diff --git a/examples/main.ll b/examples/main.ll
index 6897f35aa..f59eaf147 100644
--- a/examples/main.ll
+++ b/examples/main.ll
@@ -1,574 +1,83 @@
 %.string = type {i8*, i64} ; Basic_string
 %.rawptr = type i8* ; Basic_rawptr
 
-%Window = type {i64, i64, %WNDCLASSEXA, %.rawptr, %.rawptr, %.rawptr, %.rawptr, i8*}
-%HANDLE = type %.rawptr
-%HWND = type %.rawptr
-%HDC = type %.rawptr
-%HINSTANCE = type %.rawptr
-%HICON = type %.rawptr
-%HCURSOR = type %.rawptr
-%HMENU = type %.rawptr
-%HBRUSH = type %.rawptr
-%WPARAM = type i64
-%LPARAM = type i64
-%LRESULT = type i64
-%ATOM = type i16
-%POINT = type {i32, i32}
-%BOOL = type i32
-%WNDPROC = type %LRESULT (%HWND, i32, %WPARAM, %LPARAM)*
-%WNDCLASSEXA = type {i32, i32, %LRESULT (%HWND, i32, %WPARAM, %LPARAM)*, i32, i32, %.rawptr, %.rawptr, %.rawptr, %.rawptr, i8*, i8*, %.rawptr}
-%MSG = type {%.rawptr, i32, i64, i64, i32, %POINT}
-%HGLRC = type %.rawptr
-%PROC = type void ()*
-%wglCreateContextAttribsARBType = type %HGLRC (%HDC, %.rawptr, i32*)*
-%PIXELFORMATDESCRIPTOR = type {i32, i32, i32, i8, i8, i8, i8, i8, i8, i8, i8, i8, i8, i8, i8, i8, i8, i8, i8, i8, i8, i8, i8, i32, i32, i32}
-%Bitmap = type {i32, i32, i32, {i8*, i64, i64}}
 declare void @llvm.memmove.p0i8.p0i8.i64(i8*, i8*, i64, i32, i1) argmemonly nounwind 
 
