mirror of
https://github.com/c64scene-ar/llvm-6502.git
synced 2024-08-22 10:29:35 +00:00
One unusual feature of the z architecture is that the result of a previous load can be reused indefinitely for subsequent loads, even if a cache-coherent store to that location is performed by another CPU. A special serializing instruction must be used if you want to force a load to be reattempted. Since volatile loads are not supposed to be omitted in this way, we should insert a serializing instruction before each such load. The same goes for atomic loads. The patch implements this at the IR->DAG boundary, in a similar way to atomic fences. It is a no-op for targets other than SystemZ. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@196906 91177308-0d34-0410-b5e6-96231b3b80d8
215 lines
5.7 KiB
LLVM
215 lines
5.7 KiB
LLVM
; Test 64-bit conditional stores that are presented as selects.
|
|
;
|
|
; RUN: llc < %s -mtriple=s390x-linux-gnu -mcpu=z10 | FileCheck %s
|
|
|
|
declare void @foo(i64 *)
|
|
|
|
; Test with the loaded value first.
|
|
define void @f1(i64 *%ptr, i64 %alt, i32 %limit) {
|
|
; CHECK-LABEL: f1:
|
|
; CHECK-NOT: %r2
|
|
; CHECK: jl [[LABEL:[^ ]*]]
|
|
; CHECK-NOT: %r2
|
|
; CHECK: stg %r3, 0(%r2)
|
|
; CHECK: [[LABEL]]:
|
|
; CHECK: br %r14
|
|
%cond = icmp ult i32 %limit, 420
|
|
%orig = load i64 *%ptr
|
|
%res = select i1 %cond, i64 %orig, i64 %alt
|
|
store i64 %res, i64 *%ptr
|
|
ret void
|
|
}
|
|
|
|
; ...and with the loaded value second
|
|
define void @f2(i64 *%ptr, i64 %alt, i32 %limit) {
|
|
; CHECK-LABEL: f2:
|
|
; CHECK-NOT: %r2
|
|
; CHECK: jhe [[LABEL:[^ ]*]]
|
|
; CHECK-NOT: %r2
|
|
; CHECK: stg %r3, 0(%r2)
|
|
; CHECK: [[LABEL]]:
|
|
; CHECK: br %r14
|
|
%cond = icmp ult i32 %limit, 420
|
|
%orig = load i64 *%ptr
|
|
%res = select i1 %cond, i64 %alt, i64 %orig
|
|
store i64 %res, i64 *%ptr
|
|
ret void
|
|
}
|
|
|
|
; Check the high end of the aligned STG range.
|
|
define void @f3(i64 *%base, i64 %alt, i32 %limit) {
|
|
; CHECK-LABEL: f3:
|
|
; CHECK-NOT: %r2
|
|
; CHECK: jl [[LABEL:[^ ]*]]
|
|
; CHECK-NOT: %r2
|
|
; CHECK: stg %r3, 524280(%r2)
|
|
; CHECK: [[LABEL]]:
|
|
; CHECK: br %r14
|
|
%ptr = getelementptr i64 *%base, i64 65535
|
|
%cond = icmp ult i32 %limit, 420
|
|
%orig = load i64 *%ptr
|
|
%res = select i1 %cond, i64 %orig, i64 %alt
|
|
store i64 %res, i64 *%ptr
|
|
ret void
|
|
}
|
|
|
|
; Check the next doubleword up, which needs separate address logic.
|
|
; Other sequences besides this one would be OK.
|
|
define void @f4(i64 *%base, i64 %alt, i32 %limit) {
|
|
; CHECK-LABEL: f4:
|
|
; CHECK-NOT: %r2
|
|
; CHECK: jl [[LABEL:[^ ]*]]
|
|
; CHECK-NOT: %r2
|
|
; CHECK: agfi %r2, 524288
|
|
; CHECK: stg %r3, 0(%r2)
|
|
; CHECK: [[LABEL]]:
|
|
; CHECK: br %r14
|
|
%ptr = getelementptr i64 *%base, i64 65536
|
|
%cond = icmp ult i32 %limit, 420
|
|
%orig = load i64 *%ptr
|
|
%res = select i1 %cond, i64 %orig, i64 %alt
|
|
store i64 %res, i64 *%ptr
|
|
ret void
|
|
}
|
|
|
|
; Check the low end of the STG range.
|
|
define void @f5(i64 *%base, i64 %alt, i32 %limit) {
|
|
; CHECK-LABEL: f5:
|
|
; CHECK-NOT: %r2
|
|
; CHECK: jl [[LABEL:[^ ]*]]
|
|
; CHECK-NOT: %r2
|
|
; CHECK: stg %r3, -524288(%r2)
|
|
; CHECK: [[LABEL]]:
|
|
; CHECK: br %r14
|
|
%ptr = getelementptr i64 *%base, i64 -65536
|
|
%cond = icmp ult i32 %limit, 420
|
|
%orig = load i64 *%ptr
|
|
%res = select i1 %cond, i64 %orig, i64 %alt
|
|
store i64 %res, i64 *%ptr
|
|
ret void
|
|
}
|
|
|
|
; Check the next doubleword down, which needs separate address logic.
|
|
; Other sequences besides this one would be OK.
|
|
define void @f6(i64 *%base, i64 %alt, i32 %limit) {
|
|
; CHECK-LABEL: f6:
|
|
; CHECK-NOT: %r2
|
|
; CHECK: jl [[LABEL:[^ ]*]]
|
|
; CHECK-NOT: %r2
|
|
; CHECK: agfi %r2, -524296
|
|
; CHECK: stg %r3, 0(%r2)
|
|
; CHECK: [[LABEL]]:
|
|
; CHECK: br %r14
|
|
%ptr = getelementptr i64 *%base, i64 -65537
|
|
%cond = icmp ult i32 %limit, 420
|
|
%orig = load i64 *%ptr
|
|
%res = select i1 %cond, i64 %orig, i64 %alt
|
|
store i64 %res, i64 *%ptr
|
|
ret void
|
|
}
|
|
|
|
; Check that STG allows an index.
