Bubble up -1 option to the flag description.

lib/Makefile

@@ -172,7 +172,7 @@ DEFINES+=$(USER_DEFINES)
 
 DEFINES+=-DDEFAULT_HARDWARE='"$(HARDWARE_DESC)"'
 INCDIR=../include
-CFLAGS=-W -Wall -Wextra -Wno-unused-parameter -O3 -g -fPIC $(DEFINES)
+CFLAGS=-W -Wall -Wextra -Wno-unused-parameter -O3 -g -fPIC $(DEFINES) -march=native
 CXXFLAGS=$(CFLAGS) -fno-exceptions -std=c++11
 
 all : $(TARGET).a $(TARGET).so.1

lib/gpio.h

@@ -20,6 +20,12 @@
 
 #include <vector>
 
+#if __ARM_ARCH >= 7
+#define LED_MATRIX_ALLOW_BARRIER_DELAY 1
+#else
+#define LED_MATRIX_ALLOW_BARRIER_DELAY 0
+#endif
+
 // Putting this in our namespace to not collide with other things called like
 // this.
 namespace rgb_matrix {
@@ -73,19 +79,17 @@ class GPIO {
 
 private:
   inline void delay() const {
-    switch(slowdown_) {
-      case -1:
-#if __ARM_ARCH >= 7
+#if LED_MATRIX_ALLOW_BARRIER_DELAY
+    if (slowdown_ == -1) {
         asm volatile("dsb\tst");
+        return;
+    }
 #endif
-        break;
-      case 0:
-        break;
-      default:
-        for (int n = 0; n < slowdown_; n++)
-          *gpio_clr_bits_low_ = 0;
+    for (int n = 0; n < slowdown_; n++) {
+      *gpio_clr_bits_low_ = 0;
     }
   }
+
   inline gpio_bits_t ReadRegisters() const {
     return (static_cast<gpio_bits_t>(*gpio_read_bits_low_)
 #ifdef ENABLE_WIDE_GPIO_COMPUTE_MODULE

lib/led-matrix.cc

@@ -659,7 +659,7 @@ RGBMatrix *RGBMatrix::CreateFromOptions(const RGBMatrix::Options &options,
 
   // For the Pi4, we might need 2, maybe up to 4. Let's open up to 5.
   // on supproted architectures, -1 will emit memory barier (DSB ST) after GPIO write
-  if (runtime_options.gpio_slowdown < (__ARM_ARCH >= 7 ? -1 : 0)
+  if (runtime_options.gpio_slowdown < (LED_MATRIX_ALLOW_BARRIER_DELAY ? -1 : 0)
       || runtime_options.gpio_slowdown > 5) {
     fprintf(stderr, "--led-slowdown-gpio=%d is outside usable range\n",
             runtime_options.gpio_slowdown);

lib/options-initialize.cc

@@ -361,9 +361,13 @@ void PrintMatrixFlags(FILE *out, const RGBMatrix::Options &d,
           !d.disable_busy_waiting ? "no-" : "",
           !d.disable_busy_waiting ? "Don't u" : "U");
 
-  fprintf(out, "\t--led-slowdown-gpio=<0..4>: "
+  fprintf(out,
+          "\t--led-slowdown-gpio=<%d..4>: "
           "Slowdown GPIO. Needed for faster Pis/slower panels "
-          "(Default: %d (2 on Pi4, 1 other)).\n", r.gpio_slowdown);
+          "(Default: %d (2 on Pi4, 1 other)%s).\n",
+          (LED_MATRIX_ALLOW_BARRIER_DELAY ? -1 : 0), r.gpio_slowdown,
+          LED_MATRIX_ALLOW_BARRIER_DELAY ? "Use -1 for memory barrier approach"
+                                         : "");
   if (r.daemon >= 0) {
     const bool on = (r.daemon > 0);
     fprintf(out,