-@win32_perf_count_freq = global i64 zeroinitializer
-
-define double @time_now() {
-entry.-.0:
-	%0 = load i64, i64* @win32_perf_count_freq, align 8
-	%1 = icmp eq i64 %0, 0
-	br i1 %1, label %if.then.-.1, label %if.done.-.2
-
-if.then.-.1:
-	call void @llvm.debugtrap()
-	br label %if.done.-.2
-
-if.done.-.2:
-	%2 = alloca i64, align 8 ; counter
-	store i64 zeroinitializer, i64* %2
-	%3 = getelementptr inbounds i64, i64* %2
-	%4 = call i32 @QueryPerformanceCounter(i64* %3)
-	%5 = alloca double, align 8 ; result
-	store double zeroinitializer, double* %5
-	%6 = load i64, i64* @win32_perf_count_freq, align 8
-	%7 = sitofp i64 %6 to double
-	%8 = load i64, i64* %2, align 8
-	%9 = sitofp i64 %8 to double
-	%10 = fdiv double %9, %7
-	store double %10, double* %5
-	%11 = load double, double* %5, align 8
-	ret double %11
-}
-
-define void @win32_print_last_error() {
-entry.-.0:
-	%0 = alloca i64, align 8 ; err_code
-	store i64 zeroinitializer, i64* %0
-	%1 = call i32 @GetLastError()
-	%2 = zext i32 %1 to i64
-	store i64 %2, i64* %0
-	%3 = load i64, i64* %0, align 8
-	%4 = icmp ne i64 %3, 0
-	br i1 %4, label %if.then.-.1, label %if.done.-.2
-
-if.then.-.1:
-	%5 = getelementptr inbounds [14 x i8], [14 x i8]* @.str0, i64 0, i64 0
-	%6 = alloca %.string, align 8 
-	store %.string zeroinitializer, %.string* %6
-	%7 = getelementptr inbounds %.string, %.string* %6, i64 0, i32 0
-	%8 = getelementptr inbounds %.string, %.string* %6, i64 0, i32 1
-	store i8* %5, i8** %7
-	store i64 14, i64* %8
-	%9 = load %.string, %.string* %6, align 8
-	call void @print_string(%.string %9)
-	%10 = load i64, i64* %0, align 8
-	call void @print_int(i64 %10)
-	%11 = getelementptr inbounds [1 x i8], [1 x i8]* @.str1, i64 0, i64 0
-	%12 = alloca %.string, align 8 
-	store %.string zeroinitializer, %.string* %12
-	%13 = getelementptr inbounds %.string, %.string* %12, i64 0, i32 0
-	%14 = getelementptr inbounds %.string, %.string* %12, i64 0, i32 1
-	store i8* %11, i8** %13
-	store i64 1, i64* %14
-	%15 = load %.string, %.string* %12, align 8
-	call void @print_string(%.string %15)
-	br label %if.done.-.2
-
-if.done.-.2:
-	ret void
-}
-
-define i8* @to_c_string(%.string %s) {
-entry.-.0:
-	%0 = alloca %.string, align 8 ; s
-	store %.string zeroinitializer, %.string* %0
-	store %.string %s, %.string* %0
-	%1 = alloca i8*, align 8 ; c_str
-	store i8* zeroinitializer, i8** %1
-	%2 = getelementptr inbounds %.string, %.string* %0, i64 0, i32 1
-	%3 = load i64, i64* %2, align 8
-	%4 = add i64 %3, 1
-	%5 = call %.rawptr @malloc(i64 %4)
-	%6 = bitcast %.rawptr %5 to i8*
-	store i8* %6, i8** %1
-	%7 = load i8*, i8** %1, align 8
-	%8 = getelementptr inbounds %.string, %.string* %0, i64 0, i32 0
-	%9 = load i8*, i8** %8, align 8
-	%10 = getelementptr i8, i8* %9, i64 0
-	%11 = getelementptr inbounds i8, i8* %10
-	%12 = getelementptr inbounds %.string, %.string* %0, i64 0, i32 1
-	%13 = load i64, i64* %12, align 8
-	%14 = call i32 @memcpy(%.rawptr %7, %.rawptr %11, i64 %13)
-	%15 = load i8*, i8** %1, align 8
-	%16 = getelementptr inbounds %.string, %.string* %0, i64 0, i32 1
-	%17 = load i64, i64* %16, align 8
-	%18 = getelementptr i8, i8* %15, i64 %17
-	store i8 0, i8* %18
-	%19 = load i8*, i8** %1, align 8
-	ret i8* %19
-}
-
-define %LRESULT @win32_proc(%HWND %hwnd, i32 %msg, %WPARAM %wparam, %LPARAM %lparam) noinline {
-entry.-.0:
-	%0 = alloca %HWND, align 8 ; hwnd
-	store %HWND zeroinitializer, %HWND* %0
-	store %HWND %hwnd, %HWND* %0
-	%1 = alloca i32, align 4 ; msg
-	store i32 zeroinitializer, i32* %1
-	store i32 %msg, i32* %1
-	%2 = alloca %WPARAM, align 8 ; wparam
-	store %WPARAM zeroinitializer, %WPARAM* %2
-	store %WPARAM %wparam, %WPARAM* %2
-	%3 = alloca %LPARAM, align 8 ; lparam
-	store %LPARAM zeroinitializer, %LPARAM* %3
-	store %LPARAM %lparam, %LPARAM* %3
-	%4 = load i32, i32* %1, align 4
-	%5 = icmp eq i32 %4, 2
-	br i1 %5, label %if.then.-.1, label %cmp-or.-.3
-
-if.then.-.1:
-	call void @ExitProcess(i32 0)
-	ret %LRESULT 0
-
-cmp-or.-.2:
-	%6 = load i32, i32* %1, align 4
-	%7 = icmp eq i32 %6, 18
-	br i1 %7, label %if.then.-.1, label %if.done.-.4
-
-cmp-or.-.3:
-	%8 = load i32, i32* %1, align 4
-	%9 = icmp eq i32 %8, 16
-	br i1 %9, label %if.then.-.1, label %cmp-or.-.2
-
-if.done.-.4:
-	%10 = load %HWND, %HWND* %0, align 8
-	%11 = load i32, i32* %1, align 4
-	%12 = load %WPARAM, %WPARAM* %2, align 8
-	%13 = load %LPARAM, %LPARAM* %3, align 8
-	%14 = call %LRESULT @DefWindowProcA(%HWND %10, i32 %11, %WPARAM %12, %LPARAM %13)
-	ret i64 %14
-}
-
-define {%Window, i1} @make_window(%.string %title, i64 %msg, i64 %height) {
-entry.-.0:
-	%0 = alloca %.string, align 8 ; title
-	store %.string zeroinitializer, %.string* %0
-	store %.string %title, %.string* %0
-	%1 = alloca i64, align 8 ; msg
-	store i64 zeroinitializer, i64* %1
-	store i64 %msg, i64* %1
-	%2 = alloca i64, align 8 ; height
-	store i64 zeroinitializer, i64* %2
-	store i64 %height, i64* %2
-	%3 = alloca %Window, align 8 ; w
-	store %Window zeroinitializer, %Window* %3
-	%4 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 0
-	%5 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 1
-	%6 = load i64, i64* %1, align 8
-	%7 = load i64, i64* %2, align 8
-	store i64 %6, i64* %4
-	store i64 %7, i64* %5
-	%8 = alloca %.string, align 8 ; class_name
-	store %.string zeroinitializer, %.string* %8
-	%9 = getelementptr inbounds [18 x i8], [18 x i8]* @.str2, i64 0, i64 0
-	%10 = alloca %.string, align 8 
-	store %.string zeroinitializer, %.string* %10
-	%11 = getelementptr inbounds %.string, %.string* %10, i64 0, i32 0
-	%12 = getelementptr inbounds %.string, %.string* %10, i64 0, i32 1
-	store i8* %9, i8** %11
-	store i64 18, i64* %12
-	%13 = load %.string, %.string* %10, align 8
-	store %.string %13, %.string* %8
-	%14 = alloca i8*, align 8 ; c_class_name
-	store i8* zeroinitializer, i8** %14
-	%15 = getelementptr inbounds %.string, %.string* %8, i64 0, i32 0
-	%16 = load i8*, i8** %15, align 8
-	%17 = getelementptr i8, i8* %16, i64 0
-	%18 = getelementptr inbounds i8, i8* %17
-	store i8* %18, i8** %14
-	%19 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 7
-	%20 = load %.string, %.string* %0, align 8
-	%21 = call i8* @to_c_string(%.string %20)
-	store i8* %21, i8** %19
-	%22 = alloca %HINSTANCE, align 8 ; instance
-	store %HINSTANCE zeroinitializer, %HINSTANCE* %22
-	%23 = call %HINSTANCE @GetModuleHandleA(i8* null)
-	store %HINSTANCE %23, %HINSTANCE* %22
-	%24 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 2
-	%25 = alloca %WNDCLASSEXA, align 8 
-	store %WNDCLASSEXA zeroinitializer, %WNDCLASSEXA* %25
-	%26 = getelementptr inbounds %WNDCLASSEXA, %WNDCLASSEXA* %25, i64 0, i32 0
-	store i32 80, i32* %26
-	%27 = getelementptr inbounds %WNDCLASSEXA, %WNDCLASSEXA* %25, i64 0, i32 1
-	store i32 3, i32* %27
-	%28 = load %HINSTANCE, %HINSTANCE* %22, align 8
-	%29 = getelementptr inbounds %WNDCLASSEXA, %WNDCLASSEXA* %25, i64 0, i32 5
-	store %HINSTANCE %28, %HINSTANCE* %29
-	%30 = load i8*, i8** %14, align 8
-	%31 = getelementptr inbounds %WNDCLASSEXA, %WNDCLASSEXA* %25, i64 0, i32 10
-	store i8* %30, i8** %31
-	%32 = bitcast %LRESULT (%HWND, i32, %WPARAM, %LPARAM)* @win32_proc to %LRESULT (%HWND, i32, %WPARAM, %LPARAM)*
-	%33 = getelementptr inbounds %WNDCLASSEXA, %WNDCLASSEXA* %25, i64 0, i32 2
-	store %WNDPROC %32, %WNDPROC* %33
-	%34 = load %WNDCLASSEXA, %WNDCLASSEXA* %25, align 8
-	store %WNDCLASSEXA %34, %WNDCLASSEXA* %24
-	%35 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 2
-	%36 = getelementptr inbounds %WNDCLASSEXA, %WNDCLASSEXA* %35
-	%37 = call %ATOM @RegisterClassExA(%WNDCLASSEXA* %36)
-	%38 = icmp eq i16 %37, 0
-	br i1 %38, label %if.then.-.1, label %if.done.-.2
-
-if.then.-.1:
-	%39 = alloca {%Window, i1}, align 8 
-	store {%Window, i1} zeroinitializer, {%Window, i1}* %39
-	%40 = load %Window, %Window* %3, align 8
-	%41 = getelementptr inbounds {%Window, i1}, {%Window, i1}* %39, i64 0, i32 0
-	store %Window %40, %Window* %41
-	%42 = getelementptr inbounds {%Window, i1}, {%Window, i1}* %39, i64 0, i32 1
-	store i1 false, i1* %42
-	%43 = load {%Window, i1}, {%Window, i1}* %39, align 8
-	ret {%Window, i1} %43
-
-if.done.-.2:
-	%44 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 4
-	%45 = load i8*, i8** %14, align 8
-	%46 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 7
-	%47 = load i8*, i8** %46, align 8
-	%48 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 0
-	%49 = load i64, i64* %48, align 8
-	%50 = trunc i64 %49 to i32
-	%51 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 1
-	%52 = load i64, i64* %51, align 8
-	%53 = trunc i64 %52 to i32
-	%54 = load %HINSTANCE, %HINSTANCE* %22, align 8
-	%55 = call %HWND @CreateWindowExA(i32 0, i8* %45, i8* %47, i32 281673728, i32 2147483648, i32 2147483648, i32 %50, i32 %53, %HWND null, %HMENU null, %HINSTANCE %54, %.rawptr null)
-	store %HWND %55, %HWND* %44
-	%56 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 4
-	%57 = load %HWND, %HWND* %56, align 8
-	%58 = icmp eq %.rawptr %57, null
-	br i1 %58, label %if.then.-.3, label %if.done.-.4
-
-if.then.-.3:
-	call void @win32_print_last_error()
-	%59 = alloca {%Window, i1}, align 8 
-	store {%Window, i1} zeroinitializer, {%Window, i1}* %59
-	%60 = load %Window, %Window* %3, align 8
-	%61 = getelementptr inbounds {%Window, i1}, {%Window, i1}* %59, i64 0, i32 0
-	store %Window %60, %Window* %61
-	%62 = getelementptr inbounds {%Window, i1}, {%Window, i1}* %59, i64 0, i32 1
-	store i1 false, i1* %62
-	%63 = load {%Window, i1}, {%Window, i1}* %59, align 8
-	ret {%Window, i1} %63
-
-if.done.-.4:
-	%64 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 3
-	%65 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 4
-	%66 = load %HWND, %HWND* %65, align 8
-	%67 = call %HDC @GetDC(%HANDLE %66)
-	store %HDC %67, %HDC* %64
-	%68 = alloca %PIXELFORMATDESCRIPTOR, align 4 ; pfd
-	store %PIXELFORMATDESCRIPTOR zeroinitializer, %PIXELFORMATDESCRIPTOR* %68
-	%69 = alloca %PIXELFORMATDESCRIPTOR, align 4 
-	store %PIXELFORMATDESCRIPTOR zeroinitializer, %PIXELFORMATDESCRIPTOR* %69
-	%70 = getelementptr inbounds %PIXELFORMATDESCRIPTOR, %PIXELFORMATDESCRIPTOR* %69, i64 0, i32 0
-	store i32 44, i32* %70
-	%71 = getelementptr inbounds %PIXELFORMATDESCRIPTOR, %PIXELFORMATDESCRIPTOR* %69, i64 0, i32 1
-	store i32 1, i32* %71
-	%72 = getelementptr inbounds %PIXELFORMATDESCRIPTOR, %PIXELFORMATDESCRIPTOR* %69, i64 0, i32 2
-	store i32 37, i32* %72
-	%73 = getelementptr inbounds %PIXELFORMATDESCRIPTOR, %PIXELFORMATDESCRIPTOR* %69, i64 0, i32 3
-	store i8 0, i8* %73
-	%74 = getelementptr inbounds %PIXELFORMATDESCRIPTOR, %PIXELFORMATDESCRIPTOR* %69, i64 0, i32 4
-	store i8 32, i8* %74
-	%75 = getelementptr inbounds %PIXELFORMATDESCRIPTOR, %PIXELFORMATDESCRIPTOR* %69, i64 0, i32 11
-	store i8 8, i8* %75
-	%76 = getelementptr inbounds %PIXELFORMATDESCRIPTOR, %PIXELFORMATDESCRIPTOR* %69, i64 0, i32 18
-	store i8 24, i8* %76
-	%77 = getelementptr inbounds %PIXELFORMATDESCRIPTOR, %PIXELFORMATDESCRIPTOR* %69, i64 0, i32 19
-	store i8 8, i8* %77
-	%78 = getelementptr inbounds %PIXELFORMATDESCRIPTOR, %PIXELFORMATDESCRIPTOR* %69, i64 0, i32 21
-	store i8 0, i8* %78
-	%79 = load %PIXELFORMATDESCRIPTOR, %PIXELFORMATDESCRIPTOR* %69, align 4
-	store %PIXELFORMATDESCRIPTOR %79, %PIXELFORMATDESCRIPTOR* %68
-	%80 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 3
-	%81 = load %HDC, %HDC* %80, align 8
-	%82 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 3
-	%83 = load %HDC, %HDC* %82, align 8
-	%84 = getelementptr inbounds %PIXELFORMATDESCRIPTOR, %PIXELFORMATDESCRIPTOR* %68
-	%85 = call i32 @ChoosePixelFormat(%HDC %83, %PIXELFORMATDESCRIPTOR* %84)
-	%86 = call %BOOL @SetPixelFormat(%HDC %81, i32 %85, %PIXELFORMATDESCRIPTOR* null)
-	%87 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 5
-	%88 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 3
-	%89 = load %HDC, %HDC* %88, align 8
-	%90 = call %HGLRC @wglCreateContext(%HDC %89)
-	store %HGLRC %90, %HGLRC* %87
-	%91 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 3
-	%92 = load %HDC, %HDC* %91, align 8
-	%93 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 5
-	%94 = load %HGLRC, %HGLRC* %93, align 8
-	%95 = call %BOOL @wglMakeCurrent(%HDC %92, %HGLRC %94)
-	%96 = alloca [8 x i32], align 4 ; attribs
-	store [8 x i32] zeroinitializer, [8 x i32]* %96
-	%97 = alloca [8 x i32], align 4 
-	store [8 x i32] zeroinitializer, [8 x i32]* %97
-	%98 = getelementptr inbounds [8 x i32], [8 x i32]* %97, i64 0, i32 0
-	store i32 8337, i32* %98
-	%99 = getelementptr inbounds [8 x i32], [8 x i32]* %97, i64 0, i32 1
-	store i32 2, i32* %99
-	%100 = getelementptr inbounds [8 x i32], [8 x i32]* %97, i64 0, i32 2
-	store i32 8338, i32* %100
-	%101 = getelementptr inbounds [8 x i32], [8 x i32]* %97, i64 0, i32 3
-	store i32 1, i32* %101
-	%102 = getelementptr inbounds [8 x i32], [8 x i32]* %97, i64 0, i32 4
-	store i32 37158, i32* %102
-	%103 = getelementptr inbounds [8 x i32], [8 x i32]* %97, i64 0, i32 5
-	store i32 2, i32* %103
-	%104 = getelementptr inbounds [8 x i32], [8 x i32]* %97, i64 0, i32 6
-	store i32 0, i32* %104
-	%105 = load [8 x i32], [8 x i32]* %97, align 4
-	store [8 x i32] %105, [8 x i32]* %96
-	%106 = alloca %.string, align 8 ; wgl_string
-	store %.string zeroinitializer, %.string* %106
-	%107 = getelementptr inbounds [27 x i8], [27 x i8]* @.str3, i64 0, i64 0
-	%108 = alloca %.string, align 8 
-	store %.string zeroinitializer, %.string* %108
-	%109 = getelementptr inbounds %.string, %.string* %108, i64 0, i32 0
-	%110 = getelementptr inbounds %.string, %.string* %108, i64 0, i32 1
-	store i8* %107, i8** %109
-	store i64 27, i64* %110
-	%111 = load %.string, %.string* %108, align 8
-	store %.string %111, %.string* %106
-	%112 = alloca i8*, align 8 ; c_wgl_string
-	store i8* zeroinitializer, i8** %112
-	%113 = getelementptr inbounds %.string, %.string* %106, i64 0, i32 0
-	%114 = load i8*, i8** %113, align 8
-	%115 = getelementptr i8, i8* %114, i64 0
-	%116 = getelementptr inbounds i8, i8* %115
-	store i8* %116, i8** %112
-	%117 = alloca %wglCreateContextAttribsARBType, align 8 ; wglCreateContextAttribsARB
-	store %wglCreateContextAttribsARBType zeroinitializer, %wglCreateContextAttribsARBType* %117
-	%118 = load i8*, i8** %112, align 8
-	%119 = call %PROC @wglGetProcAddress(i8* %118)
-	%120 = bitcast void ()* %119 to %HGLRC (%HDC, %.rawptr, i32*)*
-	%121 = bitcast %HGLRC (%HDC, %.rawptr, i32*)* %120 to %HGLRC (%HDC, %.rawptr, i32*)*
-	store %wglCreateContextAttribsARBType %121, %wglCreateContextAttribsARBType* %117
-	%122 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 6
-	%123 = load %wglCreateContextAttribsARBType, %wglCreateContextAttribsARBType* %117, align 8
-	%124 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 3
-	%125 = load %HDC, %HDC* %124, align 8
-	%126 = getelementptr inbounds [8 x i32], [8 x i32]* %96, i64 0, i64 0
-	%127 = getelementptr i32, i32* %126, i64 0
-	%128 = getelementptr inbounds i32, i32* %127
-	%129 = call %HGLRC %123(%HDC %125, %.rawptr null, i32* %128)
-	store %HGLRC %129, %HGLRC* %122
-	%130 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 3
-	%131 = load %HDC, %HDC* %130, align 8
-	%132 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 6
-	%133 = load %HGLRC, %HGLRC* %132, align 8
-	%134 = call %BOOL @wglMakeCurrent(%HDC %131, %HGLRC %133)
-	%135 = getelementptr inbounds %Window, %Window* %3, i64 0, i32 3
-	%136 = load %HDC, %HDC* %135, align 8
-	%137 = call %BOOL @SwapBuffers(%HDC %136)
-	%138 = alloca {%Window, i1}, align 8 
-	store {%Window, i1} zeroinitializer, {%Window, i1}* %138
-	%139 = load %Window, %Window* %3, align 8
-	%140 = getelementptr inbounds {%Window, i1}, {%Window, i1}* %138, i64 0, i32 0
-	store %Window %139, %Window* %140
-	%141 = getelementptr inbounds {%Window, i1}, {%Window, i1}* %138, i64 0, i32 1
-	store i1 true, i1* %141
-	%142 = load {%Window, i1}, {%Window, i1}* %138, align 8
-	ret {%Window, i1} %142
-}
-
-define void @destroy_window(%Window* %w) {
-entry.-.0:
-	%0 = alloca %Window*, align 8 ; w
-	store %Window* zeroinitializer, %Window** %0
-	store %Window* %w, %Window** %0
-	%1 = load %Window*, %Window** %0, align 8
-	%2 = getelementptr %Window, %Window* %1, i64 0
-	%3 = getelementptr inbounds %Window, %Window* %2, i64 0, i32 7
-	%4 = load i8*, i8** %3, align 8
-	call void @free(%.rawptr %4)
-	ret void
-}
-
-define i1 @update_window(%Window* %w) {
-entry.-.0:
-	%0 = alloca %Window*, align 8 ; w
-	store %Window* zeroinitializer, %Window** %0
-	store %Window* %w, %Window** %0
-	%1 = alloca %MSG, align 8 ; msg
-	store %MSG zeroinitializer, %MSG* %1
-	br label %for.body.-.1
-
-for.body.-.1:
-	%2 = alloca i1, align 1 ; ok
-	store i1 zeroinitializer, i1* %2
-	%3 = getelementptr inbounds %MSG, %MSG* %1
-	%4 = call %BOOL @PeekMessageA(%MSG* %3, %HWND null, i32 0, i32 0, i32 1)
-	%5 = icmp ne i32 %4, 0
-	store i1 %5, i1* %2
-	%6 = load i1, i1* %2, align 1
-	br i1 %6, label %if.done.-.3, label %if.then.-.2
-
-if.then.-.2:
-	br label %for.done.-.6
-
-if.done.-.3:
-	%7 = getelementptr inbounds %MSG, %MSG* %1, i64 0, i32 1
-	%8 = load i32, i32* %7, align 4
-	%9 = icmp eq i32 %8, 18
-	br i1 %9, label %if.then.-.4, label %if.done.-.5
-
-if.then.-.4:
-	ret i1 true
-
-if.done.-.5:
-	%10 = getelementptr inbounds %MSG, %MSG* %1
-	%11 = call %BOOL @TranslateMessage(%MSG* %10)
-	%12 = getelementptr inbounds %MSG, %MSG* %1
-	%13 = call %LRESULT @DispatchMessageA(%MSG* %12)
-	br label %for.body.-.1
-
-for.done.-.6:
-	ret i1 false
-}
-
-define void @display_window(%Window* %w) {
-entry.-.0:
-	%0 = alloca %Window*, align 8 ; w
-	store %Window* zeroinitializer, %Window** %0
-	store %Window* %w, %Window** %0
-	%1 = load %Window*, %Window** %0, align 8
-	%2 = getelementptr %Window, %Window* %1, i64 0
-	%3 = getelementptr inbounds %Window, %Window* %2, i64 0, i32 3
-	%4 = load %HDC, %HDC* %3, align 8
-	%5 = call %BOOL @SwapBuffers(%HDC %4)
-	ret void
-}
 
