6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2024-12-30 02:32:08 +00:00
Code Issues Projects Releases Wiki Activity
3cd73a5ded
llvm-6502/test/CodeGen/R600/kernel-args.ll

474 lines
14 KiB
LLVM
Raw Normal View History

R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
; RUN: llc < %s -march=r600 -mcpu=redwood | FileCheck %s --check-prefix=EG-CHECK
; RUN: llc < %s -march=r600 -mcpu=cayman | FileCheck %s --check-prefix=EG-CHECK
; RUN: llc < %s -march=r600 -mcpu=SI -verify-machineinstrs | FileCheck %s --check-prefix=SI-CHECK
; EG-CHECK-LABEL: @i8_arg
; EG-CHECK: MOV {{[ *]*}}T{{[0-9]+\.[XYZW]}}, KC0[2].Z
; SI-CHECK-LABEL: @i8_arg
; SI-CHECK: BUFFER_LOAD_UBYTE
define void @i8_arg(i32 addrspace(1)* nocapture %out, i8 %in) nounwind {
entry:
%0 = zext i8 %in to i32
store i32 %0, i32 addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @i8_zext_arg
; EG-CHECK: MOV {{[ *]*}}T{{[0-9]+\.[XYZW]}}, KC0[2].Z
; SI-CHECK-LABEL: @i8_zext_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORD s{{[0-9]}}, s[0:1], 0xb
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @i8_zext_arg(i32 addrspace(1)* nocapture %out, i8 zeroext %in) nounwind {
entry:
%0 = zext i8 %in to i32
store i32 %0, i32 addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @i8_sext_arg
; EG-CHECK: MOV {{[ *]*}}T{{[0-9]+\.[XYZW]}}, KC0[2].Z
; SI-CHECK-LABEL: @i8_sext_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORD s{{[0-9]}}, s[0:1], 0xb
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @i8_sext_arg(i32 addrspace(1)* nocapture %out, i8 signext %in) nounwind {
entry:
%0 = sext i8 %in to i32
store i32 %0, i32 addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @i16_arg
; EG-CHECK: MOV {{[ *]*}}T{{[0-9]+\.[XYZW]}}, KC0[2].Z
; SI-CHECK-LABEL: @i16_arg
; SI-CHECK: BUFFER_LOAD_USHORT
define void @i16_arg(i32 addrspace(1)* nocapture %out, i16 %in) nounwind {
entry:
%0 = zext i16 %in to i32
store i32 %0, i32 addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @i16_zext_arg
; EG-CHECK: MOV {{[ *]*}}T{{[0-9]+\.[XYZW]}}, KC0[2].Z
; SI-CHECK-LABEL: @i16_zext_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORD s{{[0-9]}}, s[0:1], 0xb
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @i16_zext_arg(i32 addrspace(1)* nocapture %out, i16 zeroext %in) nounwind {
entry:
%0 = zext i16 %in to i32
store i32 %0, i32 addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @i16_sext_arg
; EG-CHECK: MOV {{[ *]*}}T{{[0-9]+\.[XYZW]}}, KC0[2].Z
; SI-CHECK-LABEL: @i16_sext_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORD s{{[0-9]}}, s[0:1], 0xb
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @i16_sext_arg(i32 addrspace(1)* nocapture %out, i16 signext %in) nounwind {
entry:
%0 = sext i16 %in to i32
store i32 %0, i32 addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @i32_arg
; EG-CHECK: T{{[0-9]\.[XYZW]}}, KC0[2].Z
; SI-CHECK-LABEL: @i32_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; S_LOAD_DWORD s{{[0-9]}}, s[0:1], 0xb
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @i32_arg(i32 addrspace(1)* nocapture %out, i32 %in) nounwind {
entry:
store i32 %in, i32 addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @f32_arg
; EG-CHECK: T{{[0-9]\.[XYZW]}}, KC0[2].Z
; SI-CHECK-LABEL: @f32_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; S_LOAD_DWORD s{{[0-9]}}, s[0:1], 0xb
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @f32_arg(float addrspace(1)* nocapture %out, float %in) nounwind {
entry:
store float %in, float addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @v2i8_arg
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; SI-CHECK-LABEL: @v2i8_arg
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
define void @v2i8_arg(<2 x i8> addrspace(1)* %out, <2 x i8> %in) {
entry:
store <2 x i8> %in, <2 x i8> addrspace(1)* %out
ret void
}
; EG-CHECK-LABEL: @v2i16_arg
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; SI-CHECK-LABEL: @v2i16_arg
; SI-CHECK-DAG: BUFFER_LOAD_USHORT
; SI-CHECK-DAG: BUFFER_LOAD_USHORT
define void @v2i16_arg(<2 x i16> addrspace(1)* %out, <2 x i16> %in) {
entry:
store <2 x i16> %in, <2 x i16> addrspace(1)* %out
ret void
}
; EG-CHECK-LABEL: @v2i32_arg
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].X
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[2].W
; SI-CHECK-LABEL: @v2i32_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORDX2 s{{\[[0-9]:[0-9]\]}}, s[0:1], 0xb
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @v2i32_arg(<2 x i32> addrspace(1)* nocapture %out, <2 x i32> %in) nounwind {
entry:
store <2 x i32> %in, <2 x i32> addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @v2f32_arg
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].X
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[2].W
; SI-CHECK-LABEL: @v2f32_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORDX2 s{{\[[0-9]:[0-9]\]}}, s[0:1], 0xb
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @v2f32_arg(<2 x float> addrspace(1)* nocapture %out, <2 x float> %in) nounwind {
entry:
store <2 x float> %in, <2 x float> addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @v3i8_arg
; VTX_READ_8 T{{[0-9]}}.X, T{{[0-9]}}.X, 40
