mirror of
https://github.com/RPCSX/llvm.git
synced 2024-12-14 15:39:00 +00:00
59ad77e46e
ARMv8.2-A adds 16-bit floating point versions of all existing SIMD floating-point instructions. This is an optional extension, so all of these instructions require the FeatureFullFP16 subtarget feature. Note that VFP without SIMD is not a valid combination for any version of ARMv8-A, but I have ensured that these instructions all depend on both FeatureNEON and FeatureFullFP16 for consistency. The ".2h" vector type specifier is now legal (for the scalar pairwise reduction instructions), so some unrelated tests have been modified as different error messages are emitted. This is not a problem as the invalid operands are still caught. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255010 91177308-0d34-0410-b5e6-96231b3b80d8
127 lines
6.2 KiB
ArmAsm
127 lines
6.2 KiB
ArmAsm
// RUN: llvm-mc -triple aarch64-none-linux-gnu -mattr=+neon,+fullfp16 -show-encoding < %s | FileCheck %s
|
|
|
|
// Check that the assembler can handle the documented syntax for AArch64
|
|
|
|
//----------------------------------------------------------------------
|
|
// Vector Maximum Pairwise (Signed and Unsigned Integer)
|
|
//----------------------------------------------------------------------
|
|
smaxp v0.8b, v1.8b, v2.8b
|
|
smaxp v0.16b, v1.16b, v2.16b
|
|
smaxp v0.4h, v1.4h, v2.4h
|
|
smaxp v0.8h, v1.8h, v2.8h
|
|
smaxp v0.2s, v1.2s, v2.2s
|
|
smaxp v0.4s, v1.4s, v2.4s
|
|
|
|
// CHECK: smaxp v0.8b, v1.8b, v2.8b // encoding: [0x20,0xa4,0x22,0x0e]
|
|
// CHECK: smaxp v0.16b, v1.16b, v2.16b // encoding: [0x20,0xa4,0x22,0x4e]
|
|
// CHECK: smaxp v0.4h, v1.4h, v2.4h // encoding: [0x20,0xa4,0x62,0x0e]
|
|
// CHECK: smaxp v0.8h, v1.8h, v2.8h // encoding: [0x20,0xa4,0x62,0x4e]
|
|
// CHECK: smaxp v0.2s, v1.2s, v2.2s // encoding: [0x20,0xa4,0xa2,0x0e]
|
|
// CHECK: smaxp v0.4s, v1.4s, v2.4s // encoding: [0x20,0xa4,0xa2,0x4e]
|
|
|
|
umaxp v0.8b, v1.8b, v2.8b
|
|
umaxp v0.16b, v1.16b, v2.16b
|
|
umaxp v0.4h, v1.4h, v2.4h
|
|
umaxp v0.8h, v1.8h, v2.8h
|
|
umaxp v0.2s, v1.2s, v2.2s
|
|
umaxp v0.4s, v1.4s, v2.4s
|
|
|
|
// CHECK: umaxp v0.8b, v1.8b, v2.8b // encoding: [0x20,0xa4,0x22,0x2e]
|
|
// CHECK: umaxp v0.16b, v1.16b, v2.16b // encoding: [0x20,0xa4,0x22,0x6e]
|
|
// CHECK: umaxp v0.4h, v1.4h, v2.4h // encoding: [0x20,0xa4,0x62,0x2e]
|
|
// CHECK: umaxp v0.8h, v1.8h, v2.8h // encoding: [0x20,0xa4,0x62,0x6e]
|
|
// CHECK: umaxp v0.2s, v1.2s, v2.2s // encoding: [0x20,0xa4,0xa2,0x2e]
|
|
// CHECK: umaxp v0.4s, v1.4s, v2.4s // encoding: [0x20,0xa4,0xa2,0x6e]
|
|
|
|
//----------------------------------------------------------------------
|
|
// Vector Minimum Pairwise (Signed and Unsigned Integer)
|
|
//----------------------------------------------------------------------
|
|
sminp v0.8b, v1.8b, v2.8b
|
|
sminp v0.16b, v1.16b, v2.16b
|
|
sminp v0.4h, v1.4h, v2.4h
|
|
sminp v0.8h, v1.8h, v2.8h
|
|
sminp v0.2s, v1.2s, v2.2s
|
|
sminp v0.4s, v1.4s, v2.4s
|
|
|
|
// CHECK: sminp v0.8b, v1.8b, v2.8b // encoding: [0x20,0xac,0x22,0x0e]
|
|
// CHECK: sminp v0.16b, v1.16b, v2.16b // encoding: [0x20,0xac,0x22,0x4e]
|
|
// CHECK: sminp v0.4h, v1.4h, v2.4h // encoding: [0x20,0xac,0x62,0x0e]
|
|
// CHECK: sminp v0.8h, v1.8h, v2.8h // encoding: [0x20,0xac,0x62,0x4e]
|
|
// CHECK: sminp v0.2s, v1.2s, v2.2s // encoding: [0x20,0xac,0xa2,0x0e]
|
|
// CHECK: sminp v0.4s, v1.4s, v2.4s // encoding: [0x20,0xac,0xa2,0x4e]
|
|
|
|
uminp v0.8b, v1.8b, v2.8b
|
|
uminp v0.16b, v1.16b, v2.16b
|
|
uminp v0.4h, v1.4h, v2.4h
|
|
uminp v0.8h, v1.8h, v2.8h
|
|
