[llvm] AMDGPU: Cleanup FP atomicrmw tests and cover fmin/fmax (PR #95131)
via llvm-commits
llvm-commits at lists.llvm.org
Tue Jun 11 08:01:50 PDT 2024
llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT-->
@llvm/pr-subscribers-backend-amdgpu
Author: Matt Arsenault (arsenm)
<details>
<summary>Changes</summary>
We apparently are missing codegen support for atomicrmw fmin/fmax. Also clean up FP atomicrmw tests
to be more consistent and comprehensively test the relevant cases
---
Patch is 7.89 MiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/95131.diff
15 Files Affected:
- (modified) llvm/test/CodeGen/AMDGPU/flat-atomicrmw-fadd.ll (+13916-4167)
- (added) llvm/test/CodeGen/AMDGPU/flat-atomicrmw-fmax.ll (+15756)
- (added) llvm/test/CodeGen/AMDGPU/flat-atomicrmw-fmin.ll (+15756)
- (added) llvm/test/CodeGen/AMDGPU/flat-atomicrmw-fsub.ll (+15075)
- (added) llvm/test/CodeGen/AMDGPU/global-atomicrmw-fadd-wrong-subtarget.ll (+75)
- (added) llvm/test/CodeGen/AMDGPU/global-atomicrmw-fadd.ll (+16785)
- (added) llvm/test/CodeGen/AMDGPU/global-atomicrmw-fmax.ll (+18237)
- (added) llvm/test/CodeGen/AMDGPU/global-atomicrmw-fmin.ll (+18237)
- (added) llvm/test/CodeGen/AMDGPU/global-atomicrmw-fsub.ll (+17517)
- (removed) llvm/test/CodeGen/AMDGPU/global-atomics-fp.ll (-8090)
- (added) llvm/test/CodeGen/AMDGPU/local-atomicrmw-fadd.ll (+9102)
- (added) llvm/test/CodeGen/AMDGPU/local-atomicrmw-fmax.ll (+8456)
- (added) llvm/test/CodeGen/AMDGPU/local-atomicrmw-fmin.ll (+8456)
- (added) llvm/test/CodeGen/AMDGPU/local-atomicrmw-fsub.ll (+8226)
- (removed) llvm/test/CodeGen/AMDGPU/local-atomics-fp.ll (-3931)
``````````diff
diff --git a/llvm/test/CodeGen/AMDGPU/flat-atomicrmw-fadd.ll b/llvm/test/CodeGen/AMDGPU/flat-atomicrmw-fadd.ll
index 1e7cf0e702a03..af02db95cee95 100644
--- a/llvm/test/CodeGen/AMDGPU/flat-atomicrmw-fadd.ll
+++ b/llvm/test/CodeGen/AMDGPU/flat-atomicrmw-fadd.ll
@@ -1,82 +1,125 @@
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx700 < %s | FileCheck -check-prefix=GFX7 %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx803 < %s | FileCheck -check-prefix=GFX8 %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx908 < %s | FileCheck -check-prefixes=GFX90,GFX908 %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a < %s | FileCheck -check-prefixes=GFX90,GFX90A %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 < %s | FileCheck -check-prefix=GFX940 %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 < %s | FileCheck -check-prefix=GFX10 %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 < %s | FileCheck -check-prefix=GFX11 %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1200 < %s | FileCheck -check-prefix=GFX12 %s
