diff --git a/source/gui/video.cpp b/source/gui/video.cpp
index 1acc4f98..b5edc25e 100644
--- a/source/gui/video.cpp
+++ b/source/gui/video.cpp
@@ -6,7 +6,7 @@
 #include "memory/mem2.hpp"
 #include "video.hpp"
 #include "pngu.h"
-#include "Gekko.h"
+#include "hw/Gekko.h"
 #include "gecko/gecko.hpp"
 #include "loader/sys.h"
 #include "loader/utils.h"
@@ -526,8 +526,6 @@ void CVideo::_showWaitMessages(CVideo *m)
 
 	m->prepare();
 	m->setup2DProjection();
-	wiiLightSetLevel(0);
-	wiiLightOn();
 
 	//gprintf("Wait Message Thread: Start\nDelay: %d, Images: %d\n", waitFrames, m->m_waitMessages.size());
 	while(m->m_showWaitMessage)
@@ -558,7 +556,6 @@ void CVideo::_showWaitMessages(CVideo *m)
 			VIDEO_WaitVSync();
 		waitFrames--;
 	}
-	wiiLightOff();
 	//gprintf("Wait Message Thread: End\n");
 	m->m_showingWaitMessages = false;
 }
@@ -571,6 +568,7 @@ void CVideo::hideWaitMessage()
 	m_showWaitMessage = false;
 	if(waitThread != LWP_THREAD_NULL)
 	{
+		/* end animation */
 		if(LWP_ThreadIsSuspended(waitThread))
 			LWP_ResumeThread(waitThread);
 		while(m_showingWaitMessages)
@@ -579,6 +577,8 @@ void CVideo::hideWaitMessage()
 		if(waitMessageStack != NULL)
 			MEM2_free(waitMessageStack);
 		waitMessageStack = NULL;
+		/* end light thread */
+		wiiLightEndThread();
 		m_WaitThreadRunning = false;
 	}
 	waitThread = LWP_THREAD_NULL;
@@ -623,6 +623,10 @@ void CVideo::waitMessage(const vector<TexData> &tex, float delay)
 		waitMessage(m_waitMessages[0]);
 	else if(m_waitMessages.size() > 1)
 	{
+		/* changing light */
+		wiiLightSetLevel(0);
+		wiiLightStartThread();
+		/* onscreen animation */
 		m_showWaitMessage = true;
 		if(waitMessageStack == NULL)
 			waitMessageStack = (u8*)MEM2_memalign(32, waitMessageStackSize);
diff --git a/source/gui/Gekko.c b/source/hw/Gekko.c
similarity index 91%
rename from source/gui/Gekko.c
rename to source/hw/Gekko.c
index b3e1ebfa..bf7f8eda 100644
--- a/source/gui/Gekko.c
+++ b/source/hw/Gekko.c
@@ -28,8 +28,6 @@
 #include "Gekko.h"
 #include "memory/memory.h"
 
-#define DISC_SLOT_LED		0x20
-
 lwp_t light_thread = LWP_THREAD_NULL;
 
 void *light_loop();
@@ -39,7 +37,18 @@ u8 light_level = 0;
 struct timespec light_timeon;
 struct timespec light_timeoff;
 
+// Wii disc slot light routines
 void wiiLightOn()
+{
+	*HW_GPIOB_OUT |= DISC_SLOT_LED;
+}
+
+void wiiLightOff()
+{
+	*HW_GPIOB_OUT &= ~DISC_SLOT_LED;
+}
+
+void wiiLightStartThread()
 {
 	light_on = true;
 	light_level = 0;
@@ -47,13 +56,14 @@ void wiiLightOn()
 		LWP_CreateThread(&light_thread, light_loop, NULL, NULL, 0, LWP_PRIO_HIGHEST);
 }
 
-void wiiLightOff()
+void wiiLightEndThread()
 {
 	light_on = false;
 	light_level = 0;
-	LWP_JoinThread(light_thread, NULL);
+	if(light_thread != LWP_THREAD_NULL)
+		LWP_JoinThread(light_thread, NULL);
 	light_thread = LWP_THREAD_NULL;
-	*HW_GPIOB_OUT &= ~DISC_SLOT_LED;
+	wiiLightOff();
 }
 
 void wiiLightSetLevel(int level)
@@ -89,11 +99,11 @@ void *light_loop()
 		timeon = light_timeon;
 		timeoff = light_timeoff;
 		// Turn on the light and sleep for a bit
-		*HW_GPIOB_OUT |= DISC_SLOT_LED;
+		wiiLightOn();
 		nanosleep(&timeon);
 		// Turn off the light (if required) and sleep for a bit
 		if(timeoff.tv_nsec > 0)
-			*HW_GPIOB_OUT &= ~DISC_SLOT_LED;
+			wiiLightOff();
 		nanosleep(&timeoff);
 	}
 	return NULL;
diff --git a/source/gui/Gekko.h b/source/hw/Gekko.h
similarity index 92%
rename from source/gui/Gekko.h
rename to source/hw/Gekko.h
index 60a65d44..b946721a 100644
--- a/source/gui/Gekko.h
+++ b/source/hw/Gekko.h
@@ -27,9 +27,13 @@
 extern "C" {
 #endif
 
+static const u32 DISC_SLOT_LED = 0x20;
+
 // Wii disc slot light routines
 void wiiLightOn();
 void wiiLightOff();
+void wiiLightStartThread();
+void wiiLightEndThread();
 void wiiLightSetLevel(int level);
 
 #ifdef __cplusplus
diff --git a/source/libwbfs/libwbfs.c b/source/libwbfs/libwbfs.c
index 5e758c0c..4ca2c6af 100644
--- a/source/libwbfs/libwbfs.c
+++ b/source/libwbfs/libwbfs.c
@@ -40,7 +40,7 @@ static u8 size_to_shift(u32 size)
 wbfs_t *wbfs_open_hd(rw_sector_callback_t read_hdsector, rw_sector_callback_t write_hdsector, void *callback_data, int hd_sector_size, int num_hd_sector __attribute((unused)), int reset)
 {
 	int i=num_hd_sector,ret;
-	u8 *ptr,*tmp_buffer = wbfs_ioalloc(hd_sector_size);
+	u8 *ptr,*tmp_buffer = wbfs_malloc(hd_sector_size);
 	u8 part_table[16*4];
 	ret = read_hdsector(callback_data,0,1,tmp_buffer);
 	if(ret)
@@ -57,11 +57,11 @@ wbfs_t *wbfs_open_hd(rw_sector_callback_t read_hdsector, rw_sector_callback_t wr
 		if (head->magic == wbfs_htonl(WBFS_MAGIC))
 		{
 			wbfs_t *p = wbfs_open_partition(read_hdsector, write_hdsector, callback_data, hd_sector_size, 0, part_lba,reset);
-			wbfs_iofree(tmp_buffer);
+			wbfs_free(tmp_buffer);
 			return p;
 		}
 	}
-	wbfs_iofree(tmp_buffer);
+	wbfs_free(tmp_buffer);
 	if(reset)// XXX make a empty hd partition..
 	{
 	}
@@ -73,7 +73,7 @@ wbfs_t *wbfs_open_partition(rw_sector_callback_t read_hdsector, rw_sector_callba
 {
 	wbfs_t *p = wbfs_malloc(sizeof(wbfs_t));
 
-	wbfs_head_t *head = wbfs_ioalloc( hd_sector_size ? hd_sector_size : 512 );
+	wbfs_head_t *head = wbfs_malloc( hd_sector_size ? hd_sector_size : 512 );
 
 	//constants, but put here for consistancy
 	p->wii_sec_sz = 0x8000;
@@ -131,20 +131,20 @@ wbfs_t *wbfs_open_partition(rw_sector_callback_t read_hdsector, rw_sector_callba
 	else
 	{
 		// init with all free blocks
-		p->freeblks = wbfs_ioalloc(ALIGN_LBA( p->n_wbfs_sec / 8));
+		p->freeblks = wbfs_malloc(ALIGN_LBA( p->n_wbfs_sec / 8));
 		wbfs_memset(p->freeblks, 0xff, p->n_wbfs_sec / 8);
 	}
 	p->max_disc = (p->freeblks_lba - 1) / (p->disc_info_sz >> p->hd_sec_sz_s);
 	if(p->max_disc > p->hd_sec_sz - sizeof(wbfs_head_t)) 
 		p->max_disc = p->hd_sec_sz - sizeof(wbfs_head_t);
 
-	p->tmp_buffer = wbfs_ioalloc(p->hd_sec_sz);
+	p->tmp_buffer = wbfs_malloc(p->hd_sec_sz);
 	p->n_disc_open = 0;
 	return p;
 
 error:
 	wbfs_free(p);
-	wbfs_iofree(head);
+	wbfs_free(head);
 
 	return 0;
 }
@@ -167,10 +167,10 @@ void wbfs_close(wbfs_t *p)
 	if(p->n_disc_open)
 		ERROR("trying to close wbfs while discs still open\n");
 
-	wbfs_iofree(p->head);
-	wbfs_iofree(p->tmp_buffer);
+	wbfs_free(p->head);
+	wbfs_free(p->tmp_buffer);
 	if(p->freeblks)
-		wbfs_iofree(p->freeblks);
+		wbfs_free(p->freeblks);
 	
 	wbfs_free(p);
 	
