RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-19 19:12:56 +02:00
Code Issues Projects Releases Wiki Activity
2699008644
llvm-mirror/test/CodeGen/AMDGPU/private-memory-r600.ll
Matt Arsenault 2699008644 AMDGPU: Fix emitting invalid workitem intrinsics for HSA 
The AMDGPUPromoteAlloca pass was emitting the read.local.size
calls, which with HSA was incorrectly selected to reading from
the offset mesa uses off of the kernarg pointer.

Error on intrinsics which aren't supported by HSA, and start
emitting the correct IR to read the workgroup size
out of the dispatch pointer.

Also initialize the pass so it can be tested with opt, and
start moving towards not depending on the subtarget as an
argument.

Start emitting errors for the intrinsics not handled with HSA.

llvm-svn: 259297
2016-01-30 05:19:45 +00:00

299 lines
11 KiB
LLVM
Raw Blame History

; RUN: llc -march=r600 -mcpu=redwood < %s | FileCheck %s -check-prefix=R600 -check-prefix=FUNC
; RUN: opt -S -mtriple=r600-unknown-unknown -mcpu=redwood -amdgpu-promote-alloca < %s | FileCheck -check-prefix=OPT %s
declare i32 @llvm.r600.read.tidig.x() nounwind readnone
; FUNC-LABEL: {{^}}mova_same_clause:
; R600: LDS_WRITE
; R600: LDS_WRITE
; R600: LDS_READ
; R600: LDS_READ
; OPT: call i32 @llvm.r600.read.local.size.y(), !range !0
; OPT: call i32 @llvm.r600.read.local.size.z(), !range !0
; OPT: call i32 @llvm.r600.read.tidig.x(), !range !0
; OPT: call i32 @llvm.r600.read.tidig.y(), !range !0
; OPT: call i32 @llvm.r600.read.tidig.z(), !range !0
define void @mova_same_clause(i32 addrspace(1)* nocapture %out, i32 addrspace(1)* nocapture %in) {
entry:
%stack = alloca [5 x i32], align 4
%0 = load i32, i32 addrspace(1)* %in, align 4
%arrayidx1 = getelementptr inbounds [5 x i32], [5 x i32]* %stack, i32 0, i32 %0
store i32 4, i32* %arrayidx1, align 4
%arrayidx2 = getelementptr inbounds i32, i32 addrspace(1)* %in, i32 1
%1 = load i32, i32 addrspace(1)* %arrayidx2, align 4
%arrayidx3 = getelementptr inbounds [5 x i32], [5 x i32]* %stack, i32 0, i32 %1
store i32 5, i32* %arrayidx3, align 4
%arrayidx10 = getelementptr inbounds [5 x i32], [5 x i32]* %stack, i32 0, i32 0
%2 = load i32, i32* %arrayidx10, align 4
store i32 %2, i32 addrspace(1)* %out, align 4
%arrayidx12 = getelementptr inbounds [5 x i32], [5 x i32]* %stack, i32 0, i32 1
%3 = load i32, i32* %arrayidx12
%arrayidx13 = getelementptr inbounds i32, i32 addrspace(1)* %out, i32 1
store i32 %3, i32 addrspace(1)* %arrayidx13
ret void
}
; This test checks that the stack offset is calculated correctly for structs.
; All register loads/stores should be optimized away, so there shouldn't be
; any MOVA instructions.
;
; XXX: This generated code has unnecessary MOVs, we should be able to optimize
; this.
