diff --git a/Core/HLE/sceKernel.cpp b/Core/HLE/sceKernel.cpp
index 07c0257d9e..4e0b48e738 100644
--- a/Core/HLE/sceKernel.cpp
+++ b/Core/HLE/sceKernel.cpp
@@ -627,7 +627,7 @@ const HLEFunction ThreadManForUser[] =
 	{0xBED27435,WrapI_IUUU<sceKernelAllocateVpl>,"sceKernelAllocateVpl"},
 	{0xEC0A693F,WrapI_IUUU<sceKernelAllocateVplCB>,"sceKernelAllocateVplCB"},
 	{0xAF36D708,WrapI_IUU<sceKernelTryAllocateVpl>,"sceKernelTryAllocateVpl"},
-	{0xB736E9FF,sceKernelFreeVpl,"sceKernelFreeVpl"},
+	{0xB736E9FF,WrapI_IU<sceKernelFreeVpl>,"sceKernelFreeVpl"},
 	{0x1D371B8A,sceKernelCancelVpl,"sceKernelCancelVpl"},
 	{0x39810265,sceKernelReferVplStatus,"sceKernelReferVplStatus"},
 
diff --git a/Core/HLE/sceKernelMemory.cpp b/Core/HLE/sceKernelMemory.cpp
index cd676ee4dd..7a2b5e23c6 100644
--- a/Core/HLE/sceKernelMemory.cpp
+++ b/Core/HLE/sceKernelMemory.cpp
@@ -15,6 +15,7 @@
 // Official git repository and contact information can be found at
 // https://github.com/hrydgard/ppsspp and http://www.ppsspp.org/.
 
+#include <algorithm>
 #include "HLE.h"
 #include "../System.h"
 #include "../MIPS/MIPS.h"
@@ -130,12 +131,15 @@ struct VPL : public KernelObject
 	{
 		p.Do(nv);
 		p.Do(address);
+		SceUID dv = 0;
+		p.Do(waitingThreads, dv);
 		alloc.DoState(p);
 		p.DoMarker("VPL");
 	}
 
 	SceKernelVplInfo nv;
 	u32 address;
+	std::vector<SceUID> waitingThreads;
 	BlockAllocator alloc;
 };
 
@@ -894,8 +898,14 @@ int sceKernelAllocateVpl(SceUID uid, u32 size, u32 addrPtr, u32 timeoutPtr)
 		{
 			VPL *vpl = kernelObjects.Get<VPL>(uid, error);
 			if (vpl)
+			{
 				vpl->nv.numWaitThreads++;
 
+				SceUID threadID = __KernelGetCurThread();
+				if (std::find(vpl->waitingThreads.begin(), vpl->waitingThreads.end(), threadID) == vpl->waitingThreads.end())
+					vpl->waitingThreads.push_back(threadID);
+			}
+
 			__KernelSetVplTimeout(timeoutPtr);
 			__KernelWaitCurThread(WAITTYPE_VPL, uid, size, timeoutPtr, false);
 		}
@@ -914,8 +924,14 @@ int sceKernelAllocateVplCB(SceUID uid, u32 size, u32 addrPtr, u32 timeoutPtr)
 		{
 			VPL *vpl = kernelObjects.Get<VPL>(uid, error);
 			if (vpl)
+			{
 				vpl->nv.numWaitThreads++;
 
+				SceUID threadID = __KernelGetCurThread();
+				if (std::find(vpl->waitingThreads.begin(), vpl->waitingThreads.end(), threadID) == vpl->waitingThreads.end())
+					vpl->waitingThreads.push_back(threadID);
+			}
+
 			__KernelSetVplTimeout(timeoutPtr);
 			__KernelWaitCurThread(WAITTYPE_VPL, uid, size, timeoutPtr, true);
 		}
@@ -930,26 +946,32 @@ int sceKernelTryAllocateVpl(SceUID uid, u32 size, u32 addrPtr)
 	return error;
 }
 
