BOOL
GX2RCreateSurfaceUserMemory(virt_ptr<GX2Surface> surface,
virt_ptr<uint8_t> image,
virt_ptr<uint8_t> mipmap,
GX2RResourceFlags flags)
{
GX2CalcSurfaceSizeAndAlignment(surface);
surface->resourceFlags = flags;
surface->resourceFlags &= ~GX2RResourceFlags::Locked;
surface->resourceFlags &= ~GX2RResourceFlags::Gx2rAllocated;
surface->image = image;
surface->mipmaps = mipmap;
if ((surface->resourceFlags & GX2RResourceFlags::UsageGpuWrite) ||
(surface->resourceFlags & GX2RResourceFlags::UsageDmaWrite)) {
DCInvalidateRange(virt_cast<virt_addr>(surface->image),
surface->imageSize);
if (surface->mipmaps) {
DCInvalidateRange(virt_cast<virt_addr>(surface->mipmaps),
surface->mipmapSize);
}
}
return true;
}
/**
* ARAMFetch
*
* This function will move data from ARAM to MAIN memory
*/
void
ARAMFetch (char *dst, char *src, int len)
{
DCInvalidateRange(dst, len);
AR_StartDMA( ARAM_READ, (u32) dst, (u32) src, len);
while (AR_GetDMAStatus());
}
virt_ptr<void>
GX2RLockSurfaceEx(virt_ptr<GX2Surface> surface,
int32_t level,
GX2RResourceFlags flags)
{
decaf_check(surface);
decaf_check(surface->resourceFlags & ~GX2RResourceFlags::Locked);
flags = surface->resourceFlags | internal::getOptionFlags(flags);
// Set Locked flag
surface->resourceFlags |= GX2RResourceFlags::Locked;
surface->resourceFlags |= flags & GX2RResourceFlags::LockedReadOnly;
// Check if we need to invalidate the surface.
if ((flags & GX2RResourceFlags::UsageGpuWrite) ||
(flags & GX2RResourceFlags::UsageDmaWrite)) {
if (!(flags & GX2RResourceFlags::DisableCpuInvalidate)) {
auto ptr = virt_ptr<void> { nullptr };
auto size = uint32_t { 0 };
internal::getSurfaceData(surface, level, &ptr, &size);
DCInvalidateRange(virt_cast<virt_addr>(ptr), size);
}
}
return surface->image;
}
s32 IOS_GetRevision()
{
u32 vercode;
u16 rev;
DCInvalidateRange((void*)0x80003140,8);
vercode = *((u32*)0x80003140);
rev = vercode & 0xFFFF;
if(vercode == 0 || rev == 0) return IOS_EBADVERSION;
return rev;
}
s32 IOS_GetVersion()
{
u32 vercode;
u16 version;
DCInvalidateRange((void*)0x80003140,8);
vercode = *((u32*)0x80003140);
version = vercode >> 16;
if(version == 0) return IOS_EBADVERSION;
if(version > 0xff) return IOS_EBADVERSION;
return version;
}
static void build_disp_list(void)
{
DCInvalidateRange(display_list, sizeof(display_list));
GX_BeginDispList(display_list, sizeof(display_list));
GX_Begin(GX_QUADS, GX_VTXFMT0, 4);
for (unsigned i = 0; i < 4; i++)
{
GX_Position1x8(i);
GX_TexCoord1x8(i);
}
GX_End();
display_list_size = GX_EndDispList();
}
int main(int argc, char **argv)
{
// Exit after 10 seconds if there is an error
__exception_setreload(10);
CheckForGecko();
DCInvalidateRange(loader_stub, 0x1800);
memcpy(loader_stub, (void*)0x80001800, 0x1800);
DCFlushRange(loader_stub, 0x1800);
if( !IsWiiU() )
{
gprintf("Nintendont Loader\r\n");
gprintf("Built : %s %s\r\n", __DATE__, __TIME__ );
gprintf("Version : %d.%d\r\n", NIN_VERSION>>16, NIN_VERSION&0xFFFF );
}
/****************************************************************************
* dvd_read
*
* The only DVD function we need - you gotta luv gc-linux self-boots!
* returns: 1 - ok ; 0 - error
***************************************************************************/
int
dvd_read (void *dst, unsigned int len, u64 offset)
{
unsigned char *buffer = (unsigned char *) (unsigned int) DVDreadbuffer;
if (len > 2048)
return 0; /*** We only allow 2k reads **/
DCInvalidateRange ((void *) buffer, len);
// don't read past the end of the DVD (1.5 GB for GC DVD, 4.7 GB for DVD)
if(offset < 0x57057C00 || (isWii && offset < 0x118244F00LL))
{
#ifdef HW_DOL
dvd[0] = 0x2E;
dvd[1] = 0;
dvd[2] = 0xA8000000;
dvd[3] = (u32)(offset >> 2);
dvd[4] = len;
dvd[5] = (u32) buffer;
dvd[6] = len;
dvd[7] = 3; /*** Enable reading with DMA ***/
while (dvd[7] & 1);
memcpy (dst, buffer, len);
if (dvd[0] & 0x4) /* Ensure it has completed */
return 0;
return 1;
#elif WII_DVD
int ret = 1;
ret = DI_ReadDVD(dst, len >> 11, (u32)(offset >> 11));
if (ret==0)
return 1;
else
return 0;
#endif
}
/*
DVD_LowRead64(void* dst, unsigned int len, uint64_t offset)
Read Raw, needs to be on sector boundaries
Has 8,796,093,020,160 byte limit (8 TeraBytes)
Synchronous function.
return -1 if offset is out of range
*/
int DVD_LowRead64(void* dst, unsigned int len, uint64_t offset)
{
if(offset>>2 > 0xFFFFFFFF)
return -1;
if ((((int)dst) & 0xC0000000) == 0x80000000) // cached?
dvd[0] = 0x2E;
dvd[1] = 0;
dvd[2] = read_cmd;
dvd[3] = read_cmd == DVDR ? offset>>11 : offset >> 2;
dvd[4] = read_cmd == DVDR ? len>>11 : len;
dvd[5] = (unsigned long)dst;
dvd[6] = len;
dvd[7] = 3; // enable reading!
