[llvm] [AMDGPU] Allocate i1 argument to SGPRs (PR #72461)

[Matt Arsenault via llvm-commits]

================
@@ -0,0 +1,819 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 2
+; RUN: llc -mtriple=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GFX9 %s
+; RUN: llc -mtriple=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GFX11 %s
+
+define void @void_func_i1(i1 %arg0) {
+; GFX9-LABEL: void_func_i1:
+; GFX9:       ; %bb.0:
+; GFX9-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT:    v_cndmask_b32_e64 v0, 0, 1, s[4:5]
+; GFX9-NEXT:    global_store_byte v[0:1], v0, off
+; GFX9-NEXT:    s_waitcnt vmcnt(0)
+; GFX9-NEXT:    s_setpc_b64 s[30:31]
+;
+; GFX11-LABEL: void_func_i1:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-NEXT:    v_cndmask_b32_e64 v0, 0, 1, s0
+; GFX11-NEXT:    global_store_b8 v[0:1], v0, off
+; GFX11-NEXT:    s_setpc_b64 s[30:31]
+  store i1 %arg0, ptr addrspace(1) undef
+  ret void
+}
+
+define void @test_call_void_func_i1() {
+; GFX9-LABEL: test_call_void_func_i1:
+; GFX9:    ; %bb.0:
+; GFX9-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT:    s_mov_b32 s8, s33
+; GFX9-NEXT:    s_mov_b32 s33, s32
+; GFX9-NEXT:    s_xor_saveexec_b64 s[4:5], -1
+; GFX9-NEXT:    buffer_store_dword v2, off, s[0:3], s33 ; 4-byte Folded Spill
+; GFX9-NEXT:    s_mov_b64 exec, s[4:5]
+; GFX9-NEXT:    global_load_ubyte v0, v[0:1], off
+; GFX9-NEXT:    s_addk_i32 s32, 0x400
+; GFX9-NEXT:    s_getpc_b64 s[4:5]
+; GFX9-NEXT:    s_add_u32 s4, s4, void_func_i1 at gotpcrel32@lo+4
+; GFX9-NEXT:    s_addc_u32 s5, s5, void_func_i1 at gotpcrel32@hi+12
+; GFX9-NEXT:    s_load_dwordx2 s[6:7], s[4:5], 0x0
+; GFX9-NEXT:    v_writelane_b32 v2, s30, 0
+; GFX9-NEXT:    v_writelane_b32 v2, s31, 1
+; GFX9-NEXT:    s_waitcnt vmcnt(0)
+; GFX9-NEXT:    v_and_b32_e32 v0, 1, v0
+; GFX9-NEXT:    v_cmp_eq_u32_e64 s[4:5], 1, v0
+; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
+; GFX9-NEXT:    s_swappc_b64 s[30:31], s[6:7]
+;
+; GFX11-LABEL: test_call_void_func_i1:
+; GFX11:     ; %bb.0:
+; GFX11-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-NEXT:    s_mov_b32 s4, s33
+; GFX11-NEXT:    s_mov_b32 s33, s32
+; GFX11-NEXT:    s_xor_saveexec_b32 s0, -1
+; GFX11-NEXT:    scratch_store_b32 off, v2, s33          ; 4-byte Folded Spill
+; GFX11-NEXT:    s_mov_b32 exec_lo, s0
+; GFX11-NEXT:    global_load_u8 v0, v[0:1], off
+; GFX11-NEXT:    s_add_i32 s32, s32, 16
+; GFX11-NEXT:    s_getpc_b64 s[0:1]
+; GFX11-NEXT:    s_add_u32 s0, s0, void_func_i1 at gotpcrel32@lo+4
+; GFX11-NEXT:    s_addc_u32 s1, s1, void_func_i1 at gotpcrel32@hi+12
+; GFX11-NEXT:    v_writelane_b32 v2, s30, 0
+; GFX11-NEXT:    s_load_b64 s[2:3], s[0:1], 0x0
+; GFX11-NEXT:    v_writelane_b32 v2, s31, 1
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    v_and_b32_e32 v0, 1, v0
+; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_1)
+; GFX11-NEXT:    v_cmp_eq_u32_e64 s0, 1, v0
+; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
+; GFX11-NEXT:    s_swappc_b64 s[30:31], s[2:3]
+  %val = load i1, ptr addrspace(1) undef
----------------
arsenm wrote:

Don't use an undef pointer 

https://github.com/llvm/llvm-project/pull/72461