; FUNC-LABEL: {{^}}multiple_structs:
; R600-NOT: MOVA_INT
%struct.point = type { i32, i32 }
define void @multiple_structs(i32 addrspace(1)* %out) {
entry:
%a = alloca %struct.point
%b = alloca %struct.point
%a.x.ptr = getelementptr %struct.point, %struct.point* %a, i32 0, i32 0
%a.y.ptr = getelementptr %struct.point, %struct.point* %a, i32 0, i32 1
%b.x.ptr = getelementptr %struct.point, %struct.point* %b, i32 0, i32 0
%b.y.ptr = getelementptr %struct.point, %struct.point* %b, i32 0, i32 1
store i32 0, i32* %a.x.ptr
store i32 1, i32* %a.y.ptr
store i32 2, i32* %b.x.ptr
store i32 3, i32* %b.y.ptr
%a.indirect.ptr = getelementptr %struct.point, %struct.point* %a, i32 0, i32 0
%b.indirect.ptr = getelementptr %struct.point, %struct.point* %b, i32 0, i32 0
%a.indirect = load i32, i32* %a.indirect.ptr
%b.indirect = load i32, i32* %b.indirect.ptr
%0 = add i32 %a.indirect, %b.indirect
store i32 %0, i32 addrspace(1)* %out
ret void
}
; Test direct access of a private array inside a loop. The private array
; loads and stores should be lowered to copies, so there shouldn't be any
; MOVA instructions.
; FUNC-LABEL: {{^}}direct_loop:
; R600-NOT: MOVA_INT
define void @direct_loop(i32 addrspace(1)* %out, i32 addrspace(1)* %in) {
entry:
%prv_array_const = alloca [2 x i32]
%prv_array = alloca [2 x i32]
%a = load i32, i32 addrspace(1)* %in
%b_src_ptr = getelementptr i32, i32 addrspace(1)* %in, i32 1
%b = load i32, i32 addrspace(1)* %b_src_ptr
%a_dst_ptr = getelementptr [2 x i32], [2 x i32]* %prv_array_const, i32 0, i32 0
store i32 %a, i32* %a_dst_ptr
%b_dst_ptr = getelementptr [2 x i32], [2 x i32]* %prv_array_const, i32 0, i32 1
store i32 %b, i32* %b_dst_ptr
br label %for.body
for.body:
%inc = phi i32 [0, %entry], [%count, %for.body]
%x_ptr = getelementptr [2 x i32], [2 x i32]* %prv_array_const, i32 0, i32 0
%x = load i32, i32* %x_ptr
%y_ptr = getelementptr [2 x i32], [2 x i32]* %prv_array, i32 0, i32 0
%y = load i32, i32* %y_ptr
%xy = add i32 %x, %y
store i32 %xy, i32* %y_ptr
%count = add i32 %inc, 1
%done = icmp eq i32 %count, 4095
br i1 %done, label %for.end, label %for.body
for.end:
%value_ptr = getelementptr [2 x i32], [2 x i32]* %prv_array, i32 0, i32 0
%value = load i32, i32* %value_ptr
store i32 %value, i32 addrspace(1)* %out
ret void
}
; FUNC-LABEL: {{^}}short_array:
; R600: MOVA_INT
define void @short_array(i32 addrspace(1)* %out, i32 %index) {
entry:
%0 = alloca [2 x i16]
%1 = getelementptr [2 x i16], [2 x i16]* %0, i32 0, i32 0
%2 = getelementptr [2 x i16], [2 x i16]* %0, i32 0, i32 1
store i16 0, i16* %1
store i16 1, i16* %2
%3 = getelementptr [2 x i16], [2 x i16]* %0, i32 0, i32 %index
%4 = load i16, i16* %3
%5 = sext i16 %4 to i32
store i32 %5, i32 addrspace(1)* %out
ret void
}
; FUNC-LABEL: {{^}}char_array:
; R600: MOVA_INT
define void @char_array(i32 addrspace(1)* %out, i32 %index) {
entry:
%0 = alloca [2 x i8]
%1 = getelementptr [2 x i8], [2 x i8]* %0, i32 0, i32 0
