diff --git a/CMake/CheckAndAddFlag.cmake b/CMake/CheckAndAddFlag.cmake
index 226e570833..4fe7d5be18 100644
--- a/CMake/CheckAndAddFlag.cmake
+++ b/CMake/CheckAndAddFlag.cmake
@@ -21,6 +21,8 @@ function(check_and_add_flag var flag)
   set(genexp_config_test "1")
   if(ARGV2 STREQUAL "DEBUG_ONLY")
     set(genexp_config_test "$<CONFIG:Debug>")
+  elseif(ARGV2 STREQUAL "NO_DEBINFO_ONLY")
+    set(genexp_config_test "$<NOT:$<OR:$<CONFIG:Debug>,$<CONFIG:RelWithDebInfo>>>")
   elseif(ARGV2 STREQUAL "RELEASE_ONLY")
     set(genexp_config_test "$<NOT:$<CONFIG:Debug>>")
   elseif(ARGV2)
diff --git a/CMakeLists.txt b/CMakeLists.txt
index c837ddd9c9..4d72091752 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -372,7 +372,7 @@ else()
   check_and_add_flag(VISIBILITY_INLINES_HIDDEN -fvisibility-inlines-hidden)
   check_and_add_flag(VISIBILITY_HIDDEN -fvisibility=hidden)
 
-  check_and_add_flag(FOMIT_FRAME_POINTER -fomit-frame-pointer RELEASE_ONLY)
+  check_and_add_flag(FOMIT_FRAME_POINTER -fomit-frame-pointer NO_DEBINFO_ONLY)
 
   dolphin_compile_definitions(_DEBUG DEBUG_ONLY)
   check_and_add_flag(GGDB -ggdb DEBUG_ONLY)
diff --git a/Source/Core/Common/x64ABI.cpp b/Source/Core/Common/x64ABI.cpp
index 0c4f0770ca..9f45c132a6 100644
--- a/Source/Core/Common/x64ABI.cpp
+++ b/Source/Core/Common/x64ABI.cpp
@@ -43,17 +43,24 @@ void XEmitter::ABI_CalculateFrameSize(BitSet32 mask, size_t rsp_alignment, size_
 size_t XEmitter::ABI_PushRegistersAndAdjustStack(BitSet32 mask, size_t rsp_alignment,
                                                  size_t needed_frame_size)
 {
+  mask[RSP] = false;  // Stack pointer is never pushed
   size_t shadow, subtraction, xmm_offset;
   ABI_CalculateFrameSize(mask, rsp_alignment, needed_frame_size, &shadow, &subtraction,
                          &xmm_offset);
 
-  for (int r : mask& ABI_ALL_GPRS)
+  if (mask[RBP])
+  {
+    // Make a nice stack frame for any debuggers or profilers that might be looking at this
+    PUSH(RBP);
+    MOV(64, R(RBP), R(RSP));
+  }
+  for (int r : (mask & ABI_ALL_GPRS & ~BitSet32{RBP}))
     PUSH((X64Reg)r);
 
   if (subtraction)
     SUB(64, R(RSP), subtraction >= 0x80 ? Imm32((u32)subtraction) : Imm8((u8)subtraction));
 
-  for (int x : mask& ABI_ALL_FPRS)
+  for (int x : (mask & ABI_ALL_FPRS))
   {
     MOVAPD(MDisp(RSP, (int)xmm_offset), (X64Reg)(x - 16));
     xmm_offset += 16;
@@ -65,11 +72,12 @@ size_t XEmitter::ABI_PushRegistersAndAdjustStack(BitSet32 mask, size_t rsp_align
 void XEmitter::ABI_PopRegistersAndAdjustStack(BitSet32 mask, size_t rsp_alignment,
                                               size_t needed_frame_size)
 {
+  mask[RSP] = false;  // Stack pointer is never pushed
   size_t shadow, subtraction, xmm_offset;
   ABI_CalculateFrameSize(mask, rsp_alignment, needed_frame_size, &shadow, &subtraction,
                          &xmm_offset);
 
-  for (int x : mask& ABI_ALL_FPRS)
+  for (int x : (mask & ABI_ALL_FPRS))
   {
     MOVAPD((X64Reg)(x - 16), MDisp(RSP, (int)xmm_offset));
     xmm_offset += 16;
@@ -80,9 +88,12 @@ void XEmitter::ABI_PopRegistersAndAdjustStack(BitSet32 mask, size_t rsp_alignmen
 
   for (int r = 15; r >= 0; r--)
   {
-    if (mask[r])
+    if (r != RBP && mask[r])
       POP((X64Reg)r);
   }
+  // RSP is pushed first and popped last to make debuggers/profilers happy
+  if (mask[RBP])
+    POP(RBP);
 }
 
 void XEmitter::MOVTwo(int bits, Gen::X64Reg dst1, Gen::X64Reg src1, s32 offset1, Gen::X64Reg dst2,
diff --git a/Source/Core/VideoCommon/VertexLoaderX64.cpp b/Source/Core/VideoCommon/VertexLoaderX64.cpp
index 0c049d7e64..4e96dc80c6 100644
--- a/Source/Core/VideoCommon/VertexLoaderX64.cpp
+++ b/Source/Core/VideoCommon/VertexLoaderX64.cpp
@@ -402,6 +402,7 @@ void VertexLoaderX64::GenerateVertexLoader()
   BitSet32 regs = {src_reg,  dst_reg,       scratch1,    scratch2,
                    scratch3, remaining_reg, skipped_reg, base_reg};
   regs &= ABI_ALL_CALLEE_SAVED;
+  regs[RBP] = true;  // Give us a stack frame
   ABI_PushRegistersAndAdjustStack(regs, 0);
 
   // Backup count since we're going to count it down.