AMDGPU: Cleanup FP atomicrmw tests and cover fmin/fmax

llvm/test/CodeGen/AMDGPU/global-atomicrmw-fadd-wrong-subtarget.ll

@@ -0,0 +1,75 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
+; RUN: llc -mtriple=amdgcn-amd-amdpal < %s | FileCheck -check-prefix=GCN %s
+
+define amdgpu_kernel void @global_atomic_fadd_ret_f32_wrong_subtarget(ptr addrspace(1) %ptr) #1 {
+; GCN-LABEL: global_atomic_fadd_ret_f32_wrong_subtarget:
+; GCN:       ; %bb.0:
+; GCN-NEXT:    s_mov_b64 s[4:5], exec
+; GCN-NEXT:    v_mbcnt_lo_u32_b32 v0, s4, 0
+; GCN-NEXT:    v_mbcnt_hi_u32_b32 v0, s5, v0
+; GCN-NEXT:    v_cmp_eq_u32_e32 vcc, 0, v0
+; GCN-NEXT:    ; implicit-def: $vgpr1
+; GCN-NEXT:    s_and_saveexec_b64 s[2:3], vcc
+; GCN-NEXT:    s_cbranch_execz .LBB0_4
+; GCN-NEXT:  ; %bb.1:
+; GCN-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x0
+; GCN-NEXT:    s_bcnt1_i32_b64 s7, s[4:5]
+; GCN-NEXT:    v_cvt_f32_ubyte0_e32 v1, s7
+; GCN-NEXT:    s_mov_b64 s[4:5], 0
+; GCN-NEXT:    v_mul_f32_e32 v2, 4.0, v1
+; GCN-NEXT:    s_waitcnt lgkmcnt(0)
+; GCN-NEXT:    s_load_dword s6, s[0:1], 0x0
+; GCN-NEXT:    v_mov_b32_e32 v3, 0
+; GCN-NEXT:    s_waitcnt lgkmcnt(0)
+; GCN-NEXT:    v_mov_b32_e32 v1, s6
+; GCN-NEXT:  .LBB0_2: ; %atomicrmw.start
+; GCN-NEXT:    ; =>This Inner Loop Header: Depth=1
+; GCN-NEXT:    v_mov_b32_e32 v5, v1
+; GCN-NEXT:    v_add_f32_e32 v4, v5, v2
+; GCN-NEXT:    global_atomic_cmpswap v1, v3, v[4:5], s[0:1] glc
+; GCN-NEXT:    s_waitcnt vmcnt(0)
+; GCN-NEXT:    buffer_wbinvl1
+; GCN-NEXT:    v_cmp_eq_u32_e32 vcc, v1, v5
+; GCN-NEXT:    s_or_b64 s[4:5], vcc, s[4:5]
+; GCN-NEXT:    s_andn2_b64 exec, exec, s[4:5]
+; GCN-NEXT:    s_cbranch_execnz .LBB0_2
+; GCN-NEXT:  ; %bb.3: ; %Flow
+; GCN-NEXT:    s_or_b64 exec, exec, s[4:5]
+; GCN-NEXT:  .LBB0_4: ; %Flow2
+; GCN-NEXT:    s_or_b64 exec, exec, s[2:3]
+; GCN-NEXT:    v_readfirstlane_b32 s0, v1
+; GCN-NEXT:    v_cvt_f32_ubyte0_e32 v0, v0
+; GCN-NEXT:    v_mad_f32 v0, v0, 4.0, s0
+; GCN-NEXT:    global_store_dword v[0:1], v0, off
+; GCN-NEXT:    s_endpgm
+  %result = atomicrmw fadd ptr addrspace(1) %ptr, float 4.0 syncscope("agent") seq_cst
+  store float %result, ptr addrspace(1) undef
+  ret void
+}
+
+define amdgpu_kernel void @global_atomic_fadd_noret_f32_wrong_subtarget(ptr addrspace(1) %ptr) #1 {
+; GCN-LABEL: global_atomic_fadd_noret_f32_wrong_subtarget:
+; GCN:       ; %bb.0:
+; GCN-NEXT:    s_mov_b64 s[2:3], exec
+; GCN-NEXT:    v_mbcnt_lo_u32_b32 v0, s2, 0
+; GCN-NEXT:    v_mbcnt_hi_u32_b32 v0, s3, v0
+; GCN-NEXT:    v_cmp_eq_u32_e32 vcc, 0, v0
+; GCN-NEXT:    s_and_saveexec_b64 s[4:5], vcc
+; GCN-NEXT:    s_cbranch_execz .LBB1_2
+; GCN-NEXT:  ; %bb.1:
+; GCN-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x0
+; GCN-NEXT:    s_bcnt1_i32_b64 s2, s[2:3]
+; GCN-NEXT:    v_cvt_f32_ubyte0_e32 v1, s2
+; GCN-NEXT:    v_mov_b32_e32 v0, 0
+; GCN-NEXT:    v_mul_f32_e32 v1, 4.0, v1
+; GCN-NEXT:    s_waitcnt lgkmcnt(0)
+; GCN-NEXT:    global_atomic_add_f32 v0, v1, s[0:1]
+; GCN-NEXT:    s_waitcnt vmcnt(0)
+; GCN-NEXT:    buffer_wbinvl1
+; GCN-NEXT:  .LBB1_2:
+; GCN-NEXT:    s_endpgm
+  %result = atomicrmw fadd ptr addrspace(1) %ptr, float 4.0 syncscope("agent") seq_cst
+  ret void
+}
+
+attributes #1 = { "denormal-fp-math-f32"="preserve-sign,preserve-sign" "target-cpu"="gfx803" "target-features"="+atomic-fadd-no-rtn-insts" "amdgpu-unsafe-fp-atomics"="true" }