; VTX_READ_8 T{{[0-9]}}.X, T{{[0-9]}}.X, 41
; VTX_READ_8 T{{[0-9]}}.X, T{{[0-9]}}.X, 42
; SI-CHECK-LABEL: @v3i8_arg
define void @v3i8_arg(<3 x i8> addrspace(1)* nocapture %out, <3 x i8> %in) nounwind {
entry:
store <3 x i8> %in, <3 x i8> addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @v3i16_arg
; VTX_READ_16 T{{[0-9]}}.X, T{{[0-9]}}.X, 44
; VTX_READ_16 T{{[0-9]}}.X, T{{[0-9]}}.X, 46
; VTX_READ_16 T{{[0-9]}}.X, T{{[0-9]}}.X, 48
; SI-CHECK-LABEL: @v3i16_arg
define void @v3i16_arg(<3 x i16> addrspace(1)* nocapture %out, <3 x i16> %in) nounwind {
entry:
store <3 x i16> %in, <3 x i16> addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @v3i32_arg
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].W
; SI-CHECK-LABEL: @v3i32_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORDX4 s{{\[[0-9]:[0-9]+\]}}, s[0:1], 0xd
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @v3i32_arg(<3 x i32> addrspace(1)* nocapture %out, <3 x i32> %in) nounwind {
entry:
store <3 x i32> %in, <3 x i32> addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @v3f32_arg
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].W
; SI-CHECK-LABEL: @v3f32_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORDX4 s{{\[[0-9]:[0-9]+\]}}, s[0:1], 0xd
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @v3f32_arg(<3 x float> addrspace(1)* nocapture %out, <3 x float> %in) nounwind {
entry:
store <3 x float> %in, <3 x float> addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @v4i8_arg
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; SI-CHECK-LABEL: @v4i8_arg
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
define void @v4i8_arg(<4 x i8> addrspace(1)* %out, <4 x i8> %in) {
entry:
store <4 x i8> %in, <4 x i8> addrspace(1)* %out
ret void
}
; EG-CHECK-LABEL: @v4i16_arg
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; SI-CHECK-LABEL: @v4i16_arg
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
define void @v4i16_arg(<4 x i16> addrspace(1)* %out, <4 x i16> %in) {
entry:
store <4 x i16> %in, <4 x i16> addrspace(1)* %out
ret void
}
; EG-CHECK-LABEL: @v4i32_arg
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[4].X
; SI-CHECK-LABEL: @v4i32_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORDX4 s{{\[[0-9]:[0-9]\]}}, s[0:1], 0xd
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @v4i32_arg(<4 x i32> addrspace(1)* nocapture %out, <4 x i32> %in) nounwind {
entry:
store <4 x i32> %in, <4 x i32> addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @v4f32_arg
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[3].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[4].X
; SI-CHECK-LABEL: @v4f32_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORDX4 s{{\[[0-9]:[0-9]\]}}, s[0:1], 0xd
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @v4f32_arg(<4 x float> addrspace(1)* nocapture %out, <4 x float> %in) nounwind {
entry:
store <4 x float> %in, <4 x float> addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @v8i8_arg
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; SI-CHECK-LABEL: @v8i8_arg
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
define void @v8i8_arg(<8 x i8> addrspace(1)* %out, <8 x i8> %in) {
entry:
store <8 x i8> %in, <8 x i8> addrspace(1)* %out
ret void
}
; EG-CHECK-LABEL: @v8i16_arg
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; SI-CHECK-LABEL: @v8i16_arg
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
define void @v8i16_arg(<8 x i16> addrspace(1)* %out, <8 x i16> %in) {
entry:
store <8 x i16> %in, <8 x i16> addrspace(1)* %out
ret void
}
; EG-CHECK-LABEL: @v8i32_arg
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[4].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[4].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[4].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[5].X
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[5].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[5].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[5].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[6].X
; SI-CHECK-LABEL: @v8i32_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORDX8 s{{\[[0-9]:[0-9]+\]}}, s[0:1], 0x11
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @v8i32_arg(<8 x i32> addrspace(1)* nocapture %out, <8 x i32> %in) nounwind {
entry:
store <8 x i32> %in, <8 x i32> addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @v8f32_arg
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[4].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[4].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[4].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[5].X
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[5].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[5].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[5].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[6].X
; SI-CHECK-LABEL: @v8f32_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORDX8 s{{\[[0-9]:[0-9]+\]}}, s[0:1], 0x11
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @v8f32_arg(<8 x float> addrspace(1)* nocapture %out, <8 x float> %in) nounwind {
entry:
store <8 x float> %in, <8 x float> addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @v16i8_arg
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; EG-CHECK: VTX_READ_8
; SI-CHECK-LABEL: @v16i8_arg
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
; SI-CHECK: BUFFER_LOAD_UBYTE
define void @v16i8_arg(<16 x i8> addrspace(1)* %out, <16 x i8> %in) {
entry:
store <16 x i8> %in, <16 x i8> addrspace(1)* %out
ret void
}
; EG-CHECK-LABEL: @v16i16_arg
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; EG-CHECK: VTX_READ_16
; SI-CHECK-LABEL: @v16i16_arg
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
; SI-CHECK: BUFFER_LOAD_USHORT
define void @v16i16_arg(<16 x i16> addrspace(1)* %out, <16 x i16> %in) {
entry:
store <16 x i16> %in, <16 x i16> addrspace(1)* %out
ret void
}
; EG-CHECK-LABEL: @v16i32_arg
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[6].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[6].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[6].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[7].X
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[7].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[7].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[7].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[8].X
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[8].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[8].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[8].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[9].X
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[9].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[9].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[9].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[10].X
; SI-CHECK-LABEL: @v16i32_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORDX16 s{{\[[0-9]:[0-9]+\]}}, s[0:1], 0x19
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @v16i32_arg(<16 x i32> addrspace(1)* nocapture %out, <16 x i32> %in) nounwind {
entry:
store <16 x i32> %in, <16 x i32> addrspace(1)* %out, align 4
ret void
}
; EG-CHECK-LABEL: @v16f32_arg
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[6].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[6].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[6].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[7].X
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[7].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[7].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[7].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[8].X
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[8].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[8].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[8].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[9].X
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[9].Y
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[9].Z
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[9].W
; EG-CHECK-DAG: T{{[0-9]\.[XYZW]}}, KC0[10].X
; SI-CHECK-LABEL: @v16f32_arg
R600/SI: Print more immediates in hex format Print in decimal for inline immediates, and hex otherwise. Use hex always for offsets in addressing offsets. This approximately matches what the shader compiler does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206335 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-15 22:32:49 +00:00
; SI-CHECK: S_LOAD_DWORDX16 s{{\[[0-9]:[0-9]+\]}}, s[0:1], 0x19
R600: Fix handling of vector kernel arguments The SelectionDAGBuilder was promoting vector kernel arguments to legal types, but this won't work for R600 and SI since kernel arguments are stored in memory and can't be promoted. In order to handle vector arguments correctly we need to look at the original types from the LLVM IR function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@193215 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-23 00:44:32 +00:00
define void @v16f32_arg(<16 x float> addrspace(1)* nocapture %out, <16 x float> %in) nounwind {
entry:
store <16 x float> %in, <16 x float> addrspace(1)* %out, align 4
ret void
}
R600: Correctly set the src value offset for scalarized kernel args This for some reason fixes v1i64 kernel arguments on pre-SI. This currently breaks some other cases in the kernel-args.ll test for R600, but I'm not particularly confident in the new output. VTX_READ_* are not used for some of the scalarized cases, and the code reading from the constant buffer doesn't make much sense to me. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@215564 91177308-0d34-0410-b5e6-96231b3b80d8
2014-08-13 18:14:11 +00:00
; FUNC-LABEL: @kernel_arg_i64
; SI: S_LOAD_DWORDX2
; SI: S_LOAD_DWORDX2
; SI: BUFFER_STORE_DWORDX2
define void @kernel_arg_i64(i64 addrspace(1)* %out, i64 %a) nounwind {
store i64 %a, i64 addrspace(1)* %out, align 8
ret void
}
; XFUNC-LABEL: @kernel_arg_v1i64
; XSI: S_LOAD_DWORDX2
; XSI: S_LOAD_DWORDX2
; XSI: BUFFER_STORE_DWORDX2
; define void @kernel_arg_v1i64(<1 x i64> addrspace(1)* %out, <1 x i64> %a) nounwind {
; store <1 x i64> %a, <1 x i64> addrspace(1)* %out, align 8
; ret void
; }
Reference in New Issue Copy Permalink