aboutsummaryrefslogtreecommitdiffstats
path: root/lib/Target/R600/AMDGPUCallingConv.td
diff options
context:
space:
mode:
Diffstat (limited to 'lib/Target/R600/AMDGPUCallingConv.td')
-rw-r--r--lib/Target/R600/AMDGPUCallingConv.td18
1 files changed, 11 insertions, 7 deletions
diff --git a/lib/Target/R600/AMDGPUCallingConv.td b/lib/Target/R600/AMDGPUCallingConv.td
index 9c30515..e57b5cd 100644
--- a/lib/Target/R600/AMDGPUCallingConv.td
+++ b/lib/Target/R600/AMDGPUCallingConv.td
@@ -32,17 +32,21 @@ def CC_SI : CallingConv<[
VGPR8, VGPR9, VGPR10, VGPR11, VGPR12, VGPR13, VGPR14, VGPR15,
VGPR16, VGPR17, VGPR18, VGPR19, VGPR20, VGPR21, VGPR22, VGPR23,
VGPR24, VGPR25, VGPR26, VGPR27, VGPR28, VGPR29, VGPR30, VGPR31
- ]>>>,
+ ]>>>
+
+]>;
- // This is the default for i64 values.
- // XXX: We should change this once clang understands the CC_AMDGPU.
- CCIfType<[i64], CCAssignToRegWithShadow<
- [ SGPR0, SGPR2, SGPR4, SGPR6, SGPR8, SGPR10, SGPR12, SGPR14 ],
- [ SGPR1, SGPR3, SGPR5, SGPR7, SGPR9, SGPR11, SGPR13, SGPR15 ]
- >>
+// Calling convention for SI compute kernels
+def CC_SI_Kernel : CallingConv<[
+ CCIfType<[i64], CCAssignToStack <8, 4>>,
+ CCIfType<[i32, f32], CCAssignToStack <4, 4>>,
+ CCIfType<[i16], CCAssignToStack <2, 4>>,
+ CCIfType<[i8], CCAssignToStack <1, 4>>
]>;
def CC_AMDGPU : CallingConv<[
+ CCIf<"State.getMachineFunction().getInfo<SIMachineFunctionInfo>()->"#
+ "ShaderType == ShaderType::COMPUTE", CCDelegateTo<CC_SI_Kernel>>,
CCIf<"State.getTarget().getSubtarget<AMDGPUSubtarget>().device()"#
"->getGeneration() == AMDGPUDeviceInfo::HD7XXX", CCDelegateTo<CC_SI>>
]>;