llvm-6502/lib/Target/R600/AMDGPUCallingConv.td
Marek Olsak f6518f6034 R600/SI: allow 5 more input SGPRs to a shader
Our OpenGL driver needs 22 SGPRs (16 user SGPRs + 6 streamout non-user SGPRs).

Signed-off-by: Marek Olšák <marek.olsak@amd.com>

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@207990 91177308-0d34-0410-b5e6-96231b3b80d8
2014-05-05 19:30:54 +00:00

75 lines
3.0 KiB
TableGen

//===---- AMDCallingConv.td - Calling Conventions for Radeon GPUs ---------===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This describes the calling conventions for the AMD Radeon GPUs.
//
//===----------------------------------------------------------------------===//
// Inversion of CCIfInReg
class CCIfNotInReg<CCAction A> : CCIf<"!ArgFlags.isInReg()", A> {}
// Calling convention for SI
def CC_SI : CallingConv<[
CCIfInReg<CCIfType<[f32, i32] , CCAssignToReg<[
SGPR0, SGPR1, SGPR2, SGPR3, SGPR4, SGPR5, SGPR6, SGPR7,
SGPR8, SGPR9, SGPR10, SGPR11, SGPR12, SGPR13, SGPR14, SGPR15,
SGPR16, SGPR17, SGPR18, SGPR19, SGPR20, SGPR21
]>>>,
CCIfInReg<CCIfType<[i64] , CCAssignToRegWithShadow<
[ SGPR0, SGPR2, SGPR4, SGPR6, SGPR8, SGPR10, SGPR12, SGPR14 ],
[ SGPR1, SGPR3, SGPR5, SGPR7, SGPR9, SGPR11, SGPR13, SGPR15 ]
>>>,
CCIfNotInReg<CCIfType<[f32, i32] , CCAssignToReg<[
VGPR0, VGPR1, VGPR2, VGPR3, VGPR4, VGPR5, VGPR6, VGPR7,
VGPR8, VGPR9, VGPR10, VGPR11, VGPR12, VGPR13, VGPR14, VGPR15,
VGPR16, VGPR17, VGPR18, VGPR19, VGPR20, VGPR21, VGPR22, VGPR23,
VGPR24, VGPR25, VGPR26, VGPR27, VGPR28, VGPR29, VGPR30, VGPR31
]>>>,
CCIfByVal<CCIfType<[i64] , CCAssignToRegWithShadow<
[ SGPR0, SGPR2, SGPR4, SGPR6, SGPR8, SGPR10, SGPR12, SGPR14 ],
[ SGPR1, SGPR3, SGPR5, SGPR7, SGPR9, SGPR11, SGPR13, SGPR15 ]
>>>
]>;
// Calling convention for R600
def CC_R600 : CallingConv<[
CCIfInReg<CCIfType<[v4f32, v4i32] , CCAssignToReg<[
T0_XYZW, T1_XYZW, T2_XYZW, T3_XYZW, T4_XYZW, T5_XYZW, T6_XYZW, T7_XYZW,
T8_XYZW, T9_XYZW, T10_XYZW, T11_XYZW, T12_XYZW, T13_XYZW, T14_XYZW, T15_XYZW,
T16_XYZW, T17_XYZW, T18_XYZW, T19_XYZW, T20_XYZW, T21_XYZW, T22_XYZW,
T23_XYZW, T24_XYZW, T25_XYZW, T26_XYZW, T27_XYZW, T28_XYZW, T29_XYZW,
T30_XYZW, T31_XYZW, T32_XYZW
]>>>
]>;
// Calling convention for compute kernels
def CC_AMDGPU_Kernel : CallingConv<[
CCCustom<"allocateStack">
]>;
def CC_AMDGPU : CallingConv<[
CCIf<"State.getTarget().getSubtarget<AMDGPUSubtarget>().getGeneration() >= "
"AMDGPUSubtarget::SOUTHERN_ISLANDS && "
"State.getMachineFunction().getInfo<SIMachineFunctionInfo>()->"#
"ShaderType == ShaderType::COMPUTE", CCDelegateTo<CC_AMDGPU_Kernel>>,
CCIf<"State.getTarget().getSubtarget<AMDGPUSubtarget>().getGeneration() < "
"AMDGPUSubtarget::SOUTHERN_ISLANDS && "
"State.getMachineFunction().getInfo<R600MachineFunctionInfo>()->"
"ShaderType == ShaderType::COMPUTE", CCDelegateTo<CC_AMDGPU_Kernel>>,
CCIf<"State.getTarget().getSubtarget<AMDGPUSubtarget>()"#
".getGeneration() >= AMDGPUSubtarget::SOUTHERN_ISLANDS", CCDelegateTo<CC_SI>>,
CCIf<"State.getTarget().getSubtarget<AMDGPUSubtarget>()"#
".getGeneration() < AMDGPUSubtarget::SOUTHERN_ISLANDS", CCDelegateTo<CC_R600>>
]>;