DCInvalidateRange(dst, len);
while (dvd[7] & 1);
if (dvd[0] & 0x4)
return 1;
return 0;
}
model_t* MODEL_setup(const u8* model_bmb) {
binheader_t* header = (binheader_t*) model_bmb;
const u32 posOffset = sizeof(binheader_t);
const u32 nrmOffset = posOffset + (sizeof(f32)* header->vcount * 3);
const u32 texOffset = nrmOffset + (sizeof(f32)* header->vcount * 3);
const u32 indOffset = texOffset + (sizeof(f32)* header->vcount * 2);
f32* positions = (f32*) (model_bmb + posOffset);
f32* normals = (f32*) (model_bmb + nrmOffset);
f32* texcoords = (f32*) (model_bmb + texOffset);
u16* indices = (u16*) (model_bmb + indOffset);
/* Calculate cost */
const u32 indicesCount = header->fcount * 3;
const u32 indicesSize = indicesCount * 3 * sizeof(u16); /* 3 indices per vertex index (p,n,t) that are u16 in size */
const u32 callSize = 89; /* Size of setup var */
/* Round up to nearest 32 multiplication */
const u32 dispSize = (((indicesSize + callSize + 63) >> 5) + 1) << 5;
/* Build display list */
/* Allocate and clear */
u32 i;
void* modelList = memalign(32, dispSize);
memset(modelList, 0, dispSize);
/* Set buffer data */
DCInvalidateRange(modelList, dispSize);
GX_BeginDispList(modelList, dispSize);
//GX_InvVtxCache();
GX_ClearVtxDesc();
GX_SetVtxDesc(GX_VA_POS, GX_INDEX16);
GX_SetVtxDesc(GX_VA_NRM, GX_INDEX16);
GX_SetVtxDesc(GX_VA_TEX0, GX_INDEX16);
GX_SetVtxAttrFmt(GX_VTXFMT0, GX_VA_POS, GX_POS_XYZ, GX_F32, 0);
GX_SetVtxAttrFmt(GX_VTXFMT0, GX_VA_NRM, GX_NRM_XYZ, GX_F32, 0);
GX_SetVtxAttrFmt(GX_VTXFMT0, GX_VA_TEX0, GX_TEX_ST, GX_F32, 0);
GX_SetArray(GX_VA_POS, (void*) positions, 3 * sizeof(f32));
GX_SetArray(GX_VA_NRM, (void*) normals, 3 * sizeof(f32));
GX_SetArray(GX_VA_TEX0, (void*) texcoords, 2 * sizeof(f32));
/* Fill the list with indices */
GX_Begin(GX_TRIANGLES, GX_VTXFMT0, indicesCount);
for (i = 0; i < indicesCount; i++) {
GX_Position1x16(indices[i]);
GX_Normal1x16(indices[i]);
GX_TexCoord1x16(indices[i]);
}
GX_End();
/* Close display list */
u32 modelListSize = GX_EndDispList();
if (modelListSize == 0) {
printf("Error: Display list not big enough [%u]\n", dispSize);
return NULL;
}
/* Return model info */
model_t* model = malloc(sizeof(model_t));
model->modelList = modelList;
model->modelListSize = modelListSize;
model->modelFaceCount = header->fcount;
model->modelPositions = positions;
model->modelNormals = normals;
model->modelTexcoords = texcoords;
model->modelIndices = indices;
return model;
}
s32 Swi_Handler(u32 arg0, u32 arg1, u32 arg2, u32 arg3)
{
u8 cmd;
/* Check alignment */
SwiAddr -= (SwiAddr[-4] == 0xDF) ? 3 : 1;
/* Get command */
cmd = SwiAddr[0];
/* Check command */
if(SwiAddr[0] != 0xcc) {
if (SwiTable[cmd])
return SwiTable[cmd](arg0, arg1, arg2, arg3);
else
return arg0;
}
/* Check argument */
switch (arg0) {
/** Add SWI handler **/
case 0: {
SwiTable[arg1]= (void *)arg2;
break;
}
/** Memcpy (cached to cached) **/
case 2: {
void *src = (void *)arg2;
void *dst = (void *)arg1;
u32 len = arg3;
u32 perms;
/* Apply permissions */
perms = Perms_Read();
Perms_Write(0xFFFFFFFF);
/* Copy data */
memcpy(dst, src, len);
DCFlushRange(dst, len);
/* Restore permissions */
Perms_Write(perms);
break;
}
/** Get register **/
case 3:
return *(vu32 *)arg1;
/** Set register **/
case 4:
*(vu32 *)arg1 = arg2;
break;
/** Set register **/
case 5:
*(vu32 *)arg1 |= arg2;
break;
/** Clear register **/
case 6:
*(vu32 *)arg1 &= ~arg2;
break;
/** Memcpy (uncached to cached) **/
case 9: {
void *src = (void *)arg2;
void *dst = (void *)arg1;
u32 len = arg3;
u32 perms;
/* Apply permissions */
perms = Perms_Read();
Perms_Write(0xFFFFFFFF);
/* Copy data */
DCInvalidateRange(src, len);
memcpy(dst, src, len);
DCFlushRange(dst, len);
/* Restore permissions */
Perms_Write(perms);
