mirror of
https://github.com/c64scene-ar/llvm-6502.git
synced 2024-12-15 04:30:12 +00:00
86245071b5
Print the range of registers used with a single letter prefix. This better matches what the shader compiler produces and is overall less obnoxious than concatenating all of the subregister names together. Instead of SGPR0, it will print s0. Instead of SGPR0_SGPR1, it will print s[0:1] and so on. There doesn't appear to be a straightforward way to get the actual register info in the InstPrinter, so this parses the generated name to print with the new syntax. The required test changes are pretty nasty, and register matching regexes are now worse. Since there isn't a way to add to a variable in FileCheck, some of the tests now don't check the exact number of registers used, but I don't think that will be a real problem. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@194443 91177308-0d34-0410-b5e6-96231b3b80d8
51 lines
1.7 KiB
LLVM
51 lines
1.7 KiB
LLVM
; RUN: llc < %s -march=r600 -mcpu=redwood | FileCheck --check-prefix=EG-CHECK %s
|
|
; RUN: llc < %s -march=r600 -mcpu=verde -verify-machineinstrs | FileCheck --check-prefix=SI-CHECK %s
|
|
; RUN: llc < %s -march=r600 -mcpu=bonaire -verify-machineinstrs | FileCheck --check-prefix=CI-CHECK %s
|
|
|
|
@local_memory.local_mem = internal addrspace(3) unnamed_addr global [128 x i32] zeroinitializer, align 4
|
|
|
|
; EG-CHECK-LABEL: @local_memory
|
|
; SI-CHECK-LABEL: @local_memory
|
|
; CI-CHECK-LABEL: @local_memory
|
|
|
|
; Check that the LDS size emitted correctly
|
|
; EG-CHECK: .long 166120
|
|
; EG-CHECK-NEXT: .long 128
|
|
; SI-CHECK: .long 47180
|
|
; SI-CHECK-NEXT: .long 65536
|
|
; CI-CHECK: .long 47180
|
|
; CI-CHECK-NEXT: .long 32768
|
|
|
|
; EG-CHECK: LDS_WRITE
|
|
; SI-CHECK_NOT: S_WQM_B64
|
|
; SI-CHECK: DS_WRITE_B32 0
|
|
|
|
; GROUP_BARRIER must be the last instruction in a clause
|
|
; EG-CHECK: GROUP_BARRIER
|
|
; EG-CHECK-NEXT: ALU clause
|
|
; SI-CHECK: S_BARRIER
|
|
|
|
; EG-CHECK: LDS_READ_RET
|
|
; SI-CHECK: DS_READ_B32 {{v[0-9]+}}, 0
|
|
|
|
define void @local_memory(i32 addrspace(1)* %out) {
|
|
entry:
|
|
%y.i = call i32 @llvm.r600.read.tidig.x() #0
|
|
%arrayidx = getelementptr inbounds [128 x i32] addrspace(3)* @local_memory.local_mem, i32 0, i32 %y.i
|
|
store i32 %y.i, i32 addrspace(3)* %arrayidx, align 4
|
|
%add = add nsw i32 %y.i, 1
|
|
%cmp = icmp eq i32 %add, 16
|
|
%.add = select i1 %cmp, i32 0, i32 %add
|
|
call void @llvm.AMDGPU.barrier.local()
|
|
%arrayidx1 = getelementptr inbounds [128 x i32] addrspace(3)* @local_memory.local_mem, i32 0, i32 %.add
|
|
%0 = load i32 addrspace(3)* %arrayidx1, align 4
|
|
%arrayidx2 = getelementptr inbounds i32 addrspace(1)* %out, i32 %y.i
|
|
store i32 %0, i32 addrspace(1)* %arrayidx2, align 4
|
|
ret void
|
|
}
|
|
|
|
declare i32 @llvm.r600.read.tidig.x() #0
|
|
declare void @llvm.AMDGPU.barrier.local()
|
|
|
|
attributes #0 = { readnone }
|