Merge pull request #1 from gnzlbg/p32blo_sse41

P32blo sse41

Author: p32blo

src/lib.rs

@@ -169,14 +169,20 @@ mod v32 {
 
     define_ty! { i16x2, i16, i16 }
     define_impl! { i16x2, i16, 2, i16x2, x0, x1 }
+    define_ty! { u16x2, u16, u16 }
+    define_impl! { u16x2, u16, 2, i16x2, x0, x1 }
 
     define_ty! { i8x4, i8, i8, i8, i8 }
     define_impl! { i8x4, i8, 4, i8x4, x0, x1, x2, x3 }
-
     define_ty! { u8x4, u8, u8, u8, u8 }
     define_impl! { u8x4, u8, 4, i8x4, x0, x1, x2, x3 }
 
-    define_casts!((i8x4, i32x4, as_i32x4), (i16x2, i64x2, as_i64x2));
+    define_casts!(
+        (i16x2, i64x2, as_i64x2),
+        (u16x2, i64x2, as_i64x2),
+        (i8x4, i32x4, as_i32x4),
+        (u8x4, i32x4, as_i32x4)
+    );
 }
 
 /// 16-bit wide vector tpyes
@@ -185,8 +191,10 @@ mod v16 {
 
     define_ty! { i8x2, i8, i8 }
     define_impl! { i8x2, i8, 2, i8x2, x0, x1 }
+    define_ty! { u8x2, u8, u8 }
+    define_impl! { u8x2, u8, 2, i8x2, x0, x1 }
 
-    define_casts!((i8x2, i64x2, as_i64x2));
+    define_casts!((i8x2, i64x2, as_i64x2), (u8x2, i64x2, as_i64x2));
 }
 
 #[cfg(any(target_arch = "x86", target_arch = "x86_64"))]

src/macros.rs

@@ -485,7 +485,7 @@ macro_rules! test_arithmetic_ {
 
 #[cfg(test)]
 #[macro_export]
-    macro_rules! test_neg_ {
+macro_rules! test_neg_ {
         ($tn:ident, $zero:expr, $one:expr, $two:expr, $four:expr) => {
             {
                 let z = $tn::splat($zero);
@@ -573,7 +573,7 @@ macro_rules! test_bit_arithmetic_ {
 
 #[cfg(test)]
 #[macro_export]
-    macro_rules! test_ops_si {
+macro_rules! test_ops_si {
         ($($tn:ident),+) => {
             $(
                 test_arithmetic_!($tn, 0, 1, 2, 4);
@@ -585,7 +585,7 @@ macro_rules! test_bit_arithmetic_ {
 
 #[cfg(test)]
 #[macro_export]
-    macro_rules! test_ops_ui {
+macro_rules! test_ops_ui {
         ($($tn:ident),+) => {
             $(
                 test_arithmetic_!($tn, 0, 1, 2, 4);
@@ -596,7 +596,7 @@ macro_rules! test_bit_arithmetic_ {
 
 #[cfg(test)]
 #[macro_export]
-    macro_rules! test_ops_f {
+macro_rules! test_ops_f {
         ($($tn:ident),+)  => {
             $(
                 test_arithmetic_!($tn, 0., 1., 2., 4.);

src/v64.rs

@@ -60,11 +60,14 @@ define_casts!(
     (u8x8, i8x8, as_i8x8),
     (i8x8, u8x8, as_u8x8),
     (i8x8, i16x8, as_i16x8),
+    (u8x8, i16x8, as_i16x8),
     (i16x4, i32x4, as_i32x4),
     (i32x2, i64x2, as_i64x2),
     (u8x8, u16x8, as_u16x8),
     (u16x4, u32x4, as_u32x4),
-    (u32x2, u64x2, as_u64x2)
+    (u16x4, i32x4, as_i32x4),
+    (u32x2, u64x2, as_u64x2),
+    (u32x2, i64x2, as_i64x2)
 );
 
 #[cfg(test)]

src/x86/macros.rs

@@ -328,6 +328,22 @@ macro_rules! constify_imm4 {
     }
 }
 
+macro_rules! constify_imm3 {
+    ($imm8:expr, $expand:ident) => {
+        #[allow(overflowing_literals)]
+        match $imm8 & 0b111 {
+            0 => $expand!(0),
+            1 => $expand!(1),
+            2 => $expand!(2),
+            3 => $expand!(3),
+            4 => $expand!(4),
+            5 => $expand!(5),
+            6 => $expand!(6),
+            _ => $expand!(7),
+        }
+    }
+}
+
 macro_rules! constify_imm2 {
     ($imm8:expr, $expand:ident) => {
         #[allow(overflowing_literals)]