break;
}
/** Call function **/
case 16: {
s32 (*Function)(void *in, void *out);
/* Set function */
Function = (void *)arg1;
/* Call function */
return Function((void *)arg2, (void *)arg3);
}
/** Get syscall base */
case 17:
return ios.syscall;
/** Get IOS info **/
case 18: {
iosInfo *buffer = (iosInfo *)arg1;
/* Copy IOS info */
memcpy(buffer, &ios, sizeof(ios));
DCFlushRange(buffer, sizeof(ios));
break;
}
/** Get MLOAD version **/
case 19:
return (MLOAD_VER << 4) | MLOAD_SUBVER;
/** Led on **/
case 128:
*(vu32 *)0x0d8000c0 |= 0x20;
break;
/** Led off **/
case 129:
*(vu32 *)0x0d8000c0 &=~ 0x20;
break;
/** Led blink **/
case 130:
*(vu32 *)0x0d8000c0 ^= 0x20;
break;
default:
break;
}
return 0;
}
s32 LoadKernel()
{
u32 TMDSize;
u32 i,u;
s32 r = ES_GetStoredTMDSize( 0x000000010000003aLL, &TMDSize );
if( r < 0 )
{
gprintf("ES_GetStoredTMDSize():%d\r\n", r );
return r;
}
gprintf("TMDSize:%u\r\n", TMDSize );
TitleMetaData *TMD = (TitleMetaData*)memalign( 32, TMDSize );
if( TMD == (TitleMetaData*)NULL )
{
gprintf("Failed to alloc:%u\r\n", TMDSize );
return r; //todo errors are < 0 r still has >= 0 from previous call
}
r = ES_GetStoredTMD( 0x000000010000003aLL, (signed_blob *)TMD, TMDSize );
if( r < 0 )
{
gprintf("ES_GetStoredTMD():%d\r\n", r );
free(TMD);
return r;
}
//Look for boot index
for( i=0; i < TMD->ContentCount; ++i )
{
if( TMD->BootIndex == TMD->Contents[i].Index )
break;
}
gprintf("BootIndex:%u\r\n", i );
s32 cfd = IOS_Open( "/shared1/content.map", 1 );
if( cfd < 0 )
{
gprintf("IOS_Open():%d\r\n", cfd );
free(TMD);
return cfd;
}
for( u=0;; u+=0x1C )
{
if( IOS_Read( cfd, Entry, 0x1C ) != 0x1C )
{
gprintf("Hash not found in content.map\r\n");
free(TMD);
return -2;
}
if( memcmp( (char*)(Entry+8), TMD->Contents[i].SHA1, 0x14 ) == 0 )
break;
}
FoundVersion = ((TMD->TitleID & 0xFFFF) << 16) | (TMD->TitleVersion);
free(TMD);
IOS_Close( cfd );
u32 *IOSVersion = (u32*)0x93003000;
DCInvalidateRange(IOSVersion, 0x20);
*IOSVersion = FoundVersion;
gprintf("IOS Version: 0x%08X\n", FoundVersion);
DCFlushRange(IOSVersion, 0x20);
//char Path[32];
char *Path = (char*)0x93003020;
DCInvalidateRange(Path, 1024);
memset(Path, 0, 1024);
sprintf( Path, "/shared1/%.8s.app", Entry );
gprintf("Kernel:\"%s\"\r\n", Path );
DCFlushRange(Path, 1024);
s32 kfd = IOS_Open( Path, 1 );
if( kfd < 0 )
{
gprintf("IOS_Open():%d\r\n", kfd );
return kfd;
}
KernelSize = IOS_Seek( kfd, 0, SEEK_END );
IOS_Seek( kfd, 0, 0);
gprintf("KernelSize:%u\r\n", KernelSize );
if( IOS_Read( kfd, Kernel, KernelSize ) != KernelSize )
{
gprintf("IOS_Read() failed\r\n");
IOS_Close(kfd);
return -1;
}
IOS_Close(kfd);
return 0;
}
static void __ARCheckSize()
{
u32 i,arsize,arszflag;
static u32 test_data[8] ATTRIBUTE_ALIGN(32);
static u32 dummy_data[8] ATTRIBUTE_ALIGN(32);
static u32 buffer[8] ATTRIBUTE_ALIGN(32);
#ifdef _AR_DEBUG
printf("__ARCheckSize()\n");
#endif
while(!(_dspReg[11]&0x0001));
__ARSize = __ARInternalSize = arsize = 0x1000000;
_dspReg[9] = (_dspReg[9]&~0x3f)|0x23;
for(i=0;i<8;i++) {
test_data[i] = 0xBAD1BAD0;
dummy_data[i] = 0xDEADBEEF;
}
DCFlushRange(test_data,32);
DCFlushRange(dummy_data,32);
__ARExpansionSize = 0;
__ARWriteDMA((u32)test_data,0x1000000,32);
__ARWriteDMA((u32)test_data,0x1200000,32);
__ARWriteDMA((u32)test_data,0x2000000,32);
__ARWriteDMA((u32)test_data,0x1000200,32);
__ARWriteDMA((u32)test_data,0x1400000,32);
memset(buffer,0,32);
DCFlushRange(buffer,32);
__ARWriteDMA((u32)dummy_data,0x1000000,32);
DCInvalidateRange(buffer,32);
__ARReadDMA((u32)buffer,0x1000000,32);
_nop();
arszflag = 0x03;
if(buffer[0]==dummy_data[0]) {
memset(buffer,0,32);
DCFlushRange(buffer,32);
__ARReadDMA((u32)buffer,0x1200000,32);
_nop();
if(buffer[0]==dummy_data[0]) {
__ARExpansionSize = 0x200000;
arsize += 0x200000;