 define void @main() {
 entry.-.0:
 	call void @__$startup_runtime()
-	%0 = alloca %Window, align 8 ; window
-	store %Window zeroinitializer, %Window* %0
-	%1 = alloca i1, align 1 ; window_success
-	store i1 zeroinitializer, i1* %1
-	%2 = getelementptr inbounds [18 x i8], [18 x i8]* @.str4, i64 0, i64 0
-	%3 = alloca %.string, align 8 
-	store %.string zeroinitializer, %.string* %3
-	%4 = getelementptr inbounds %.string, %.string* %3, i64 0, i32 0
-	%5 = getelementptr inbounds %.string, %.string* %3, i64 0, i32 1
-	store i8* %2, i8** %4
-	store i64 18, i64* %5
-	%6 = load %.string, %.string* %3, align 8
-	%7 = call {%Window, i1} @make_window(%.string %6, i64 854, i64 480)
-	%8 = extractvalue {%Window, i1} %7, 0
-	%9 = extractvalue {%Window, i1} %7, 1
-	store %Window %8, %Window* %0
-	store i1 %9, i1* %1
-	%10 = load i1, i1* %1, align 1
-	br i1 %10, label %if.done.-.2, label %if.then.-.1
+	%0 = alloca <4 x float>, align 8 ; a
+	store <4 x float> zeroinitializer, <4 x float>* %0
+	%1 = alloca <4 x float>, align 8 
+	store <4 x float> zeroinitializer, <4 x float>* %1
+	%2 = load <4 x float>, <4 x float>* %1, align 8
+	%3 = insertelement <4 x float> %2, float 0x3ff0000000000000, i64 0
+	%4 = shufflevector <4 x float> %3, <4 x float> undef, <4 x i32> <i32 0, i32 0, i32 0, i32 0>
+	store <4 x float> %4, <4 x float>* %0
+	%5 = alloca <4 x float>, align 8 
+	store <4 x float> zeroinitializer, <4 x float>* %5
+	%6 = load <4 x float>, <4 x float>* %5, align 8
+	%7 = insertelement <4 x float> %6, float 0x3ff0000000000000, i64 0
+	%8 = insertelement <4 x float> %7, float 0x4000000000000000, i64 1
+	%9 = insertelement <4 x float> %8, float 0x4008000000000000, i64 2
+	%10 = insertelement <4 x float> %9, float 0x4010000000000000, i64 3
+	%11 = shufflevector <4 x float> %10, <4 x float> undef, <4 x i32> <i32 1, i32 3, i32 2, i32 0>
+	store <4 x float> %11, <4 x float>* %0
+	br label %for.init.-.1
 
-if.then.-.1:
-	ret void
+for.init.-.1:
+	%12 = alloca i64, align 8 ; i
+	store i64 zeroinitializer, i64* %12
+	store i64 0, i64* %12
+	br label %for.loop.-.3
 
