caio/clang-p2996
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
12ccdd67aaf56bd05ab3375d81c58dfd5a2db02c
clang-p2996/llvm/test/CodeGen/AMDGPU/idiv-licm.ll
Jay Foad 3eb2281bc0 [AMDGPU] Aggressively fold immediates in SIFoldOperands 
Previously SIFoldOperands::foldInstOperand would only fold a
non-inlinable immediate into a single user, so as not to increase code
size by adding the same 32-bit literal operand to many instructions.

This patch removes that restriction, so that a non-inlinable immediate
will be folded into any number of users. The rationale is:
- It reduces the number of registers used for holding constant values,
  which might increase occupancy. (On the other hand, many of these
  registers are SGPRs which no longer affect occupancy on GFX10+.)
- It reduces ALU stalls between the instruction that loads a constant
  into a register, and the instruction that uses it.
- The above benefits are expected to outweigh any increase in code size.

Differential Revision: https://reviews.llvm.org/D114643
2022-05-18 10:19:35 +01:00

756 lines
29 KiB
LLVM
Raw Blame History

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GFX9 %s
; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GFX10 %s
define amdgpu_kernel void @udiv32_invariant_denom(i32 addrspace(1)* nocapture %arg, i32 %arg1) {
; GFX9-LABEL: udiv32_invariant_denom:
; GFX9: ; %bb.0: ; %bb
; GFX9-NEXT: s_load_dword s4, s[0:1], 0x2c
; GFX9-NEXT: s_mov_b64 s[2:3], 0
; GFX9-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x24
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: v_cvt_f32_u32_e32 v0, s4
; GFX9-NEXT: s_sub_i32 s5, 0, s4
; GFX9-NEXT: v_rcp_iflag_f32_e32 v0, v0
; GFX9-NEXT: v_mul_f32_e32 v0, 0x4f7ffffe, v0
; GFX9-NEXT: v_cvt_u32_f32_e32 v0, v0
; GFX9-NEXT: v_mul_lo_u32 v1, s5, v0
; GFX9-NEXT: v_mul_hi_u32 v1, v0, v1
; GFX9-NEXT: v_add_u32_e32 v0, v0, v1
; GFX9-NEXT: v_mov_b32_e32 v1, 0
; GFX9-NEXT: .LBB0_1: ; %bb3
; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT: v_mul_lo_u32 v2, s3, v0
; GFX9-NEXT: v_mul_hi_u32 v3, s2, v0
; GFX9-NEXT: v_add_u32_e32 v2, v3, v2
; GFX9-NEXT: v_mul_lo_u32 v3, s5, v2
; GFX9-NEXT: v_not_b32_e32 v5, v2
; GFX9-NEXT: v_mul_lo_u32 v5, s4, v5
; GFX9-NEXT: v_add_u32_e32 v4, 1, v2
; GFX9-NEXT: v_add_u32_e32 v3, s2, v3
; GFX9-NEXT: v_cmp_le_u32_e32 vcc, s4, v3
; GFX9-NEXT: v_cndmask_b32_e32 v2, v2, v4, vcc
; GFX9-NEXT: v_add_u32_e32 v4, s2, v5
; GFX9-NEXT: v_cndmask_b32_e32 v3, v3, v4, vcc
