limit samples to only non zero positive numbersHEADmaster

This fixes a bug with too high scales causing a division by zero

1 files changed, 4 insertions, 1 deletions

diff --git a/src/main.c b/src/main.c
index 6d37dc8..7c9a161 100644
--- a/src/main.c
+++ b/src/main.c
@@ -58,8 +58,11 @@ static uint32_t average(const uint32_t dest, const uint32_t src)
 }
 
 static void scale_image(const uint32_t* input, const signed input_width, const signed input_height,
-                        uint32_t* output, const double multiplier, const signed samples)
+                        uint32_t* output, const double multiplier, signed samples)
 {
+    if (samples <= 0)
+        samples = 1;
+
     signed output_width = input_width * multiplier;
     uint32_t pixel;