-if.done.-.2:
-	%11 = alloca double, align 8 ; start_time
-	store double zeroinitializer, double* %11
-	%12 = call double @time_now()
-	store double %12, double* %11
-	%13 = alloca i1, align 1 ; running
-	store i1 zeroinitializer, i1* %13
-	store i1 true, i1* %13
-	br label %for.loop.-.4
+for.body.-.2:
+	%13 = load i64, i64* %12, align 8
+	%14 = icmp sgt i64 %13, 0
+	br i1 %14, label %if.then.-.5, label %if.done.-.6
 
-for.body.-.3:
-	%14 = alloca double, align 8 ; curr_time
-	store double zeroinitializer, double* %14
-	%15 = call double @time_now()
-	store double %15, double* %14
-	%16 = alloca double, align 8 ; dt
-	store double zeroinitializer, double* %16
-	%17 = load double, double* %11, align 8
-	%18 = load double, double* %14, align 8
-	%19 = fsub double %18, %17
-	store double %19, double* %16
-	%20 = getelementptr inbounds %Window, %Window* %0
-	%21 = call i1 @update_window(%Window* %20)
-	br i1 %21, label %if.then.-.5, label %if.done.-.6
+for.loop.-.3:
+	%15 = load i64, i64* %12, align 8
+	%16 = icmp slt i64 %15, 4
+	br i1 %16, label %for.body.-.2, label %for.done.-.7
 
-for.loop.-.4:
-	%22 = load i1, i1* %13, align 1
-	br i1 %22, label %for.body.-.3, label %for.done.-.10
+for.post.-.4:
+	%17 = load i64, i64* %12, align 8
+	%18 = add i64 %17, 1
+	store i64 %18, i64* %12
+	br label %for.loop.-.3
 
 if.then.-.5:
-	store i1 false, i1* %13
+	%19 = getelementptr inbounds [2 x i8], [2 x i8]* @.str0, i64 0, i64 0
+	%20 = alloca %.string, align 8 
+	store %.string zeroinitializer, %.string* %20
+	%21 = getelementptr inbounds %.string, %.string* %20, i64 0, i32 0
+	%22 = getelementptr inbounds %.string, %.string* %20, i64 0, i32 1
+	store i8* %19, i8** %21
+	store i64 2, i64* %22
+	%23 = load %.string, %.string* %20, align 8
+	call void @print_string(%.string %23)
 	br label %if.done.-.6
 
 if.done.-.6:
-	call void @glClearColor(float 0x3fe0000000000000, float 0x3fe6666660000000, float 0x3ff0000000000000, float 0x3ff0000000000000)
-	call void @glClear(i32 16384)
-	call void @glBegin(i32 4)
-	call void @glColor3f(float 0x3ff0000000000000, float 0x0000000000000000, float 0x0000000000000000)
-	call void @glVertex3f(float 0x3fe0000000000000, float 0xbfe0000000000000, float 0x0000000000000000)
-	call void @glColor3f(float 0x0000000000000000, float 0x3ff0000000000000, float 0x0000000000000000)
-	call void @glVertex3f(float 0x3fe0000000000000, float 0x3fe0000000000000, float 0x0000000000000000)
-	call void @glColor3f(float 0x0000000000000000, float 0x0000000000000000, float 0x3ff0000000000000)
-	call void @glVertex3f(float 0xbfe0000000000000, float 0x3fe0000000000000, float 0x0000000000000000)
-	call void @glColor3f(float 0x0000000000000000, float 0x0000000000000000, float 0x3ff0000000000000)
-	call void @glVertex3f(float 0xbfe0000000000000, float 0x3fe0000000000000, float 0x0000000000000000)
-	call void @glColor3f(float 0x3ff0000000000000, float 0x3ff0000000000000, float 0x0000000000000000)
-	call void @glVertex3f(float 0xbfe0000000000000, float 0xbfe0000000000000, float 0x0000000000000000)
-	call void @glColor3f(float 0x3ff0000000000000, float 0x0000000000000000, float 0x0000000000000000)
-	call void @glVertex3f(float 0x3fe0000000000000, float 0xbfe0000000000000, float 0x0000000000000000)
-	br label %defer.-.7
+	%24 = load i64, i64* %12, align 8
+	%25 = load <4 x float>, <4 x float>* %0, align 8
+	%26 = extractelement <4 x float> %25, i64 %24
+	%27 = fptosi float %26 to i64
+	call void @print_int(i64 %27)
+	br label %for.post.-.4
 
-defer.-.7:
-	call void @glEnd()
-	%23 = getelementptr inbounds %Window, %Window* %0
-	call void @display_window(%Window* %23)
-	%24 = alloca i32, align 4 ; ms
-	store i32 zeroinitializer, i32* %24
-	%25 = load double, double* %16, align 8
-	%26 = fmul double %25, 0x408f400000000000
-	%27 = fsub double 0x4030000000000000, %26
-	%28 = fptosi double %27 to i32
-	store i32 %28, i32* %24
-	%29 = load i32, i32* %24, align 4
-	%30 = icmp sgt i32 %29, 0
-	br i1 %30, label %if.then.-.8, label %if.done.-.9
-
-if.then.-.8:
-	%31 = load i32, i32* %24, align 4
-	call void @sleep_ms(i32 %31)
-	br label %if.done.-.9
-
-if.done.-.9:
-	br label %for.loop.-.4
-
-for.done.-.10:
-	br label %defer.-.11
-
-defer.-.11:
-	%32 = getelementptr inbounds %Window, %Window* %0
-	call void @destroy_window(%Window* %32)
+for.done.-.7:
+	%28 = getelementptr inbounds [1 x i8], [1 x i8]* @.str1, i64 0, i64 0
+	%29 = alloca %.string, align 8 
+	store %.string zeroinitializer, %.string* %29
+	%30 = getelementptr inbounds %.string, %.string* %29, i64 0, i32 0
+	%31 = getelementptr inbounds %.string, %.string* %29, i64 0, i32 1
+	store i8* %28, i8** %30
+	store i64 1, i64* %31
+	%32 = load %.string, %.string* %29, align 8
+	call void @print_string(%.string %32)
 	ret void
 }
 
@@ -946,7 +455,7 @@ for.body.-.5:
 	%16 = getelementptr inbounds [65 x i8], [65 x i8]* %2, i64 0, i64 0
 	%17 = load i64, i64* %3, align 8
 	%18 = getelementptr i8, i8* %16, i64 %17
-	%19 = getelementptr inbounds [64 x i8], [64 x i8]* @.str5, i64 0, i64 0
+	%19 = getelementptr inbounds [64 x i8], [64 x i8]* @.str2, i64 0, i64 0
 	%20 = load i64, i64* %1, align 8
 	%21 = load i64, i64* %0, align 8
 	%22 = srem i64 %21, %20
@@ -1088,7 +597,7 @@ for.body.-.5:
 	%16 = getelementptr inbounds [65 x i8], [65 x i8]* %2, i64 0, i64 0
 	%17 = load i64, i64* %3, align 8
 	%18 = getelementptr i8, i8* %16, i64 %17
-	%19 = getelementptr inbounds [64 x i8], [64 x i8]* @.str6, i64 0, i64 0
+	%19 = getelementptr inbounds [64 x i8], [64 x i8]* @.str3, i64 0, i64 0
 	%20 = load i64, i64* %1, align 8
 	%21 = load i64, i64* %0, align 8
 	%22 = urem i64 %21, %20
@@ -1180,7 +689,7 @@ entry.-.0:
 	br i1 %1, label %if.then.-.1, label %if.else.-.2
 
 if.then.-.1:
-	%2 = getelementptr inbounds [4 x i8], [4 x i8]* @.str7, i64 0, i64 0
+	%2 = getelementptr inbounds [4 x i8], [4 x i8]* @.str4, i64 0, i64 0
 	%3 = alloca %.string, align 8 
 	store %.string zeroinitializer, %.string* %3
 	%4 = getelementptr inbounds %.string, %.string* %3, i64 0, i32 0
@@ -1192,7 +701,7 @@ if.then.-.1:
 	br label %if.done.-.3
 
