1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-26 04:32:44 +01:00
llvm-mirror/test/CodeGen/BPF/xadd_legal.ll
Jiong Wang c5d6744168 bpf: enable sub-register code-gen for XADD
Support sub-register code-gen for XADD is like supporting any other Load
and Store patterns.

No new instruction is introduced.

  lock *(u32 *)(r1 + 0) += w2

has exactly the same underlying insn as:

  lock *(u32 *)(r1 + 0) += r2

BPF_W width modifier has guaranteed they behave the same at runtime. This
patch merely teaches BPF back-end that BPF_W width modifier could work
GPR32 register class and that's all needed for sub-register code-gen
support for XADD.

test/CodeGen/BPF/xadd.ll updated to include sub-register code-gen tests.

A new testcase test/CodeGen/BPF/xadd_legal.ll is added to make sure the
legal case could pass on all code-gen modes. It could also test dead Def
check on GPR32. If there is no proper handling like what has been done
inside BPFMIChecking.cpp:hasLivingDefs, then this testcase will fail.

Acked-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
llvm-svn: 355126
2019-02-28 19:21:28 +00:00

27 lines
1.0 KiB
LLVM

; RUN: llc -march=bpfel < %s 2>&1 | FileCheck --check-prefix=CHECK-64 %s
; RUN: llc -march=bpfeb < %s 2>&1 | FileCheck --check-prefix=CHECK-64 %s
; RUN: llc -march=bpfel -mattr=+alu32 < %s 2>&1 | FileCheck --check-prefix=CHECK-32 %s
; RUN: llc -march=bpfeb -mattr=+alu32 < %s 2>&1 | FileCheck --check-prefix=CHECK-32 %s
; This file is generated with the source command and source
; $ clang -target bpf -O2 -S -emit-llvm t.c
; $ cat t.c
; int test(int *ptr, unsigned long long a) {
; __sync_fetch_and_add(ptr, a);
; return *ptr;
; }
;
; NOTE: passing unsigned long long as the second operand of __sync_fetch_and_add
; could effectively create sub-register reference coming from indexing a full
; register which could then exerceise hasLivingDefs inside BPFMIChecker.cpp.
define dso_local i32 @test(i32* nocapture %ptr, i64 %a) {
entry:
%conv = trunc i64 %a to i32
%0 = atomicrmw add i32* %ptr, i32 %conv seq_cst
; CHECK-64: lock *(u32 *)(r1 + 0) += r2
; CHECK-32: lock *(u32 *)(r1 + 0) += w2
%1 = load i32, i32* %ptr, align 4
ret i32 %1
}