; GFX9-NEXT: s_add_u32 s2, s2, 1
; GFX9-NEXT: v_add_u32_e32 v4, 1, v2
; GFX9-NEXT: v_cmp_le_u32_e32 vcc, s4, v3
; GFX9-NEXT: s_addc_u32 s3, s3, 0
; GFX9-NEXT: v_cndmask_b32_e32 v2, v2, v4, vcc
; GFX9-NEXT: global_store_dword v1, v2, s[0:1]
; GFX9-NEXT: s_add_u32 s0, s0, 4
; GFX9-NEXT: s_addc_u32 s1, s1, 0
; GFX9-NEXT: s_cmpk_eq_i32 s2, 0x400
; GFX9-NEXT: s_cbranch_scc0 .LBB0_1
; GFX9-NEXT: ; %bb.2: ; %bb2
; GFX9-NEXT: s_endpgm
;
; GFX10-LABEL: udiv32_invariant_denom:
; GFX10: ; %bb.0: ; %bb
; GFX10-NEXT: s_load_dword s4, s[0:1], 0x2c
; GFX10-NEXT: s_mov_b64 s[2:3], 0
; GFX10-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x24
; GFX10-NEXT: s_waitcnt lgkmcnt(0)
; GFX10-NEXT: v_cvt_f32_u32_e32 v0, s4
; GFX10-NEXT: s_sub_i32 s5, 0, s4
; GFX10-NEXT: v_rcp_iflag_f32_e32 v0, v0
; GFX10-NEXT: v_mul_f32_e32 v0, 0x4f7ffffe, v0
; GFX10-NEXT: v_cvt_u32_f32_e32 v0, v0
; GFX10-NEXT: v_mul_lo_u32 v1, s5, v0
; GFX10-NEXT: v_mul_hi_u32 v1, v0, v1
; GFX10-NEXT: v_add_nc_u32_e32 v0, v0, v1
; GFX10-NEXT: v_mov_b32_e32 v1, 0
; GFX10-NEXT: .LBB0_1: ; %bb3
; GFX10-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX10-NEXT: v_mul_lo_u32 v2, s3, v0
; GFX10-NEXT: v_mul_hi_u32 v3, s2, v0
; GFX10-NEXT: v_add_nc_u32_e32 v2, v3, v2
; GFX10-NEXT: v_not_b32_e32 v3, v2
; GFX10-NEXT: v_mul_lo_u32 v4, s5, v2
; GFX10-NEXT: v_add_nc_u32_e32 v5, 1, v2
; GFX10-NEXT: v_mul_lo_u32 v3, s4, v3
; GFX10-NEXT: v_add_nc_u32_e32 v4, s2, v4
; GFX10-NEXT: v_add_nc_u32_e32 v3, s2, v3
; GFX10-NEXT: v_cmp_le_u32_e32 vcc_lo, s4, v4
; GFX10-NEXT: s_add_u32 s2, s2, 1
; GFX10-NEXT: s_addc_u32 s3, s3, 0
; GFX10-NEXT: v_cndmask_b32_e32 v2, v2, v5, vcc_lo
; GFX10-NEXT: v_cndmask_b32_e32 v3, v4, v3, vcc_lo
; GFX10-NEXT: v_add_nc_u32_e32 v4, 1, v2
; GFX10-NEXT: v_cmp_le_u32_e32 vcc_lo, s4, v3
; GFX10-NEXT: v_cndmask_b32_e32 v2, v2, v4, vcc_lo
; GFX10-NEXT: global_store_dword v1, v2, s[0:1]
; GFX10-NEXT: s_waitcnt_depctr 0xffe3
; GFX10-NEXT: s_add_u32 s0, s0, 4
; GFX10-NEXT: s_addc_u32 s1, s1, 0
; GFX10-NEXT: s_cmpk_eq_i32 s2, 0x400
; GFX10-NEXT: s_cbranch_scc0 .LBB0_1
; GFX10-NEXT: ; %bb.2: ; %bb2
; GFX10-NEXT: s_endpgm
bb:
br label %bb3
bb2: ; preds = %bb3
ret void
bb3: ; preds = %bb3, %bb
%tmp = phi i32 [ 0, %bb ], [ %tmp7, %bb3 ]
%tmp4 = udiv i32 %tmp, %arg1
%tmp5 = zext i32 %tmp to i64
%tmp6 = getelementptr inbounds i32, i32 addrspace(1)* %arg, i64 %tmp5
store i32 %tmp4, i32 addrspace(1)* %tmp6, align 4