+; RUN: llc -mtriple=amdgcn -mcpu=gfx1200 < %s | FileCheck -check-prefix=GFX12 %s
+; RUN: llc -mtriple=amdgcn -mcpu=gfx940 < %s | FileCheck -check-prefix=GFX940 %s
+; RUN: llc -mtriple=amdgcn -mcpu=gfx1100 < %s | FileCheck -check-prefix=GFX11 %s
+; RUN: llc -mtriple=amdgcn -mcpu=gfx1010 < %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -mtriple=amdgcn -mcpu=gfx90a < %s | FileCheck -check-prefix=GFX90A %s
+; RUN: llc -mtriple=amdgcn -mcpu=gfx908 < %s | FileCheck -check-prefix=GFX908 %s
+; RUN: llc -mtriple=amdgcn -mcpu=tonga < %s | FileCheck -check-prefix=GFX8 %s
+; RUN: llc -mtriple=amdgcn -mcpu=hawaii < %s | FileCheck -check-prefix=GFX7 %s
-define <2 x half> @flat_agent_atomic_fadd_ret_v2f16(ptr %ptr, <2 x half> %val) {
-; GFX7-LABEL: flat_agent_atomic_fadd_ret_v2f16:
-; GFX7: ; %bb.0:
-; GFX7-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GFX7-NEXT: flat_load_dword v5, v[0:1]
-; GFX7-NEXT: v_cvt_f16_f32_e32 v3, v3
-; GFX7-NEXT: v_cvt_f16_f32_e32 v6, v2
-; GFX7-NEXT: s_mov_b64 s[4:5], 0
-; GFX7-NEXT: v_cvt_f32_f16_e32 v4, v3
-; GFX7-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
-; GFX7-NEXT: v_lshrrev_b32_e32 v3, 16, v5
-; GFX7-NEXT: v_cvt_f32_f16_e32 v2, v5
-; GFX7-NEXT: v_cvt_f32_f16_e32 v3, v3
-; GFX7-NEXT: v_cvt_f32_f16_e32 v5, v6
-; GFX7-NEXT: .LBB0_1: ; %atomicrmw.start
-; GFX7-NEXT: ; =>This Inner Loop Header: Depth=1
-; GFX7-NEXT: v_cvt_f16_f32_e32 v3, v3
-; GFX7-NEXT: v_cvt_f16_f32_e32 v2, v2
-; GFX7-NEXT: v_cvt_f32_f16_e32 v6, v3
-; GFX7-NEXT: v_cvt_f32_f16_e32 v7, v2
-; GFX7-NEXT: v_lshlrev_b32_e32 v3, 16, v3
-; GFX7-NEXT: v_add_f32_e32 v6, v6, v4
-; GFX7-NEXT: v_add_f32_e32 v7, v7, v5
-; GFX7-NEXT: v_cvt_f16_f32_e32 v6, v6
-; GFX7-NEXT: v_cvt_f16_f32_e32 v8, v7
-; GFX7-NEXT: v_or_b32_e32 v7, v2, v3
-; GFX7-NEXT: v_lshlrev_b32_e32 v2, 16, v6
-; GFX7-NEXT: v_or_b32_e32 v6, v8, v2
-; GFX7-NEXT: flat_atomic_cmpswap v6, v[0:1], v[6:7] glc
-; GFX7-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
-; GFX7-NEXT: buffer_wbinvl1_vol
-; GFX7-NEXT: v_lshrrev_b32_e32 v3, 16, v6
-; GFX7-NEXT: v_cvt_f32_f16_e32 v2, v6
-; GFX7-NEXT: v_cvt_f32_f16_e32 v3, v3
-; GFX7-NEXT: v_cmp_eq_u32_e32 vcc, v6, v7
-; GFX7-NEXT: s_or_b64 s[4:5], vcc, s[4:5]
-; GFX7-NEXT: s_andn2_b64 exec, exec, s[4:5]
-; GFX7-NEXT: s_cbranch_execnz .LBB0_1
-; GFX7-NEXT: ; %bb.2: ; %atomicrmw.end
-; GFX7-NEXT: s_or_b64 exec, exec, s[4:5]
-; GFX7-NEXT: v_mov_b32_e32 v0, v2
-; GFX7-NEXT: v_mov_b32_e32 v1, v3
-; GFX7-NEXT: s_setpc_b64 s[30:31]
+; --------------------------------------------------------------------
+; float
+; --------------------------------------------------------------------
+
