(a: gvec2f, b: gvec2f) = gatan2{gvec2f}(a, b); +export fn atan2{T}(a: gvec2f, b: T) = gatan2{gvec2f}(a, b.gvec2f); +export fn atan2{T}(a: T, b: gvec2f) = gatan2{gvec2f}(a.gvec2f, b); +export fn atan2(a: gvec3f, b: gvec3f) = gatan2{gvec3f}(a, b); +export fn atan2{T}(a: gvec3f, b: T) = gatan2{gvec3f}(a, b.gvec3f); +export fn atan2{T}(a: T, b: gvec3f) = gatan2{gvec3f}(a.gvec3f, b); +export fn atan2(a: gvec4f, b: gvec4f) = gatan2{gvec4f}(a, b); +export fn atan2{T}(a: gvec4f, b: T) = gatan2{gvec4f}(a, b.gvec4f); +export fn atan2{T}(a: T, b: gvec4f) = gatan2{gvec4f}(a.gvec4f, b); + +fn gfloor{I}(v: I) { + let varName = 'floor('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn floor(v: gf32) = gfloor{gf32}(v); +export fn floor(v: gvec2f) = gfloor{gvec2f}(v); +export fn floor(v: gvec3f) = gfloor{gvec3f}(v); +export fn floor(v: gvec4f) = gfloor{gvec4f}(v); + +fn gceil{I}(v: I) { + let varName = 'ceil('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn ceil(v: gf32) = gceil{gf32}(v); +export fn ceil(v: gvec2f) = gceil{gvec2f}(v); +export fn ceil(v: gvec3f) = gceil{gvec3f}(v); +export fn ceil(v: gvec4f) = gceil{gvec4f}(v); + +fn gclamp{I}(v: I, l: I, h: I) { + let varName = 'clamp('.concat(v.varName).concat(', ').concat(l.varName).concat(', ').concat(h.varName).concat(')'); + let statements = v.statements.concat(l.statements).concat(h.statements); + let buffers = v.buffers.union(l.buffers).union(h.buffers); + return {I}(varName, statements, buffers); +} +export fn clamp(v: gf32, l: gf32, h: gf32) = gclamp{gf32}(v, l, h); +export fn clamp{T}(v: gf32, l: T, h: T) = gclamp{gf32}(v, l.gf32, h.gf32); +export fn clamp{T}(v: gf32, l: gf32, h: T) = gclamp{gf32}(v, l, h.gf32); +export fn clamp{T}(v: gf32, l: T, h: gf32) = gclamp{gf32}(v, l.gf32, h); +export fn clamp(v: gvec2f, l: gvec2f, h: gvec2f) = gclamp{gvec2f}(v, l, h); +export fn clamp{T}(v: gvec2f, l: T, h: T) = gclamp{gvec2f}(v, l.gvec2f, h.gvec2f); +export fn clamp{T}(v: gvec2f, l: gvec2f, h: T) = gclamp{gvec2f}(v, l, h.gvec2f); +export fn clamp{T}(v: gvec2f, l: T, h: gvec2f) = gclamp{gvec2f}(v, l.gvec2f, h); +export fn clamp(v: gvec3f, l: gvec3f, h: gvec3f) = gclamp{gvec3f}(v, l, h); +export fn clamp{T}(v: gvec3f, l: T, h: T) = gclamp{gvec3f}(v, l.gvec3f, h.gvec3f); +export fn clamp{T}(v: gvec3f, l: gvec3f, h: T) = gclamp{gvec3f}(v, l, h.gvec3f); +export fn clamp{T}(v: gvec3f, l: T, h: gvec3f) = gclamp{gvec3f}(v, l.gvec3f, h); +export fn clamp(v: gvec4f, l: gvec4f, h: gvec4f) = gclamp{gvec4f}(v, l, h); +export fn clamp{T}(v: gvec4f, l: T, h: T) = gclamp{gvec4f}(v, l.gvec4f, h.gvec4f); +export fn clamp{T}(v: gvec4f, l: gvec4f, h: T) = gclamp{gvec4f}(v, l, h.gvec4f); +export fn clamp{T}(v: gvec4f, l: T, h: gvec4f) = gclamp{gvec4f}(v, l.gvec4f, h); + +fn gexp{I}(v: I) { + let varName = 'exp('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn exp(v: gf32) = gexp{gf32}(v); +export fn exp(v: gvec2f) = gexp{gvec2f}(v); +export fn exp(v: gvec3f) = gexp{gvec3f}(v); +export fn exp(v: gvec4f) = gexp{gvec4f}(v); + +fn gln{I}(v: I) { + let varName = 'log('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn ln(v: gf32) = gln{gf32}(v); +export fn ln(v: gvec2f) = gln{gvec2f}(v); +export fn ln(v: gvec3f) = gln{gvec3f}(v); +export fn ln(v: gvec4f) = gln{gvec4f}(v); + +fn glog2{I}(v: I) { + let varName = 'log2('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn log2(v: gf32) = glog2{gf32}(v); +export fn log2(v: gvec2f) = glog2{gvec2f}(v); +export fn log2(v: gvec3f) = glog2{gvec3f}(v); +export fn log2(v: gvec4f) = glog2{gvec4f}(v); + +fn glog10{I}(v: I) { + let varName = '(log('.concat(v.varName).concat(') / log(10.0))'); + return {I}(varName, v.statements, v.buffers); +} +export fn log10(v: gf32) = glog10{gf32}(v); +export fn log10(v: gvec2f) = glog10{gvec2f}(v); +export fn log10(v: gvec3f) = glog10{gvec3f}(v); +export fn log10(v: gvec4f) = glog10{gvec4f}(v); + +fn gcos{I}(v: I) { + let varName = 'cos('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn cos(v: gf32) = gcos{gf32}(v); +export fn cos(v: gvec2f) = gcos{gvec2f}(v); +export fn cos(v: gvec3f) = gcos{gvec3f}(v); +export fn cos(v: gvec4f) = gcos{gvec4f}(v); + +fn gcosh{I}(v: I) { + let varName = 'cosh('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn cosh(v: gf32) = gcosh{gf32}(v); +export fn cosh(v: gvec2f) = gcosh{gvec2f}(v); +export fn cosh(v: gvec3f) = gcosh{gvec3f}(v); +export fn cosh(v: gvec4f) = gcosh{gvec4f}(v); + +fn gsin{I}(v: I) { + let varName = 'sin('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn sin(v: gf32) = gsin{gf32}(v); +export fn sin(v: gvec2f) = gsin{gvec2f}(v); +export fn sin(v: gvec3f) = gsin{gvec3f}(v); +export fn sin(v: gvec4f) = gsin{gvec4f}(v); + +fn gsinh{I}(v: I) { + let varName = 'sinh('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn sinh(v: gf32) = gsinh{gf32}(v); +export fn sinh(v: gvec2f) = gsinh{gvec2f}(v); +export fn sinh(v: gvec3f) = gsinh{gvec3f}(v); +export fn sinh(v: gvec4f) = gsinh{gvec4f}(v); + +fn gtan{I}(v: I) { + let varName = 'tan('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn tan(v: gf32) = gtan{gf32}(v); +export fn tan(v: gvec2f) = gtan{gvec2f}(v); +export fn tan(v: gvec3f) = gtan{gvec3f}(v); +export fn tan(v: gvec4f) = gtan{gvec4f}(v); + +fn gtanh{I}(v: I) { + let varName = 'tanh('.concat(v.varName).concat(')'); + return {I}(varName, v.statements, v.buffers); +} +export fn tanh(v: gf32) = gtanh{gf32}(v); +export fn tanh(v: gvec2f) = gtanh{gvec2f}(v); +export fn tanh(v: gvec3f) = gtanh{gvec3f}(v); +export fn tanh(v: gvec4f) = gtanh{gvec4f}(v); + +export fn sec(v: gf32) = 1.0.div(cos(v)); +export fn sec(v: gvec2f) = gvec2f(1.0).div(cos(v)); +export fn sec(v: gvec3f) = gvec3f(1.0).div(cos(v)); +export fn sec(v: gvec4f) = gvec4f(1.0).div(cos(v)); + +export fn csc(v: gf32) = 1.0.div(sin(v)); +export fn csc(v: gvec2f) = gvec2f(1.0).div(sin(v)); +export fn csc(v: gvec3f) = gvec3f(1.0).div(sin(v)); +export fn csc(v: gvec4f) = gvec4f(1.0).div(sin(v)); + +export fn cot(v: gf32) = 1.0.div(tan(v)); +export fn cot(v: gvec2f) = gvec2f(1.0).div(tan(v)); +export fn cot(v: gvec3f) = gvec3f(1.0).div(tan(v)); +export fn cot(v: gvec4f) = gvec4f(1.0).div(tan(v)); + +export fn asec(v: gf32) = acos(1.0.div(v)); +export fn asec(v: gvec2f) = acos(gvec2f(1.0).div(v)); +export fn asec(v: gvec3f) = acos(gvec3f(1.0).div(v)); +export fn asec(v: gvec4f) = acos(gvec4f(1.0).div(v)); + +export fn acsc(v: gf32) = asin(1.0.div(v)); +export fn acsc(v: gvec2f) = asin(gvec2f(1.0).div(v)); +export fn acsc(v: gvec3f) = asin(gvec3f(1.0).div(v)); +export fn acsc(v: gvec4f) = asin(gvec4f(1.0).div(v)); + +export fn acot(v: gf32) = pi.div(2.0).sub(atan(v)); +export fn acot(v: gvec2f) = gvec2f(pi.div(2.0), pi.div(2.0)).sub(atan(v)); +export fn acot(v: gvec3f) = gvec3f(pi.div(2.0), pi.div(2.0), pi.div(2.0)).sub(atan(v)); +export fn acot(v: gvec4f) = gvec4f(pi.div(2.0), pi.div(2.0), pi.div(2.0), pi.div(2.0)).sub(atan(v)); + +export fn sech(v: gf32) = 1.0.div(cosh(v)); +export fn sech(v: gvec2f) = gvec2f(1.0).div(cosh(v)); +export fn sech(v: gvec3f) = gvec3f(1.0).div(cosh(v)); +export fn sech(v: gvec4f) = gvec4f(1.0).div(cosh(v)); + +export fn csch(v: gf32) = 1.0.div(sinh(v)); +export fn csch(v: gvec2f) = gvec2f(1.0).div(sinh(v)); +export fn csch(v: gvec3f) = gvec3f(1.0).div(sinh(v)); +export fn csch(v: gvec4f) = gvec4f(1.0).div(sinh(v)); + +export fn coth(v: gf32) = 1.0.div(tanh(v)); +export fn coth(v: gvec2f) = gvec2f(1.0).div(tanh(v)); +export fn coth(v: gvec3f) = gvec3f(1.0).div(tanh(v)); +export fn coth(v: gvec4f) = gvec4f(1.0).div(tanh(v)); + +export fn asech(v: gf32) = ln(1.0.add(sqrt(1.0.sub(v.pow(2.0)))).div(v)); +export fn asech(v: gvec2f) = ln(gvec2f(1.0).add(sqrt(gvec2f(1.0).sub(v.pow(gvec2f(2.0))))).div(v)); +export fn asech(v: gvec3f) = ln(gvec3f(1.0).add(sqrt(gvec3f(1.0).sub(v.pow(gvec3f(2.0))))).div(v)); +export fn asech(v: gvec4f) = ln(gvec4f(1.0).add(sqrt(gvec4f(1.0).sub(v.pow(gvec4f(2.0))))).div(v)); + +export fn acsch(x: gf32) = ln(1.0.div(x).add(sqrt(1.0.div(x.pow(2.0)).add(1.0)))); +export fn acsch(x: gvec2f) = ln(gvec2f(1.0).div(x).add(sqrt(gvec2f(1.0).div(x.pow(gvec2f(2.0))).add(gvec2f(1.0))))); +export fn acsch(x: gvec3f) = ln(gvec3f(1.0).div(x).add(sqrt(gvec3f(1.0).div(x.pow(gvec3f(2.0))).add(gvec3f(1.0))))); +export fn acsch(x: gvec4f) = ln(gvec4f(1.0).div(x).add(sqrt(gvec4f(1.0).div(x.pow(gvec4f(2.0))).add(gvec4f(1.0))))); + +export fn acoth(x: gf32) = ln(x.add(1.0).div(x.sub(1.0))).div(2.0); +export fn acoth(x: gvec2f) = ln(x.add(gvec2f(1.0)).div(x.sub(gvec2f(1.0)))).div(gvec2f(2.0)); +export fn acoth(x: gvec3f) = ln(x.add(gvec3f(1.0)).div(x.sub(gvec3f(1.0)))).div(gvec3f(2.0)); +export fn acoth(x: gvec4f) = ln(x.add(gvec4f(1.0)).div(x.sub(gvec4f(1.0)))).div(gvec4f(2.0)); + // GPU Comparison methods fn geq{I, O}(a: I, b: I) { @@ -4312,6 +4740,14 @@ export fn map{G, G2}(gb: GBuffer, f: G -> G2) { compute.build.run; return out; }*/ +export fn map(gb: GBuffer, f: gu32 -> gu32) { + let idx = gFor(gb.len); + let val = gb[idx].asU32; + let out = GBuffer{u32}(gb.len); + let compute = out[idx].store(f(val).asI32); + compute.build.run; + return out; +} export fn map(gb: GBuffer, f: gi32 -> gi32) { let idx = gFor(gb.len); let val = gb[idx]; @@ -4320,6 +4756,14 @@ export fn map(gb: GBuffer, f: gi32 -> gi32) { compute.build.run; return out; } +export fn map(gb: GBuffer, f: gf32 -> gf32) { + let idx = gFor(gb.len); + let val = gb[idx].asF32; + let out = GBuffer{f32}(gb.len); + let compute = out[idx].store(f(val).asI32); + compute.build.run; + return out; +} // This one technically only works for i32/u32/f32 by chance. Once the issue above is fixed, this // one can be fixed, too. export fn map{G, G2}(gb: GBuffer, f: (G, gu32) -> G2) { From 6334b55487eeb6841d9a78ed5ba6068c5d39fa06 Mon Sep 17 00:00:00 2001 From: David Ellis Date: Mon, 11 Nov 2024 15:05:22 -0600 Subject: [PATCH 2/4] Reduce precision on gpu trig tests --- alan/src/compile/integration_tests.rs | 116 +++++++++++++------------- 1 file changed, 58 insertions(+), 58 deletions(-) diff --git a/alan/src/compile/integration_tests.rs b/alan/src/compile/integration_tests.rs index 285b8e69..7e116f38 100644 --- a/alan/src/compile/integration_tests.rs +++ b/alan/src/compile/integration_tests.rs @@ -3107,78 +3107,78 @@ test_gpgpu!(gpu_trig => r#" export fn main { 'Logarithms and e^x'.print; // Contrived way to get the GPU to do this work, don't follow this pattern for real GPU usage - GBuffer([e.f32]).map(fn (v: gf32) = exp(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([e.f32]).map(fn (v: gf32) = ln(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([e.f32]).map(fn (v: gf32) = log10(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([e.f32]).map(fn (v: gf32) = log2(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([e.f32]).map(fn (v: gf32) = exp(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([e.f32]).map(fn (v: gf32) = ln(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([e.f32]).map(fn (v: gf32) = log10(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([e.f32]).map(fn (v: gf32) = log2(v)).read{f32}[0].getOrExit.string(3).print; 'Basic Trig functions'.print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sin(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cos(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = tan(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sec(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = csc(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cot(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sin(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cos(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = tan(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sec(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = csc(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cot(v)).read{f32}[0].getOrExit.string(3).print; 'Inverse Trig functions'.print; - GBuffer([0.0.f32]).map(fn (v: gf32) = asin(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([1.0.f32]).map(fn (v: gf32) = acos(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([0.0.f32]).map(fn (v: gf32) = atan(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([1.0.f32]).map(fn (v: gf32) = atan2(v, 2.0)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = asec(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acsc(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acot(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([0.0.f32]).map(fn (v: gf32) = asin(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([1.0.f32]).map(fn (v: gf32) = acos(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([0.0.f32]).map(fn (v: gf32) = atan(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([1.0.f32]).map(fn (v: gf32) = atan2(v, 2.0)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = asec(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acsc(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acot(v)).read{f32}[0].getOrExit.string(3).print; 'Hyperbolic Trig functions'.print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sinh(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cosh(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = tanh(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sech(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = csch(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = coth(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sinh(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cosh(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = tanh(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sech(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = csch(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = coth(v)).read{f32}[0].getOrExit.string(3).print; 'Inverse Hyperbolic Trig functions'.print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = asinh(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acosh(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([pi.f32 / 6.0.f32]).map(fn (v: gf32) = atanh(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([0.5.f32]).map(fn (v: gf32) = asech(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acsch(v)).read{f32}[0].getOrExit.string(4).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acoth(v)).read{f32}[0].getOrExit.string(4).