%tmp7 = add nuw nsw i32 %tmp, 1
%tmp8 = icmp eq i32 %tmp7, 1024
br i1 %tmp8, label %bb2, label %bb3
}
define amdgpu_kernel void @urem32_invariant_denom(i32 addrspace(1)* nocapture %arg, i32 %arg1) {
; GFX9-LABEL: urem32_invariant_denom:
; GFX9: ; %bb.0: ; %bb
; GFX9-NEXT: s_load_dword s4, s[0:1], 0x2c
; GFX9-NEXT: s_mov_b64 s[2:3], 0
; GFX9-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x24
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: v_cvt_f32_u32_e32 v0, s4
; GFX9-NEXT: s_sub_i32 s5, 0, s4
; GFX9-NEXT: v_rcp_iflag_f32_e32 v0, v0
; GFX9-NEXT: v_mul_f32_e32 v0, 0x4f7ffffe, v0
; GFX9-NEXT: v_cvt_u32_f32_e32 v0, v0
; GFX9-NEXT: v_mul_lo_u32 v1, s5, v0
; GFX9-NEXT: v_mul_hi_u32 v1, v0, v1
; GFX9-NEXT: v_add_u32_e32 v0, v0, v1
; GFX9-NEXT: v_mov_b32_e32 v1, 0
; GFX9-NEXT: .LBB1_1: ; %bb3
; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT: v_mul_lo_u32 v2, s3, v0
; GFX9-NEXT: v_mul_hi_u32 v3, s2, v0
; GFX9-NEXT: v_add_u32_e32 v2, v3, v2
; GFX9-NEXT: v_mul_lo_u32 v3, s5, v2
; GFX9-NEXT: v_not_b32_e32 v2, v2
; GFX9-NEXT: v_mul_lo_u32 v2, s4, v2
; GFX9-NEXT: v_add_u32_e32 v3, s2, v3
; GFX9-NEXT: v_cmp_le_u32_e32 vcc, s4, v3
; GFX9-NEXT: v_add_u32_e32 v2, s2, v2
; GFX9-NEXT: v_cndmask_b32_e32 v2, v3, v2, vcc
; GFX9-NEXT: s_add_u32 s2, s2, 1
; GFX9-NEXT: v_subrev_u32_e32 v3, s4, v2
; GFX9-NEXT: v_cmp_le_u32_e32 vcc, s4, v2
; GFX9-NEXT: s_addc_u32 s3, s3, 0
; GFX9-NEXT: v_cndmask_b32_e32 v2, v2, v3, vcc
; GFX9-NEXT: global_store_dword v1, v2, s[0:1]
; GFX9-NEXT: s_add_u32 s0, s0, 4
; GFX9-NEXT: s_addc_u32 s1, s1, 0
; GFX9-NEXT: s_cmpk_eq_i32 s2, 0x400
; GFX9-NEXT: s_cbranch_scc0 .LBB1_1
; GFX9-NEXT: ; %bb.2: ; %bb2
; GFX9-NEXT: s_endpgm
;
; GFX10-LABEL: urem32_invariant_denom:
; GFX10: ; %bb.0: ; %bb
; GFX10-NEXT: s_load_dword s4, s[0:1], 0x2c
; GFX10-NEXT: s_mov_b64 s[2:3], 0
; GFX10-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x24
; GFX10-NEXT: s_waitcnt lgkmcnt(0)
; GFX10-NEXT: v_cvt_f32_u32_e32 v0, s4
; GFX10-NEXT: s_sub_i32 s5, 0, s4
; GFX10-NEXT: v_rcp_iflag_f32_e32 v0, v0
; GFX10-NEXT: v_mul_f32_e32 v0, 0x4f7ffffe, v0
; GFX10-NEXT: v_cvt_u32_f32_e32 v0, v0
; GFX10-NEXT: v_mul_lo_u32 v1, s5, v0
; GFX10-NEXT: v_mul_hi_u32 v1, v0, v1
; GFX10-NEXT: v_add_nc_u32_e32 v0, v0, v1
; GFX10-NEXT: v_mov_b32_e32 v1, 0
; GFX10-NEXT: .LBB1_1: ; %bb3
; GFX10-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX10-NEXT: v_mul_lo_u32 v2, s3, v0
; GFX10-NEXT: v_mul_hi_u32 v3, s2, v0
