Auto merge of #33460 - shepmaster:16-bit-pointers, r=Aatch

Support 16-bit pointers as well as i/usize

I'm opening this pull request to get some feedback from the community.

Although Rust doesn't support any platforms with a native 16-bit pointer at the moment, the [AVR-Rust][ar] fork is working towards that goal. Keeping this forked logic up-to-date with the changes in master has been onerous so I'd like to merge these changes so that they get carried along when refactoring happens. I do not believe this should increase the maintenance burden.

This is based on the original work of Dylan McKay (@dylanmckay).

[ar]: https://github.com/avr-rust/rust

Author: bors

src/liballoc/raw_vec.rs

@@ -578,9 +578,9 @@ impl<T> Drop for RawVec<T> {
 // * We don't overflow `usize::MAX` and actually allocate too little
 //
 // On 64-bit we just need to check for overflow since trying to allocate
-// `> isize::MAX` bytes will surely fail. On 32-bit we need to add an extra
-// guard for this in case we're running on a platform which can use all 4GB in
-// user-space. e.g. PAE or x32
+// `> isize::MAX` bytes will surely fail. On 32-bit and 16-bit we need to add
+// an extra guard for this in case we're running on a platform which can use
+// all 4GB in user-space. e.g. PAE or x32
 
 #[inline]
 fn alloc_guard(alloc_size: usize) {

src/libcore/fmt/num.rs

@@ -29,6 +29,7 @@ trait Int: Zero + PartialEq + PartialOrd + Div<Output=Self> + Rem<Output=Self> +
            Sub<Output=Self> + Copy {
     fn from_u8(u: u8) -> Self;
     fn to_u8(&self) -> u8;
+    fn to_u16(&self) -> u16;
     fn to_u32(&self) -> u32;
     fn to_u64(&self) -> u64;
 }
@@ -37,6 +38,7 @@ macro_rules! doit {
     ($($t:ident)*) => ($(impl Int for $t {
         fn from_u8(u: u8) -> $t { u as $t }
         fn to_u8(&self) -> u8 { *self as u8 }
+        fn to_u16(&self) -> u16 { *self as u16 }
         fn to_u32(&self) -> u32 { *self as u32 }
         fn to_u64(&self) -> u64 { *self as u64 }
     })*)
@@ -256,6 +258,8 @@ macro_rules! impl_Display {
 
 impl_Display!(i8, u8, i16, u16, i32, u32: to_u32);
 impl_Display!(i64, u64: to_u64);
+#[cfg(target_pointer_width = "16")]
+impl_Display!(isize, usize: to_u16);
 #[cfg(target_pointer_width = "32")]
 impl_Display!(isize, usize: to_u32);
 #[cfg(target_pointer_width = "64")]

src/libcore/mem.rs

@@ -518,6 +518,10 @@ pub fn replace<T>(dest: &mut T, mut src: T) -> T {
 #[stable(feature = "rust1", since = "1.0.0")]
 pub fn drop<T>(_x: T) { }
 
+macro_rules! repeat_u8_as_u16 {
+    ($name:expr) => { (($name as u16) <<  8 |
+                       ($name as u16)) }
+}
 macro_rules! repeat_u8_as_u32 {
     ($name:expr) => { (($name as u32) << 24 |
                        ($name as u32) << 16 |
@@ -543,11 +547,18 @@ macro_rules! repeat_u8_as_u64 {
 pub const POST_DROP_U8: u8 = 0x1d;
 #[unstable(feature = "filling_drop", issue = "5016")]
 #[allow(missing_docs)]
+pub const POST_DROP_U16: u16 = repeat_u8_as_u16!(POST_DROP_U8);
+#[unstable(feature = "filling_drop", issue = "5016")]
+#[allow(missing_docs)]
 pub const POST_DROP_U32: u32 = repeat_u8_as_u32!(POST_DROP_U8);
 #[unstable(feature = "filling_drop", issue = "5016")]
 #[allow(missing_docs)]
 pub const POST_DROP_U64: u64 = repeat_u8_as_u64!(POST_DROP_U8);
 
+#[cfg(target_pointer_width = "16")]
+#[unstable(feature = "filling_drop", issue = "5016")]
+#[allow(missing_docs)]
+pub const POST_DROP_USIZE: usize = POST_DROP_U16 as usize;
 #[cfg(target_pointer_width = "32")]
 #[unstable(feature = "filling_drop", issue = "5016")]
 #[allow(missing_docs)]

src/libcore/num/isize.rs

@@ -14,6 +14,8 @@
 
 #![stable(feature = "rust1", since = "1.0.0")]
 
+#[cfg(target_pointer_width = "16")]
+int_module! { isize, 16 }
 #[cfg(target_pointer_width = "32")]
 int_module! { isize, 32 }
 #[cfg(target_pointer_width = "64")]

src/libcore/num/mod.rs

@@ -1176,6 +1176,15 @@ impl i64 {
         intrinsics::mul_with_overflow }
 }
 
+#[cfg(target_pointer_width = "16")]
+#[lang = "isize"]
+impl isize {
+    int_impl! { i16, u16, 16,
+        intrinsics::add_with_overflow,
+        intrinsics::sub_with_overflow,
+        intrinsics::mul_with_overflow }
+}
+
 #[cfg(target_pointer_width = "32")]
 #[lang = "isize"]
 impl isize {
@@ -2188,6 +2197,18 @@ impl u64 {
         intrinsics::mul_with_overflow }
 }
 