 if.else.-.2:
-	%7 = getelementptr inbounds [5 x i8], [5 x i8]* @.str8, i64 0, i64 0
+	%7 = getelementptr inbounds [5 x i8], [5 x i8]* @.str5, i64 0, i64 0
 	%8 = alloca %.string, align 8 
 	store %.string zeroinitializer, %.string* %8
 	%9 = getelementptr inbounds %.string, %.string* %8, i64 0, i32 0
@@ -1206,234 +715,7 @@ if.else.-.2:
 if.done.-.3:
 	ret void
 }
-declare %HANDLE @GetStdHandle(i32 %h) ; foreign
-declare i32 @CloseHandle(%HANDLE %h) ; foreign
-declare i32 @WriteFileA(%HANDLE %h, %.rawptr %buf, i32 %len, i32* %written_result, %.rawptr %overlapped) ; foreign
-declare i32 @GetLastError() ; foreign
-declare void @ExitProcess(i32 %exit_code) ; foreign
-declare %HWND @GetDesktopWindow() ; foreign
-declare i32 @GetCursorPos(%POINT* %p) ; foreign
-declare i32 @ScreenToClient(%HWND %h, %POINT* %p) ; foreign
-declare %HINSTANCE @GetModuleHandleA(i8* %module_name) ; foreign
-declare i32 @QueryPerformanceFrequency(i64* %result) ; foreign
-declare i32 @QueryPerformanceCounter(i64* %result) ; foreign
-
-define void @sleep_ms(i32 %ms) {
-entry.-.0:
-	%0 = alloca i32, align 4 ; ms
-	store i32 zeroinitializer, i32* %0
-	store i32 %ms, i32* %0
-	%1 = load i32, i32* %0, align 4
-	%2 = call i32 @Sleep(i32 %1)
-	ret void
-}
-declare i32 @Sleep(i32 %ms) declare void @OutputDebugStringA(i8* %c_str) ; foreign
-declare %ATOM @RegisterClassExA(%WNDCLASSEXA* %wc) ; foreign
-declare %HWND @CreateWindowExA(i32 %ex_style, i8* %class_name, i8* %title, i32 %style, i32 %x, i32 %y, i32 %w, i32 %h, %HWND %parent, %HMENU %menu, %HINSTANCE %instance, %.rawptr %param) ; foreign
-declare %BOOL @ShowWindow(%HWND %hwnd, i32 %cmd_show) ; foreign
-declare %BOOL @UpdateWindow(%HWND %hwnd) ; foreign
-declare %BOOL @PeekMessageA(%MSG* %msg, %HWND %hwnd, i32 %msg_filter_min, i32 %msg_filter_max, i32 %remove_msg) ; foreign
-declare %BOOL @TranslateMessage(%MSG* %msg) ; foreign
-declare %LRESULT @DispatchMessageA(%MSG* %msg) ; foreign
-declare %LRESULT @DefWindowProcA(%HWND %hwnd, i32 %msg, %WPARAM %wparam, %LPARAM %lparam) ; foreign
-
-define i64 @GetQueryPerformanceFrequency() {
-entry.-.0:
-	%0 = alloca i64, align 8 ; r
-	store i64 zeroinitializer, i64* %0
-	%1 = getelementptr inbounds i64, i64* %0
-	%2 = call i32 @QueryPerformanceFrequency(i64* %1)
-	%3 = load i64, i64* %0, align 8
-	ret i64 %3
-}
-declare %HDC @GetDC(%HANDLE %h) ; foreign
-declare %BOOL @SetPixelFormat(%HDC %hdc, i32 %pixel_format, %PIXELFORMATDESCRIPTOR* %pfd) ; foreign
-declare i32 @ChoosePixelFormat(%HDC %hdc, %PIXELFORMATDESCRIPTOR* %pfd) ; foreign
-declare %BOOL @SwapBuffers(%HDC %hdc) ; foreign
-declare %HGLRC @wglCreateContext(%HDC %hdc) ; foreign
-declare %BOOL @wglMakeCurrent(%HDC %hdc, %HGLRC %hglrc) ; foreign
-declare %PROC @wglGetProcAddress(i8* %c_str) ; foreign
-declare %BOOL @wglDeleteContext(%HGLRC %hglrc) ; foreign
-declare void @glClear(i32 %mask) ; foreign
-declare void @glClearColor(float %r, float %g, float %b, float %a) ; foreign
-declare void @glBegin(i32 %mode) ; foreign
-declare void @glEnd() ; foreign
-declare void @glColor3f(float %r, float %g, float %b) ; foreign
-declare void @glColor4f(float %r, float %g, float %b, float %a) ; foreign
-declare void @glVertex3f(float %x, float %y, float %z) ; foreign
-declare void @glTexCoord2f(float %u, float %v) ; foreign
-declare void @glLoadIdentity() ; foreign
-declare void @glOrtho(double %left, double %right, double %bottom, double %top, double %near, double %far) ; foreign
-declare void @glBlendFunc(i32 %sfactor, i32 %dfactor) ; foreign
-declare void @glEnable(i32 %cap) ; foreign
-declare void @glDisable(i32 %cap) ; foreign
-declare void @glGenTextures(i32 %count, i32* %result) ; foreign
-declare void @glTexParameteri(i32 %target, i32 %pname, i32 %param) ; foreign
-declare void @glTexParameterf(i32 %target, i32 %pname, float %param) ; foreign
-declare void @glBindTexture(i32 %target, i32 %texture) ; foreign
-declare void @glTexImage2D(i32 %target, i32 %level, i32 %internal_format, i32 %width, i32 %height, i32 %border, i32 %format, i32 %_type, %.rawptr %pixels) ; foreign
-
-define %Bitmap @make_bitmap(%.string %filename) {
-entry.-.0:
-	%0 = alloca %.string, align 8 ; filename
-	store %.string zeroinitializer, %.string* %0
-	store %.string %filename, %.string* %0
-	%1 = alloca [1024 x i8], align 1 ; c_buf
-	store [1024 x i8] zeroinitializer, [1024 x i8]* %1
-	%2 = alloca {i8*, i64, i64}, align 8 ; bytes
-	store {i8*, i64, i64} zeroinitializer, {i8*, i64, i64}* %2
-	%3 = load %.string, %.string* %0, align 8
-	%4 = alloca %.string, align 8 
-	store %.string zeroinitializer, %.string* %4
-	store %.string %3, %.string* %4
-	%5 = getelementptr inbounds %.string, %.string* %4, i64 0, i32 0
-	%6 = load i8*, i8** %5, align 8
-	%7 = alloca i8*, align 8 
-	store i8* zeroinitializer, i8** %7
-	store i8* %6, i8** %7
-	%8 = getelementptr inbounds %.string, %.string* %4, i64 0, i32 1
-	%9 = load i64, i64* %8, align 8
-	%10 = sub i64 %9, 0
-	%11 = sub i64 %9, 0
-	%12 = load i8*, i8** %7, align 8
-	%13 = getelementptr i8, i8* %12, i64 0
-	%14 = alloca {i8*, i64, i64}, align 8 
-	store {i8*, i64, i64} zeroinitializer, {i8*, i64, i64}* %14
-	%15 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %14, i64 0, i32 0
-	store i8* %13, i8** %15
-	%16 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %14, i64 0, i32 1
-	store i64 %10, i64* %16
-	%17 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %14, i64 0, i32 2
-	store i64 %11, i64* %17
-	%18 = load {i8*, i64, i64}, {i8*, i64, i64}* %14, align 8
-	store {i8*, i64, i64} %18, {i8*, i64, i64}* %2
-	%19 = alloca i64, align 8 ; str_len
-	store i64 zeroinitializer, i64* %19
-	%20 = sub i64 1024, 0
-	%21 = sub i64 1024, 0
-	%22 = getelementptr inbounds [1024 x i8], [1024 x i8]* %1, i64 0, i64 0
-	%23 = getelementptr i8, i8* %22, i64 0
-	%24 = alloca {i8*, i64, i64}, align 8 
-	store {i8*, i64, i64} zeroinitializer, {i8*, i64, i64}* %24
-	%25 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %24, i64 0, i32 0
-	store i8* %23, i8** %25
-	%26 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %24, i64 0, i32 1
-	store i64 %20, i64* %26
-	%27 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %24, i64 0, i32 2
-	store i64 %21, i64* %27
-	%28 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %24, i64 0, i32 0
-	%29 = load i8*, i8** %28, align 8
-	%30 = bitcast i8* %29 to %.rawptr
-	%31 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %2, i64 0, i32 0
-	%32 = load i8*, i8** %31, align 8
-	%33 = bitcast i8* %32 to %.rawptr
-	%34 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %24, i64 0, i32 1
-	%35 = load i64, i64* %34, align 8
-	%36 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %2, i64 0, i32 1
-	%37 = load i64, i64* %36, align 8
-	%38 = icmp slt i64 %35, %37
-	%39 = select i1 %38, i64 %35, i64 %37
-	%40 = mul i64 %39, 1
-	call void @llvm.memmove.p0i8.p0i8.i64(i8* %30, i8* %33, i64 %40, i32 1, i1 false)
-	store i64 %39, i64* %19
-	%41 = alloca %Bitmap, align 8 ; b
-	store %Bitmap zeroinitializer, %Bitmap* %41
-	%42 = alloca i8*, align 8 ; pixels
-	store i8* zeroinitializer, i8** %42
-	%43 = getelementptr inbounds [1024 x i8], [1024 x i8]* %1, i64 0, i64 0
-	%44 = getelementptr i8, i8* %43, i64 0
-	%45 = getelementptr inbounds i8, i8* %44
-	%46 = getelementptr inbounds %Bitmap, %Bitmap* %41, i64 0, i32 0
-	%47 = getelementptr inbounds i32, i32* %46
-	%48 = getelementptr inbounds %Bitmap, %Bitmap* %41, i64 0, i32 1
-	%49 = getelementptr inbounds i32, i32* %48
-	%50 = getelementptr inbounds %Bitmap, %Bitmap* %41, i64 0, i32 2
-	%51 = getelementptr inbounds i32, i32* %50
-	%52 = call i8* @stbi_load(i8* %45, i32* %47, i32* %49, i32* %51, i32 4)
-	store i8* %52, i8** %42
-	%53 = alloca i64, align 8 ; len
-	store i64 zeroinitializer, i64* %53
-	%54 = getelementptr inbounds %Bitmap, %Bitmap* %41, i64 0, i32 2
-	%55 = load i32, i32* %54, align 4
-	%56 = getelementptr inbounds %Bitmap, %Bitmap* %41, i64 0, i32 1
-	%57 = load i32, i32* %56, align 4
-	%58 = getelementptr inbounds %Bitmap, %Bitmap* %41, i64 0, i32 0
-	%59 = load i32, i32* %58, align 4
-	%60 = mul i32 %59, %57
-	%61 = mul i32 %60, %55
-	%62 = zext i32 %61 to i64
-	store i64 %62, i64* %53
-	%63 = getelementptr inbounds %Bitmap, %Bitmap* %41, i64 0, i32 3
-	%64 = load i64, i64* %53, align 8
-	%65 = sub i64 %64, 0
-	%66 = sub i64 %64, 0
-	%67 = load i8*, i8** %42, align 8
-	%68 = getelementptr i8, i8* %67, i64 0
-	%69 = alloca {i8*, i64, i64}, align 8 
-	store {i8*, i64, i64} zeroinitializer, {i8*, i64, i64}* %69
-	%70 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %69, i64 0, i32 0
-	store i8* %68, i8** %70
-	%71 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %69, i64 0, i32 1
-	store i64 %65, i64* %71
-	%72 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %69, i64 0, i32 2
-	store i64 %66, i64* %72
-	%73 = load {i8*, i64, i64}, {i8*, i64, i64}* %69, align 8
-	store {i8*, i64, i64} %73, {i8*, i64, i64}* %63
-	%74 = load %Bitmap, %Bitmap* %41, align 8
-	ret %Bitmap %74
-}
-declare i8* @stbi_load(i8* %filename, i32* %x, i32* %y, i32* %comp, i32 %req_comp) 
-define void @destroy_bitmap(%Bitmap* %b) {
-entry.-.0:
-	%0 = alloca %Bitmap*, align 8 ; b
-	store %Bitmap* zeroinitializer, %Bitmap** %0
-	store %Bitmap* %b, %Bitmap** %0
-	%1 = load %Bitmap*, %Bitmap** %0, align 8
-	%2 = getelementptr %Bitmap, %Bitmap* %1, i64 0
-	%3 = getelementptr inbounds %Bitmap, %Bitmap* %2, i64 0, i32 3
-	%4 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %3, i64 0, i32 0
-	%5 = load i8*, i8** %4, align 8
-	%6 = getelementptr i8, i8* %5, i64 0
-	%7 = getelementptr inbounds i8, i8* %6
-	call void @stbi_image_free(%.rawptr %7)
-	%8 = load %Bitmap*, %Bitmap** %0, align 8
-	%9 = getelementptr %Bitmap, %Bitmap* %8, i64 0
-	%10 = getelementptr inbounds %Bitmap, %Bitmap* %9, i64 0, i32 3
-	%11 = load %Bitmap*, %Bitmap** %0, align 8
-	%12 = getelementptr %Bitmap, %Bitmap* %11, i64 0
-	%13 = getelementptr inbounds %Bitmap, %Bitmap* %12, i64 0, i32 3
-	%14 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %13, i64 0, i32 2
-	%15 = load i64, i64* %14, align 8
-	%16 = sub i64 0, 0
-	%17 = sub i64 %15, 0
-	%18 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %13, i64 0, i32 0
-	%19 = load i8*, i8** %18, align 8
-	%20 = getelementptr i8, i8* %19, i64 0
-	%21 = alloca {i8*, i64, i64}, align 8 
-	store {i8*, i64, i64} zeroinitializer, {i8*, i64, i64}* %21
-	%22 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %21, i64 0, i32 0
-	store i8* %20, i8** %22
-	%23 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %21, i64 0, i32 1
-	store i64 %16, i64* %23
-	%24 = getelementptr inbounds {i8*, i64, i64}, {i8*, i64, i64}* %21, i64 0, i32 2
-	store i64 %17, i64* %24
-	%25 = load {i8*, i64, i64}, {i8*, i64, i64}* %21, align 8
-	store {i8*, i64, i64} %25, {i8*, i64, i64}* %10
-	%26 = load %Bitmap*, %Bitmap** %0, align 8
-	%27 = getelementptr %Bitmap, %Bitmap* %26, i64 0
-	%28 = getelementptr inbounds %Bitmap, %Bitmap* %27, i64 0, i32 0
-	store i32 0, i32* %28
-	%29 = load %Bitmap*, %Bitmap** %0, align 8
-	%30 = getelementptr %Bitmap, %Bitmap* %29, i64 0
-	%31 = getelementptr inbounds %Bitmap, %Bitmap* %30, i64 0, i32 1
-	store i32 0, i32* %31
-	%32 = load %Bitmap*, %Bitmap** %0, align 8
-	%33 = getelementptr %Bitmap, %Bitmap* %32, i64 0
-	%34 = getelementptr inbounds %Bitmap, %Bitmap* %33, i64 0, i32 2
-	store i32 0, i32* %34
-	ret void
-}
-declare void @stbi_image_free(%.rawptr %retval_from_stbi_load) declare i32 @putchar(i32 %c) ; foreign
+declare i32 @putchar(i32 %c) ; foreign
 declare %.rawptr @malloc(i64 %sz) ; foreign
 declare void @free(%.rawptr %ptr) ; foreign
 declare i32 @memcmp(%.rawptr %dst, %.rawptr %src, i64 %len) ; foreign
@@ -1679,19 +961,14 @@ entry.-.0:
 	%5 = icmp sge i64 %4, 0
 	ret i1 %5
 }
-@.str0 = global [14 x i8] c"GetLastError\3A\20"
+@.str0 = global [2 x i8] c"\2C\20"
 @.str1 = global [1 x i8] c"\0A"
-@.str2 = global [18 x i8] c"Win32-Odin-Window\00"
-@.str3 = global [27 x i8] c"wglCreateContextAttribsARB\00"
-@.str4 = global [18 x i8] c"Odin\20Language\20Demo"
-@.str5 = global [64 x i8] c"0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz\40$"
-@.str6 = global [64 x i8] c"0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz\40$"
-@.str7 = global [4 x i8] c"true"
-@.str8 = global [5 x i8] c"false"
+@.str2 = global [64 x i8] c"0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz\40$"
+@.str3 = global [64 x i8] c"0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz\40$"
+@.str4 = global [4 x i8] c"true"
+@.str5 = global [5 x i8] c"false"
 