uminp v0.2s, v1.2s, v2.2s
|
|
uminp v0.4s, v1.4s, v2.4s
|
|
|
|
// CHECK: uminp v0.8b, v1.8b, v2.8b // encoding: [0x20,0xac,0x22,0x2e]
|
|
// CHECK: uminp v0.16b, v1.16b, v2.16b // encoding: [0x20,0xac,0x22,0x6e]
|
|
// CHECK: uminp v0.4h, v1.4h, v2.4h // encoding: [0x20,0xac,0x62,0x2e]
|
|
// CHECK: uminp v0.8h, v1.8h, v2.8h // encoding: [0x20,0xac,0x62,0x6e]
|
|
// CHECK: uminp v0.2s, v1.2s, v2.2s // encoding: [0x20,0xac,0xa2,0x2e]
|
|
// CHECK: uminp v0.4s, v1.4s, v2.4s // encoding: [0x20,0xac,0xa2,0x6e]
|
|
|
|
//----------------------------------------------------------------------
|
|
// Vector Maximum Pairwise (Floating Point)
|
|
//----------------------------------------------------------------------
|
|
fmaxp v0.4h, v1.4h, v2.4h
|
|
fmaxp v31.8h, v15.8h, v16.8h
|
|
fmaxp v0.2s, v1.2s, v2.2s
|
|
fmaxp v31.4s, v15.4s, v16.4s
|
|
fmaxp v7.2d, v8.2d, v25.2d
|
|
|
|
// CHECK: fmaxp v0.4h, v1.4h, v2.4h // encoding: [0x20,0x34,0x42,0x2e]
|
|
// CHECK: fmaxp v31.8h, v15.8h, v16.8h // encoding: [0xff,0x35,0x50,0x6e]
|
|
// CHECK: fmaxp v0.2s, v1.2s, v2.2s // encoding: [0x20,0xf4,0x22,0x2e]
|
|
// CHECK: fmaxp v31.4s, v15.4s, v16.4s // encoding: [0xff,0xf5,0x30,0x6e]
|
|
// CHECK: fmaxp v7.2d, v8.2d, v25.2d // encoding: [0x07,0xf5,0x79,0x6e]
|
|
|
|
//----------------------------------------------------------------------
|
|
// Vector Minimum Pairwise (Floating Point)
|
|
//----------------------------------------------------------------------
|
|
fminp v10.4h, v15.4h, v22.4h
|
|
fminp v3.8h, v5.8h, v6.8h
|
|
fminp v10.2s, v15.2s, v22.2s
|
|
fminp v3.4s, v5.4s, v6.4s
|
|
fminp v17.2d, v13.2d, v2.2d
|
|
|
|
// CHECK: fminp v10.4h, v15.4h, v22.4h // encoding: [0xea,0x35,0xd6,0x2e]
|
|
// CHECK: fminp v3.8h, v5.8h, v6.8h // encoding: [0xa3,0x34,0xc6,0x6e]
|
|
// CHECK: fminp v10.2s, v15.2s, v22.2s // encoding: [0xea,0xf5,0xb6,0x2e]
|
|
// CHECK: fminp v3.4s, v5.4s, v6.4s // encoding: [0xa3,0xf4,0xa6,0x6e]
|
|
// CHECK: fminp v17.2d, v13.2d, v2.2d // encoding: [0xb1,0xf5,0xe2,0x6e]
|
|
|
|
//----------------------------------------------------------------------
|
|
// Vector maxNum Pairwise (Floating Point)
|
|
//----------------------------------------------------------------------
|
|
fmaxnmp v0.4h, v1.4h, v2.4h
|
|
fmaxnmp v31.8h, v15.8h, v16.8h
|
|
fmaxnmp v0.2s, v1.2s, v2.2s
|
|
fmaxnmp v31.4s, v15.4s, v16.4s
|
|
fmaxnmp v7.2d, v8.2d, v25.2d
|
|
|
|
// CHECK: fmaxnmp v0.4h, v1.4h, v2.4h // encoding: [0x20,0x04,0x42,0x2e]
|
|
// CHECK: fmaxnmp v31.8h, v15.8h, v16.8h // encoding: [0xff,0x05,0x50,0x6e]
|
|
// CHECK: fmaxnmp v0.2s, v1.2s, v2.2s // encoding: [0x20,0xc4,0x22,0x2e]
|
|
// CHECK: fmaxnmp v31.4s, v15.4s, v16.4s // encoding: [0xff,0xc5,0x30,0x6e]
|
|
// CHECK: fmaxnmp v7.2d, v8.2d, v25.2d // encoding: [0x07,0xc5,0x79,0x6e]
|
|
|
|
//----------------------------------------------------------------------
|
|
// Vector minNum Pairwise (Floating Point)
|
|
//----------------------------------------------------------------------
|
|
fminnmp v10.4h, v15.4h, v22.4h
|
|
fminnmp v3.8h, v5.8h, v6.8h
|
|
fminnmp v10.2s, v15.2s, v22.2s
|
|
fminnmp v3.4s, v5.4s, v6.4s
|
|
fminnmp v17.2d, v13.2d, v2.2d
|
|
|
|
// CHECK: fminnmp v10.4h, v15.4h, v22.4h // encoding: [0xea,0x05,0xd6,0x2e]
|
|
// CHECK: fminnmp v3.8h, v5.8h, v6.8h // encoding: [0xa3,0x04,0xc6,0x6e]
|
|
// CHECK: fminnmp v10.2s, v15.2s, v22.2s // encoding: [0xea,0xc5,0xb6,0x2e]
|
|
// CHECK: fminnmp v3.4s, v5.4s, v6.4s // encoding: [0xa3,0xc4,0xa6,0x6e]
|
|
// CHECK: fminnmp v17.2d, v13.2d, v2.2d // encoding: [0xb1,0xc5,0xe2,0x6e]
|
|
|