%2 = getelementptr [2 x i8], [2 x i8]* %0, i32 0, i32 1
store i8 0, i8* %1
store i8 1, i8* %2
%3 = getelementptr [2 x i8], [2 x i8]* %0, i32 0, i32 %index
%4 = load i8, i8* %3
%5 = sext i8 %4 to i32
store i32 %5, i32 addrspace(1)* %out
ret void
}
; Make sure we don't overwrite workitem information with private memory
; FUNC-LABEL: {{^}}work_item_info:
; R600-NOT: MOV T0.X
; Additional check in case the move ends up in the last slot
; R600-NOT: MOV * TO.X
define void @work_item_info(i32 addrspace(1)* %out, i32 %in) {
entry:
%0 = alloca [2 x i32]
%1 = getelementptr [2 x i32], [2 x i32]* %0, i32 0, i32 0
%2 = getelementptr [2 x i32], [2 x i32]* %0, i32 0, i32 1
store i32 0, i32* %1
store i32 1, i32* %2
%3 = getelementptr [2 x i32], [2 x i32]* %0, i32 0, i32 %in
%4 = load i32, i32* %3
%5 = call i32 @llvm.r600.read.tidig.x()
%6 = add i32 %4, %5
store i32 %6, i32 addrspace(1)* %out
ret void
}
; Test that two stack objects are not stored in the same register
; The second stack object should be in T3.X
; FUNC-LABEL: {{^}}no_overlap:
; R600_CHECK: MOV
; R600_CHECK: [[CHAN:[XYZW]]]+
; R600-NOT: [[CHAN]]+
define void @no_overlap(i32 addrspace(1)* %out, i32 %in) {
entry:
%0 = alloca [3 x i8], align 1
%1 = alloca [2 x i8], align 1
%2 = getelementptr [3 x i8], [3 x i8]* %0, i32 0, i32 0
%3 = getelementptr [3 x i8], [3 x i8]* %0, i32 0, i32 1
%4 = getelementptr [3 x i8], [3 x i8]* %0, i32 0, i32 2
%5 = getelementptr [2 x i8], [2 x i8]* %1, i32 0, i32 0
%6 = getelementptr [2 x i8], [2 x i8]* %1, i32 0, i32 1
store i8 0, i8* %2
store i8 1, i8* %3
store i8 2, i8* %4
store i8 1, i8* %5
store i8 0, i8* %6
%7 = getelementptr [3 x i8], [3 x i8]* %0, i32 0, i32 %in
%8 = getelementptr [2 x i8], [2 x i8]* %1, i32 0, i32 %in
%9 = load i8, i8* %7
%10 = load i8, i8* %8
%11 = add i8 %9, %10
%12 = sext i8 %11 to i32
store i32 %12, i32 addrspace(1)* %out
ret void
}
define void @char_array_array(i32 addrspace(1)* %out, i32 %index) {
entry:
%alloca = alloca [2 x [2 x i8]]
%gep0 = getelementptr [2 x [2 x i8]], [2 x [2 x i8]]* %alloca, i32 0, i32 0, i32 0
%gep1 = getelementptr [2 x [2 x i8]], [2 x [2 x i8]]* %alloca, i32 0, i32 0, i32 1
store i8 0, i8* %gep0
store i8 1, i8* %gep1
%gep2 = getelementptr [2 x [2 x i8]], [2 x [2 x i8]]* %alloca, i32 0, i32 0, i32 %index
%load = load i8, i8* %gep2
%sext = sext i8 %load to i32
store i32 %sext, i32 addrspace(1)* %out
ret void
}
define void @i32_array_array(i32 addrspace(1)* %out, i32 %index) {
entry:
%alloca = alloca [2 x [2 x i32]]
%gep0 = getelementptr [2 x [2 x i32]], [2 x [2 x i32]]* %alloca, i32 0, i32 0, i32 0
%gep1 = getelementptr [2 x [2 x i32]], [2 x [2 x i32]]* %alloca, i32 0, i32 0, i32 1
store i32 0, i32* %gep0
store i32 1, i32* %gep1
%gep2 = getelementptr [2 x [2 x i32]], [2 x [2 x i32]]* %alloca, i32 0, i32 0, i32 %index
%load = load i32, i32* %gep2
store i32 %load, i32 addrspace(1)* %out
ret void
}