 define void @__$startup_runtime() noinline {
 entry.-.0:
-	%0 = call i64 @GetQueryPerformanceFrequency()
-	store i64 %0, i64* @win32_perf_count_freq
 	ret void
 }
diff --git a/examples/main.odin b/examples/main.odin
index 89bdceddd..a844fe070 100644
--- a/examples/main.odin
+++ b/examples/main.odin
@@ -1,7 +1,24 @@
 #load "basic.odin"
+main :: proc() {
+	a := {4}f32{1}; // {1, 1, 1, 1} broadcasts to all
+	a = swizzle({4}f32{1, 2, 3, 4}, 1, 3, 2, 0);
+
+	for i := 0; i < len(a); i++ {
+		if i > 0 {
+			print_string(", ");
+		}
+
+		print_int(a[i] as int);
+	}
+	print_string("\n");
+
+}
+
+/*
 #load "win32.odin"
 #load "opengl.odin"
 #load "stb_image.odin"
+#load "math.odin"
 
 win32_perf_count_freq := GetQueryPerformanceFrequency();
 time_now :: proc() -> f64 {
@@ -148,48 +165,54 @@ display_window :: proc(w: ^Window) {
 
 
 
-main :: proc() {
-	WINDOW_WIDTH  :: 854;
-	WINDOW_HEIGHT :: 480;
 
-	window, window_success := make_window("Odin Language Demo", WINDOW_WIDTH, WINDOW_HEIGHT);
+main :: proc() {
+	window, window_success := make_window("Odin Language Demo", 854, 480);
 	if !window_success {
 		return;
 	}
 	defer destroy_window(^window);
 
-	start_time := time_now();
+	{
+		v := Vec2{1, 2};
+		c := v * 2;
+	}
+
+
+	prev_time := time_now();
 	running := true;
 	for running {
 		curr_time := time_now();
-		dt := curr_time - start_time;
+		dt := (curr_time - prev_time) as f32;
+		prev_time = curr_time;
 
 		if update_window(^window) {
 			running = false;
 		}
 
-
 		glClearColor(0.5, 0.7, 1.0, 1.0);
 		glClear(GL_COLOR_BUFFER_BIT);
-		// glOrtho(0, WINDOW_WIDTH, 0, WINDOW_HEIGHT, -1, +1);
 
-		{
+		glLoadIdentity();
+		glOrtho(0, window.width as f64,
+		        0, window.height as f64, 0, 1);
+		draw_rect :: proc(x, y, w, h: f32) {
 			glBegin(GL_TRIANGLES);
-			defer glEnd();
+			glColor3f(1, 0, 0); glVertex3f(x,   y,   0);
+			glColor3f(0, 1, 0); glVertex3f(x+w, y,   0);
+			glColor3f(0, 0, 1); glVertex3f(x+w, y+h, 0);
 
-			glColor3f(1, 0, 0); glVertex3f(+0.5, -0.5, 0);
-			glColor3f(0, 1, 0); glVertex3f(+0.5, +0.5, 0);
-			glColor3f(0, 0, 1); glVertex3f(-0.5, +0.5, 0);
+			glColor3f(0, 0, 1); glVertex3f(x+w, y+h, 0);
+			glColor3f(1, 1, 0); glVertex3f(x,   y+h, 0);
+			glColor3f(1, 0, 0); glVertex3f(x,   y,   0);
 
-			glColor3f(0, 0, 1); glVertex3f(-0.5, +0.5, 0);
-			glColor3f(1, 1, 0); glVertex3f(-0.5, -0.5, 0);
-			glColor3f(1, 0, 0); glVertex3f(+0.5, -0.5, 0);
+			glEnd();
 		}
 
+		x, y : f32 = 100, 100;
+		draw_rect(x, y, 50, 50);
+
 		display_window(^window);
-		{
-			ms := (16 - dt*1000) as i32;
-			if ms > 0 { sleep_ms(ms); }
-		}
 	}
 }
+*/
diff --git a/examples/opengl.odin b/examples/opengl.odin
index 9002c1e61..edc7cb083 100644
--- a/examples/opengl.odin
+++ b/examples/opengl.odin
@@ -31,6 +31,7 @@ glBegin         :: proc(mode: i32) #foreign
 glEnd           :: proc() #foreign
 glColor3f       :: proc(r, g, b: f32) #foreign
 glColor4f       :: proc(r, g, b, a: f32) #foreign
+glVertex2f      :: proc(x, y: f32) #foreign
 glVertex3f      :: proc(x, y, z: f32) #foreign
 glTexCoord2f    :: proc(u, v: f32) #foreign
 glLoadIdentity  :: proc() #foreign
diff --git a/src/checker/checker.cpp b/src/checker/checker.cpp
index e75da4c87..b37effb8f 100644
--- a/src/checker/checker.cpp
+++ b/src/checker/checker.cpp
@@ -123,6 +123,7 @@ enum BuiltinProcId {
 	BuiltinProc_cap,
 	BuiltinProc_copy,
 	BuiltinProc_append,
+	BuiltinProc_swizzle,
 	BuiltinProc_print,
 	BuiltinProc_println,
 