+define float @flat_agent_atomic_fadd_ret_f32(ptr %ptr, float %val) #0 {
+; GFX12-LABEL: flat_agent_atomic_fadd_ret_f32:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_wait_storecnt 0x0
+; GFX12-NEXT: flat_atomic_add_f32 v0, v[0:1], v2 th:TH_ATOMIC_RETURN
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: global_inv scope:SCOPE_DEV
+; GFX12-NEXT: s_setpc_b64 s[30:31]
;
-; GFX8-LABEL: flat_agent_atomic_fadd_ret_v2f16:
-; GFX8: ; %bb.0:
-; GFX8-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GFX8-NEXT: flat_load_dword v3, v[0:1]
-; GFX8-NEXT: s_mov_b64 s[4:5], 0
-; GFX8-NEXT: .LBB0_1: ; %atomicrmw.start
-; GFX8-NEXT: ; =>This Inner Loop Header: Depth=1
-; GFX8-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
-; GFX8-NEXT: v_mov_b32_e32 v4, v3
-; GFX8-NEXT: v_add_f16_sdwa v3, v4, v2 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:WORD_1
-; GFX8-NEXT: v_add_f16_e32 v5, v4, v2
-; GFX8-NEXT: v_or_b32_e32 v3, v5, v3
-; GFX8-NEXT: flat_atomic_cmpswap v3, v[0:1], v[3:4] glc
-; GFX8-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
-; GFX8-NEXT: buffer_wbinvl1_vol
-; GFX8-NEXT: v_cmp_eq_u32_e32 vcc, v3, v4
-; GFX8-NEXT: s_or_b64 s[4:5], vcc, s[4:5]
-; GFX8-NEXT: s_andn2_b64 exec, exec, s[4:5]
-; GFX8-NEXT: s_cbranch_execnz .LBB0_1
-; GFX8-NEXT: ; %bb.2: ; %atomicrmw.end
-; GFX8-NEXT: s_or_b64 exec, exec, s[4:5]
-; GFX8-NEXT: v_mov_b32_e32 v0, v3
-; GFX8-NEXT: s_setpc_b64 s[30:31]
+; GFX940-LABEL: flat_agent_atomic_fadd_ret_f32:
+; GFX940: ; %bb.0:
+; GFX940-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX940-NEXT: buffer_wbl2 sc1
+; GFX940-NEXT: flat_atomic_add_f32 v0, v[0:1], v2 sc0
+; GFX940-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX940-NEXT: buffer_inv sc1
+; GFX940-NEXT: s_setpc_b64 s[30:31]
;
-; GFX908-LABEL: flat_agent_atomic_fadd_ret_v2f16:
+; GFX11-LABEL: flat_agent_atomic_fadd_ret_f32:
+; GFX11: ; %bb.0:
+; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-NEXT: s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT: flat_atomic_add_f32 v0, v[0:1], v2 glc
+; GFX11-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX11-NEXT: buffer_gl1_inv
+; GFX11-NEXT: buffer_gl0_inv
+; GFX11-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX10-LABEL: flat_agent_atomic_fadd_ret_f32:
+; GFX10: ; %bb.0:
+; GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX10-NEXT: flat_load_dword v3, v[0:1]
+; GFX10-NEXT: s_mov_b32 s4, 0
+; GFX10-NEXT: .LBB0_1: ; %atomicrmw.start
+; GFX10-NEXT: ; =>This Inner Loop Header: Depth=1
+; GFX10-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX10-NEXT: v_mov_b32_e32 v4, v3
+; GFX10-NEXT: v_add_f32_e32 v3, v4, v2