goto end_check; //not nice but fast
}
memset(buffer,0,32);
DCFlushRange(buffer,32);
__ARReadDMA((u32)buffer,0x2000000,32);
_nop();
if(buffer[0]==dummy_data[0]) {
__ARExpansionSize = 0x400000;
arsize += 0x400000;
arszflag |= 0x08;
goto end_check; //not nice but fast
}
memset(buffer,0,32);
DCFlushRange(buffer,32);
__ARReadDMA((u32)buffer,0x1400000,32);
_nop();
if(buffer[0]==dummy_data[0]) {
__ARExpansionSize = 0x1000000;
arsize += 0x1000000;
arszflag |= 0x18;
} else {
__ARExpansionSize = 0x2000000;
arsize += 0x2000000;
arszflag |= 0x20;
}
end_check:
_dspReg[9] = (_dspReg[9]&~0x3f)|arszflag;
}
#ifdef _AR_DEBUG
printf("__ARCheckSize(%d)\n",arsize);
#endif
*(u32*)0x800000d0 = arsize;
__ARSize = arsize;
}
void _start()
{
/* Load a good stack */
asm(
"lis %r1, 0x1ab5 ;"
"ori %r1, %r1, 0xd138 ;"
);
/* Get a handle to coreinit.rpl, dmae.rpl, and nsysnet.rpl */
unsigned int coreinit_handle, dmae_handle, nsysnet_handle;
OSDynLoad_Acquire("coreinit.rpl", &coreinit_handle);
OSDynLoad_Acquire("dmae.rpl", &dmae_handle);
OSDynLoad_Acquire("nsysnet.rpl", &nsysnet_handle);
/* Cache, DMA, and socket functions */
void (*DCFlushRange)(void *addr, unsigned int length);
void (*DCInvalidateRange)(void *addr, unsigned int length);
unsigned long long (*DMAECopyMem)(void *dst, void *src, unsigned int nwords, int endian);
int (*DMAEWaitDone)(unsigned long long ret);
int (*socket)(int family, int type, int proto);
int (*connect)(int fd, struct sockaddr *addr, int addrlen);
int (*send)(int fd, const void *buffer, int len, int flags);
/* Read the addresses of the functions */
OSDynLoad_FindExport(coreinit_handle, 0, "DCFlushRange", &DCFlushRange);
OSDynLoad_FindExport(coreinit_handle, 0, "DCInvalidateRange", &DCInvalidateRange);
OSDynLoad_FindExport(dmae_handle, 0, "DMAECopyMem", &DMAECopyMem);
OSDynLoad_FindExport(dmae_handle, 0, "DMAEWaitDone", &DMAEWaitDone);
OSDynLoad_FindExport(nsysnet_handle, 0, "socket", &socket);
OSDynLoad_FindExport(nsysnet_handle, 0, "connect", &connect);
OSDynLoad_FindExport(nsysnet_handle, 0, "send", &send);
/* Set up our socket address structure */
struct sockaddr sin;
sin.sin_family = AF_INET;
sin.sin_port = 12345;
sin.sin_addr.s_addr = PC_IP;
int i;
for (i = 0; i < 8; i++)
{
sin.sin_zero[i] = 0;
}
/* Connect to the PC */
int pc = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP);
int status = connect(pc, &sin, 0x10);
if (status) OSFatal("Error connecting to PC server");
/* Cache stuff */
DCFlushRange(0x01000000, 0x20);
DCInvalidateRange(0x1dd7b820, 0x20);
/* Do the copy */
int success = DMAEWaitDone(DMAECopyMem(0x1dd7b820, 0x01000000, 2, 0));
if (success)
{
send(pc, "Success", 8, 0);
}
else
{
send(pc, "Fail", 5, 0);
}
while(1);
}
//addr == addr of 4kb block of PowerPC_block ptrs to pull out from ARAM
void ARAM_ReadBlock(u32 addr, int block_num)
{
ARQ_PostRequest(&ARQ_request_blocks, 0x2EAD, AR_ARAMTOMRAM, ARQ_PRIO_LO,
(int)(BLOCKS_CACHE_ADDR + addr), (int)&cached_block[block_num][0], CACHED_BLOCK_SIZE);
DCInvalidateRange((void*)&cached_block[block_num][0], CACHED_BLOCK_SIZE);
}
int BuildLists(GXTexObj texture) {
// Make the new display list
// For display lists, each command has an associated "cost" in bytes.
// Add all these up to calculate the size of your display list before rounding up.
// eke-eke says GX_Begin() costs 3 bytes (u8 + u16)
// According to my research:
// GX_Position3f32() is 12 bytes (f32*3)
// GX_Normal3f32() is 12 bytes (f32*3)
// GX_Color3f32() is actually 3 bytes ((f32 -> u8) * 3)
// GX_TexCoord2f32() is 8 bytes (f32*2)
// GX_End() seems to cost zero (there's no actual code in it)
// Size -must- be multiple of 32, so (12*24) + (12*24) + (3*24) + (8*24) + 3 = 843
// Rounded up to the nearest 32 is 864.