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = asinh(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acosh(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([pi.f32 / 6.0.f32]).map(fn (v: gf32) = atanh(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([0.5.f32]).map(fn (v: gf32) = asech(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acsch(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acoth(v)).read{f32}[0].getOrExit.string(3).print; }"#; stdout r#"Logarithms and e^x -15.1543 -1.0000 -0.4343 -1.4427 +15.154 +1.000 +0.434 +1.443 Basic Trig functions -0.8660 -0.5000 -1.7321 -2.0000 -1.1547 -0.5774 +0.866 +0.500 +1.732 +2.000 +1.155 +0.577 Inverse Trig functions -0.0000 -0.0000 -0.0000 -0.4636 -0.3013 -1.2698 -0.7623 +0.000 +0.000 +0.000 +0.464 +0.301 +1.270 +0.762 Hyperbolic Trig functions -1.2494 -1.6003 -0.7807 -0.6249 -0.8004 -1.2809 +1.249 +1.600 +0.781 +0.625 +0.800 +1.281 Inverse Hyperbolic Trig functions -0.9144 -0.3060 -0.5813 -1.3170 -0.8491 -1.8849 +0.914 +0.306 +0.581 +1.317 +0.849 +1.885 "#; ); From a3864a03d5d7e859173845b4c4bd2ae7ef3d9ef6 Mon Sep 17 00:00:00 2001 From: David Ellis Date: Mon, 11 Nov 2024 16:24:10 -0600 Subject: [PATCH 3/4] Do we at least have cent-level accuracy across GPU/OS combos? --- alan/src/compile/integration_tests.rs | 116 +++++++++++++------------- 1 file changed, 58 insertions(+), 58 deletions(-) diff --git a/alan/src/compile/integration_tests.rs b/alan/src/compile/integration_tests.rs index 7e116f38..d4f97be7 100644 --- a/alan/src/compile/integration_tests.rs +++ b/alan/src/compile/integration_tests.rs @@ -3107,78 +3107,78 @@ test_gpgpu!(gpu_trig => r#" export fn main { 'Logarithms and e^x'.print; // Contrived way to get the GPU to do this work, don't follow this pattern for real GPU usage - GBuffer([e.f32]).map(fn (v: gf32) = exp(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([e.f32]).map(fn (v: gf32) = ln(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([e.f32]).map(fn (v: gf32) = log10(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([e.f32]).map(fn (v: gf32) = log2(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([e.f32]).map(fn (v: gf32) = exp(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([e.f32]).map(fn (v: gf32) = ln(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([e.f32]).map(fn (v: gf32) = log10(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([e.f32]).map(fn (v: gf32) = log2(v)).read{f32}[0].getOrExit.string(2).print; 'Basic Trig functions'.print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sin(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cos(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = tan(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sec(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = csc(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cot(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sin(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cos(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = tan(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sec(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = csc(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cot(v)).read{f32}[0].getOrExit.string(2).print; 'Inverse Trig functions'.print; - GBuffer([0.0.f32]).map(fn (v: gf32) = asin(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([1.0.f32]).map(fn (v: gf32) = acos(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([0.0.f32]).map(fn (v: gf32) = atan(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([1.0.f32]).map(fn (v: gf32) = atan2(v, 2.0)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = asec(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acsc(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acot(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([0.0.f32]).map(fn (v: gf32) = asin(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([1.0.f32]).map(fn (v: gf32) = acos(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([0.0.f32]).map(fn (v: gf32) = atan(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([1.0.f32]).map(fn (v: gf32) = atan2(v, 2.0)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = asec(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acsc(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acot(v)).read{f32}[0].getOrExit.string(2).print; 'Hyperbolic Trig functions'.print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sinh(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cosh(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = tanh(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sech(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = csch(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = coth(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sinh(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = cosh(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = tanh(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = sech(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = csch(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = coth(v)).read{f32}[0].getOrExit.string(2).print; 'Inverse Hyperbolic Trig functions'.print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = asinh(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acosh(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([pi.f32 / 6.0.f32]).map(fn (v: gf32) = atanh(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([0.5.f32]).map(fn (v: gf32) = asech(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acsch(v)).read{f32}[0].getOrExit.string(3).print; - GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acoth(v)).read{f32}[0].getOrExit.string(3).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = asinh(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acosh(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([pi.f32 / 6.0.f32]).map(fn (v: gf32) = atanh(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([0.5.f32]).map(fn (v: gf32) = asech(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acsch(v)).read{f32}[0].getOrExit.string(2).print; + GBuffer([tau.f32 / 6.0.f32]).map(fn (v: gf32) = acoth(v)).read{f32}[0].getOrExit.string(2).print; }"#; stdout r#"Logarithms and e^x -15.154 -1.000 -0.434 -1.443 +15.15 +1.00 +0.43 +1.44 Basic Trig functions -0.866 -0.500 -1.732 -2.000 -1.155 -0.577 +0.87 +0.50 +1.73 +2.00 +1.15 +0.58 Inverse Trig functions -0.000 -0.000 -0.000 -0.464 -0.301 -1.270 -0.762 +0.00 +0.00 +0.00 +0.46 +0.30 +1.27 +0.76 Hyperbolic Trig functions -1.249 -1.600 -0.781 -0.625 -0.800 -1.281 +1.25 +1.60 +0.78 +0.62 +0.80 +1.28 Inverse Hyperbolic Trig functions -0.914 -0.306 -0.581 -1.317 -0.849 -1.885 +0.91 +0.31 +0.58 +1.32 +0.85 +1.88 "#; ); From 897dfaf889ad5d4383dda00fa90fe80d538479d7 Mon Sep 17 00:00:00 2001 From: David Ellis Date: Tue, 12 Nov 2024 10:38:33 -0600 Subject: [PATCH 4/4] It looks like the default thread stack is too small on ARM64 for my tests --- .github/workflows/rust.yml | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/.github/workflows/rust.yml b/.github/workflows/rust.yml index 3a8b0674..0d7d7c5b 100644 --- a/.github/workflows/rust.yml +++ b/.github/workflows/rust.yml @@ -59,7 +59,7 @@ jobs: run: cargo build --verbose - if: ${{ github.ref_name == 'main' }} name: Run tests - run: cargo test --verbose -- --include-ignored + run: RUST_MIN_STACK=8388608 cargo test --verbose -- --include-ignored - if: ${{ github.ref_name != 'main' }} name: Run tests - run: cargo test --verbose + run: RUST_MIN_STACK=8388608 cargo test --verbose