@@ -195,7 +195,7 @@ wbfs_disc_t *wbfs_open_disc(wbfs_t* p, const u8 *discid)
 					ERROR("allocating memory\n");
 				d->p = p;
 				d->i = i;
-				d->header = wbfs_ioalloc(p->disc_info_sz);
+				d->header = wbfs_malloc(p->disc_info_sz);
 				if(!d->header)
 					ERROR("allocating memory\n");
 				p->read_hdsector(p->callback_data, p->part_lba + 1 + i * disc_info_sz_lba, disc_info_sz_lba, d->header);
@@ -209,13 +209,13 @@ wbfs_disc_t *wbfs_open_disc(wbfs_t* p, const u8 *discid)
 	return 0;
 
 error:
-	if(d) wbfs_iofree(d);
+	if(d) wbfs_free(d);
 	return 0;
 }
 void wbfs_close_disc(wbfs_disc_t*d)
 {
 	d->p->n_disc_open --;
-	wbfs_iofree(d->header);
+	wbfs_free(d->header);
 	wbfs_free(d);
 }
 // offset is pointing 32bit words to address the whole dvd, although len is in bytes
@@ -349,10 +349,10 @@ u32 wbfs_get_disc_info(wbfs_t *p, u32 index,u8 *header,int header_size,u32 *size
 				memcpy(header, p->tmp_buffer, header_size);
 				if(size)
 				{
-					u8 *header = wbfs_ioalloc(p->disc_info_sz);
+					u8 *header = wbfs_malloc(p->disc_info_sz);
 					p->read_hdsector(p->callback_data, p->part_lba + 1 + i * disc_info_sz_lba, disc_info_sz_lba, header);
 					u32 sec_used = wbfs_sector_used(p,(wbfs_disc_info_t *)header);
-					wbfs_iofree(header);
+					wbfs_free(header);
 					*size = sec_used<<(p->wbfs_sec_sz_s-2);
 				}
 				return 0;
@@ -367,7 +367,7 @@ static void load_freeblocks(wbfs_t *p)
 	if(p->freeblks)
 		return;
 	// XXX should handle malloc error..
-	p->freeblks = wbfs_ioalloc(ALIGN_LBA(p->n_wbfs_sec/8));
+	p->freeblks = wbfs_malloc(ALIGN_LBA(p->n_wbfs_sec/8));
 	p->read_hdsector(p->callback_data, p->part_lba + p->freeblks_lba, ALIGN_LBA(p->n_wbfs_sec / 8) >> p->hd_sec_sz_s, p->freeblks);
 }
 
@@ -476,10 +476,10 @@ u32 wbfs_add_disc(wbfs_t *p, read_wiidisc_callback_t read_src_wii_disc, void *ca
 	load_freeblocks(p);
 
 	// build disc info
-	info = wbfs_ioalloc(p->disc_info_sz);
+	info = wbfs_malloc(p->disc_info_sz);
 	read_src_wii_disc(callback_data, 0, 0x100, info->disc_header_copy);
 
-	copy_buffer = wbfs_ioalloc(p->wii_sec_sz);
+	copy_buffer = wbfs_malloc(p->wii_sec_sz);
 	if(!copy_buffer)
 		ERROR("alloc memory\n");
 	tot = 0;
@@ -568,9 +568,9 @@ error:
 	if(used)
 		wbfs_free(used);
 	if(info)
-		wbfs_iofree(info);
+		wbfs_free(info);
 	if(copy_buffer)
-		wbfs_iofree(copy_buffer);
+		wbfs_free(copy_buffer);
 
 	// init with all free blocks
 	return retval;
@@ -622,7 +622,7 @@ u32 wbfs_size_disc(wbfs_t *p,read_wiidisc_callback_t read_src_wii_disc,
 	u32 wii_sec_per_wbfs_sect = 1 << (p->wbfs_sec_sz_s - p->wii_sec_sz_s);
        wiidisc_t *d = 0;
 
-	u8 *used = wbfs_ioalloc(p->n_wii_sec_per_disc);
+	u8 *used = wbfs_malloc(p->n_wii_sec_per_disc);
 	if(!used) 
 		ERROR("unable to alloc memory\n");
 
diff --git a/source/libwbfs/libwbfs_os.h b/source/libwbfs/libwbfs_os.h
index c9844abb..4307800a 100644
--- a/source/libwbfs/libwbfs_os.h
+++ b/source/libwbfs/libwbfs_os.h
@@ -14,8 +14,22 @@
 #include "loader/utils.h"
 #include "memory/mem2.hpp"
 
-#define wbfs_fatal(x)		do { gprintf(x); wd_last_error = 1; } while(0)
-#define wbfs_error(x)		do { gprintf(x); wd_last_error = 2; } while(0)
+#ifdef __cplusplus
+extern "C" {
+#endif /* __cplusplus */
+
+extern int wd_last_error;
+static inline void wbfs_fatal(const char *x)
+{
+	gprintf(x);
+	wd_last_error = 1;
+}
+
+static inline void wbfs_error(const char *x)
+{
+	gprintf(x);
+	wd_last_error = 2;
+}
 
 static inline void *wbfs_malloc(size_t size)
 {
@@ -29,8 +43,9 @@ static inline void wbfs_free(void *ptr)
 	MEM2_free(ptr);
 }
 
-#define wbfs_ioalloc(x)		wbfs_malloc(x)
-#define wbfs_iofree(x)		wbfs_free(x)
+#ifdef __cplusplus
+}
+#endif /* __cplusplus */
 