-void sceKernelFreeVpl()
+int sceKernelFreeVpl(SceUID uid, u32 addr)
 {
-	SceUID id = PARAM(0);
-	u32 blockPtr = PARAM(1);
-	DEBUG_LOG(HLE,"sceKernelFreeVpl(%i, %08x)", id, blockPtr);
+	if (addr && !Memory::IsValidAddress(addr))
+	{
+		WARN_LOG(HLE, "%08x=sceKernelFreeVpl(%i, %08x): Invalid address", SCE_KERNEL_ERROR_ILLEGAL_ADDR, uid, addr);
+		return SCE_KERNEL_ERROR_ILLEGAL_ADDR;
+	}
+
+	DEBUG_LOG(HLE, "sceKernelFreeVpl(%i, %08x)", uid, addr);
 	u32 error;
-	VPL *vpl = kernelObjects.Get<VPL>(id, error);
+	VPL *vpl = kernelObjects.Get<VPL>(uid, error);
 	if (vpl)
 	{
-		if (vpl->alloc.Free(blockPtr)) {
-			RETURN(0);
+		if (vpl->alloc.FreeExact(addr))
+		{
 			// Should trigger waiting threads
-		} else {
-			ERROR_LOG(HLE, "sceKernelFreeVpl: Error freeing %08x", blockPtr);
-			RETURN(-1);
+			return 0;
+		}
+		else
+		{
+			WARN_LOG(HLE, "%08x=sceKernelFreeVpl(%i, %08x): Unable to free", SCE_KERNEL_ERROR_ILLEGAL_MEMBLOCK, uid, addr);
+			return SCE_KERNEL_ERROR_ILLEGAL_MEMBLOCK;
 		}
 	}
-	else {
-		RETURN(error);
-	}
+	else
+		return error;
 }
 
 void sceKernelCancelVpl()
diff --git a/Core/HLE/sceKernelMemory.h b/Core/HLE/sceKernelMemory.h
index a175d02025..163d8ff9ef 100644
--- a/Core/HLE/sceKernelMemory.h
+++ b/Core/HLE/sceKernelMemory.h
@@ -41,7 +41,7 @@ void sceKernelDeleteVpl();
 int sceKernelAllocateVpl(SceUID uid, u32 size, u32 addrPtr, u32 timeoutPtr);
 int sceKernelAllocateVplCB(SceUID uid, u32 size, u32 addrPtr, u32 timeoutPtr);
 int sceKernelTryAllocateVpl(SceUID uid, u32 size, u32 addrPtr);
-void sceKernelFreeVpl();
+int sceKernelFreeVpl(SceUID uid, u32 addr);
 void sceKernelCancelVpl();
 void sceKernelReferVplStatus();
 
diff --git a/Core/Util/BlockAllocator.cpp b/Core/Util/BlockAllocator.cpp
index c4c426aa90..2085781e70 100644
--- a/Core/Util/BlockAllocator.cpp
+++ b/Core/Util/BlockAllocator.cpp
@@ -219,6 +219,18 @@ bool BlockAllocator::Free(u32 position)
 	}
 }
 
+bool BlockAllocator::FreeExact(u32 position)
+{
+	BlockAllocator::Block *b = GetBlockFromAddress(position);
+	if (b && b->taken && b->start == position)
+		return Free(position);
+	else
+	{
+		ERROR_LOG(HLE, "BlockAllocator : invalid free %08x", position);
+		return false;
+	}
+}
+
 void BlockAllocator::CheckBlocks()
 {
 	for (std::list<Block>::iterator iter = blocks.begin(); iter != blocks.end(); iter++)
diff --git a/Core/Util/BlockAllocator.h b/Core/Util/BlockAllocator.h
index 498468daf6..70f38e7c6a 100644
--- a/Core/Util/BlockAllocator.h
+++ b/Core/Util/BlockAllocator.h
@@ -29,6 +29,7 @@ public:
 	u32 AllocAt(u32 position, u32 size, const char *tag = 0);
 
 	bool Free(u32 position);
+	bool FreeExact(u32 position);
 	bool IsBlockFree(u32 position) {
 		Block *b = GetBlockFromAddress(position);
 		if (b)