@@ -149,8 +150,12 @@ gb_global BuiltinProc builtin_procs[BuiltinProc_Count] = {
 	{STR_LIT("cap"),              1, false, Expression_Expression},
 	{STR_LIT("copy"),             2, false, Expression_Expression},
 	{STR_LIT("append"),           2, false, Expression_Expression},
+
+	{STR_LIT("swizzle"),          1, true,  Expression_Expression},
+
 	{STR_LIT("print"),            1, true,  Expression_Statement},
 	{STR_LIT("println"),          1, true,  Expression_Statement},
+
 };
 
 struct CheckerContext {
diff --git a/src/checker/expr.cpp b/src/checker/expr.cpp
index c74d9e332..f03de5cbe 100644
--- a/src/checker/expr.cpp
+++ b/src/checker/expr.cpp
@@ -409,10 +409,7 @@ Type *check_type(Checker *c, AstNode *e, Type *named_type) {
 	case_ast_node(pt, ProcType, e);
 		type = alloc_type(c->allocator, Type_Proc);
 		set_base_type(named_type, type);
-		CheckerContext context = c->context;
-		c->context.scope = make_scope(c->context.scope, c->allocator);
 		check_procedure_type(c, type, e);
-		c->context = context;
 		goto end;
 	case_end;
 
@@ -979,7 +976,10 @@ void check_binary_expr(Checker *c, Operand *x, AstNode *node) {
 		return;
 	}
 
-	if (token_is_shift(be->op)) {
+	Token op = be->op;
+
+
+	if (token_is_shift(op)) {
 		check_shift(c, x, y, node);
 		return;
 	}
@@ -992,7 +992,6 @@ void check_binary_expr(Checker *c, Operand *x, AstNode *node) {
 		return;
 	}
 
-	Token op = be->op;
 	if (token_is_comparison(op)) {
 		check_comparison(c, x, y, op);
 		return;
@@ -1616,6 +1615,54 @@ b32 check_builtin_procedure(Checker *c, Operand *operand, AstNode *call, i32 id)
 		operand->mode = Addressing_Value;
 	} break;
 
+	case BuiltinProc_swizzle: {
+		// swizzle :: proc(v: {N}T, T...) -> {M}T
+		Type *vector_type = get_base_type(operand->type);
+		if (!is_type_vector(vector_type)) {
+			gbString type_str = type_to_string(operand->type);
+			defer (gb_string_free(type_str));
+			error(&c->error_collector, ast_node_token(call),
+			      "You can only `swizzle` a vector, got `%s`",
+			      type_str);
+			return false;
+		}
+
+		isize max_count = vector_type->vector.count;
+		isize arg_count = 0;
+		for (AstNode *arg = ce->arg_list->next; arg != NULL; arg = arg->next) {
+			Operand op = {};
+			check_expr(c, &op, arg);
+			if (op.mode == Addressing_Invalid)
+				return false;
+			Type *arg_type = get_base_type(op.type);
+			if (!is_type_integer(arg_type) || op.mode != Addressing_Constant) {
+				error(&c->error_collector, ast_node_token(op.expr), "Indices to `swizzle` must be constant integers");
+				return false;
+			}
+
+			if (op.value.value_integer < 0) {
+				error(&c->error_collector, ast_node_token(op.expr), "Negative `swizzle` index");
+				return false;
+			}
+
+			if (max_count <= op.value.value_integer) {
+				error(&c->error_collector, ast_node_token(op.expr), "`swizzle` index exceeds vector length");
+				return false;
+			}
+
+			arg_count++;
+		}
+
+		if (arg_count > max_count) {
+			error(&c->error_collector, ast_node_token(call), "Too many `swizzle` indices, %td > %td", arg_count, max_count);
+			return false;
+		}
+
+		Type *elem_type = vector_type->vector.elem;
+		operand->type = make_type_vector(c->allocator, elem_type, arg_count);
+		operand->mode = Addressing_Value;
+	}
+
 	case BuiltinProc_print:
 	case BuiltinProc_println: {
 		for (AstNode *arg = ce->arg_list; arg != NULL; arg = arg->next) {
@@ -1958,7 +2005,12 @@ ExpressionKind check__expr_base(Checker *c, Operand *o, AstNode *node, Type *typ
 				if (t->kind == Type_Array &&
 				    t->array.count >= 0 &&
 				    index >= t->array.count) {
-					error(&c->error_collector, ast_node_token(elem), "Index %lld is out of bounds (>= %lld)", index, t->array.count);
+					error(&c->error_collector, ast_node_token(elem), "Index %lld is out of bounds (>= %lld) for array literal", index, t->array.count);
+				}
+				if (t->kind == Type_Vector &&
+				    t->vector.count >= 0 &&
+				    index >= t->vector.count) {
+					error(&c->error_collector, ast_node_token(elem), "Index %lld is out of bounds (>= %lld) for vector literal", index, t->vector.count);
 				}
 
 				Operand o = {};
@@ -1968,6 +2020,13 @@ ExpressionKind check__expr_base(Checker *c, Operand *o, AstNode *node, Type *typ
 			if (max < index)
 				max = index;
 
+			if (t->kind == Type_Vector) {
+				if (t->vector.count > 1 && gb_is_between(index, 2, t->vector.count-1)) {
+					error(&c->error_collector, ast_node_token(cl->elem_list),
+					      "Expected either 1 (broadcast) or %td elements in vector literal, got %td", t->vector.count, index);
+				}
+			}
+
 			if (t->kind == Type_Array && ellipsis_array) {
 				t->array.count = max;
 			}
diff --git a/src/checker/stmt.cpp b/src/checker/stmt.cpp
index 95d25616e..3a059b2ab 100644
--- a/src/checker/stmt.cpp
+++ b/src/checker/stmt.cpp
@@ -369,18 +369,6 @@ void check_type_decl(Checker *c, Entity *e, AstNode *type_expr, Type *named_type
 	set_base_type(named, get_base_type(get_base_type(named)));
 }
 
-void check_alias_decl(Checker *c, Entity *e, AstNode *type_expr, Type *alias_type) {
-	GB_ASSERT(e->type == NULL);
-	Type *named = make_type_alias(c->allocator, e->token.string, NULL, e);
-	named->alias.alias_name = e;
-	set_base_type(alias_type, named);
-	e->type = named;
-
-	check_type(c, type_expr, named);
-
-	set_base_type(named, get_base_type(get_base_type(named)));
-}
-
 void check_proc_body(Checker *c, Token token, DeclInfo *decl, Type *type, AstNode *body) {
 	GB_ASSERT(body->kind == AstNode_BlockStmt);
 
@@ -496,9 +484,6 @@ void check_entity_decl(Checker *c, Entity *e, DeclInfo *d, Type *named_type) {
 	case Entity_TypeName:
 		check_type_decl(c, e, d->type_expr, named_type);
 		break;
-	case Entity_AliasName:
-		check_alias_decl(c, e, d->type_expr, named_type);
-		break;
 	case Entity_Procedure:
 		check_proc_decl(c, e, d, true);
 		break;
diff --git a/src/checker/type.cpp b/src/checker/type.cpp
index 0931774e5..650143ef5 100644
--- a/src/checker/type.cpp
+++ b/src/checker/type.cpp
@@ -64,7 +64,6 @@ struct BasicType {
 	TYPE_KIND(Structure), \
 	TYPE_KIND(Pointer), \
 	TYPE_KIND(Named), \
-	TYPE_KIND(Alias), \
 	TYPE_KIND(Tuple), \
 	TYPE_KIND(Proc), \
 	TYPE_KIND(Count),
@@ -110,11 +109,6 @@ struct Type {
 			Type *  base;
 			Entity *type_name; // Entity_TypeName
 		} named;
-		struct {
-			String  name;
-			Type *  base;
-			Entity *alias_name; // Entity_AliasName
-		} alias;
 		struct {
 			Entity **variables; // Entity_Variable
 			isize    variable_count;
@@ -130,12 +124,8 @@ struct Type {
 };
 
 Type *get_base_type(Type *t) {
-	while (t->kind == Type_Named || t->kind == Type_Alias) {
-		if (t->kind == Type_Named) {
-			t = t->named.base;
-		} else {
-			t = t->alias.base;
-		}
+	while (t->kind == Type_Named) {
+		t = t->named.base;
 	}
 	return t;
 }
@@ -143,8 +133,6 @@ Type *get_base_type(Type *t) {
 void set_base_type(Type *t, Type *base) {
 	if (t && t->kind == Type_Named) {
 		t->named.base = base;
-	} else if (t && t->kind == Type_Alias) {
-		t->alias.base = base;
 	}
 }
 
@@ -201,14 +189,6 @@ Type *make_type_named(gbAllocator a, String name, Type *base, Entity *type_name)
 	return t;
 }
 
-Type *make_type_alias(gbAllocator a, String name, Type *base, Entity *alias_name) {
-	Type *t = alloc_type(a, Type_Alias);
-	t->alias.name = name;
-	t->alias.base = base;
-	t->alias.alias_name = alias_name;
-	return t;
-}
-
 Type *make_type_tuple(gbAllocator a) {
 	Type *t = alloc_type(a, Type_Tuple);
 	return t;
@@ -465,16 +445,11 @@ b32 are_types_identical(Type *x, Type *y) {
 			return are_types_identical(x->pointer.elem, y->pointer.elem);
 		break;
 