 #define wbfs_be16(x)		(*((u16*)(x)))
 #define wbfs_be32(x)		(*((u32*)(x)))
diff --git a/source/libwbfs/rijndael.c b/source/libwbfs/rijndael.c
new file mode 100644
index 00000000..104d97df
--- /dev/null
+++ b/source/libwbfs/rijndael.c
@@ -0,0 +1,445 @@
+/* Rijndael Block Cipher - rijndael.c
+
+ Written by Mike Scott 21st April 1999
+ mike@compapp.dcu.ie
+
+ Permission for free direct or derivative use is granted subject
+ to compliance with any conditions that the originators of the
+ algorithm place on its exploitation.
+
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include "gecko/gecko.hpp"
+#include "loader/utils.h"
+#define u8 unsigned char	   /* 8 bits  */
+#define u32 unsigned long	   /* 32 bits */
+#define u64 unsigned long long
+
+/* rotates x one bit to the left */
+
+#define ROTL(x) (((x)>>7)|((x)<<1))
+
+/* Rotates 32-bit word left by 1, 2 or 3 byte  */
+
+#define ROTL8(x) (((x)<<8)|((x)>>24))
+#define ROTL16(x) (((x)<<16)|((x)>>16))
+#define ROTL24(x) (((x)<<24)|((x)>>8))
+
+/* Fixed Data */
+
+static u8 InCo[4] = { 0xB, 0xD, 0x9, 0xE }; /* Inverse Coefficients */
+
+static u8 fbsub[256];
+static u8 rbsub[256];
+static u8 ptab[256], ltab[256];
+static u32 ftable[256];
+static u32 rtable[256];
+static u32 rco[30];
+
+/* Parameter-dependent data */
+
+int Nk, Nb, Nr;
+u8 fi[24], ri[24];
+u32 fkey[120];
+u32 rkey[120];
+
+static u32 pack(u8 *b)
+{ /* pack bytes into a 32-bit Word */
+	return ((u32 ) b[3] << 24) | ((u32 ) b[2] << 16) | ((u32 ) b[1] << 8) | (u32 ) b[0];
+}
+
+static void unpack(u32 a, u8 *b)
+{ /* unpack bytes from a word */
+	b[0] = (u8 ) a;
+	b[1] = (u8 ) (a >> 8);
+	b[2] = (u8 ) (a >> 16);
+	b[3] = (u8 ) (a >> 24);
+}
+
+static u8 xtime(u8 a)
+{
+	u8 b;
+	if (a & 0x80)
+		b = 0x1B;
+	else b = 0;
+	a <<= 1;
+	a ^= b;
+	return a;
+}
+
+static u8 bmul(u8 x, u8 y)
+{ /* x.y= AntiLog(Log(x) + Log(y)) */
+	if (x && y)
+		return ptab[(ltab[x] + ltab[y]) % 255];
+	else return 0;
+}
+
+static u32 SubByte(u32 a)
+{
+	u8 b[4];
+	unpack(a, b);
+	b[0] = fbsub[b[0]];
+	b[1] = fbsub[b[1]];
+	b[2] = fbsub[b[2]];
+	b[3] = fbsub[b[3]];
+	return pack(b);
+}
+
+static u8 product(u32 x, u32 y)
+{ /* dot product of two 4-byte arrays */
+	u8 xb[4], yb[4];
+	unpack(x, xb);
+	unpack(y, yb);
+	return bmul(xb[0], yb[0]) ^ bmul(xb[1], yb[1]) ^ bmul(xb[2], yb[2]) ^ bmul(xb[3], yb[3]);
+}
+
+static u32 InvMixCol(u32 x)
+{ /* matrix Multiplication */
+	u32 y, m;
+	u8 b[4];
+
+	m = pack(InCo);
+	b[3] = product(m, x);
+	m = ROTL24( m );
+	b[2] = product(m, x);
+	m = ROTL24( m );
+	b[1] = product(m, x);
+	m = ROTL24( m );
+	b[0] = product(m, x);
+	y = pack(b);
+	return y;
+}
+
+u8 ByteSub(u8 x)
+{
+	u8 y = ptab[255 - ltab[x]]; /* multiplicative inverse */
+	x = y;
+	x = ROTL( x );
+	y ^= x;
+	x = ROTL( x );
+	y ^= x;
+	x = ROTL( x );
+	y ^= x;
+	x = ROTL( x );
+	y ^= x;
+	y ^= 0x63;
+	return y;
+}
+
+void gentables(void)
+{ /* generate tables */
+	int i;
+	u8 y, b[4];
+
+	/* use 3 as primitive root to generate power and log tables */
+
+	ltab[0] = 0;
+	ptab[0] = 1;
+	ltab[1] = 0;
+	ptab[1] = 3;
+	ltab[3] = 1;
+	for (i = 2; i < 256; i++)
+	{
+		ptab[i] = ptab[i - 1] ^ xtime(ptab[i - 1]);
+		ltab[ptab[i]] = i;
+	}
+
+	/* affine transformation:- each bit is xored with itself shifted one bit */
+
+	fbsub[0] = 0x63;
+	rbsub[0x63] = 0;
+	for (i = 1; i < 256; i++)
+	{
+		y = ByteSub((u8 ) i);
+		fbsub[i] = y;
+		rbsub[y] = i;
+	}
+
+	for (i = 0, y = 1; i < 30; i++)
+	{
+		rco[i] = y;
+		y = xtime(y);
+	}
+
+	/* calculate forward and reverse tables */
+	for (i = 0; i < 256; i++)
+	{
+		y = fbsub[i];
+		b[3] = y ^ xtime(y);
+		b[2] = y;
+		b[1] = y;
+		b[0] = xtime(y);
+		ftable[i] = pack(b);
+
+		y = rbsub[i];
+		b[3] = bmul(InCo[0], y);
+		b[2] = bmul(InCo[1], y);
+		b[1] = bmul(InCo[2], y);
+		b[0] = bmul(InCo[3], y);
+		rtable[i] = pack(b);
+	}
+}
+
+void gkey(int nb, int nk, char *key)
+{ /* blocksize=32*nb bits. Key=32*nk bits */
+	/* currently nb,bk = 4, 6 or 8		  */
+	/* key comes as 4*Nk bytes			  */
+	/* Key Scheduler. Create expanded encryption key */
+	int i, j, k, m, N;
+	int C1, C2, C3;
+	u32 CipherKey[8];
+
+	Nb = nb;
+	Nk = nk;
+
+	/* Nr is number of rounds */
+	if (Nb >= Nk)
+		Nr = 6 + Nb;
+	else Nr = 6 + Nk;
+
+	C1 = 1;
+	if (Nb < 8)
+	{
+		C2 = 2;
+		C3 = 3;
+	}
+	else
+	{
+		C2 = 3;
+		C3 = 4;
+	}
+
+	/* pre-calculate forward and reverse increments */
+	for (m = j = 0; j < nb; j++, m += 3)
+	{
+		fi[m] = (j + C1) % nb;
+		fi[m + 1] = (j + C2) % nb;
+		fi[m + 2] = (j + C3) % nb;
+		ri[m] = (nb + j - C1) % nb;
+		ri[m + 1] = (nb + j - C2) % nb;
+		ri[m + 2] = (nb + j - C3) % nb;
+	}
+
+	N = Nb * (Nr + 1);
+
+	for (i = j = 0; i < Nk; i++, j += 4)
+	{
+		CipherKey[i] = pack((u8 *) &key[j]);
+	}
+	for (i = 0; i < Nk; i++)
+		fkey[i] = CipherKey[i];
+	for (j = Nk, k = 0; j < N; j += Nk, k++)
+	{
+		fkey[j] = fkey[j - Nk] ^ SubByte(ROTL24( fkey[j-1] )) ^ rco[k];
+		if (Nk <= 6)
+		{
+			for (i = 1; i < Nk && (i + j) < N; i++)
+				fkey[i + j] = fkey[i + j - Nk] ^ fkey[i + j - 1];
+		}
+		else
+		{
+			for (i = 1; i < 4 && (i + j) < N; i++)
+				fkey[i + j] = fkey[i + j - Nk] ^ fkey[i + j - 1];
+			if ((j + 4) < N) fkey[j + 4] = fkey[j + 4 - Nk] ^ SubByte(fkey[j + 3]);
+			for (i = 5; i < Nk && (i + j) < N; i++)
+				fkey[i + j] = fkey[i + j - Nk] ^ fkey[i + j - 1];
+		}
+
+	}
+
+	/* now for the expanded decrypt key in reverse order */
+
+	for (j = 0; j < Nb; j++)
+		rkey[j + N - Nb] = fkey[j];
+	for (i = Nb; i < N - Nb; i += Nb)
+	{
+		k = N - Nb - i;
+		for (j = 0; j < Nb; j++)
+			rkey[k + j] = InvMixCol(fkey[i + j]);
+	}
+	for (j = N - Nb; j < N; j++)
+		rkey[j - N + Nb] = fkey[j];
+}
+
+/* There is an obvious time/space trade-off possible here.	 *
+ * Instead of just one ftable[], I could have 4, the other	 *
+ * 3 pre-rotated to save the ROTL8, ROTL16 and ROTL24 overhead */
+
+void encrypt(char *buff)
+{
+	int i, j, k, m;
+	u32 a[8], b[8], *x, *y, *t;
+
+	for (i = j = 0; i < Nb; i++, j += 4)
+	{
+		a[i] = pack((u8 *) &buff[j]);
+		a[i] ^= fkey[i];
+	}
+	k = Nb;
+	x = a;
+	y = b;
+
+	/* State alternates between a and b */
+	for (i = 1; i < Nr; i++)
+	{ /* Nr is number of rounds. May be odd. */
+
+		/* if Nb is fixed - unroll this next
+		 loop and hard-code in the values of fi[]  */
+
+		for (m = j = 0; j < Nb; j++, m += 3)
+		{ /* deal with each 32-bit element of the State */
+			/* This is the time-critical bit */
+			y[j] = fkey[k++] ^ ftable[(u8 ) x[j]] ^ ROTL8( ftable[( u8 )( x[fi[m]] >> 8 )] )
+					^ ROTL16( ftable[( u8 )( x[fi[m+1]] >> 16 )] ) ^ ROTL24( ftable[x[fi[m+2]] >> 24] );
+		}
+		t = x;
+		x = y;
+		y = t; /* swap pointers */
+	}
+
+	/* Last Round - unroll if possible */
+	for (m = j = 0; j < Nb; j++, m += 3)
+	{
+		y[j] = fkey[k++] ^ (u32 ) fbsub[(u8 ) x[j]] ^ ROTL8( ( u32 )fbsub[( u8 )( x[fi[m]] >> 8 )] )
+				^ ROTL16( ( u32 )fbsub[( u8 )( x[fi[m+1]] >> 16 )] ) ^ ROTL24( ( u32 )fbsub[x[fi[m+2]] >> 24] );
+	}
+	for (i = j = 0; i < Nb; i++, j += 4)
+	{
+		unpack(y[i], (u8 *) &buff[j]);
+		x[i] = y[i] = 0; /* clean up stack */
+	}
+	return;
+}
+
+void decrypt(char *buff)
+{
+	int i, j, k, m;
+	u32 a[8], b[8], *x, *y, *t;
+
+	for (i = j = 0; i < Nb; i++, j += 4)
+	{
+		a[i] = pack((u8 *) &buff[j]);
+		a[i] ^= rkey[i];
+	}
+	k = Nb;
+	x = a;
+	y = b;
+
+	/* State alternates between a and b */
+	for (i = 1; i < Nr; i++)
+	{ /* Nr is number of rounds. May be odd. */
+
+		/* if Nb is fixed - unroll this next
+		 loop and hard-code in the values of ri[]  */
+
+		for (m = j = 0; j < Nb; j++, m += 3)
+		{ /* This is the time-critical bit */
+			y[j] = rkey[k++] ^ rtable[(u8 ) x[j]] ^ ROTL8( rtable[( u8 )( x[ri[m]] >> 8 )] )
+					^ ROTL16( rtable[( u8 )( x[ri[m+1]] >> 16 )] ) ^ ROTL24( rtable[x[ri[m+2]] >> 24] );
+		}
+		t = x;
+		x = y;
+		y = t; /* swap pointers */
+	}
+
+	/* Last Round - unroll if possible */
+	for (m = j = 0; j < Nb; j++, m += 3)
+	{
+		y[j] = rkey[k++] ^ (u32 ) rbsub[(u8 ) x[j]] ^ ROTL8( ( u32 )rbsub[( u8 )( x[ri[m]] >> 8 )] )
+				^ ROTL16( ( u32 )rbsub[( u8 )( x[ri[m+1]] >> 16 )] ) ^ ROTL24( ( u32 )rbsub[x[ri[m+2]] >> 24] );
+	}
+	for (i = j = 0; i < Nb; i++, j += 4)
+	{
+		unpack(y[i], (u8 *) &buff[j]);
+		x[i] = y[i] = 0; /* clean up stack */
+	}
+	return;
+}
+
+void aes_set_key(const u8 *key)
+{
+	gentables();
+	gkey(4, 4, (char*) key);
+}
+
+// CBC mode decryption
+void aes_decrypt(u8 *iv, u8 *inbuf, u8 *outbuf, u64 len)
+{
+	u8 block[16];
+	u32 blockno = 0, i;
+
+	//printf("aes_decrypt(%p, %p, %p, %lld)\n", iv, inbuf, outbuf, len);
+
+	for (blockno = 0; blockno <= (len / sizeof(block)); blockno++)
+	{
+		u32 fraction;
+		if (blockno == (len / sizeof(block))) // last block
+		{
+			fraction = len % sizeof(block);
+			if (fraction == 0) break;
+			memset(block, 0, sizeof(block));
+		}
+		else fraction = 16;
+
+		//	debug_printf("block %d: fraction = %d\n", blockno, fraction);
+		memcpy(block, inbuf + blockno * sizeof(block), fraction);
+		decrypt((char*) block);
+		u8 *ctext_ptr;
+		if (blockno == 0)
+			ctext_ptr = iv;
+		else ctext_ptr = inbuf + (blockno - 1) * sizeof(block);
+
+		for (i = 0; i < fraction; i++)
+			outbuf[blockno * sizeof(block) + i] = ctext_ptr[i] ^ block[i];
+		//	debug_printf("Block %d output: ", blockno);
+		//	hexdump(outbuf + blockno*sizeof(block), 16);
+	}
+}
+
+// CBC mode encryption
+void aes_encrypt(u8 *iv, u8 *inbuf, u8 *outbuf, u64 len)
+{
+	u8 block[16];
+	u32 blockno = 0, i;
+
+	//  debug_printf("aes_decrypt(%p, %p, %p, %lld)\n", iv, inbuf, outbuf, len);
+
+	for (blockno = 0; blockno <= (len / sizeof(block)); blockno++)
+	{
+		u32 fraction;
+		if (blockno == (len / sizeof(block))) // last block
+		{
+			fraction = len % sizeof(block);
+			if (fraction == 0) break;
+			memset(block, 0, sizeof(block));
+		}
+		else fraction = 16;
+
+		//	debug_printf("block %d: fraction = %d\n", blockno, fraction);
+		memcpy(block, inbuf + blockno * sizeof(block), fraction);
+
+		for (i = 0; i < fraction; i++)
+			block[i] = inbuf[blockno * sizeof(block) + i] ^ iv[i];
+
+		encrypt((char*) block);
+		memcpy(iv, block, sizeof(block));
+		memcpy(outbuf + blockno * sizeof(block), block, sizeof(block));
+		//	debug_printf("Block %d output: ", blockno);
+		//	hexdump(outbuf + blockno*sizeof(block), 16);
+	}
+}
+
+
+// CBC mode decryption
+#define WAD_BUF 0x10000
+
+void aes_decrypt_partial(u8 *inbuf, u8 *outbuf, u8 block[16], u8 *ctext_ptr, u32 tmp_blockno)
+{
+	memcpy(block, inbuf + tmp_blockno * 16, 16);
+	decrypt((char*)block);
+	u32 i;
+	for(i = 0; i < 16; i++)
+		outbuf[tmp_blockno * 16 + i] = ctext_ptr[i] ^ block[i];
+}
diff --git a/source/libwbfs/wiidisc.c b/source/libwbfs/wiidisc.c
index ca26e039..666912d8 100644
--- a/source/libwbfs/wiidisc.c
+++ b/source/libwbfs/wiidisc.c
@@ -6,6 +6,10 @@
 #include "wiidisc.h"
 #include "hw/aes.h"
 #include "loader/nk.h"
+#include "loader/sys.h"
+
+void aes_set_key(const u8 *key);
+void aes_decrypt(u8 *iv, u8 *inbuf, u8 *outbuf, u64 len);
 