; GFX10-NEXT: v_add_nc_u32_e32 v2, v3, v2
; GFX10-NEXT: v_not_b32_e32 v3, v2
; GFX10-NEXT: v_mul_lo_u32 v2, s5, v2
; GFX10-NEXT: v_mul_lo_u32 v3, s4, v3
; GFX10-NEXT: v_add_nc_u32_e32 v2, s2, v2
; GFX10-NEXT: v_add_nc_u32_e32 v3, s2, v3
; GFX10-NEXT: v_cmp_le_u32_e32 vcc_lo, s4, v2
; GFX10-NEXT: s_add_u32 s2, s2, 1
; GFX10-NEXT: s_addc_u32 s3, s3, 0
; GFX10-NEXT: v_cndmask_b32_e32 v2, v2, v3, vcc_lo
; GFX10-NEXT: v_subrev_nc_u32_e32 v3, s4, v2
; GFX10-NEXT: v_cmp_le_u32_e32 vcc_lo, s4, v2
; GFX10-NEXT: v_cndmask_b32_e32 v2, v2, v3, vcc_lo
; GFX10-NEXT: global_store_dword v1, v2, s[0:1]
; GFX10-NEXT: s_waitcnt_depctr 0xffe3
; GFX10-NEXT: s_add_u32 s0, s0, 4
; GFX10-NEXT: s_addc_u32 s1, s1, 0
; GFX10-NEXT: s_cmpk_eq_i32 s2, 0x400
; GFX10-NEXT: s_cbranch_scc0 .LBB1_1
; GFX10-NEXT: ; %bb.2: ; %bb2
; GFX10-NEXT: s_endpgm
bb:
br label %bb3
bb2: ; preds = %bb3
ret void
bb3: ; preds = %bb3, %bb
%tmp = phi i32 [ 0, %bb ], [ %tmp7, %bb3 ]
%tmp4 = urem i32 %tmp, %arg1
%tmp5 = zext i32 %tmp to i64
%tmp6 = getelementptr inbounds i32, i32 addrspace(1)* %arg, i64 %tmp5
store i32 %tmp4, i32 addrspace(1)* %tmp6, align 4
%tmp7 = add nuw nsw i32 %tmp, 1
%tmp8 = icmp eq i32 %tmp7, 1024
br i1 %tmp8, label %bb2, label %bb3
}
define amdgpu_kernel void @sdiv32_invariant_denom(i32 addrspace(1)* nocapture %arg, i32 %arg1) {
; GFX9-LABEL: sdiv32_invariant_denom:
; GFX9: ; %bb.0: ; %bb
; GFX9-NEXT: s_load_dword s3, s[0:1], 0x2c
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: s_ashr_i32 s2, s3, 31
; GFX9-NEXT: s_add_i32 s3, s3, s2
; GFX9-NEXT: s_xor_b32 s3, s3, s2
; GFX9-NEXT: v_cvt_f32_u32_e32 v0, s3
; GFX9-NEXT: s_sub_i32 s4, 0, s3
; GFX9-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x24
; GFX9-NEXT: v_rcp_iflag_f32_e32 v0, v0
; GFX9-NEXT: v_mul_f32_e32 v0, 0x4f7ffffe, v0
; GFX9-NEXT: v_cvt_u32_f32_e32 v0, v0
; GFX9-NEXT: v_mul_lo_u32 v1, s4, v0
; GFX9-NEXT: s_mov_b32 s4, 0
; GFX9-NEXT: v_mul_hi_u32 v1, v0, v1
; GFX9-NEXT: v_add_u32_e32 v0, v0, v1
; GFX9-NEXT: v_mov_b32_e32 v1, 0
; GFX9-NEXT: .LBB2_1: ; %bb3
; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT: v_mul_hi_u32 v2, s4, v0
; GFX9-NEXT: v_mul_lo_u32 v3, v2, s3
; GFX9-NEXT: v_add_u32_e32 v4, 1, v2
; GFX9-NEXT: v_sub_u32_e32 v3, s4, v3
; GFX9-NEXT: v_cmp_le_u32_e32 vcc, s3, v3
; GFX9-NEXT: v_cndmask_b32_e32 v2, v2, v4, vcc
; GFX9-NEXT: v_subrev_u32_e32 v4, s3, v3
; GFX9-NEXT: v_cndmask_b32_e32 v3, v3, v4, vcc
; GFX9-NEXT: v_add_u32_e32 v4, 1, v2
