math: add dummy tgamma and tgammaf implementations

src/math/tgamma.c

+#include <math.h>
+
+// FIXME: use lanczos approximation
+
+double __lgamma_r(double, int *);
+
+double tgamma(double x)
+{
+	int sign;
+	double y;
+
+	y = exp(__lgamma_r(x, &sign));
+	if (sign < 0)
+		y = -y;
+	return y;
+}

src/math/tgammaf.c

+#include <math.h>
+
+// FIXME: use lanczos approximation
+
+float __lgammaf_r(float, int *);
+
+float tgammaf(float x)
+{
+	int sign;
+	float y;
+
+	y = exp(__lgammaf_r(x, &sign));
+	if (sign < 0)
+		y = -y;
+	return y;
+}