 int wd_last_error = 0;
 
@@ -32,11 +36,19 @@ void decrypt_title_key(u8 *tik, u8 *title_key)
 	wbfs_memset(iv, 0, sizeof iv);
 	wbfs_memcpy(iv, tik + 0x01dc, 8);
 
-	AES_ResetEngine();
 	//check byte 0x1f1 in ticket to determine whether or not to use Korean Common Key
 	//if value = 0x01, use Korean Common Key, else just use regular one
-	AES_EnableDecrypt((tik[0x01f1] == 1) ? korean_key : common_key, iv);
-	AES_Decrypt(tik + 0x01bf, title_key, 1);
+	if(Sys_HW_Access())
+	{
+		AES_ResetEngine();
+		AES_EnableDecrypt((tik[0x01f1] == 1) ? korean_key : common_key, iv);
+		AES_Decrypt(tik + 0x01bf, title_key, 1);
+	}
+	else
+	{
+		aes_set_key((tik[0x01f1] == 1) ? korean_key : common_key);
+		aes_decrypt(iv, tik + 0x01bf, title_key, 16);
+	}
 }
 
 static u32 _be32(const u8 *p)
@@ -44,19 +56,22 @@ static u32 _be32(const u8 *p)
 	return (p[0] << 24) | (p[1] << 16) | (p[2] << 8) | p[3];
 }
 
-static void disc_read(wiidisc_t *d, u32 offset, u8 *data, u32 len)
+static int disc_read(wiidisc_t *d, u32 offset, u8 *data, u32 len)
 {
 	if (data)
 	{
 		int ret = 0;
 		if (len == 0)
-			return;
+			return -1;
 
 		ret = d->read(d->fp, offset, len, data);
 		if(ret)
+		{
 			wbfs_fatal("error reading disc\n");
+			return -1;
+		}
 	}
-	if (d->sector_usage_table)
+	if(d->sector_usage_table)
 	{
 		u32 blockno = offset >> 13;
 		do
@@ -66,35 +81,39 @@ static void disc_read(wiidisc_t *d, u32 offset, u8 *data, u32 len)
 			if (len > 0x8000) len -= 0x8000;
 		} while(len > 0x8000);
 	}
+	return 0;
 }
 
-static void partition_raw_read(wiidisc_t *d, u32 offset, u8 *data, u32 len)
+static int partition_raw_read(wiidisc_t *d, u32 offset, u8 *data, u32 len)
 {
-	disc_read(d, d->partition_raw_offset + offset, data, len);
+	return disc_read(d, d->partition_raw_offset + offset, data, len);
 }
 
-static void partition_read_block(wiidisc_t *d, u32 blockno, u8 *block)
+static int partition_read_block(wiidisc_t *d, u32 blockno, u8 *block)
 {
 	u8*raw = d->tmp_buffer;
 	u8 iv[16];
 	u32 offset;
 	if (d->sector_usage_table) d->sector_usage_table[d->partition_block+blockno] = 1;
 	offset = d->partition_data_offset + ((0x8000 >> 2) * blockno);
-	partition_raw_read(d,offset, raw, 0x8000);
+	if(partition_raw_read(d,offset, raw, 0x8000) < 0)
+		return -1;
 
 	memcpy(iv, raw + 0x3d0, 16);
-	
+
 	// decrypt data
-	if(neek2o())
+	if(Sys_HW_Access())
 	{
-		NKAESDecryptBlock(raw, block);
-	}
-	else
-	{
-		AES_ResetEngine();	
+		AES_ResetEngine();
 		AES_EnableDecrypt(d->disc_key, iv);
 		AES_Decrypt(raw + 0x400, block, 0x7c0);
 	}
+	else
+	{
+		aes_set_key(d->disc_key);
+		aes_decrypt(iv, raw + 0x400, block, 0x7c00);
+	}
+	return 0;
 }
 
 static void partition_read(wiidisc_t *d, u32 offset, u8 *data, u32 len, int fake)
@@ -111,7 +130,8 @@ static void partition_read(wiidisc_t *d, u32 offset, u8 *data, u32 len, int fake
 		if (len_in_block > len) len_in_block = len;
 		if (!fake)
 		{
-			partition_read_block(d,offset / (0x7c00 >> 2), block);
+			if(partition_read_block(d,offset / (0x7c00 >> 2), block) < 0)
+				break;
 			wbfs_memcpy(data, block + (offset_in_block << 2), len_in_block);
 		}
 		else d->sector_usage_table[d->partition_block + (offset / (0x7c00 >> 2))] = 1;
@@ -151,7 +171,7 @@ static u32 do_fst(wiidisc_t *d, u8 *fst, const char *names, u32 i)
 		offset = _be32(fst + 12 * i + 4);
 		if (d->extract_pathname && strcasecmp(name, d->extract_pathname) == 0)
 		{
-			d->extracted_buffer = wbfs_ioalloc(size);
+			d->extracted_buffer = wbfs_malloc(size);
 			if (d->extracted_buffer != 0)
 			{
 				d->extracted_size = size;
@@ -165,13 +185,13 @@ static u32 do_fst(wiidisc_t *d, u8 *fst, const char *names, u32 i)
 
 static void do_files(wiidisc_t*d)
 {
-	u8 *b = wbfs_ioalloc(0x480); // XXX: determine actual header size
+	u8 *b = wbfs_malloc(0x480); // XXX: determine actual header size
 	u32 dol_offset;
 	u32 fst_offset;
 	u32 fst_size;
 	u32 apl_offset;
 	u32 apl_size;
-	u8 *apl_header = wbfs_ioalloc(0x20);
+	u8 *apl_header = wbfs_malloc(0x20);
 	u8 *fst;
 	u32 n_files;
 	partition_read(d, 0, b, 0x480, 0);
@@ -189,7 +209,7 @@ static void do_files(wiidisc_t*d)
 
 	if (fst_size)
 	{
-		fst = wbfs_ioalloc(fst_size);
+		fst = wbfs_malloc(fst_size);
 		if (fst == 0)
 			wbfs_fatal("malloc fst\n");
 		partition_read(d, fst_offset, fst, fst_size,0);
@@ -209,16 +229,16 @@ static void do_files(wiidisc_t*d)
 		if (12 * n_files <= fst_size)
 			if (n_files > 1) do_fst(d, fst, (char *)fst + 12 * n_files, 0);
 		