; GFX9-NEXT: v_cmp_le_u32_e32 vcc, s3, v3
; GFX9-NEXT: v_cndmask_b32_e32 v2, v2, v4, vcc
; GFX9-NEXT: v_xor_b32_e32 v2, s2, v2
; GFX9-NEXT: s_add_i32 s4, s4, 1
; GFX9-NEXT: v_subrev_u32_e32 v2, s2, v2
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: global_store_dword v1, v2, s[0:1]
; GFX9-NEXT: s_add_u32 s0, s0, 4
; GFX9-NEXT: s_addc_u32 s1, s1, 0
; GFX9-NEXT: s_cmpk_eq_i32 s4, 0x400
; GFX9-NEXT: s_cbranch_scc0 .LBB2_1
; GFX9-NEXT: ; %bb.2: ; %bb2
; GFX9-NEXT: s_endpgm
;
; GFX10-LABEL: sdiv32_invariant_denom:
; GFX10: ; %bb.0: ; %bb
; GFX10-NEXT: s_load_dword s3, s[0:1], 0x2c
; GFX10-NEXT: s_waitcnt lgkmcnt(0)
; GFX10-NEXT: s_ashr_i32 s2, s3, 31
; GFX10-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x24
; GFX10-NEXT: s_add_i32 s3, s3, s2
; GFX10-NEXT: s_xor_b32 s3, s3, s2
; GFX10-NEXT: v_cvt_f32_u32_e32 v0, s3
; GFX10-NEXT: s_sub_i32 s4, 0, s3
; GFX10-NEXT: v_rcp_iflag_f32_e32 v0, v0
; GFX10-NEXT: v_mul_f32_e32 v0, 0x4f7ffffe, v0
; GFX10-NEXT: v_cvt_u32_f32_e32 v0, v0
; GFX10-NEXT: v_mul_lo_u32 v1, s4, v0
; GFX10-NEXT: s_mov_b32 s4, 0
; GFX10-NEXT: v_mul_hi_u32 v1, v0, v1
; GFX10-NEXT: v_add_nc_u32_e32 v0, v0, v1
; GFX10-NEXT: v_mov_b32_e32 v1, 0
; GFX10-NEXT: .LBB2_1: ; %bb3
; GFX10-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX10-NEXT: v_mul_hi_u32 v2, s4, v0
; GFX10-NEXT: v_mul_lo_u32 v3, v2, s3
; GFX10-NEXT: v_add_nc_u32_e32 v4, 1, v2
; GFX10-NEXT: v_sub_nc_u32_e32 v3, s4, v3
; GFX10-NEXT: s_add_i32 s4, s4, 1
; GFX10-NEXT: v_subrev_nc_u32_e32 v5, s3, v3
; GFX10-NEXT: v_cmp_le_u32_e32 vcc_lo, s3, v3
; GFX10-NEXT: v_cndmask_b32_e32 v2, v2, v4, vcc_lo
; GFX10-NEXT: v_cndmask_b32_e32 v3, v3, v5, vcc_lo
; GFX10-NEXT: v_add_nc_u32_e32 v4, 1, v2
; GFX10-NEXT: v_cmp_le_u32_e32 vcc_lo, s3, v3
; GFX10-NEXT: v_cndmask_b32_e32 v2, v2, v4, vcc_lo
; GFX10-NEXT: v_xor_b32_e32 v2, s2, v2
; GFX10-NEXT: v_subrev_nc_u32_e32 v2, s2, v2
; GFX10-NEXT: s_waitcnt lgkmcnt(0)
; GFX10-NEXT: global_store_dword v1, v2, s[0:1]
; GFX10-NEXT: s_waitcnt_depctr 0xffe3
; GFX10-NEXT: s_add_u32 s0, s0, 4
; GFX10-NEXT: s_addc_u32 s1, s1, 0
; GFX10-NEXT: s_cmpk_eq_i32 s4, 0x400
; GFX10-NEXT: s_cbranch_scc0 .LBB2_1
; GFX10-NEXT: ; %bb.2: ; %bb2
; GFX10-NEXT: s_endpgm
bb:
br label %bb3
bb2: ; preds = %bb3
ret void
bb3: ; preds = %bb3, %bb
%tmp = phi i32 [ 0, %bb ], [ %tmp7, %bb3 ]
%tmp4 = sdiv i32 %tmp, %arg1
%tmp5 = zext i32 %tmp to i64
%tmp6 = getelementptr inbounds i32, i32 addrspace(1)* %arg, i64 %tmp5
