GBA Memory: Back out a hack that introduces incorrect behavior, even though it fixes a game

This commit is contained in:
Jeffrey Pfau 2015-05-13 22:07:20 -07:00
parent 3dfa8bc547
commit daf3ce76e7
2 changed files with 0 additions and 7 deletions

View File

@ -15,7 +15,6 @@ Bugfixes:
- GBA: Fix bounds-checking on EEPROM access
- GBA Audio: FIFOs should not poll DMAs that are not scheduled for audio
- GBA BIOS: Initialize a variable that may be uninitialized in very rare cases
- GBA Memory: Ensure changing the timing of a DMA reschedules it
- GBA Memory: Allow SRAM to be 64kB
- GBA Memory: Fix 32-bit loads from unaddress cartridge space
- GBA Memory: Fix jumping to invalid memory when switching from Thumb to ARM

View File

@ -1256,12 +1256,6 @@ uint16_t GBAMemoryWriteDMACNT_HI(struct GBA* gba, int dma, uint16_t control) {
struct GBAMemory* memory = &gba->memory;
struct GBADMA* currentDma = &memory->dma[dma];
int wasEnabled = GBADMARegisterIsEnable(currentDma->reg);
int oldTiming = GBADMARegisterGetTiming(currentDma->reg);
int newTiming = GBADMARegisterGetTiming(control);
// This is probably a huge hack...verify what this does on hardware
if (oldTiming && oldTiming != DMA_TIMING_CUSTOM && oldTiming != newTiming) {
wasEnabled = false;
}
currentDma->reg = control;
if (GBADMARegisterIsDRQ(currentDma->reg)) {