-		if (fst != d->extracted_buffer) wbfs_iofree( fst );
+		if (fst != d->extracted_buffer) wbfs_free( fst );
 	}
-	wbfs_iofree(b);
-	wbfs_iofree(apl_header);
+	wbfs_free(b);
+	wbfs_free(apl_header);
 }
 
 static void do_partition(wiidisc_t*d)
 {
-	u8 *tik = wbfs_ioalloc(0x2a4);
-	u8 *b = wbfs_ioalloc(0x1c);
+	u8 *tik = wbfs_malloc(0x2a4);
+	u8 *b = wbfs_malloc(0x1c);
 	u64 tmd_offset;
 	u32 tmd_size;
 	u8 *tmd;
@@ -238,7 +258,7 @@ static void do_partition(wiidisc_t*d)
 	h3_offset = _be32(b + 0x10);
 	d->partition_data_offset = _be32(b + 0x14);
 	d->partition_block = (d->partition_raw_offset + d->partition_data_offset) >> 13;
-	tmd = wbfs_ioalloc(tmd_size);
+	tmd = wbfs_malloc(tmd_size);
 	if (tmd == 0)
 		wbfs_fatal("malloc tmd\n");
 	partition_raw_read(d, tmd_offset, tmd, tmd_size);
@@ -249,24 +269,20 @@ static void do_partition(wiidisc_t*d)
 		d->extracted_size = tmd_size;
 	}
 
-	cert = wbfs_ioalloc(cert_size);
+	cert = wbfs_malloc(cert_size);
 	if (cert == 0)
 		wbfs_fatal("malloc cert\n");
 	partition_raw_read(d, cert_offset, cert, cert_size);
 
-	
-	if(neek2o())
-		NKKeyCreate(tik);
-	else
-		decrypt_title_key(tik, d->disc_key);
+	decrypt_title_key(tik, d->disc_key);
 
 	partition_raw_read(d, h3_offset, 0, 0x18000);
 
-	wbfs_iofree(b);
-	wbfs_iofree(tik);
-	wbfs_iofree(cert);
+	wbfs_free(b);
+	wbfs_free(tik);
+	wbfs_free(cert);
 	if(tmd != d->extracted_buffer)
-		wbfs_iofree( tmd );
+		wbfs_free( tmd );
 
 	do_files(d);
 
@@ -284,22 +300,25 @@ static int test_parition_skip(u32 partition_type, partition_selector_t part_sel)
 		default:
 			return (partition_type != part_sel);
 	}
-} 
+}
+
 static void do_disc(wiidisc_t *d)
 {
-	u8 *b = wbfs_ioalloc(0x100);
+	u8 *b = wbfs_malloc(0x100);
+	if(b == NULL)
+		goto out;
 	u64 partition_offset[32]; // XXX: don't know the real maximum
 	u64 partition_type[32]; // XXX: don't know the real maximum
 	u32 n_partitions;
 	u32 magic;
 	u32 i;
-	disc_read(d, 0, b, 0x100);
+	if(disc_read(d, 0, b, 0x100) < 0)
+		goto out;
 	magic = _be32(b + 24);
 	if (magic != WII_MAGIC)
 	{
-		wbfs_iofree(b);
 		wbfs_error("not a wii disc");
-		return ;
+		goto out;
 	}
 	disc_read(d, 0x40000 >> 2, b, 0x100);
 	n_partitions = _be32(b);
@@ -314,7 +333,8 @@ static void do_disc(wiidisc_t *d)
 		d->partition_raw_offset = partition_offset[i];
 		if (!test_parition_skip(partition_type[i], d->part_sel)) do_partition(d);
 	}
-	wbfs_iofree(b);
+out:
+	wbfs_free(b);
 }
 
 wiidisc_t *wd_open_disc(read_wiidisc_callback_t read, void *fp)
@@ -325,19 +345,19 @@ wiidisc_t *wd_open_disc(read_wiidisc_callback_t read, void *fp)
 	d->read = read;
 	d->fp = fp;
 	d->part_sel = ALL_PARTITIONS;
-	d->tmp_buffer = wbfs_ioalloc(0x8000);
-	d->tmp_buffer2 = wbfs_ioalloc(0x8000);
+	d->tmp_buffer = wbfs_malloc(0x8000);
+	d->tmp_buffer2 = wbfs_malloc(0x8000);
 	return d;
 }
 
 void wd_close_disc(wiidisc_t *d)
 {
-	wbfs_iofree(d->tmp_buffer);
+	wbfs_free(d->tmp_buffer);
 	wbfs_free(d->tmp_buffer2);
 	wbfs_free(d);
 }
 
-// returns a buffer allocated with wbfs_ioalloc() or NULL if not found of alloc error
+// returns a buffer allocated with wbfs_malloc() or NULL if not found of alloc error
 // XXX pathname not implemented. files are extracted by their name. 
 // first file found with that name is returned.
 u8 *wd_extract_file(wiidisc_t *d, u32 *size, partition_selector_t partition_type, const char *pathname)
diff --git a/source/loader/gc_disc_dump.cpp b/source/loader/gc_disc_dump.cpp
index 6085ff78..78e601f1 100644
--- a/source/loader/gc_disc_dump.cpp
+++ b/source/loader/gc_disc_dump.cpp
@@ -36,7 +36,7 @@
 #include "devicemounter/DeviceHandler.hpp"
 #include "fileOps/fileOps.h"
 #include "gecko/gecko.hpp"
-#include "gui/Gekko.h"
+#include "hw/Gekko.h"
 #include "gui/text.hpp"
 #include "memory/mem2.hpp"
 #include "menu/menu.hpp"
@@ -69,7 +69,6 @@ void GCDump::__AnalizeMultiDisc()
 s32 GCDump::__DiscReadRaw(void *outbuf, u64 offset, u32 length)
 {
 	length = ALIGN32(length);
-	wiiLightSetLevel(255);
 	wiiLightOn();
 	while(1)
 	{
@@ -179,7 +178,6 @@ bool GCDump::__WaitForDisc(u8 dsc, u32 msg)
 		mainMenu.GC_Messenger(msg, dsc+1, minfo);
 		while(1)
 		{
-			wiiLightSetLevel(255);
 			wiiLightOn();
 			usleep(1000000);
 			wiiLightOff();
@@ -191,7 +189,6 @@ bool GCDump::__WaitForDisc(u8 dsc, u32 msg)
 
 		while(1)
 		{
-			wiiLightSetLevel(255);
 			wiiLightOn();
 			usleep(1000000);
 			wiiLightOff();
diff --git a/source/loader/nk.c b/source/loader/nk.c
index ec4bddb4..b8055bc8 100644
--- a/source/loader/nk.c
+++ b/source/loader/nk.c
@@ -29,12 +29,13 @@
 #include <malloc.h>
 
 #include "nk.h"
+#include "sys.h"
 #include "armboot.h"
 #include "fileOps/fileOps.h"
 #include "memory/mem2.hpp"
 #include "gecko/gecko.hpp"
 
-static u32 KeyID;
+//static u32 KeyID;
 bool checked = false;
 bool neek = false;
 u32 kernelSize = 0;
@@ -45,6 +46,8 @@ void check_neek2o(void)
 	if(checked == true)
 		return;
 	checked = true;
+	if(Sys_DolphinMode() == true)
+		return;
 
 	s32 ESHandle = IOS_Open("/dev/es", 0);
 	neek = (IOS_Ioctlv(ESHandle, 0xA2, 0, 0, NULL) == 0x666c6f77);
@@ -133,7 +136,7 @@ s32 Launch_nk(u64 TitleID, const char *nandpath, u64 ReturnTo)
 	MEM1_free(mini);
 	return 1;
 }
-
+/*
 void NKKeyCreate(u8 *tik)
 {	
 	u32 *TKeyID = (u32*)MEM1_memalign(32, sizeof(u32));	
@@ -185,3 +188,4 @@ void NKAESDecryptBlock(u8 *in, u8 *out)
 	IOS_Ioctlv(ESHandle, 0x2D, 3, 2, (ioctlv *)v);	
 	IOS_Close(ESHandle);
 }
+*/
\ No newline at end of file
diff --git a/source/loader/nk.h b/source/loader/nk.h
index a02677ef..4073a0e8 100644
--- a/source/loader/nk.h
+++ b/source/loader/nk.h
@@ -29,8 +29,8 @@ s32 Launch_nk(u64 TitleID, const char *nandpath, u64 ReturnTo);
 bool Load_Neek2o_Kernel();
 void check_neek2o(void);
 bool neek2o(void);
-void NKKeyCreate(u8 *TIK);
-void NKAESDecryptBlock(u8 *in, u8 *out);
+/*void NKKeyCreate(u8 *TIK);
+void NKAESDecryptBlock(u8 *in, u8 *out);*/
 