store i32 %tmp4, i32 addrspace(1)* %tmp6, align 4
%tmp7 = add nuw nsw i32 %tmp, 1
%tmp8 = icmp eq i32 %tmp7, 1024
br i1 %tmp8, label %bb2, label %bb3
}
define amdgpu_kernel void @srem32_invariant_denom(i32 addrspace(1)* nocapture %arg, i32 %arg1) {
; GFX9-LABEL: srem32_invariant_denom:
; GFX9: ; %bb.0: ; %bb
; GFX9-NEXT: s_load_dword s2, s[0:1], 0x2c
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: s_ashr_i32 s3, s2, 31
; GFX9-NEXT: s_add_i32 s2, s2, s3
; GFX9-NEXT: s_xor_b32 s2, s2, s3
; GFX9-NEXT: v_cvt_f32_u32_e32 v0, s2
; GFX9-NEXT: s_sub_i32 s3, 0, s2
; GFX9-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x24
; GFX9-NEXT: v_rcp_iflag_f32_e32 v0, v0
; GFX9-NEXT: v_mul_f32_e32 v0, 0x4f7ffffe, v0
; GFX9-NEXT: v_cvt_u32_f32_e32 v0, v0
; GFX9-NEXT: v_mul_lo_u32 v1, s3, v0
; GFX9-NEXT: s_mov_b32 s3, 0
; GFX9-NEXT: v_mul_hi_u32 v1, v0, v1
; GFX9-NEXT: v_add_u32_e32 v0, v0, v1
; GFX9-NEXT: v_mov_b32_e32 v1, 0
; GFX9-NEXT: .LBB3_1: ; %bb3
; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT: v_mul_hi_u32 v2, s3, v0
; GFX9-NEXT: v_mul_lo_u32 v2, v2, s2
; GFX9-NEXT: v_sub_u32_e32 v2, s3, v2
; GFX9-NEXT: v_subrev_u32_e32 v3, s2, v2
; GFX9-NEXT: v_cmp_le_u32_e32 vcc, s2, v2
; GFX9-NEXT: v_cndmask_b32_e32 v2, v2, v3, vcc
; GFX9-NEXT: v_subrev_u32_e32 v3, s2, v2
; GFX9-NEXT: v_cmp_le_u32_e32 vcc, s2, v2
; GFX9-NEXT: s_add_i32 s3, s3, 1
; GFX9-NEXT: v_cndmask_b32_e32 v2, v2, v3, vcc
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: global_store_dword v1, v2, s[0:1]
; GFX9-NEXT: s_add_u32 s0, s0, 4
; GFX9-NEXT: s_addc_u32 s1, s1, 0
; GFX9-NEXT: s_cmpk_eq_i32 s3, 0x400
; GFX9-NEXT: s_cbranch_scc0 .LBB3_1
; GFX9-NEXT: ; %bb.2: ; %bb2
; GFX9-NEXT: s_endpgm
;
; GFX10-LABEL: srem32_invariant_denom:
; GFX10: ; %bb.0: ; %bb
; GFX10-NEXT: s_load_dword s2, s[0:1], 0x2c
; GFX10-NEXT: s_waitcnt lgkmcnt(0)
; GFX10-NEXT: s_ashr_i32 s3, s2, 31
; GFX10-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x24
; GFX10-NEXT: s_add_i32 s2, s2, s3
; GFX10-NEXT: s_xor_b32 s2, s2, s3
; GFX10-NEXT: v_cvt_f32_u32_e32 v0, s2
; GFX10-NEXT: s_sub_i32 s3, 0, s2
; GFX10-NEXT: v_rcp_iflag_f32_e32 v0, v0
; GFX10-NEXT: v_mul_f32_e32 v0, 0x4f7ffffe, v0
; GFX10-NEXT: v_cvt_u32_f32_e32 v0, v0
; GFX10-NEXT: v_mul_lo_u32 v1, s3, v0
; GFX10-NEXT: s_mov_b32 s3, 0
; GFX10-NEXT: v_mul_hi_u32 v1, v0, v1
; GFX10-NEXT: v_add_nc_u32_e32 v0, v0, v1
; GFX10-NEXT: v_mov_b32_e32 v1, 0
; GFX10-NEXT: .LBB3_1: ; %bb3
