devkitPro

That's a bummer, now it has to be inline functions with inline ASM?

This is a minimal C program that triggers an "internal compiler error" if built with the -Os switch.

I'm building it with the command "arm-none-eabi-gcc test.c -Os"

The code: (Save as "test.c")

int main()
{
{
char *s=(char*)0x0203fc08;
//advance past first null
while (*s++ != 0);
//advance ...

I just wanted to give a heads up to anyone experiencing this. I only caught this because of warnings from the NO$GBA debugger.

In past versions of DevKitArm, you could write code like this:

void ByteSet(char *p, char value, int size)
{
while (size)
{
*p++ = value;
size--;
}
}

In past ...

The problem with this question is that "Memory Management" is a very broad term. It could refer to how RAM is physically allocated (static arrays vs malloc/new, or even allocating from areans), or how objects are assigned to each other. (Always Deep Copy, Reference Counting, or Garbage Collection).

keysDown() tells you what keys have *changed state* from not-pressed to pressed.
keysHeld() tells you what keys are currently in a pressed state.

Just because there's a bios function doesn't mean it will be any faster.
Compiling code in ARM mode instead of THUMB mode will make it run faster. Then profile it to see what the "hot" loops are, and rewrite those in ASM if they need it. Placing code in ITCM makes it faster, but loops end up getting ...

Big congratulations here at getting rid of all that exceptions stuff. *applauds*
Only thing is I was getting exception-handling code added when using the "new" operator. Even when I tried using the "nothrow" version, it was also including the code.
When I switched to malloc and placement new, the ...

The "cp" command in MSYS appears to have broken.
Not sure if reverting back to MSYS 1.0.15 is the best idea or not, but this means that the newer version has bugs in it.

Here's what I'm doing here:
I have created a FAT32 disk image, and mounted it using FileDisk, making a V:\ drive.
I am copying ...

There are other things you can do to optimize code.
First, if you are using any significant amount of multiplication or division, build as ARM code instead of THUMB code. GCC generates really bad THUMB code for math heavy code, but it's better for ARM mode.
Second, there's a trick to write a ...

Good.

For now, I'm just using dummy functions (blank functions with the same name) until the bug is fixed.