+; GFX10-NEXT: s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT: flat_atomic_cmpswap v3, v[0:1], v[3:4] glc
+; GFX10-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX10-NEXT: buffer_gl1_inv
+; GFX10-NEXT: buffer_gl0_inv
+; GFX10-NEXT: v_cmp_eq_u32_e32 vcc_lo, v3, v4
+; GFX10-NEXT: s_or_b32 s4, vcc_lo, s4
+; GFX10-NEXT: s_andn2_b32 exec_lo, exec_lo, s4
+; GFX10-NEXT: s_cbranch_execnz .LBB0_1
+; GFX10-NEXT: ; %bb.2: ; %atomicrmw.end
+; GFX10-NEXT: s_or_b32 exec_lo, exec_lo, s4
+; GFX10-NEXT: v_mov_b32_e32 v0, v3
+; GFX10-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX90A-LABEL: flat_agent_atomic_fadd_ret_f32:
+; GFX90A: ; %bb.0:
+; GFX90A-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX90A-NEXT: s_mov_b64 s[4:5], src_shared_base
+; GFX90A-NEXT: v_cmp_ne_u32_e32 vcc, s5, v1
+; GFX90A-NEXT: ; implicit-def: $vgpr3
+; GFX90A-NEXT: s_and_saveexec_b64 s[4:5], vcc
+; GFX90A-NEXT: s_xor_b64 s[4:5], exec, s[4:5]
+; GFX90A-NEXT: s_cbranch_execz .LBB0_6
+; GFX90A-NEXT: ; %bb.1: ; %atomicrmw.check.private
+; GFX90A-NEXT: s_mov_b64 s[6:7], src_private_base
+; GFX90A-NEXT: v_cmp_ne_u32_e32 vcc, s7, v1
+; GFX90A-NEXT: ; implicit-def: $vgpr3
+; GFX90A-NEXT: s_and_saveexec_b64 s[6:7], vcc
+; GFX90A-NEXT: s_xor_b64 s[6:7], exec, s[6:7]
+; GFX90A-NEXT: s_cbranch_execz .LBB0_3
+; GFX90A-NEXT: ; %bb.2: ; %atomicrmw.global
+; GFX90A-NEXT: global_atomic_add_f32 v3, v[0:1], v2, off glc
+; GFX90A-NEXT: s_waitcnt vmcnt(0)
+; GFX90A-NEXT: buffer_wbinvl1_vol
+; GFX90A-NEXT: ; implicit-def: $vgpr0_vgpr1
+; GFX90A-NEXT: ; implicit-def: $vgpr2
+; GFX90A-NEXT: .LBB0_3: ; %Flow
+; GFX90A-NEXT: s_andn2_saveexec_b64 s[6:7], s[6:7]
+; GFX90A-NEXT: s_cbranch_execz .LBB0_5
+; GFX90A-NEXT: ; %bb.4: ; %atomicrmw.private
+; GFX90A-NEXT: v_cmp_ne_u64_e32 vcc, 0, v[0:1]
+; GFX90A-NEXT: v_cndmask_b32_e32 v0, -1, v0, vcc
+; GFX90A-NEXT: buffer_load_dword v3, v0, s[0:3], 0 offen
+; GFX90A-NEXT: s_waitcnt vmcnt(0)
+; GFX90A-NEXT: v_add_f32_e32 v1, v3, v2
+; GFX90A-NEXT: buffer_store_dword v1, v0, s[0:3], 0 offen
+; GFX90A-NEXT: .LBB0_5: ; %Flow1
+; GFX90A-NEXT: s_or_b64 exec, exec, s[6:7]
+; GFX90A-NEXT: ; implicit-def: $vgpr0_vgpr1
+; GFX90A-NEXT: ; implicit-def: $vgpr2
+; GFX90A-NEXT: .LBB0_6: ; %Flow2
+; GFX90A-NEXT: s_andn2_saveexec_b64 s[4:5], s[4:5]
+; GFX90A-NEXT: s_cbranch_execz .LBB0_8
+; GFX90A-NEXT: ; %bb.7: ; %atomicrmw.shared
+; GFX90A-NEXT: v_cmp_ne_u64_e32 vcc, 0, v[0:1]
+; GFX90A-NEXT: v_cndmask_b32_e32 v0, -1, v0, vcc
+; GFX90A-NEXT: ds_add_rtn_f32 v3, v0, v2
+; GFX90A-NEXT: s_waitcnt lgkmcnt(0)
+; GFX90A-NEXT: .LBB0_8: ; %atomicrmw.phi
+; GFX90A-NEXT: s_or_b64 exec, exec, s[4:5]
+; GFX90A-NEXT: v_mov_b32_e32 v0, v3
+; GFX90A-NEXT: s_waitcnt vmcnt(0)