+#[cfg(target_pointer_width = "16")]
+#[lang = "usize"]
+impl usize {
+    uint_impl! { u16, 16,
+        intrinsics::ctpop,
+        intrinsics::ctlz,
+        intrinsics::cttz,
+        intrinsics::bswap,
+        intrinsics::add_with_overflow,
+        intrinsics::sub_with_overflow,
+        intrinsics::mul_with_overflow }
+}
 #[cfg(target_pointer_width = "32")]
 #[lang = "usize"]
 impl usize {

src/libcore/num/usize.rs

@@ -14,6 +14,8 @@
 
 #![stable(feature = "rust1", since = "1.0.0")]
 
+#[cfg(target_pointer_width = "16")]
+uint_module! { usize, 16 }
 #[cfg(target_pointer_width = "32")]
 uint_module! { usize, 32 }
 #[cfg(target_pointer_width = "64")]

src/libcore/num/wrapping.rs

@@ -292,6 +292,12 @@ wrapping_impl! { usize u8 u16 u32 u64 isize i8 i16 i32 i64 }
 mod shift_max {
     #![allow(non_upper_case_globals)]
 
+    #[cfg(target_pointer_width = "16")]
+    mod platform {
+        pub const usize: u32 = super::u16;
+        pub const isize: u32 = super::i16;
+    }
+
     #[cfg(target_pointer_width = "32")]
     mod platform {
         pub const usize: u32 = super::u32;

src/libcoretest/mem.rs

@@ -18,6 +18,13 @@ fn size_of_basic() {
     assert_eq!(size_of::<u64>(), 8);
 }
 
+#[test]
+#[cfg(target_pointer_width = "16")]
+fn size_of_16() {
+    assert_eq!(size_of::<usize>(), 2);
+    assert_eq!(size_of::<*const usize>(), 2);
+}
+
 #[test]
 #[cfg(target_pointer_width = "32")]
 fn size_of_32() {
@@ -47,6 +54,13 @@ fn align_of_basic() {
     assert_eq!(align_of::<u32>(), 4);
 }
 
+#[test]
+#[cfg(target_pointer_width = "16")]
+fn align_of_16() {
+    assert_eq!(align_of::<usize>(), 2);
+    assert_eq!(align_of::<*const usize>(), 2);
+}
+
 #[test]
 #[cfg(target_pointer_width = "32")]
 fn align_of_32() {

src/librustc/session/config.rs

@@ -796,6 +796,7 @@ pub fn build_target_config(opts: &Options, sp: &Handler) -> Config {
     };
 
     let (int_type, uint_type) = match &target.target_pointer_width[..] {
+        "16" => (ast::IntTy::I16, ast::UintTy::U16),
         "32" => (ast::IntTy::I32, ast::UintTy::U32),
         "64" => (ast::IntTy::I64, ast::UintTy::U64),
         w    => panic!(sp.fatal(&format!("target specification was invalid: \

src/librustc/ty/layout.rs

@@ -170,6 +170,7 @@ impl TargetDataLayout {
     /// address space on 64-bit ARMv8 and x86_64.
     pub fn obj_size_bound(&self) -> u64 {
         match self.pointer_size.bits() {
+            16 => 1 << 15,
             32 => 1 << 31,
             64 => 1 << 47,
             bits => bug!("obj_size_bound: unknown pointer bit size {}", bits)
@@ -178,6 +179,7 @@ impl TargetDataLayout {
 
     pub fn ptr_sized_integer(&self) -> Integer {
         match self.pointer_size.bits() {
+            16 => I16,
             32 => I32,
             64 => I64,
             bits => bug!("ptr_sized_integer: unknown pointer bit size {}", bits)

src/librustc/ty/util.rs

@@ -62,6 +62,7 @@ impl IntTypeExt for attr::IntType {
             SignedInt(ast::IntTy::I32)   => ConstInt::I32(0),
             SignedInt(ast::IntTy::I64)   => ConstInt::I64(0),
             SignedInt(ast::IntTy::Is) => match tcx.sess.target.int_type {
+                ast::IntTy::I16 => ConstInt::Isize(ConstIsize::Is16(0)),
                 ast::IntTy::I32 => ConstInt::Isize(ConstIsize::Is32(0)),
                 ast::IntTy::I64 => ConstInt::Isize(ConstIsize::Is64(0)),
                 _ => bug!(),
@@ -71,6 +72,7 @@ impl IntTypeExt for attr::IntType {
             UnsignedInt(ast::UintTy::U32) => ConstInt::U32(0),
             UnsignedInt(ast::UintTy::U64) => ConstInt::U64(0),
             UnsignedInt(ast::UintTy::Us) => match tcx.sess.target.uint_type {
+                ast::UintTy::U16 => ConstInt::Usize(ConstUsize::Us16(0)),
                 ast::UintTy::U32 => ConstInt::Usize(ConstUsize::Us32(0)),
                 ast::UintTy::U64 => ConstInt::Usize(ConstUsize::Us64(0)),
                 _ => bug!(),

src/librustc_const_eval/eval.rs

@@ -583,6 +583,9 @@ pub fn eval_const_expr_partial<'a, 'tcx>(tcx: TyCtxt<'a, 'tcx, 'tcx>,
                     (&LitKind::Int(n, Unsuffixed), Some(&ty::TyInt(IntTy::Is))) |
                     (&LitKind::Int(n, Signed(IntTy::Is)), _) => {
                         match tcx.sess.target.int_type {
+                            IntTy::I16 => if n == I16_OVERFLOW {
+                                return Ok(Integral(Isize(Is16(::std::i16::MIN))));
+                            },
                             IntTy::I32 => if n == I32_OVERFLOW {
                                 return Ok(Integral(Isize(Is32(::std::i32::MIN))));
                             },