 #ifdef __cplusplus
 }
diff --git a/source/loader/splits.c b/source/loader/splits.c
index 910e4176..12d3e158 100644
--- a/source/loader/splits.c
+++ b/source/loader/splits.c
@@ -18,7 +18,10 @@
 #define FMT_llu "%llu"
 #define FMT_lld "%lld"
 
-#define split_error(x)		do { gprintf("\nsplit error: %s\n\n",x); } while(0)
+static inline void split_error(const char *x)
+{
+	gprintf("\nsplit error: %s\n\n",x);
+}
 
 // 1 cluster less than 4gb
 u64 OPT_split_size = (u64)4LL * 1024 * 1024 * 1024 - 32 * 1024;
diff --git a/source/loader/sys.c b/source/loader/sys.c
index 0d27eefa..99dcd026 100644
--- a/source/loader/sys.c
+++ b/source/loader/sys.c
@@ -205,5 +205,18 @@ bool Sys_DolphinMode(void)
 	return DolphinMode;
 }
 
+bool hw_check = false;
+bool on_hw = false;
+bool Sys_HW_Access(void)
+{
+	if(hw_check == true)
+		return on_hw;
+
+	check_neek2o();
+	on_hw = AHBRPOT_Patched() && (!Sys_DolphinMode() && !neek2o());
+	hw_check = true;
+	return on_hw;
+}
+
 /* KILL IT */
 s32 __IOS_LoadStartupIOS() { return 0; }
diff --git a/source/loader/sys.h b/source/loader/sys.h
index 8ed56fd7..5551d45b 100644
--- a/source/loader/sys.h
+++ b/source/loader/sys.h
@@ -32,6 +32,7 @@ enum
 /* Prototypes */
 void Sys_Init(void);
 bool Sys_DolphinMode(void);
+bool Sys_HW_Access(void);
 bool Sys_Exiting(void);
 void Sys_Exit(void);
 void Sys_ExitTo(int);
diff --git a/source/memory/mem2.cpp b/source/memory/mem2.cpp
index 0644ffb0..7d9d11ea 100644
--- a/source/memory/mem2.cpp
+++ b/source/memory/mem2.cpp
@@ -2,25 +2,28 @@
 #include <malloc.h>
 #include <string.h>
 #include <ogc/system.h>
+#include <ogc/machine/processor.h>
 
 #include "mem2.hpp"
-#include "mem2alloc.hpp"
+#include "mem_manager.hpp"
 #include "gecko/gecko.hpp"
 #include "loader/utils.h"
 
-#define MEM2_PRIORITY_SIZE	0x1000
+static const u32 MEM2_PRIORITY_SIZE = 0x1000;
 
 // Forbid the use of MEM2 through malloc
 u32 MALLOC_MEM2 = 0;
 
-void *MEM1_lo_start = (void*)0x80004000;
-void *MEM1_lo_end = (void*)0x80620000;
+u8 *MEM1_lo_start = (u8*)0x80004000;
+u8 *MEM1_lo_end = (u8*)0x80620000;
+u8 *MEM1_lo_list = (u8*)0x93280000;
 
-void *MEM2_start = (void*)0x90200000;
-void *MEM2_end = (void*)0x93100000;
+u8 *MEM2_start = (u8*)0x90200000;
+u8 *MEM2_end = (u8*)0x93200000;
+u8 *MEM2_list = (u8*)0x93200000;
 
-static CMEM2Alloc g_mem1lo;
-static CMEM2Alloc g_mem2gp;
+MemManager g_mem1lo;
+MemManager g_mem2gp;
 
 extern "C"
 {
@@ -34,23 +37,25 @@ extern __typeof(malloc_usable_size) __real_malloc_usable_size;
 
 void MEM_init()
 {
-	g_mem1lo.init(MEM1_lo_start, MEM1_lo_end); //about 6mb
-	g_mem1lo.clear();
+	MemMutexInit();
 
-	g_mem2gp.init(MEM2_start, MEM2_end); //about 47mb
-	g_mem2gp.clear();
+	g_mem1lo.Init(MEM1_lo_start, MEM1_lo_list, (u32)(MEM1_lo_end-MEM1_lo_start)); //about 6mb
+	g_mem1lo.ClearMem();
+
+	g_mem2gp.Init(MEM2_start, MEM2_list, (u32)(MEM2_end-MEM2_start)); //about 48mb
+	g_mem2gp.ClearMem();
 }
 
 void *MEM1_lo_alloc(unsigned int s)
 {
-	return g_mem1lo.allocate(s);
+	return g_mem1lo.Alloc(s);
 }
 
 void MEM1_lo_free(void *p)
 {
 	if(!p)
 		return;
-	g_mem1lo.release(p);
+	g_mem1lo.Free(p);
 }
 
 unsigned int MEM1_lo_freesize()
@@ -91,28 +96,28 @@ void MEM2_free(void *p)
 {
 	if(!p)
 		return;
-	g_mem2gp.release(p);
+	g_mem2gp.Free(p);
 }
 
 void *MEM2_alloc(unsigned int s)
 {
-	return g_mem2gp.allocate(s);
+	return g_mem2gp.Alloc(s);
 }
 
 /* Placeholder, will be needed with new memory manager */
 void *MEM2_memalign(unsigned int /* alignment */, unsigned int s)
 {
-	return MEM2_alloc(s);
+	return g_mem2gp.Alloc(s);
 }
 
 void *MEM2_realloc(void *p, unsigned int s)
 {
-	return g_mem2gp.reallocate(p, s);
+	return g_mem2gp.ReAlloc(p, s);
 }
 
 unsigned int MEM2_usableSize(void *p)
 {
-	return g_mem2gp.usableSize(p);
+	return g_mem2gp.MemBlockSize(p);
 }
 
 unsigned int MEM2_freesize()
@@ -125,7 +130,7 @@ void *__wrap_malloc(size_t size)
 	void *p;
 	if(size >= MEM2_PRIORITY_SIZE)
 	{
-		p = g_mem2gp.allocate(size);
+		p = g_mem2gp.Alloc(size);
 		if(p != 0) 
 			return p;
 		return __real_malloc(size);
@@ -133,7 +138,7 @@ void *__wrap_malloc(size_t size)
 	p = __real_malloc(size);
 	if(p != 0)
 		return p;
-	return g_mem2gp.allocate(size);
+	return g_mem2gp.Alloc(size);
 }
 
 void *__wrap_calloc(size_t n, size_t size)
@@ -141,7 +146,7 @@ void *__wrap_calloc(size_t n, size_t size)
 	void *p;
 	if((n * size) >= MEM2_PRIORITY_SIZE)
 	{
-		p = g_mem2gp.allocate(n * size);
+		p = g_mem2gp.Alloc(n * size);
 		if (p != 0)
 		{
 			memset(p, 0, n * size);
@@ -153,7 +158,7 @@ void *__wrap_calloc(size_t n, size_t size)
 	p = __real_calloc(n, size);
 	if (p != 0) return p;
 
-	p = g_mem2gp.allocate(n * size);
+	p = g_mem2gp.Alloc(n * size);
 	if (p != 0)
 		memset(p, 0, n * size);
 	return p;
@@ -166,7 +171,7 @@ void *__wrap_memalign(size_t a, size_t size)
 	{
 		if(a <= 32 && 32 % a == 0)
 		{
-			p = g_mem2gp.allocate(size);
+			p = g_mem2gp.Alloc(size);
 			if (p != 0)
 				return p;
 		}
@@ -176,7 +181,7 @@ void *__wrap_memalign(size_t a, size_t size)
 	if(p != 0 || a > 32 || 32 % a != 0)
 		return p;
 
-	return g_mem2gp.allocate(size);
+	return g_mem2gp.Alloc(size);
 }
 
 void __wrap_free(void *p)
@@ -187,9 +192,9 @@ void __wrap_free(void *p)
 	if(((u32)p & 0x10000000) != 0)
 	{
 		//if(p > MEM2_start)
-		g_mem2gp.release(p);
+		g_mem2gp.Free(p);
 		//else
-		//g_mem2lo_gp.release(p);
+		//g_mem2lo_gp.Free(p);
 	}
 	else
 		MEM1_free(p);
@@ -201,7 +206,7 @@ void *__wrap_realloc(void *p, size_t size)
 	// ptr from mem2
 	if(((u32)p & 0x10000000) != 0 || (p == 0 && size > MEM2_PRIORITY_SIZE))
 	{
-		n = g_mem2gp.reallocate(p, size);
+		n = g_mem2gp.ReAlloc(p, size);
 		if(n != 0)
 			return n;
 		n = __real_malloc(size);
@@ -210,7 +215,7 @@ void *__wrap_realloc(void *p, size_t size)
 		if(p != 0)
 		{
 			memcpy(n, p, MEM2_usableSize(p) < size ? MEM2_usableSize(p) : size);
-			g_mem2gp.release(p);
+			g_mem2gp.Free(p);
 		}
 		return n;
 	}
@@ -218,7 +223,7 @@ void *__wrap_realloc(void *p, size_t size)
 	n = __real_realloc(p, size);
 	if(n != 0)
 		return n;
-	n = g_mem2gp.allocate(size);
+	n = g_mem2gp.Alloc(size);
 	if(n == 0)
 		return 0;
 	if(p != 0)
@@ -232,7 +237,7 @@ void *__wrap_realloc(void *p, size_t size)
 size_t __wrap_malloc_usable_size(void *p)
 {
 	if(((u32)p & 0x10000000) != 0)
-		return CMEM2Alloc::usableSize(p);
+		return g_mem2gp.MemBlockSize(p);
 	return __real_malloc_usable_size(p);
 }
 
diff --git a/source/memory/mem2.hpp b/source/memory/mem2.hpp
index 0ec7e5eb..eb61ca96 100644
--- a/source/memory/mem2.hpp
+++ b/source/memory/mem2.hpp
@@ -1,5 +1,3 @@
-// 1 MEM2 allocator, one for general purpose
-// Aligned and padded to 32 bytes, as required by many functions
 