+; GFX90A-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX908-LABEL: flat_agent_atomic_fadd_ret_f32:
; GFX908: ; %bb.0:
; GFX908-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX908-NEXT: flat_load_dword v3, v[0:1]
@@ -85,7 +128,7 @@ define <2 x half> @flat_agent_atomic_fadd_ret_v2f16(ptr %ptr, <2 x half> %val) {
; GFX908-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX908-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX908-NEXT: v_mov_b32_e32 v4, v3
-; GFX908-NEXT: v_pk_add_f16 v3, v4, v2
+; GFX908-NEXT: v_add_f32_e32 v3, v4, v2
; GFX908-NEXT: flat_atomic_cmpswap v3, v[0:1], v[3:4] glc
; GFX908-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX908-NEXT: buffer_wbinvl1_vol
@@ -98,1336 +141,1326 @@ define <2 x half> @flat_agent_atomic_fadd_ret_v2f16(ptr %ptr, <2 x half> %val) {
; GFX908-NEXT: v_mov_b32_e32 v0, v3
; GFX908-NEXT: s_setpc_b64 s[30:31]
;
-; GFX90A-LABEL: flat_agent_atomic_fadd_ret_v2f16:
-; GFX90A: ; %bb.0:
-; GFX90A-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GFX90A-NEXT: flat_load_dword v3, v[0:1]
-; GFX90A-NEXT: s_mov_b64 s[4:5], 0
-; GFX90A-NEXT: .LBB0_1: ; %atomicrmw.start
-; GFX90A-NEXT: ; =>This Inner Loop Header: Depth=1
-; GFX90A-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
-; GFX90A-NEXT: v_mov_b32_e32 v5, v3
-; GFX90A-NEXT: v_pk_add_f16 v4, v5, v2
-; GFX90A-NEXT: flat_atomic_cmpswap v3, v[0:1], v[4:5] glc
-; GFX90A-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
-; GFX90A-NEXT: buffer_wbinvl1_vol
-; GFX90A-NEXT: v_cmp_eq_u32_e32 vcc, v3, v5
-; GFX90A-NEXT: s_or_b64 s[4:5], vcc, s[4:5]
-; GFX90A-NEXT: s_andn2_b64 exec, exec, s[4:5]
-; GFX90A-NEXT: s_cbranch_execnz .LBB0_1
-; GFX90A-NEXT: ; %bb.2: ; %atomicrmw.end
-; GFX90A-NEXT: s_or_b64 exec, exec, s[4:5]
-; GFX90A-NEXT: v_mov_b32_e32 v0, v3
-; GFX90A-NEXT: s_setpc_b64 s[30:31]
+; GFX8-LABEL: flat_agent_atomic_fadd_ret_f32:
+; GFX8: ; %bb.0:
+; GFX8-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX8-NEXT: flat_load_dword v3, v[0:1]
+; GFX8-NEXT: s_mov_b64 s[4:5], 0
+; GFX8-NEXT: .LBB0_1: ; %atomicrmw.start
+; GFX8-NEXT: ; =>This Inner Loop Header: Depth=1
+; GFX8-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX8-NEXT: v_mov_b32_e32 v4, v3
+; GFX8-NEXT: v_add_f32_e32 v3, v4, v2
+; GFX8-NEXT: flat_atomic_cmpswap v3, v[0:1], v[3:4] glc
+; GFX8-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX8-NEXT: buffer_wbinvl1_vol
+; GFX8-NEXT: v_cmp_eq_u32_e32 vcc, v3, v4
+; GFX8-NEXT: s_or_b64 s[4:5], vcc, s[4:5]
+; GFX8-NEXT: s_andn2_b64 exec, exec, s[4:5]
+; GFX8-NEXT: s_cbranch_execnz .LBB0_1
+; GFX8-NEXT: ; %bb.2: ; %atomicrmw.end
+; GFX8-NEXT: s_or_b64 exec, exec, s[4:5]