// NOTE: Actual size may be up to 63 bytes -larger- than you calculate it to be due to padding and cache alignment.
for (int i=0; i<5;i++) {
boxList[i] = memalign(32,896);
memset(boxList[i],0,896);
DCInvalidateRange(boxList[i],896);
GX_BeginDispList(boxList[i],896);
GX_Begin(GX_QUADS,GX_VTXFMT0,24); // Start drawing
// Bottom face
GX_Position3f32(-1.0f,-1.0f,-1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(1.0f,1.0f); // Top right
GX_Position3f32( 1.0f,-1.0f,-1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(0.0f,1.0f); // Top left
GX_Position3f32( 1.0f,-1.0f, 1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(0.0f,0.0f); // Bottom left
GX_Position3f32(-1.0f,-1.0f, 1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(1.0f,0.0f); // Bottom right
// Front face
GX_Position3f32(-1.0f,-1.0f, 1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(0.0f,0.0f); // Bottom left
GX_Position3f32( 1.0f,-1.0f, 1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(1.0f,0.0f); // Bottom right
GX_Position3f32( 1.0f, 1.0f, 1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(1.0f,1.0f); // Top right
GX_Position3f32(-1.0f, 1.0f, 1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(0.0f,1.0f); // Top left
// Back face
GX_Position3f32(-1.0f,-1.0f,-1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(1.0f,0.0f); // Bottom right
GX_Position3f32(-1.0f, 1.0f,-1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(1.0f,1.0f); // Top right
GX_Position3f32( 1.0f, 1.0f,-1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(0.0f,1.0f); // Top left
GX_Position3f32( 1.0f,-1.0f,-1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(0.0f,0.0f); // Bottom left
// Right face
GX_Position3f32( 1.0f,-1.0f,-1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(1.0f,0.0f); // Bottom right
GX_Position3f32( 1.0f, 1.0f,-1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(1.0f,1.0f); // Top right
GX_Position3f32( 1.0f, 1.0f, 1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(0.0f,1.0f); // Top left
GX_Position3f32( 1.0f,-1.0f, 1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(0.0f,0.0f); // Bottom left
// Left face
GX_Position3f32(-1.0f,-1.0f,-1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(0.0f,0.0f); // Bottom right
GX_Position3f32(-1.0f,-1.0f, 1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(1.0f,0.0f); // Top right
GX_Position3f32(-1.0f, 1.0f, 1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(1.0f,1.0f); // Top left
GX_Position3f32(-1.0f, 1.0f,-1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(BoxColors[i][0],BoxColors[i][1],BoxColors[i][2]); GX_TexCoord2f32(0.0f,1.0f); // Bottom left
// Top face
GX_Position3f32(-1.0f, 1.0f,-1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(TopColors[i][0],TopColors[i][1],TopColors[i][2]); GX_TexCoord2f32(0.0f,1.0f); // Top left
GX_Position3f32(-1.0f, 1.0f, 1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(TopColors[i][0],TopColors[i][1],TopColors[i][2]); GX_TexCoord2f32(0.0f,0.0f); // Bottom left
GX_Position3f32( 1.0f, 1.0f, 1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(TopColors[i][0],TopColors[i][1],TopColors[i][2]); GX_TexCoord2f32(1.0f,0.0f); // Bottom rught
GX_Position3f32( 1.0f, 1.0f,-1.0f); GX_Normal3f32((f32)0,(f32)0,(f32)1);
GX_Color3f32(TopColors[i][0],TopColors[i][1],TopColors[i][2]); GX_TexCoord2f32(1.0f,1.0f); // Top right
GX_End(); // Done drawing quads
// GX_EndDispList() returns the size of the display list, so store that value and use it with GX_CallDispList().
boxSize[i] = GX_EndDispList(); // Done building the box list
if (boxSize[i] == 0) return 1;
}
// setup texture coordinate generation
// args: texcoord slot 0-7, matrix type, source to generate texture coordinates from, matrix to use
GX_SetTexCoordGen(GX_TEXCOORD0, GX_TG_MTX2x4, GX_TG_TEX0, GX_IDENTITY);
// Set up TEV to paint the textures properly.
GX_SetTevOp(GX_TEVSTAGE0,GX_MODULATE);
GX_SetTevOrder(GX_TEVSTAGE0, GX_TEXCOORD0, GX_TEXMAP0, GX_COLOR0A0);
// Load up the textures (just one this time).
