mirror of
https://github.com/c64scene-ar/llvm-6502.git
synced 2024-11-06 21:05:51 +00:00
95eb45c5d9
R600 was using a clamped version of rsq, but SI was not. Add a new rsq_clamped intrinsic and use them consistently. It's unclear to me from the documentation what behavior the R600 instructions have, so I assume they have the legacy behavior described by the SI documents. For R600, use RECIPSQRT_IEEE for both llvm.AMDGPU.rsq.legacy and llvm.AMDGPU.rsq. R600 also has RECIPSQRT_FF, which I'm not sure how it fits in here. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@211637 91177308-0d34-0410-b5e6-96231b3b80d8
179 lines
6.3 KiB
TableGen
179 lines
6.3 KiB
TableGen
//===-- AMDGPUInstrInfo.td - AMDGPU DAG nodes --------------*- tablegen -*-===//
|
|
//
|
|
// The LLVM Compiler Infrastructure
|
|
//
|
|
// This file is distributed under the University of Illinois Open Source
|
|
// License. See LICENSE.TXT for details.
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
//
|
|
// This file contains DAG node defintions for the AMDGPU target.
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
//===----------------------------------------------------------------------===//
|
|
// AMDGPU DAG Profiles
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
def AMDGPUDTIntTernaryOp : SDTypeProfile<1, 3, [
|
|
SDTCisSameAs<0, 1>, SDTCisSameAs<0, 2>, SDTCisInt<0>, SDTCisInt<3>
|
|
]>;
|
|
|
|
def AMDGPUTrigPreOp : SDTypeProfile<1, 2,
|
|
[SDTCisSameAs<0, 1>, SDTCisFP<0>, SDTCisInt<2>]
|
|
>;
|
|
|
|
def AMDGPUDivScaleOp : SDTypeProfile<2, 3,
|
|
[SDTCisFP<0>, SDTCisInt<1>, SDTCisSameAs<0, 2>, SDTCisSameAs<0, 3>, SDTCisSameAs<0, 4>]
|
|
>;
|
|
|
|
//===----------------------------------------------------------------------===//
|
|
// AMDGPU DAG Nodes
|
|
//
|
|
|
|
// This argument to this node is a dword address.
|
|
def AMDGPUdwordaddr : SDNode<"AMDGPUISD::DWORDADDR", SDTIntUnaryOp>;
|
|
|
|
// out = a - floor(a)
|
|
def AMDGPUfract : SDNode<"AMDGPUISD::FRACT", SDTFPUnaryOp>;
|
|
|
|
// out = 1.0 / a
|
|
def AMDGPUrcp : SDNode<"AMDGPUISD::RCP", SDTFPUnaryOp>;
|
|
|
|
// out = 1.0 / sqrt(a)
|
|
def AMDGPUrsq : SDNode<"AMDGPUISD::RSQ", SDTFPUnaryOp>;
|
|
|
|
// out = 1.0 / sqrt(a)
|
|
def AMDGPUrsq_legacy : SDNode<"AMDGPUISD::RSQ_LEGACY", SDTFPUnaryOp>;
|
|
|
|
// out = 1.0 / sqrt(a) result clamped to +/- max_float.
|
|
def AMDGPUrsq_clamped : SDNode<"AMDGPUISD::RSQ_CLAMPED", SDTFPUnaryOp>;
|
|
|
|
// out = max(a, b) a and b are floats
|
|
def AMDGPUfmax : SDNode<"AMDGPUISD::FMAX", SDTFPBinOp,
|
|
[SDNPCommutative, SDNPAssociative]
|
|
>;
|
|
|
|
def AMDGPUclamp : SDNode<"AMDGPUISD::CLAMP", SDTFPTernaryOp, []>;
|
|
|
|
// out = max(a, b) a and b are signed ints
|
|
def AMDGPUsmax : SDNode<"AMDGPUISD::SMAX", SDTIntBinOp,
|
|
[SDNPCommutative, SDNPAssociative]
|
|
>;
|
|
|
|
// out = max(a, b) a and b are unsigned ints
|
|
def AMDGPUumax : SDNode<"AMDGPUISD::UMAX", SDTIntBinOp,
|
|
[SDNPCommutative, SDNPAssociative]
|
|
>;
|
|
|
|
// out = min(a, b) a and b are floats
|
|
def AMDGPUfmin : SDNode<"AMDGPUISD::FMIN", SDTFPBinOp,
|
|
[SDNPCommutative, SDNPAssociative]
|
|
>;
|
|
|
|
// out = min(a, b) a snd b are signed ints
|
|
def AMDGPUsmin : SDNode<"AMDGPUISD::SMIN", SDTIntBinOp,
|
|
[SDNPCommutative, SDNPAssociative]
|
|
>;
|
|
|
|
// out = min(a, b) a and b are unsigned ints
|
|
def AMDGPUumin : SDNode<"AMDGPUISD::UMIN", SDTIntBinOp,
|
|
[SDNPCommutative, SDNPAssociative]
|
|
>;
|
|
|
|
|
|
def AMDGPUcvt_f32_ubyte0 : SDNode<"AMDGPUISD::CVT_F32_UBYTE0",
|
|
SDTIntToFPOp, []>;
|
|
def AMDGPUcvt_f32_ubyte1 : SDNode<"AMDGPUISD::CVT_F32_UBYTE1",
|
|
SDTIntToFPOp, []>;
|
|
def AMDGPUcvt_f32_ubyte2 : SDNode<"AMDGPUISD::CVT_F32_UBYTE2",
|
|
SDTIntToFPOp, []>;
|
|
def AMDGPUcvt_f32_ubyte3 : SDNode<"AMDGPUISD::CVT_F32_UBYTE3",
|
|
SDTIntToFPOp, []>;
|
|
|
|
|
|
// urecip - This operation is a helper for integer division, it returns the
|
|
// result of 1 / a as a fractional unsigned integer.