+; GFX8-NEXT: v_mov_b32_e32 v0, v3
+; GFX8-NEXT: s_setpc_b64 s[30:31]
;
-; GFX940-LABEL: flat_agent_atomic_fadd_ret_v2f16:
+; GFX7-LABEL: flat_agent_atomic_fadd_ret_f32:
+; GFX7: ; %bb.0:
+; GFX7-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX7-NEXT: flat_load_dword v3, v[0:1]
+; GFX7-NEXT: s_mov_b64 s[4:5], 0
+; GFX7-NEXT: .LBB0_1: ; %atomicrmw.start
+; GFX7-NEXT: ; =>This Inner Loop Header: Depth=1
+; GFX7-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX7-NEXT: v_mov_b32_e32 v4, v3
+; GFX7-NEXT: v_add_f32_e32 v3, v4, v2
+; GFX7-NEXT: flat_atomic_cmpswap v3, v[0:1], v[3:4] glc
+; GFX7-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX7-NEXT: buffer_wbinvl1_vol
+; GFX7-NEXT: v_cmp_eq_u32_e32 vcc, v3, v4
+; GFX7-NEXT: s_or_b64 s[4:5], vcc, s[4:5]
+; GFX7-NEXT: s_andn2_b64 exec, exec, s[4:5]
+; GFX7-NEXT: s_cbranch_execnz .LBB0_1
+; GFX7-NEXT: ; %bb.2: ; %atomicrmw.end
+; GFX7-NEXT: s_or_b64 exec, exec, s[4:5]
+; GFX7-NEXT: v_mov_b32_e32 v0, v3
+; GFX7-NEXT: s_setpc_b64 s[30:31]
+ %result = atomicrmw fadd ptr %ptr, float %val syncscope("agent") seq_cst
+ ret float %result
+}
+
+define float @flat_agent_atomic_fadd_ret_f32__offset12b_pos(ptr %ptr, float %val) #0 {
+; GFX12-LABEL: flat_agent_atomic_fadd_ret_f32__offset12b_pos:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_wait_storecnt 0x0
+; GFX12-NEXT: flat_atomic_add_f32 v0, v[0:1], v2 offset:2044 th:TH_ATOMIC_RETURN
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: global_inv scope:SCOPE_DEV
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX940-LABEL: flat_agent_atomic_fadd_ret_f32__offset12b_pos:
; GFX940: ; %bb.0:
; GFX940-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GFX940-NEXT: flat_load_dword v3, v[0:1]
-; GFX940-NEXT: s_mov_b64 s[0:1], 0
-; GFX940-NEXT: .LBB0_1: ; %atomicrmw.start
-; GFX940-NEXT: ; =>This Inner Loop Header: Depth=1
-; GFX940-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
-; GFX940-NEXT: v_mov_b32_e32 v5, v3
-; GFX940-NEXT: v_pk_add_f16 v4, v5, v2
; GFX940-NEXT: buffer_wbl2 sc1
-; GFX940-NEXT: flat_atomic_cmpswap v3, v[0:1], v[4:5] sc0
+; GFX940-NEXT: flat_atomic_add_f32 v0, v[0:1], v2 offset:2044 sc0
; GFX940-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX940-NEXT: buffer_inv sc1
-; GFX940-NEXT: v_cmp_eq_u32_e32 vcc, v3, v5
-; GFX940-NEXT: s_or_b64 s[0:1], vcc, s[0:1]
-; GFX940-NEXT: s_andn2_b64 exec, exec, s[0:1]
-; GFX940-NEXT: s_cbranch_execnz .LBB0_1
-; GFX940-NEXT: ; %bb.2: ; %atomicrmw.end
-; GFX940-NEXT: s_or_b64 exec, exec, s[0:1]
-; GFX940-NEXT: v_mov_b32_e32 v0, v3
; GFX940-NEXT: s_setpc_b64 s[30:31]
;
-; GFX10-LABEL: flat_agent_atomic_fadd_ret_v2f16:
+; GFX11-LABEL: flat_agent_atomic_fadd_ret_f32__offset12b_pos:
+; GFX11: ; %bb.0:
+; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-NEXT: s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT: flat_atomic_add_f32 v0, v[0:1], v2 offset:2044 glc
+; GFX11-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX11-NEXT: buffer_gl1_inv
+; GFX11-NEXT: buffer_gl0_inv
+; GFX11-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX10-LABEL: flat_agent_atomic_fadd_ret_f32__offset12b_pos:
; GFX10: ; %bb.0:
; GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GFX10-NEXT: flat_load_dword v3, v[0:1]
+; GFX10-NEXT: v_add_co_u32 v3, vcc_lo, 0x7fc, v0
+; GFX10-NEXT: v_add_co_ci_u32_e32 v4, vcc_lo, 0, v1, vcc_lo
; GFX10-NEXT: s_mov_b32 s4, 0
-; GFX10-NEXT: .LBB0_1: ; %atomicrmw.start
+; GFX10-NEXT: flat_load_dword v0, v[3:4]
+; GFX10-NEXT: .LBB1_1: ; %atomicrmw.start
; GFX10-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX10-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
-; GFX10-NEXT: v_mov_b32_e32 v4, v3
-; GFX10-NEXT: v_pk_add_f16 v3, v4, v2
+; GFX10-NEXT: v_mov_b32_e32 v1, v0
+; GFX10-NEXT: v_add_f32_e32 v0, v1, v2
; GFX10-NEXT: s_waitcnt_vscnt null, 0x0
-; GFX10-NEXT: flat_atomic_cmpswap v3, v[0:1], v[3:4] glc
+; GFX10-NEXT: flat_atomic_cmpswap v0, v[3:4], v[0:1] glc
; GFX10-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX10-NEXT: buffer_gl1_inv
; GFX10-NEXT: buffer_gl0_inv
-; GFX10-NEXT: v_cmp_eq_u32_e32 vcc_lo, v3, v4
+; GFX10-NEXT: v_cmp_eq_u32_e32 vcc_lo, v0, v1
; GFX10-NEXT: s_or_b32 s4, vcc_lo, s4
; GFX10-NEXT: s_andn2_b32 exec_lo, exec_lo, s4
-; GFX10-NEXT: s_cbranch_execnz .LBB0_1
+; GFX10-NEXT: s_cbranch_execnz .LBB1_1
; GFX10-NEXT: ; %bb.2: ; %atomicrmw.end
; GFX10-NEXT: s_or_b32 exec_lo, exec_lo, s4
-; GFX10-NEXT: v_mov_b32_e32 v0, v3
; GFX10-NEXT: s_setpc_b64 s[30:31]
;
-; GFX11-LABEL: flat_agent_atomic_fadd_ret_v2f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GFX11-NEXT: flat_load_b32 v3, v[0:1]
-; GFX11-NEXT: s_mov_b32 s0, 0
-; GFX11-NEXT: .LBB0_1: ; %atomicrmw.start
-; GFX11-NEXT: ; =>This Inner Loop Header: Depth=1
-; GFX11-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
-; GFX11-NEXT: v_mov_b32_e32 v4, v3
-; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1)
-; GFX11-NEXT: v_pk_add_f16 v3, v4, v2
-; GFX11-NEXT: s_waitcnt_vscnt null, 0x0
-; GFX11-NEXT: flat_atomic_cmpswap_b32 v3, v[0:1], v[3:4] glc
-; GFX11-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
-; GFX11-NEXT: buffer_gl1_inv
-; GFX11-NEXT: buffer_gl0_inv
-; GFX11-NEXT: v_cmp_eq_u32_e32 vcc_lo, v3, v4
-; GFX11-NEXT: s_or_b32 s0, vcc_lo, s0
-; GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1)