 #ifndef __MEM2_HPP
 #define __MEM2_HPP
diff --git a/source/memory/mem2alloc.cpp b/source/memory/mem2alloc.cpp
deleted file mode 100644
index d9c82003..00000000
--- a/source/memory/mem2alloc.cpp
+++ /dev/null
@@ -1,230 +0,0 @@
-#include "mem2alloc.hpp"
-
-#include <ogc/system.h>
-#include <ogc/cache.h>
-#include <algorithm>
-#include <string.h>
-#include "lockMutex.hpp"
-#include "gecko/gecko.hpp"
-void CMEM2Alloc::init(void *addr, void *end)
-{
-	m_baseAddress = (SBlock *)(((u32)addr + 31) & ~31);
-	m_endAddress = (SBlock *)((u32)end & ~31);
-	LWP_MutexInit(&m_mutex, 0);
-}
-
-void CMEM2Alloc::cleanup(void)
-{
-	LWP_MutexDestroy(m_mutex);
-	m_mutex = 0;
-	m_first = 0;
-	// Try to release the range we took through SYS functions
-	/*if (SYS_GetArena2Lo() == m_endAddress)
-		SYS_SetArena2Lo(m_baseAddress);*/
-	m_baseAddress = 0;
-	m_endAddress = 0;
-}
-
-void CMEM2Alloc::clear(void)
-{
-	m_first = 0;
-	u32 Size = (u32)m_endAddress - (u32)m_baseAddress;
-	memset(m_baseAddress, 0, Size);
-	DCFlushRange(m_baseAddress, Size);
-}
-
-unsigned int CMEM2Alloc::usableSize(void *p)
-{
-	return p == 0 ? 0 : ((SBlock *)p - 1)->s * sizeof (SBlock);
-}
-
-bool CMEM2Alloc::is_valid(const SBlock *block)
-{
-	if((u32)block >= (u32)m_baseAddress && (u32)block < (u32)m_endAddress)
-		return true;
-	gprintf("WARNING: Found invalid memory location!\n");
-	return false;
-}
-
-void *CMEM2Alloc::allocate(unsigned int s)
-{
-	if (s == 0)
-		s = 1;
-	//
-	LockMutex lock(m_mutex);
-	//
-	s = (s - 1) / sizeof (SBlock) + 1;
-	// First block
-	if (m_first == 0)
-	{
-		if (m_baseAddress + s + 1 >= m_endAddress)
-			return 0;
-		m_first = m_baseAddress;
-		m_first->next = 0;
-		m_first->prev = 0;
-		m_first->s = s;
-		m_first->f = false;
-		return (void *)(m_first + 1);
-	}
-	// Search for a free block
-	SBlock *i;
-	SBlock *j = m_first;
-	for (i = m_first; i != 0; i = i->next)
-	{
-		if (i->f && i->s >= s)
-			break;
-		j = i;
-	}
-	// Create a new block
-	if (i == 0)
-	{
-		i = j + j->s + 1;
-		if (i + s + 1 >= m_endAddress)
-			return 0;
-		j->next = i;
-		i->prev = j;
-		i->next = 0;
-		i->s = s;
-		i->f = false;
-		return (void *)(i + 1);
-	}
-	// Reuse a free block
-	i->f = false;
-	// Split it
-	if (i->s > s + 1)
-	{
-		j = i + s + 1;
-		j->f = true;
-		j->s = i->s - s - 1;
-		i->s = s;
-		j->next = i->next;
-		j->prev = i;
-		i->next = j;
-		if(j->next != 0 && is_valid(j->next))
-			j->next->prev = j;
-	}
-	return (void *)(i + 1);
-}
-
-void CMEM2Alloc::release(void *p)
-{
-	if (p == 0)
-		return;
-
-	LockMutex lock(m_mutex);
-	SBlock *i = (SBlock *)p - 1;
-	i->f = true;
-
-	// If there are no other blocks following yet,
-	// set the remaining size to free size. - Dimok
-	if(i->next == 0)
-		i->s = m_endAddress - i - 1;
-
-	// Merge with previous block
-	if (i->prev != 0 && is_valid(i->prev) && is_valid(i->prev->next) && i->prev->f)
-	{
-		i = i->prev;
-		i->s += i->next->s + 1;
-		if(i->next->next != 0 && is_valid(i->next->next))
-		{
-			i->next = i->next->next;
-			i->next->prev = i;
-		}
-	}
-	// Merge with next block
-	if (i->next != 0 && is_valid(i->next) && i->next->f)
-	{
-		i->s += i->next->s + 1;
-		i->next = i->next->next;
-		if(i->next != 0 && is_valid(i->next))
-			i->next->prev = i;
-	}
-}
-
-void *CMEM2Alloc::reallocate(void *p, unsigned int s)
-{
-	SBlock *i;
-	SBlock *j;
-	void *n;
-
-	if (s == 0)
-		s = 1;
-	if (p == 0)
-		return allocate(s);
-
-	i = (SBlock *)p - 1;
-	s = (s - 1) / sizeof (SBlock) + 1;
-
-	LockMutex lock(m_mutex);
-
-	//out of memory /* Dimok */
-	if (i + s + 1 >= m_endAddress)
-	{
-		return 0;
-	}
-
-	// Last block
-	if (i->next == 0 && i + s + 1 < m_endAddress)
-	{
-		i->s = s;
-		return p;
-	}
-	// Size <= current size + next block
-	if (i->next != 0 && is_valid(i->next) && i->s < s && i->next->f && i->s + i->next->s + 1 >= s)
-	{
-		// Merge
-		i->s += i->next->s + 1;
-		i->next = i->next->next;
-		if(i->next != 0 && is_valid(i->next))
-			i->next->prev = i;
-	}
-	// Size <= current size
-	if (i->s >= s)
-	{
-		// Split
-		if (i->s > s + 1)
-		{
-			j = i + s + 1;
-			j->f = true;
-			j->s = i->s - s - 1;
-			i->s = s;
-			j->next = i->next;
-			j->prev = i;
-			i->next = j;
-			if(j->next != 0 && is_valid(j->next))
-				j->next->prev = j;
-		}
-		return p;
-	}
-
-	// Size > current size
-	n = allocate(s * sizeof (SBlock));
-	if (n == 0)
-		return 0;
-	memcpy(n, p, i->s * sizeof (SBlock));
-	release(p);
-	return n;
-}
-
-unsigned int CMEM2Alloc::FreeSize()
-{
-	LockMutex lock(m_mutex);
-
-	if (m_first == 0)
-		return (u32)m_endAddress - (u32)m_baseAddress;
-
-	SBlock *i;
-	unsigned int size = 0;
-
-	for(i = m_first; i != 0; i = i->next)
-	{
-		if(i->f && i->next != 0 && is_valid(i->next))
-			size += i->s;
-		else if(i->f && i->next == 0)
-			size += m_endAddress - i - 1;
-		else if(!i->f && i->next == 0)
-			size += m_endAddress - i - i->s - 1;
-	}
-
-	return size*sizeof(SBlock);
-}
diff --git a/source/memory/mem2alloc.hpp b/source/memory/mem2alloc.hpp
deleted file mode 100644
index 209fd7af..00000000
--- a/source/memory/mem2alloc.hpp
+++ /dev/null
@@ -1,43 +0,0 @@
-// MEM2 allocator
-// Made as a class so i can have 2 sections, one being dedicated to the covers
-
-#ifndef __MEM2ALLOC_HPP
-#define __MEM2ALLOC_HPP
-
-#include <ogc/mutex.h>
-
-class CMEM2Alloc
-{
-public:
-	void *allocate(unsigned int s);
-	void release(void *p);
-	void *reallocate(void *p, unsigned int s);
-	void init(unsigned int size);
-	void init(void *addr, void *end);
-	void cleanup(void);
-	void clear(void);
-	static unsigned int usableSize(void *p);
-	void forceEndAddress(void *newAddr) { m_endAddress = (SBlock *)newAddr; }
-	void *getEndAddress(void) const { return m_endAddress; }
-	void info(void *&address, unsigned int &size) const { address = m_baseAddress; size = (const char *)m_endAddress - (const char *)m_baseAddress; }
-	unsigned int FreeSize();
-	// 
-	CMEM2Alloc(void) : m_baseAddress(0), m_endAddress(0), m_first(0), m_mutex(0) { }
-private:
-	struct SBlock
-	{
-		unsigned int s;
-		SBlock *next;
-		SBlock *prev;
-		bool f;
-	} __attribute__((aligned(32)));
-	bool is_valid(const SBlock *block);
-	SBlock *m_baseAddress;
-	SBlock *m_endAddress;
-	SBlock *m_first;
-	mutex_t m_mutex;
-private:
-	CMEM2Alloc(const CMEM2Alloc &);
-};
-
-#endif // !defined(__MEM2ALLOC_HPP)
diff --git a/source/memory/mem_manager.cpp b/source/memory/mem_manager.cpp
new file mode 100644
index 00000000..c143930e