|
|
// out = (2^32 / a) + e
|
|
// e is rounding error
|
|
def AMDGPUurecip : SDNode<"AMDGPUISD::URECIP", SDTIntUnaryOp>;
|
|
|
|
// Special case divide preop and flags.
|
|
def AMDGPUdiv_scale : SDNode<"AMDGPUISD::DIV_SCALE", AMDGPUDivScaleOp>;
|
|
|
|
// Special case divide FMA with scale and flags (src0 = Quotient,
|
|
// src1 = Denominator, src2 = Numerator).
|
|
def AMDGPUdiv_fmas : SDNode<"AMDGPUISD::DIV_FMAS", SDTFPTernaryOp>;
|
|
|
|
// Single or double precision division fixup.
|
|
// Special case divide fixup and flags(src0 = Quotient, src1 =
|
|
// Denominator, src2 = Numerator).
|
|
def AMDGPUdiv_fixup : SDNode<"AMDGPUISD::DIV_FIXUP", SDTFPTernaryOp>;
|
|
|
|
// Look Up 2.0 / pi src0 with segment select src1[4:0]
|
|
def AMDGPUtrig_preop : SDNode<"AMDGPUISD::TRIG_PREOP", AMDGPUTrigPreOp>;
|
|
|
|
def AMDGPUregister_load : SDNode<"AMDGPUISD::REGISTER_LOAD",
|
|
SDTypeProfile<1, 2, [SDTCisPtrTy<1>, SDTCisInt<2>]>,
|
|
[SDNPHasChain, SDNPMayLoad]>;
|
|
|
|
def AMDGPUregister_store : SDNode<"AMDGPUISD::REGISTER_STORE",
|
|
SDTypeProfile<0, 3, [SDTCisPtrTy<1>, SDTCisInt<2>]>,
|
|
[SDNPHasChain, SDNPMayStore]>;
|
|
|
|
// MSKOR instructions are atomic memory instructions used mainly for storing
|
|
// 8-bit and 16-bit values. The definition is:
|
|
//
|
|
// MSKOR(dst, mask, src) MEM[dst] = ((MEM[dst] & ~mask) | src)
|
|
//
|
|
// src0: vec4(src, 0, 0, mask)
|
|
// src1: dst - rat offset (aka pointer) in dwords
|
|
def AMDGPUstore_mskor : SDNode<"AMDGPUISD::STORE_MSKOR",
|
|
SDTypeProfile<0, 2, []>,
|
|
[SDNPHasChain, SDNPMayStore, SDNPMemOperand]>;
|
|
|
|
def AMDGPUround : SDNode<"ISD::FROUND",
|
|
SDTypeProfile<1, 1, [SDTCisFP<0>, SDTCisSameAs<0,1>]>>;
|
|
|
|
def AMDGPUbfe_u32 : SDNode<"AMDGPUISD::BFE_U32", AMDGPUDTIntTernaryOp>;
|
|
def AMDGPUbfe_i32 : SDNode<"AMDGPUISD::BFE_I32", AMDGPUDTIntTernaryOp>;
|
|
def AMDGPUbfi : SDNode<"AMDGPUISD::BFI", AMDGPUDTIntTernaryOp>;
|
|
def AMDGPUbfm : SDNode<"AMDGPUISD::BFM", SDTIntBinOp>;
|
|
|
|
def AMDGPUbrev : SDNode<"AMDGPUISD::BREV", SDTIntUnaryOp>;
|
|
|
|
// Signed and unsigned 24-bit mulitply. The highest 8-bits are ignore when
|
|
// performing the mulitply. The result is a 32-bit value.
|
|
def AMDGPUmul_u24 : SDNode<"AMDGPUISD::MUL_U24", SDTIntBinOp,
|
|
[SDNPCommutative]
|
|
>;
|
|
def AMDGPUmul_i24 : SDNode<"AMDGPUISD::MUL_I24", SDTIntBinOp,
|
|
[SDNPCommutative]
|
|
>;
|
|
|
|
def AMDGPUmad_u24 : SDNode<"AMDGPUISD::MAD_U24", AMDGPUDTIntTernaryOp,
|
|
[]
|
|
>;
|
|
def AMDGPUmad_i24 : SDNode<"AMDGPUISD::MAD_I24", AMDGPUDTIntTernaryOp,
|
|
[]
|
|
>;
|
|
|
|
//===----------------------------------------------------------------------===//
|
|
// Flow Control Profile Types
|
|
//===----------------------------------------------------------------------===//
|
|
// Branch instruction where second and third are basic blocks
|
|
def SDTIL_BRCond : SDTypeProfile<0, 2, [
|
|
SDTCisVT<0, OtherVT>
|
|
]>;
|
|
|
|
//===----------------------------------------------------------------------===//
|
|
// Flow Control DAG Nodes
|
|
//===----------------------------------------------------------------------===//
|
|
def IL_brcond : SDNode<"AMDGPUISD::BRANCH_COND", SDTIL_BRCond, [SDNPHasChain]>;
|
|
|
|
//===----------------------------------------------------------------------===//
|
|
// Call/Return DAG Nodes
|
|
//===----------------------------------------------------------------------===//
|
|
def IL_retflag : SDNode<"AMDGPUISD::RET_FLAG", SDTNone,
|
|
[SDNPHasChain, SDNPOptInGlue]>;
|