std.math.sign: Return smallest integer type that fits possible values

This may be a breaking change for callers that depend on the return type
size being identical to the input. For example, `sign(i32) * i8 + i8`
could now overflow. Callers can add an explicit type annotation to avoid
that problem.

lib/std/math.zig

@@ -1772,26 +1772,33 @@ pub const F80 = struct {
     }
 };
 
+fn SignOf(T: type) type {
+    return switch (@typeInfo(T)) {
+        .comptime_int, .comptime_float => comptime_int,
+        .int => |int| switch (int.signedness) {
+            .signed => IntFittingRange(-1, 1),
+            .unsigned => IntFittingRange(0, 1),
+        },
+        .float => IntFittingRange(-1, 1),
+        .vector => |vec| @Vector(vec.len, SignOf(vec.child)),
+        else => @compileError("Expected an int, float, or a vector of one, found " ++ @typeName(T)),
+    };
+}
+
 /// Returns -1, 0, or 1.
 /// Supports integer and float types and vectors of integer and float types.
 /// Unsigned integer types will always return 0 or 1.
+/// The returned integer type is the smallest that fits the possible values.
 /// Branchless.
-pub inline fn sign(i: anytype) @TypeOf(i) {
-    const T = @TypeOf(i);
+pub inline fn sign(n: anytype) SignOf(@TypeOf(n)) {
+    const T = SignOf(@TypeOf(n));
     return switch (@typeInfo(T)) {
-        .int, .comptime_int => @as(T, @intFromBool(i > 0)) - @as(T, @intFromBool(i < 0)),
-        .float, .comptime_float => @as(T, @floatFromInt(@intFromBool(i > 0))) - @as(T, @floatFromInt(@intFromBool(i < 0))),
-        .vector => |vinfo| blk: {
-            switch (@typeInfo(vinfo.child)) {
-                .int, .float => {
-                    const zero: T = @splat(0);
-                    const one: T = @splat(1);
-                    break :blk @select(vinfo.child, i > zero, one, zero) - @select(vinfo.child, i < zero, one, zero);
-                },
-                else => @compileError("Expected vector of ints or floats, found " ++ @typeName(T)),
-            }
+        .vector => |vec| blk: {
+            const zero: T = @splat(0);
+            const one: T = @splat(1);
+            break :blk @select(vec.child, n > zero, one, zero) - @select(vec.child, n < zero, one, zero);
         },
-        else => @compileError("Expected an int, float or vector of one, found " ++ @typeName(T)),
+        else => @as(T, @intFromBool(n > 0)) - @as(T, @intFromBool(n < 0)),
     };
 }