-; GFX11-NEXT: s_and_not1_b32 exec_lo, exec_lo, s0
-; GFX11-NEXT: s_cbranch_execnz .LBB0_1
-; GFX11-NEXT: ; %bb.2: ; %atomicrmw.end
-; GFX11-NEXT: s_or_b32 exec_lo, exec_lo, s0
-; GFX11-NEXT: v_mov_b32_e32 v0, v3
-; GFX11-NEXT: s_setpc_b64 s[30:31]
-;
-; GFX12-LABEL: flat_agent_atomic_fadd_ret_v2f16:
-; GFX12: ; %bb.0:
-; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
-; GFX12-NEXT: s_wait_expcnt 0x0
-; GFX12-NEXT: s_wait_samplecnt 0x0
-; GFX12-NEXT: s_wait_bvhcnt 0x0
-; GFX12-NEXT: s_wait_kmcnt 0x0
-; GFX12-NEXT: flat_load_b32 v3, v[0:1]
-; GFX12-NEXT: s_mov_b32 s0, 0
-; GFX12-NEXT: .LBB0_1: ; %atomicrmw.start
-; GFX12-NEXT: ; =>This Inner Loop Header: Depth=1
-; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
-; GFX12-NEXT: v_mov_b32_e32 v4, v3
-; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1)
-; GFX12-NEXT: v_pk_add_f16 v3, v4, v2
-; GFX12-NEXT: s_wait_storecnt 0x0
-; GFX12-NEXT: flat_atomic_cmpswap_b32 v3, v[0:1], v[3:4] th:TH_ATOMIC_RETURN
-; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
-; GFX12-NEXT: global_inv scope:SCOPE_DEV
-; GFX12-NEXT: v_cmp_eq_u32_e32 vcc_lo, v3, v4
-; GFX12-NEXT: s_or_b32 s0, vcc_lo, s0
-; GFX12-NEXT: s_delay_alu instid0(SALU_CYCLE_1)
-; GFX12-NEXT: s_and_not1_b32 exec_lo, exec_lo, s0
-; GFX12-NEXT: s_cbranch_execnz .LBB0_1
-; GFX12-NEXT: ; %bb.2: ; %atomicrmw.end
-; GFX12-NEXT: s_or_b32 exec_lo, exec_lo, s0
-; GFX12-NEXT: v_mov_b32_e32 v0, v3
-; GFX12-NEXT: s_setpc_b64 s[30:31]
- %result = atomicrmw fadd ptr %ptr, <2 x half> %val syncscope("agent") seq_cst
- ret <2 x half> %result
-}
-
-define void @flat_agent_atomic_fadd_noret_v2f16(ptr %ptr, <2 x half> %val) {
-; GFX7-LABEL: flat_agent_atomic_fadd_noret_v2f16:
-; GFX7: ; %bb.0:
-; GFX7-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GFX7-NEXT: flat_load_dword v5, v[0:1]
-; GFX7-NEXT: v_cvt_f16_f32_e32 v3, v3
-; GFX7-NEXT: v_cvt_f16_f32_e32 v6, v2
-; GFX7-NEXT: s_mov_b64 s[4:5], 0
-; GFX7-NEXT: v_cvt_f32_f16_e32 v2, v3
-; GFX7-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
-; GFX7-NEXT: v_lshrrev_b32_e32 v3, 16, v5
-; GFX7-NEXT: v_cvt_f32_f16_e32 v4, v5
-; GFX7-NEXT: v_cvt_f32_f16_e32 v5, v3
-; GFX7-NEXT: v_cvt_f32_f16_e32 v3, v6
-; GFX7-NEXT: .LBB1_1: ; %atomicrmw.start
-; GFX7-NEXT: ; =>This Inner Loop Header: Depth=1
-; GFX7-NEXT: v_cvt_f16_f32_e32 v5, v5
-; GFX7-NEXT: v_cvt_f16_f32_e32 v4, v4
-; GFX7-NEXT: v_cvt_f32_f16_e32 v6, v5
-; GFX7-NEXT: v_cvt_f32_f16_e32 v7, v4
-; GFX7-NEXT: v_lshlrev_b32_e32 v5, 16, v5
-; GFX7-NEXT: v_add_f32_e32 v6, v6, v2
-; GFX7-NEXT: ...
[truncated]
``````````
</details>
https://github.com/llvm/llvm-project/pull/95131
More information about the llvm-commits
mailing list