; GFX10-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX10-NEXT: v_mul_hi_u32 v2, s3, v0
; GFX10-NEXT: v_mul_lo_u32 v2, v2, s2
; GFX10-NEXT: v_sub_nc_u32_e32 v2, s3, v2
; GFX10-NEXT: s_add_i32 s3, s3, 1
; GFX10-NEXT: v_subrev_nc_u32_e32 v3, s2, v2
; GFX10-NEXT: v_cmp_le_u32_e32 vcc_lo, s2, v2
; GFX10-NEXT: v_cndmask_b32_e32 v2, v2, v3, vcc_lo
; GFX10-NEXT: v_subrev_nc_u32_e32 v3, s2, v2
; GFX10-NEXT: v_cmp_le_u32_e32 vcc_lo, s2, v2
; GFX10-NEXT: v_cndmask_b32_e32 v2, v2, v3, vcc_lo
; GFX10-NEXT: s_waitcnt lgkmcnt(0)
; GFX10-NEXT: global_store_dword v1, v2, s[0:1]
; GFX10-NEXT: s_waitcnt_depctr 0xffe3
; GFX10-NEXT: s_add_u32 s0, s0, 4
; GFX10-NEXT: s_addc_u32 s1, s1, 0
; GFX10-NEXT: s_cmpk_eq_i32 s3, 0x400
; GFX10-NEXT: s_cbranch_scc0 .LBB3_1
; GFX10-NEXT: ; %bb.2: ; %bb2
; GFX10-NEXT: s_endpgm
bb:
br label %bb3
bb2: ; preds = %bb3
ret void
bb3: ; preds = %bb3, %bb
%tmp = phi i32 [ 0, %bb ], [ %tmp7, %bb3 ]
%tmp4 = srem i32 %tmp, %arg1
%tmp5 = zext i32 %tmp to i64
%tmp6 = getelementptr inbounds i32, i32 addrspace(1)* %arg, i64 %tmp5
store i32 %tmp4, i32 addrspace(1)* %tmp6, align 4
%tmp7 = add nuw nsw i32 %tmp, 1
%tmp8 = icmp eq i32 %tmp7, 1024
br i1 %tmp8, label %bb2, label %bb3
}
define amdgpu_kernel void @udiv16_invariant_denom(i16 addrspace(1)* nocapture %arg, i16 %arg1) {
; GFX9-LABEL: udiv16_invariant_denom:
; GFX9: ; %bb.0: ; %bb
; GFX9-NEXT: s_load_dword s2, s[0:1], 0x2c
; GFX9-NEXT: v_mov_b32_e32 v1, 0
; GFX9-NEXT: s_movk_i32 s4, 0x400
; GFX9-NEXT: v_mov_b32_e32 v4, 0
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: s_and_b32 s2, 0xffff, s2
; GFX9-NEXT: v_cvt_f32_u32_e32 v2, s2
; GFX9-NEXT: s_load_dwordx2 s[2:3], s[0:1], 0x24
; GFX9-NEXT: v_rcp_iflag_f32_e32 v3, v2
; GFX9-NEXT: .LBB4_1: ; %bb3
; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT: v_and_b32_e32 v0, 0xffff, v4
; GFX9-NEXT: v_cvt_f32_u32_e32 v8, v0
; GFX9-NEXT: v_lshlrev_b64 v[5:6], 1, v[0:1]
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: v_mov_b32_e32 v7, s3
; GFX9-NEXT: v_add_co_u32_e64 v5, s[0:1], s2, v5
; GFX9-NEXT: v_mul_f32_e32 v0, v8, v3
; GFX9-NEXT: v_trunc_f32_e32 v0, v0
; GFX9-NEXT: v_addc_co_u32_e64 v6, s[0:1], v7, v6, s[0:1]
; GFX9-NEXT: v_cvt_u32_f32_e32 v7, v0
; GFX9-NEXT: v_add_u16_e32 v4, 1, v4
; GFX9-NEXT: v_mad_f32 v0, -v0, v2, v8
; GFX9-NEXT: v_cmp_eq_u16_e32 vcc, s4, v4
; GFX9-NEXT: v_cmp_ge_f32_e64 s[0:1], |v0|, v2
; GFX9-NEXT: v_addc_co_u32_e64 v0, s[0:1], 0, v7, s[0:1]