-
-	case Type_Alias:
-		return are_types_identical(get_base_type(x), y);
-
 	case Type_Named:
 		if (y->kind == Type_Named)
 			return x->named.base == y->named.base;
 		break;
 
-
 	case Type_Tuple:
 		if (y->kind == Type_Tuple) {
 			if (x->tuple.variable_count == y->tuple.variable_count) {
@@ -737,15 +712,6 @@ gbString write_type_to_string(gbString str, Type *type) {
 		}
 		break;
 
-	case Type_Alias:
-		if (type->alias.alias_name != NULL) {
-			str = gb_string_append_length(str, type->alias.name.text, type->alias.name.len);
-		} else {
-			// NOTE(bill): Just in case
-			str = gb_string_appendc(str, "<alias type>");
-		}
-		break;
-
 	case Type_Tuple:
 		if (type->tuple.variable_count > 0) {
 			for (isize i = 0; i < type->tuple.variable_count; i++) {
diff --git a/src/codegen/print_llvm.cpp b/src/codegen/print_llvm.cpp
index d2394f202..f7c6d2651 100644
--- a/src/codegen/print_llvm.cpp
+++ b/src/codegen/print_llvm.cpp
@@ -164,9 +164,6 @@ void ssa_print_type(gbFile *f, BaseTypeSizes s, Type *t) {
 	case Type_Named:
 		ssa_print_encoded_local(f, t->named.name);
 		break;
-	case Type_Alias:
-		ssa_print_type(f, s, t->alias.base);
-		break;
 	case Type_Tuple:
 		if (t->tuple.variable_count == 1) {
 			ssa_print_type(f, s, t->tuple.variables[0]->type);
@@ -635,6 +632,29 @@ void ssa_print_instr(gbFile *f, ssaModule *m, ssaValue *value) {
 		ssa_fprintf(f, "\n");
 	} break;
 
+	case ssaInstr_ShuffleVector: {
+		auto *sv = &instr->shuffle_vector;
+		Type *vt = ssa_value_type(sv->vector);
+		ssa_fprintf(f, "%%%d = shufflevector ", value->id);
+
+		ssa_print_type(f, m->sizes, vt);
+		ssa_fprintf(f, " ");
+		ssa_print_value(f, m, sv->vector, vt);
+		ssa_fprintf(f, ", ");
+		ssa_print_type(f, m->sizes, vt);
+		ssa_fprintf(f, " undef,");
+
+		ssa_fprintf(f, " <%td x i32> <", sv->index_count);
+		for (isize i = 0; i < sv->index_count; i++) {
+			if (i > 0) {
+				ssa_fprintf(f, ", ");
+			}
+			ssa_fprintf(f, "i32 %d", sv->indices[i]);
+		}
+		ssa_fprintf(f, ">");
+		ssa_fprintf(f, "\n");
+	} break;
+
 
 	default: {
 		GB_PANIC("<unknown instr> %d\n", instr->kind);
diff --git a/src/codegen/ssa.cpp b/src/codegen/ssa.cpp
index fbf1f8d7f..fc927b735 100644
--- a/src/codegen/ssa.cpp
+++ b/src/codegen/ssa.cpp
@@ -208,6 +208,12 @@ struct ssaInstr {
 			ssaValue *elem;
 			ssaValue *index;
 		} insert_element;
+		struct {
+			ssaValue *vector;
+			i32 *indices;
+			isize index_count;
+			Type *type;
+		} shuffle_vector;
 
 		struct {} startup_runtime;
 	};
@@ -291,8 +297,9 @@ ssaLvalue ssa_make_lvalue_vector(ssaValue *address, ssaValue *index, AstNode *ex
 
 
 void ssa_module_init(ssaModule *m, Checker *c) {
+	// TODO(bill): Determine a decent size for the arena
 	isize token_count = c->parser->total_token_count;
-	isize arena_size = 3 * token_count * gb_size_of(ssaValue);
+	isize arena_size = 4 * token_count * gb_size_of(ssaValue);
 	gb_arena_init_from_allocator(&m->arena, gb_heap_allocator(), arena_size);
 	m->allocator = gb_arena_allocator(&m->arena);
 	m->info = &c->info;
@@ -354,6 +361,8 @@ Type *ssa_instr_type(ssaInstr *instr) {
 	} break;
 	case ssaInstr_InsertElement:
 		return ssa_value_type(instr->insert_element.vector);
+	case ssaInstr_ShuffleVector:
+		return instr->shuffle_vector.type;
 	}
 	return NULL;
 }
@@ -651,6 +660,21 @@ ssaValue *ssa_make_instr_insert_element(ssaProcedure *p, ssaValue *vector, ssaVa
 	return v;
 }
 
+ssaValue *ssa_make_instr_shuffle_vector(ssaProcedure *p, ssaValue *vector, i32 *indices, isize index_count) {
+	ssaValue *v = ssa_alloc_instr(p->module->allocator, ssaInstr_ShuffleVector);
+	v->instr.shuffle_vector.vector      = vector;
+	v->instr.shuffle_vector.indices     = indices;
+	v->instr.shuffle_vector.index_count = index_count;
+
+	Type *vt = get_base_type(ssa_value_type(vector));
+	v->instr.shuffle_vector.type = make_type_vector(p->module->allocator, vt->vector.elem, index_count);
+
+	if (p->curr_block) {
+		gb_array_append(p->curr_block->values, v);
+	}
+	return v;
+}
+
 ssaValue *ssa_make_instr_no_op(ssaProcedure *p) {
 	ssaValue *v = ssa_alloc_instr(p->module->allocator, ssaInstr_NoOp);
 	if (p->curr_block) {
@@ -1379,6 +1403,23 @@ ssaValue *ssa_emit_conv(ssaProcedure *proc, ssaValue *value, Type *t) {
 		return ssa_emit_load(proc, slice);
 	}
 
+	if (is_type_vector(dst)) {
+		Type *dst_elem = dst->vector.elem;
+		value = ssa_emit_conv(proc, value, dst_elem);
+		ssaValue *v = ssa_add_local_generated(proc, t);
+		v = ssa_emit_load(proc, v);
+		v = ssa_emit(proc, ssa_make_instr_insert_element(proc, v, value, v_zero32));
+		// NOTE(bill): Broadcast lowest value to all values
+		isize index_count = dst->vector.count;
+		i32 *indices = gb_alloc_array(proc->module->allocator, i32, index_count);
+		for (isize i = 0; i < index_count; i++) {
+			indices[i] = 0;
+		}
+
+		v = ssa_emit(proc, ssa_make_instr_shuffle_vector(proc, v, indices, index_count));
+		return v;
+	}
+
 
 	gb_printf_err("Not Identical %s != %s\n", type_to_string(src_type), type_to_string(t));
 	gb_printf_err("Not Identical %s != %s\n", type_to_string(src), type_to_string(dst));
@@ -1574,6 +1615,15 @@ ssaValue *ssa_build_single_expr(ssaProcedure *proc, AstNode *expr, TypeAndValue
 				ssaValue *i = ssa_make_value_constant(proc->module->allocator, t_int, make_exact_value_integer(index));
 				result = ssa_emit(proc, ssa_make_instr_insert_element(proc, result, ev, i));
 			}
+			if (index == 1 && base_type->vector.count > 1) {
+				isize index_count = base_type->vector.count;
+				i32 *indices = gb_alloc_array(proc->module->allocator, i32, index_count);
+				for (isize i = 0; i < index_count; i++) {
+					indices[i] = 0;
+				}
+				return ssa_emit(proc, ssa_make_instr_shuffle_vector(proc, result, indices, index_count));
+			}
+
 			return result;
 		} break;
 
@@ -1750,6 +1800,27 @@ ssaValue *ssa_build_single_expr(ssaProcedure *proc, AstNode *expr, TypeAndValue
 
 					return ssa_emit_conv(proc, cond, t_bool);
 				} break;
+
+				case BuiltinProc_swizzle: {
+					ssaValue *vector = ssa_build_expr(proc, ce->arg_list);
+					isize index_count = ce->arg_list_count-1;
+					if (index_count == 0) {
+						return vector;
+					}
+
+					i32 *indices = gb_alloc_array(proc->module->allocator, i32, index_count);
+					isize index = 0;
+					for (AstNode *arg = ce->arg_list->next; arg != NULL; arg = arg->next) {
+						TypeAndValue *tv = type_and_value_of_expression(proc->module->info, arg);
+						GB_ASSERT(is_type_integer(tv->type));
+						GB_ASSERT(tv->value.kind == ExactValue_Integer);
+						indices[index++] = cast(i32)tv->value.value_integer;
+					}
+
+					return ssa_emit(proc, ssa_make_instr_shuffle_vector(proc, vector, indices, index_count));
+
+				} break;
+
 				case BuiltinProc_print: {
 					// print :: proc(...)
 					GB_PANIC("TODO(bill): BuiltinProc_print");
diff --git a/src/main.cpp b/src/main.cpp
index 5abb99eb3..c337d1ba2 100644
--- a/src/main.cpp
+++ b/src/main.cpp
@@ -84,8 +84,8 @@ int main(int argc, char **argv) {
 				if (exit_code == 0) {
 					win32_exec_command_line_app(
 						"clang -o %.*s.exe %.*s.bc -Wno-override-module "
-						" ../c_libs/stb_image.c -DSTB_IMAGE_IMPLEMENTATION "
-						"-lkernel32.lib -luser32.lib -lgdi32.lib -lopengl32.lib",
+						"-lkernel32.lib -luser32.lib -lgdi32.lib -lopengl32.lib "
+						"-l../c_libs/stb_image.lib",
 						cast(int)base_name_len, output_name,
 						cast(int)base_name_len, output_name);
 					if (run_output) {