define void @i64_array_array(i64 addrspace(1)* %out, i32 %index) {
entry:
%alloca = alloca [2 x [2 x i64]]
%gep0 = getelementptr [2 x [2 x i64]], [2 x [2 x i64]]* %alloca, i32 0, i32 0, i32 0
%gep1 = getelementptr [2 x [2 x i64]], [2 x [2 x i64]]* %alloca, i32 0, i32 0, i32 1
store i64 0, i64* %gep0
store i64 1, i64* %gep1
%gep2 = getelementptr [2 x [2 x i64]], [2 x [2 x i64]]* %alloca, i32 0, i32 0, i32 %index
%load = load i64, i64* %gep2
store i64 %load, i64 addrspace(1)* %out
ret void
}
%struct.pair32 = type { i32, i32 }
define void @struct_array_array(i32 addrspace(1)* %out, i32 %index) {
entry:
%alloca = alloca [2 x [2 x %struct.pair32]]
%gep0 = getelementptr [2 x [2 x %struct.pair32]], [2 x [2 x %struct.pair32]]* %alloca, i32 0, i32 0, i32 0, i32 1
%gep1 = getelementptr [2 x [2 x %struct.pair32]], [2 x [2 x %struct.pair32]]* %alloca, i32 0, i32 0, i32 1, i32 1
store i32 0, i32* %gep0
store i32 1, i32* %gep1
%gep2 = getelementptr [2 x [2 x %struct.pair32]], [2 x [2 x %struct.pair32]]* %alloca, i32 0, i32 0, i32 %index, i32 0
%load = load i32, i32* %gep2
store i32 %load, i32 addrspace(1)* %out
ret void
}
define void @struct_pair32_array(i32 addrspace(1)* %out, i32 %index) {
entry:
%alloca = alloca [2 x %struct.pair32]
%gep0 = getelementptr [2 x %struct.pair32], [2 x %struct.pair32]* %alloca, i32 0, i32 0, i32 1
%gep1 = getelementptr [2 x %struct.pair32], [2 x %struct.pair32]* %alloca, i32 0, i32 1, i32 0
store i32 0, i32* %gep0
store i32 1, i32* %gep1
%gep2 = getelementptr [2 x %struct.pair32], [2 x %struct.pair32]* %alloca, i32 0, i32 %index, i32 0
%load = load i32, i32* %gep2
store i32 %load, i32 addrspace(1)* %out
ret void
}
define void @select_private(i32 addrspace(1)* %out, i32 %in) nounwind {
entry:
%tmp = alloca [2 x i32]
%tmp1 = getelementptr [2 x i32], [2 x i32]* %tmp, i32 0, i32 0
%tmp2 = getelementptr [2 x i32], [2 x i32]* %tmp, i32 0, i32 1
store i32 0, i32* %tmp1
store i32 1, i32* %tmp2
%cmp = icmp eq i32 %in, 0
%sel = select i1 %cmp, i32* %tmp1, i32* %tmp2
%load = load i32, i32* %sel
store i32 %load, i32 addrspace(1)* %out
ret void
}
; AMDGPUPromoteAlloca does not know how to handle ptrtoint. When it
; finds one, it should stop trying to promote.
; FUNC-LABEL: ptrtoint:
; SI-NOT: ds_write
; SI: buffer_store_dword v{{[0-9]+}}, v{{[0-9]+}}, s[{{[0-9]+:[0-9]+}}], s{{[0-9]+}} offen
; SI: buffer_load_dword v{{[0-9]+}}, v{{[0-9]+}}, s[{{[0-9]+:[0-9]+}}], s{{[0-9]+}} offen ;
define void @ptrtoint(i32 addrspace(1)* %out, i32 %a, i32 %b) {
%alloca = alloca [16 x i32]
%tmp0 = getelementptr [16 x i32], [16 x i32]* %alloca, i32 0, i32 %a
store i32 5, i32* %tmp0
%tmp1 = ptrtoint [16 x i32]* %alloca to i32
%tmp2 = add i32 %tmp1, 5
%tmp3 = inttoptr i32 %tmp2 to i32*
%tmp4 = getelementptr i32, i32* %tmp3, i32 %b
%tmp5 = load i32, i32* %tmp4
store i32 %tmp5, i32 addrspace(1)* %out
ret void
}
; OPT: !0 = !{i32 0, i32 2048}
Reference in New Issue View Git Blame Copy Permalink