GX_LoadTexObj(&texture, GX_TEXMAP0);
return 0;
}
/* Install an exception handler */
int _start(int argc, char **argv)
{
/* Get a handle to coreinit.rpl and nsysnet.rpl */
unsigned int coreinit_handle, nsysnet_handle;
OSDynLoad_Acquire("coreinit.rpl", &coreinit_handle);
OSDynLoad_Acquire("nsysnet.rpl", &nsysnet_handle);
/* Memory and cache functions */
unsigned int (*OSEffectiveToPhysical)(void *ptr);
void (*DCFlushRange)(void *addr, unsigned int len);
void (*DCInvalidateRange)(void *addr, unsigned int len);
OSDynLoad_FindExport(coreinit_handle, 0, "OSEffectiveToPhysical", &OSEffectiveToPhysical);
OSDynLoad_FindExport(coreinit_handle, 0, "DCFlushRange", &DCFlushRange);
OSDynLoad_FindExport(coreinit_handle, 0, "DCInvalidateRange", &DCInvalidateRange);
/* Kernel is mapped */
if (OSEffectiveToPhysical((void*)0xA0000000) == 0xC0000000)
{
/* Patches installed */
if (*((unsigned int*)(0xA0000000 + (0xFFF00B40 - 0xC0000000))) == 1)
{
/* Socket functions */
int (*socket_lib_init)();
int (*get_socket_rm_fd)();
int (*socket)(int family, int type, int proto);
int (*connect)(int fd, struct sockaddr *addr, int addrlen);
OSDynLoad_FindExport(nsysnet_handle, 0, "socket_lib_init", &socket_lib_init);
OSDynLoad_FindExport(nsysnet_handle, 0, "get_socket_rm_fd", &get_socket_rm_fd);
OSDynLoad_FindExport(nsysnet_handle, 0, "socket", &socket);
OSDynLoad_FindExport(nsysnet_handle, 0, "connect", &connect);
/* Init the sockets library */
socket_lib_init();
/* Set up our socket address structure */
struct sockaddr sin;
sin.sin_family = AF_INET;
sin.sin_port = 12345;
sin.sin_addr.s_addr = PC_IP;
int i;
for (i = 0; i < 8; i++)
{
sin.sin_zero[i] = 0;
}
/* Connect to the PC */
int pc = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP);
int status = connect(pc, &sin, 0x10);
if (status) OSFatal("Error connecting to PC");
/* Store the socket info in a place the kernel can access */
unsigned int sock_info[4] = {1, (unsigned int)get_socket_rm_fd(), (unsigned int)pc, 0};
memcpy((void*)(0xA0000000 + (0xFFF00B40 - 0xC0000000)), sock_info, 0x10);
DCFlushRange((void*)(0xA0000000 + (0xFFF00B40 - 0xC0000000)), 0x20);
DCInvalidateRange((void*)(0xA0000000 + (0xFFF00B40 - 0xC0000000)), 0x20);
}
/* Patches not yet installed, but they will be next time */
else
{
/* Signify patches installed, socket data still invalid, spinlock 0 */
*((unsigned int*)(0xA0000000 + (0xFFF00B40 - 0xC0000000))) = 1;
*((unsigned int*)(0xA0000000 + (0xFFF00B44 - 0xC0000000))) = -1;
*((unsigned int*)(0xA0000000 + (0xFFF00B48 - 0xC0000000))) = -1;
*((unsigned int*)(0xA0000000 + (0xFFF00B4C - 0xC0000000))) = 0;
DCFlushRange((void*)(0xA0000000 + (0xFFF00B40 - 0xC0000000)), 0x20);
DCInvalidateRange((void*)(0xA0000000 + (0xFFF00B40 - 0xC0000000)), 0x20);
}
}
/* Return to main() */
#if VER == 532
int (*main)(int argc, char **argv) = (int (*)(int,char**)) (*((unsigned int*)0x1005D180));
#elif VER == 550
int (*main)(int argc, char **argv) = (int (*)(int,char**)) (*((unsigned int*)0x1005E040));
#else
int (*main)(int argc, char **argv) = (int (*)(int,char**)) (*((unsigned int*)0x0));
#endif
return main(argc, argv);
}
void
InitVideo ()
{
VIDEO_Init();
vmode = VIDEO_GetPreferredMode(NULL); // get default video mode
bool pal = false;
if (vmode == &TVPal528IntDf)
pal = true;
if (CONF_GetAspectRatio() == CONF_ASPECT_16_9)
{
vmode->fbWidth = 640;
vmode->efbHeight = 456;
vmode->viWidth = 686;
if (pal)
{
vmode->xfbHeight = 542;
vmode->viHeight = 542;
}
else
{
vmode->xfbHeight = 456;
vmode->viHeight = 456;
}
}
else
{
if (pal)
vmode = &TVPal574IntDfScale;
vmode->viWidth = 672;
}
if (pal)
{
vmode->viXOrigin = (VI_MAX_WIDTH_PAL - vmode->viWidth) / 2;
vmode->viYOrigin = (VI_MAX_HEIGHT_PAL - vmode->viHeight) / 2;
}
else
{
vmode->viXOrigin = (VI_MAX_WIDTH_NTSC - vmode->viWidth) / 2;
vmode->viYOrigin = (VI_MAX_HEIGHT_NTSC - vmode->viHeight) / 2;
}
VIDEO_Configure (vmode);
// Allocate the video buffers
xfb[0] = (u32 *) SYS_AllocateFramebuffer (vmode);
xfb[1] = (u32 *) SYS_AllocateFramebuffer (vmode);
DCInvalidateRange(xfb[0], VIDEO_GetFrameBufferSize(vmode));
DCInvalidateRange(xfb[1], VIDEO_GetFrameBufferSize(vmode));
xfb[0] = (u32 *) MEM_K0_TO_K1 (xfb[0]);
xfb[1] = (u32 *) MEM_K0_TO_K1 (xfb[1]);
// Clear framebuffers etc.
VIDEO_ClearFrameBuffer (vmode, xfb[0], COLOR_BLACK);
VIDEO_ClearFrameBuffer (vmode, xfb[1], COLOR_BLACK);
VIDEO_SetNextFramebuffer (xfb[0]);
VIDEO_SetBlack (FALSE);
VIDEO_Flush ();
VIDEO_WaitVSync ();
if (vmode->viTVMode & VI_NON_INTERLACE)
VIDEO_WaitVSync ();
StartGX();
ResetVideo_Menu();
// Finally, the video is up and ready for use :)
}
int main(int argc, char **argv)
{
// Exit after 10 seconds if there is an error
__exception_setreload(10);
// u64 timeout = 0;
CheckForGecko();
DCInvalidateRange(loader_stub, 0x1800);
memcpy(loader_stub, (void*)0x80001800, 0x1800);
RAMInit();
//Meh, doesnt do anything anymore anyways
//STM_RegisterEventHandler(HandleSTMEvent);
Initialise();
// Checking for storage devices...