|
|
define void @f7(i64 %base, i64 %index, i64 %alt, i32 %limit) {
|
|
; CHECK-LABEL: f7:
|
|
; CHECK-NOT: %r2
|
|
; CHECK: jl [[LABEL:[^ ]*]]
|
|
; CHECK-NOT: %r2
|
|
; CHECK: stg %r4, 524287(%r3,%r2)
|
|
; CHECK: [[LABEL]]:
|
|
; CHECK: br %r14
|
|
%add1 = add i64 %base, %index
|
|
%add2 = add i64 %add1, 524287
|
|
%ptr = inttoptr i64 %add2 to i64 *
|
|
%cond = icmp ult i32 %limit, 420
|
|
%orig = load i64 *%ptr
|
|
%res = select i1 %cond, i64 %orig, i64 %alt
|
|
store i64 %res, i64 *%ptr
|
|
ret void
|
|
}
|
|
|
|
; Check that volatile loads are not matched.
|
|
define void @f8(i64 *%ptr, i64 %alt, i32 %limit) {
|
|
; CHECK-LABEL: f8:
|
|
; CHECK: lg {{%r[0-5]}}, 0(%r2)
|
|
; CHECK: {{jl|jnl}} [[LABEL:[^ ]*]]
|
|
; CHECK: [[LABEL]]:
|
|
; CHECK: stg {{%r[0-5]}}, 0(%r2)
|
|
; CHECK: br %r14
|
|
%cond = icmp ult i32 %limit, 420
|
|
%orig = load volatile i64 *%ptr
|
|
%res = select i1 %cond, i64 %orig, i64 %alt
|
|
store i64 %res, i64 *%ptr
|
|
ret void
|
|
}
|
|
|
|
; ...likewise stores. In this case we should have a conditional load into %r3.
|
|
define void @f9(i64 *%ptr, i64 %alt, i32 %limit) {
|
|
; CHECK-LABEL: f9:
|
|
; CHECK: jhe [[LABEL:[^ ]*]]
|
|
; CHECK: lg %r3, 0(%r2)
|
|
; CHECK: [[LABEL]]:
|
|
; CHECK: stg %r3, 0(%r2)
|
|
; CHECK: br %r14
|
|
%cond = icmp ult i32 %limit, 420
|
|
%orig = load i64 *%ptr
|
|
%res = select i1 %cond, i64 %orig, i64 %alt
|
|
store volatile i64 %res, i64 *%ptr
|
|
ret void
|
|
}
|
|
|
|
; Check that atomic loads are not matched. The transformation is OK for
|
|
; the "unordered" case tested here, but since we don't try to handle atomic
|
|
; operations at all in this context, it seems better to assert that than
|
|
; to restrict the test to a stronger ordering.
|
|
define void @f10(i64 *%ptr, i64 %alt, i32 %limit) {
|
|
; FIXME: should use a normal load instead of CSG.
|
|
; CHECK-LABEL: f10:
|
|
; CHECK: lg {{%r[0-5]}}, 0(%r2)
|
|
; CHECK: {{jl|jnl}} [[LABEL:[^ ]*]]
|
|
; CHECK: [[LABEL]]:
|
|
; CHECK: stg {{%r[0-5]}}, 0(%r2)
|
|
; CHECK: br %r14
|
|
%cond = icmp ult i32 %limit, 420
|
|
%orig = load atomic i64 *%ptr unordered, align 8
|
|
%res = select i1 %cond, i64 %orig, i64 %alt
|
|
store i64 %res, i64 *%ptr
|
|
ret void
|
|
}
|
|
|
|
; ...likewise stores.
|
|
define void @f11(i64 *%ptr, i64 %alt, i32 %limit) {
|
|
; FIXME: should use a normal store instead of CSG.
|
|
; CHECK-LABEL: f11:
|
|
; CHECK: jhe [[LABEL:[^ ]*]]
|
|
; CHECK: lg %r3, 0(%r2)
|
|
; CHECK: [[LABEL]]:
|
|
; CHECK: stg %r3, 0(%r2)
|
|
; CHECK: br %r14
|
|
%cond = icmp ult i32 %limit, 420
|
|
%orig = load i64 *%ptr
|
|
%res = select i1 %cond, i64 %orig, i64 %alt
|
|
store atomic i64 %res, i64 *%ptr unordered, align 8
|
|
ret void
|
|
}
|
|
|
|
; Try a frame index base.
|
|
define void @f12(i64 %alt, i32 %limit) {
|
|
; CHECK-LABEL: f12:
|
|
; CHECK: brasl %r14, foo@PLT
|
|
; CHECK-NOT: %r15
|
|
; CHECK: jl [[LABEL:[^ ]*]]
|
|
; CHECK-NOT: %r15
|
|
; CHECK: stg {{%r[0-9]+}}, {{[0-9]+}}(%r15)
|
|
; CHECK: [[LABEL]]:
|
|
; CHECK: brasl %r14, foo@PLT
|
|
; CHECK: br %r14
|
|
%ptr = alloca i64
|
|
call void @foo(i64 *%ptr)
|
|
%cond = icmp ult i32 %limit, 420
|
|
%orig = load i64 *%ptr
|
|
%res = select i1 %cond, i64 %orig, i64 %alt
|
|
store i64 %res, i64 *%ptr
|
|
call void @foo(i64 *%ptr)
|
|
ret void
|
|
}
|