SDL: Avoid using higher precision floating point numbers by using it in a function

SDL Commits
1 
From 69a6efcd1bc65e714d0aa228b092924c2c7576e1 Mon Sep 17 00:00:00 2001
From: Anonymous Maarten <[EMAIL REDACTED]>
Date: Sat, 14 Oct 2023 04:38:35 +0200
Subject: [PATCH] Avoid using higher precision floating point numbers by using
 it in a function

---
 test/testautomation_math.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/test/testautomation_math.c b/test/testautomation_math.c
index 1d39eaa5feb8..1e643bd0a9cb 100644
--- a/test/testautomation_math.c
+++ b/test/testautomation_math.c
@@ -138,6 +138,8 @@ helper_ddtod(const char *func_name, dd_to_d_func func,
     Uint32 i;
     for (i = 0; i < cases_size; i++) {
         const double result = func(cases[i].x_input, cases[i].y_input);
+        /* By using the result as input, the compiler is less likely to use higher precision floating point number */
+        (void)SDL_sin(result);
         SDLTest_AssertCheck(result == cases[i].expected,
                             "%s(%f,%f), expected %f, got %f",
                             func_name,