ShowMessageScreen("Checking storage devices...");
u32 u;
//Disables MEMPROT for patches
write16(MEM_PROT, 0);
//Patches FS access
for( u = 0x93A00000; u < 0x94000000; u+=2 )
{
if( memcmp( (void*)(u), FSAccessPattern, sizeof(FSAccessPattern) ) == 0 )
{
// gprintf("FSAccessPatch:%08X\r\n", u );
memcpy( (void*)u, FSAccessPatch, sizeof(FSAccessPatch) );
DCFlushRange((void*)u, sizeof(FSAccessPatch));
break;
}
}
//for BT.c
CONF_GetPadDevices((conf_pads*)0x932C0000);
DCFlushRange((void*)0x932C0000, sizeof(conf_pads));
*(vu32*)0x932C0490 = CONF_GetIRSensitivity();
*(vu32*)0x932C0494 = CONF_GetSensorBarPosition();
DCFlushRange((void*)0x932C0490, 8);
if(LoadKernel() < 0)
{
ClearScreen();
gprintf("Failed to load kernel from NAND!\r\n");
ShowMessageScreenAndExit("Failed to load kernel from NAND!", 1);
}
InsertModule((char*)kernel_bin, kernel_bin_size);
memset( (void*)0x92f00000, 0, 0x100000 );
DCFlushRange( (void*)0x92f00000, 0x100000 );
DCInvalidateRange( (void*)0x939F02F0, 0x20 );
memcpy( (void*)0x939F02F0, Boot2Patch, sizeof(Boot2Patch) );
DCFlushRange( (void*)0x939F02F0, 0x20 );
//libogc still has that, lets close it
__ES_Close();
s32 fd = IOS_Open( "/dev/es", 0 );
memset( STATUS, 0xFFFFFFFF, 0x20 );
DCFlushRange( STATUS, 0x20 );
memset( (void*)0x91000000, 0xFFFFFFFF, 0x20 );
DCFlushRange( (void*)0x91000000, 0x20 );
*(vu32*)0xD3003420 = 0; //make sure kernel doesnt reload
raw_irq_handler_t irq_handler = BeforeIOSReload();
IOS_IoctlvAsync( fd, 0x1F, 0, 0, &IOCTL_Buf, NULL, NULL );
AfterIOSReload( irq_handler, FoundVersion );
while(1)
{
DCInvalidateRange( STATUS, 0x20 );
if((STATUS_LOADING > 0 || abs(STATUS_LOADING) > 1) && STATUS_LOADING < 20)
{
gprintf("Kernel sent signal\n");
break;
}
}
/* For slow USB HDDs */
time_t timeout = time(NULL);
while(time(NULL) - timeout < 10)
{
if(__io_custom_usbstorage.startup() && __io_custom_usbstorage.isInserted())
break;
usleep(50000);
}
fatInitDefault();
gprintf("Nintendont at your service!\r\n%s\r\n", NIN_BUILD_STRING);
KernelLoaded = 1;
char* first_slash = strrchr(argv[0], '/');
if (first_slash != NULL) strncpy(launch_dir, argv[0], first_slash-argv[0]+1);
gprintf("launch_dir = %s\r\n", launch_dir);
FPAD_Init();
FPAD_Update();
/* Read IPL Font before doing any patches */
void *fontbuffer = memalign(32, 0x50000);
__SYS_ReadROM((void*)fontbuffer,0x50000,0x1AFF00);
memcpy((void*)0xD3100000, fontbuffer, 0x50000);
DCInvalidateRange( (void*)0x93100000, 0x50000 );
free(fontbuffer);
//gprintf("Font: 0x1AFF00 starts with %.4s, 0x1FCF00 with %.4s\n", (char*)0x93100000, (char*)0x93100000 + 0x4D000);
// Simple code to autoupdate the meta.xml in Nintendont's folder
FILE *meta = fopen("meta.xml", "w");
if(meta != NULL)
{
fprintf(meta, "%s\r\n<app version=\"1\">\r\n\t<name>%s</name>\r\n", META_XML, META_NAME);
fprintf(meta, "\t<coder>%s</coder>\r\n\t<version>%d.%d</version>\r\n", META_AUTHOR, NIN_VERSION>>16, NIN_VERSION&0xFFFF);
fprintf(meta, "\t<release_date>20150531000000</release_date>\r\n");
fprintf(meta, "\t<short_description>%s</short_description>\r\n", META_SHORT);
fprintf(meta, "\t<long_description>%s\r\n\r\n%s</long_description>\r\n", META_LONG1, META_LONG2);
fprintf(meta, "\t<ahb_access/>\r\n</app>");
fclose(meta);
}
void _start()
{
/* Get a handle to coreinit.rpl */
unsigned int coreinit_handle;
OSDynLoad_Acquire("coreinit.rpl", &coreinit_handle);
//OS memory functions
void* (*memcpy)(void *dest, void *src, uint32_t length);
void* (*memset)(void * dest, uint32_t value, uint32_t bytes);
void* (*OSAllocFromSystem)(uint32_t size, int align);
void (*OSFreeToSystem)(void *ptr);
OSDynLoad_FindExport(coreinit_handle, 0, "memcpy", &memcpy);
OSDynLoad_FindExport(coreinit_handle, 0, "memset", &memset);
OSDynLoad_FindExport(coreinit_handle, 0, "OSAllocFromSystem", &OSAllocFromSystem);
OSDynLoad_FindExport(coreinit_handle, 0, "OSFreeToSystem", &OSFreeToSystem);
void (*OSCoherencyBarrier)();