--- /dev/null
+++ b/source/memory/mem_manager.cpp
@@ -0,0 +1,194 @@
+/****************************************************************************
+ * Copyright (C) 2013 FIX94
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ * 
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ ****************************************************************************/
+#include <string.h>
+#include <algorithm>
+#include "mem_manager.hpp"
+#include "gecko/gecko.hpp"
+#include "loader/utils.h"
+
+static mutex_t memMutex = 0;
+static const u32 MEM_BLOCK_SIZE = 256;
+
+void MemMutexInit()
+{
+	LWP_MutexInit(&memMutex, 0);
+}
+
+void MemMutexDestroy()
+{
+	LWP_MutexDestroy(memMutex);
+	memset(&memMutex, 0, sizeof(mutex_t));
+}
+
+MemManager::MemManager()
+{
+	startAddr = NULL;
+	memList = NULL;
+	memListEnd = NULL;
+	memSize = 0;
+}
+
+void MemManager::Init(u8 *start, u8 *list, u32 size)
+{
+	LWP_MutexLock(memMutex);
+	//gprintf("Init %x %i\n", memList, memSize);
+
+	startAddr = start;
+	memList = list;
+	memSize = ALIGN(MEM_BLOCK_SIZE, size) / MEM_BLOCK_SIZE;
+	memListEnd = list + memSize;
+	ICInvalidateRange(memList, memSize+1);
+	memset(memList, MEM_FREE , memSize);
+	memset(memListEnd, MEM_END, 1); //thats the +1
+	DCFlushRange(memList, memSize+1);
+
+	LWP_MutexUnlock(memMutex);
+}
+
+void MemManager::ClearMem()
+{
+	LWP_MutexLock(memMutex);
+	//gprintf("ClearMem %x %i\n", startAddr, memSize * MEM_BLOCK_SIZE);
+
+	u32 MemFull = memSize * MEM_BLOCK_SIZE;
+	ICInvalidateRange(startAddr, MemFull);
+	memset(startAddr, 0, MemFull);
+	DCFlushRange(startAddr, MemFull);
+
+	LWP_MutexUnlock(memMutex);
+}
+
+void *MemManager::Alloc(u32 size)
+{
+	size = ALIGN(MEM_BLOCK_SIZE, size) / MEM_BLOCK_SIZE;
+	if(size > memSize)
+		return NULL;
+
+	LWP_MutexLock(memMutex);
+
+	vu8 *tmp_block;
+	u32 blocksFree;
+	for(vu8 *block = memList; block < memListEnd; block++)
+	{
+		for( ; *block & (ALLOC_USED | ALLOC_END); block++) ;
+		blocksFree = 0;
+		for(tmp_block = block; *tmp_block == MEM_FREE; tmp_block++)
+		{
+			blocksFree++;
+			if(blocksFree == size)
+			{
+				u8 *addr = (u8*)block;
+				ICInvalidateRange(addr, blocksFree);
+				memset(addr, ALLOC_USED, blocksFree - 1); //start blocks
+				memset(addr + blocksFree - 1, ALLOC_END, 1); //end block
+				DCFlushRange(addr, blocksFree);
+				void *ptr = (void*)(startAddr + ((addr - memList)*MEM_BLOCK_SIZE));
+				//gprintf("Alloc %x mem, %i blocks\n", ptr, blocksFree);
+				LWP_MutexUnlock(memMutex);
+				return ptr;
+			}
+		}
+		block = tmp_block;
+	}
+	LWP_MutexUnlock(memMutex);
+	return NULL;
+}
+
+void MemManager::Free(void *mem)
+{
+	vu8 *blockUsed = memList+(((u8*)mem-startAddr) / MEM_BLOCK_SIZE);
+	if(blockUsed < memList || blockUsed >= memListEnd || *blockUsed == 0)
+		return;
+
+	LWP_MutexLock(memMutex);
+
+	//gprintf("Free %x mem, %x block\n", mem, blockUsed);
+
+	u32 size = 0;
+	vu8 *tmp_block = blockUsed;
+	for( ; *tmp_block == ALLOC_USED; tmp_block++)
+		size++;
+	if(*tmp_block == ALLOC_END)
+		size++;
+
+	u8 *addr = (u8*)blockUsed;
+	ICInvalidateRange(addr, size);
+	memset(addr, MEM_FREE, size);
+	DCFlushRange(addr, size);
+
+	LWP_MutexUnlock(memMutex);
+}
+
+u32 MemManager::MemBlockSize(void *mem)
+{
+	vu8 *blockUsed = memList+(((u8*)mem-startAddr) / MEM_BLOCK_SIZE);
+	if(blockUsed < memList || blockUsed >= memListEnd || *blockUsed == 0)
+		return 0;
+
+	LWP_MutexLock(memMutex);
+
+	u32 size = 0;
+	for( ; *blockUsed == ALLOC_USED; blockUsed++)
+		size++;
+	if(*blockUsed == ALLOC_END)
+		size++;
+	size *= MEM_BLOCK_SIZE;
+
+	LWP_MutexUnlock(memMutex);
+
+	return size;
+}
+
+u32 MemManager::FreeSize()
+{
+	LWP_MutexLock(memMutex);
+
+	u32 free = 0;
+	for(vu8 *block = memList; block < memListEnd; block++)
+	{
+		for( ; *block == MEM_FREE; block++)
+			free++;
+	}
+	free *= MEM_BLOCK_SIZE;
+
+	LWP_MutexUnlock(memMutex);
+
+	return free;
+}
+
+void *MemManager::ReAlloc(void *mem, u32 size)
+{
+	if(mem == NULL)
+		return Alloc(size);
+
+	//gprintf("Realloc %x, %i\n", mem, size);
+	void *new_m = Alloc(size);
+	if(new_m == NULL)
+	{
+		Free(mem);
+		return NULL;
+	}
+	const u32 copysize = std::min(MemBlockSize(mem), size);
+
+	LWP_MutexLock(memMutex);
+	memcpy(new_m, mem, copysize);
+	LWP_MutexUnlock(memMutex);
+
+	Free(mem);
+
+	return new_m;
+}
diff --git a/source/memory/mem_manager.hpp b/source/memory/mem_manager.hpp
new file mode 100644
index 00000000..96fd4de1
--- /dev/null
+++ b/source/memory/mem_manager.hpp
@@ -0,0 +1,51 @@
+/****************************************************************************
+ * Copyright (C) 2013 FIX94
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ * 
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ ****************************************************************************/
+#ifndef _MEM_MANAGER_HPP_
+#define _MEM_MANAGER_HPP_
+
+#include <ogcsys.h>
+#include <ogc/mutex.h>
+
+enum mem_states
+{
+	MEM_FREE = (1<<0),
+	ALLOC_USED = (1<<1),
+	ALLOC_END = (1<<2),
+	MEM_END = (1<<3),
+};
+
+class MemManager {
+public:
+	MemManager();
+	void Init(u8 *start, u8 *list, u32 size);
+	void ClearMem();
+	void *Alloc(u32 size);
+	void Free(void *mem);
+	u32 MemBlockSize(void *mem);
+	u32 FreeSize();
+	void *ReAlloc(void *mem, u32 size);
+private:
+	u8 *startAddr;
+	u8 *memList;
+	u8 *memListEnd;
+	u32 memSize;
+};
+
+void MemMutexInit();
+void MemMutexDestroy();
+
+#endif
diff --git a/source/menu/menu.cpp b/source/menu/menu.cpp
index 489aaeef..25fbf450 100644
--- a/source/menu/menu.cpp
+++ b/source/menu/menu.cpp
@@ -14,7 +14,7 @@
 #include "channel/nand.hpp"
 #include "channel/nand_save.hpp"
 #include "gc/gc.hpp"
-#include "gui/Gekko.h"
+#include "hw/Gekko.h"
 #include "gui/GameTDB.hpp"
 #include "loader/alt_ios.h"
 #include "loader/cios.h"
@@ -255,7 +255,7 @@ void CMenu::init()
 	m_cfg.getBool(CHANNEL_DOMAIN, "disable", true);
 	/* Load cIOS Map */
 	_installed_cios.clear();
-	if(!neek2o())
+	if(!neek2o() && !Sys_DolphinMode())
 		_load_installed_cioses();
 	else
 		_installed_cios[CurrentIOS.Version] = CurrentIOS.Version;
diff --git a/source/menu/menu_game.cpp b/source/menu/menu_game.cpp
index fcadc98a..89214266 100644
--- a/source/menu/menu_game.cpp
+++ b/source/menu/menu_game.cpp
@@ -21,7 +21,6 @@
 #include "gc/gcdisc.hpp"
 #include "gui/WiiMovie.hpp"
 #include "gui/GameTDB.hpp"
-#include "gui/Gekko.h"
 #include "homebrew/homebrew.h"
 #include "loader/alt_ios.h"
 #include "loader/wdvd.h"