OSDynLoad_FindExport(coreinit_handle, 0, "OSCoherencyBarrier", &OSCoherencyBarrier);
void (*_Exit)();
OSDynLoad_FindExport(coreinit_handle, 0, "_Exit", &_Exit);
//DC memory functions
void (*DCFlushRangeNoSync)(void *buffer, uint32_t length);
void (*DCInvalidateRange)(void *buffer, uint32_t length);
OSDynLoad_FindExport(coreinit_handle, 0, "DCFlushRangeNoSync", &DCFlushRangeNoSync);
OSDynLoad_FindExport(coreinit_handle, 0, "DCInvalidateRange", &DCInvalidateRange);
//LC memory functions
void* (*LCAlloc)( uint32_t bytes );
void (*LCDealloc)();
uint32_t (*LCHardwareIsAvailable)();
uint32_t (*LCIsDMAEnabled)();
void (*LCEnableDMA)();
void (*LCDisableDMA)();
void (*LCLoadDMABlocks)(void* lc_addr, void* src_addr, uint32_t blocks);
void (*LCStoreDMABlocks)(void* dest_addr, void* lc_addr, uint32_t blocks);
void (*LCWaitDMAQueue)( uint32_t length );
OSDynLoad_FindExport(coreinit_handle, 0, "LCAlloc", &LCAlloc);
OSDynLoad_FindExport(coreinit_handle, 0, "LCDealloc", &LCDealloc);
OSDynLoad_FindExport(coreinit_handle, 0, "LCHardwareIsAvailable", &LCHardwareIsAvailable);
OSDynLoad_FindExport(coreinit_handle, 0, "LCIsDMAEnabled", &LCIsDMAEnabled);
OSDynLoad_FindExport(coreinit_handle, 0, "LCEnableDMA", &LCEnableDMA);
OSDynLoad_FindExport(coreinit_handle, 0, "LCDisableDMA", &LCDisableDMA);
OSDynLoad_FindExport(coreinit_handle, 0, "LCLoadDMABlocks", &LCLoadDMABlocks);
OSDynLoad_FindExport(coreinit_handle, 0, "LCStoreDMABlocks", &LCStoreDMABlocks);
OSDynLoad_FindExport(coreinit_handle, 0, "LCWaitDMAQueue", &LCWaitDMAQueue);
//Used for keeping track of vairables to print to screen
char output[1000];
//Alloc 64 byte alligned space
void* src_addr=OSAllocFromSystem(512,64);
void* dest_addr=OSAllocFromSystem(512,64);
//Store some debug values
__os_snprintf(output, 1000, "src_addr:%02x,",(uint32_t)src_addr);
__os_snprintf(output+strlen(output), 255, "dest_addr: %02x,", (uint32_t)dest_addr);
//Number of 32bit blocks to copy. Must be multiple of 2 between [0,127]
uint32_t blocks=2; //2 32bit blocks
//Do something to the source
memset(src_addr,1,64);
//Grab values for debug
uint32_t * src_val=src_addr;
uint32_t * dest_val=dest_addr;
__os_snprintf(output+strlen(output), 255, "Old src_val: %02x\n",src_val[0]);
__os_snprintf(output+strlen(output), 255, "Old dest_val: %02x,",dest_val[0]);
//Get some locked cache address space
void *lc_addr=LCAlloc(512); //512 Minmum. Must be multiple of 512.
//Calculate size from blocks to flush/invalidate range properly
uint32_t size;
//If blocks is set to 0, the transaction will default to 128 blocks
if(blocks==0)
{
size=32*128;
}
else
{
size=32*blocks;
}
//Flush the range at the source to ensure cache gets written back to memory.
DCFlushRangeNoSync(src_addr,size);
//Invalidate the range at the destination
DCInvalidateRange(dest_addr,size);
//Sync
OSCoherencyBarrier();
//Check to see of DMA hardware is avaliable
if(LCHardwareIsAvailable()!=1)
{
OSFatal("Hardware is not avaliable.");
}
//Gets the current state of DMA, so we can restore it after our copy
uint32_t dmaState=LCIsDMAEnabled();
//Checks to see if DMA is enabled, if not it will try to enable it.
if(dmaState!=1)
{
LCEnableDMA();
dmaState=LCIsDMAEnabled();
if(dmaState!=1)
{
OSFatal("Can't enable DMA");
}
}
//Load memory to locked cache
LCLoadDMABlocks(lc_addr,src_addr,blocks);
LCWaitDMAQueue(0);
//Store memory from locked cache
LCStoreDMABlocks(dest_addr,lc_addr,blocks);
LCWaitDMAQueue(0);
//If DMA was not previously enabled, then disable it to restore state.
if(dmaState!=1)
{
LCDisableDMA();
dmaState=LCIsDMAEnabled();
//If DMA failed to disable, return error code
if(dmaState!=1)
{
OSFatal("Couldn't Disable DMA");
}
}
__os_snprintf(output+strlen(output), 255, "New src_val: %02x,",src_val[0]);
__os_snprintf(output+strlen(output), 255, "New dest_val: %02x,",dest_val[0]);
//Cleanup
LCDealloc(lc_addr);
OSFreeToSystem(dest_addr);
OSFreeToSystem(src_addr);
OSFatal(output);
}