; GFX9-NEXT: global_store_short v[5:6], v0, off
; GFX9-NEXT: s_cbranch_vccz .LBB4_1
; GFX9-NEXT: ; %bb.2: ; %bb2
; GFX9-NEXT: s_endpgm
;
; GFX10-LABEL: udiv16_invariant_denom:
; GFX10: ; %bb.0: ; %bb
; GFX10-NEXT: s_clause 0x1
; GFX10-NEXT: s_load_dword s4, s[0:1], 0x2c
; GFX10-NEXT: s_load_dwordx2 s[2:3], s[0:1], 0x24
; GFX10-NEXT: v_mov_b32_e32 v1, 0
; GFX10-NEXT: v_mov_b32_e32 v4, 0
; GFX10-NEXT: s_waitcnt lgkmcnt(0)
; GFX10-NEXT: s_and_b32 s0, 0xffff, s4
; GFX10-NEXT: v_cvt_f32_u32_e32 v2, s0
; GFX10-NEXT: v_rcp_iflag_f32_e32 v3, v2
; GFX10-NEXT: .LBB4_1: ; %bb3
; GFX10-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX10-NEXT: v_and_b32_e32 v0, 0xffff, v4
; GFX10-NEXT: v_add_nc_u16 v4, v4, 1
; GFX10-NEXT: v_cvt_f32_u32_e32 v7, v0
; GFX10-NEXT: v_lshlrev_b64 v[5:6], 1, v[0:1]
; GFX10-NEXT: v_cmp_eq_u16_e32 vcc_lo, 0x400, v4
; GFX10-NEXT: v_mul_f32_e32 v0, v7, v3
; GFX10-NEXT: v_add_co_u32 v5, s0, s2, v5
; GFX10-NEXT: v_add_co_ci_u32_e64 v6, s0, s3, v6, s0
; GFX10-NEXT: v_trunc_f32_e32 v0, v0
; GFX10-NEXT: s_and_b32 vcc_lo, exec_lo, vcc_lo
; GFX10-NEXT: v_mad_f32 v7, -v0, v2, v7
; GFX10-NEXT: v_cvt_u32_f32_e32 v0, v0
; GFX10-NEXT: v_cmp_ge_f32_e64 s0, |v7|, v2
; GFX10-NEXT: v_add_co_ci_u32_e64 v0, s0, 0, v0, s0
; GFX10-NEXT: global_store_short v[5:6], v0, off
; GFX10-NEXT: s_cbranch_vccz .LBB4_1
; GFX10-NEXT: ; %bb.2: ; %bb2
; GFX10-NEXT: s_endpgm
bb:
br label %bb3
bb2: ; preds = %bb3
ret void
bb3: ; preds = %bb3, %bb
%tmp = phi i16 [ 0, %bb ], [ %tmp7, %bb3 ]
%tmp4 = udiv i16 %tmp, %arg1
%tmp5 = zext i16 %tmp to i64
%tmp6 = getelementptr inbounds i16, i16 addrspace(1)* %arg, i64 %tmp5
store i16 %tmp4, i16 addrspace(1)* %tmp6, align 2
%tmp7 = add nuw nsw i16 %tmp, 1
%tmp8 = icmp eq i16 %tmp7, 1024
br i1 %tmp8, label %bb2, label %bb3
}
define amdgpu_kernel void @urem16_invariant_denom(i16 addrspace(1)* nocapture %arg, i16 %arg1) {
; GFX9-LABEL: urem16_invariant_denom:
; GFX9: ; %bb.0: ; %bb
; GFX9-NEXT: s_load_dword s2, s[0:1], 0x2c
; GFX9-NEXT: s_load_dwordx2 s[4:5], s[0:1], 0x24
; GFX9-NEXT: v_mov_b32_e32 v1, 0
; GFX9-NEXT: s_movk_i32 s7, 0x400
; GFX9-NEXT: v_mov_b32_e32 v4, 0
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: s_and_b32 s6, 0xffff, s2
; GFX9-NEXT: v_cvt_f32_u32_e32 v2, s6
; GFX9-NEXT: v_rcp_iflag_f32_e32 v3, v2
; GFX9-NEXT: .LBB5_1: ; %bb3
; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT: v_and_b32_e32 v0, 0xffff, v4
; GFX9-NEXT: v_cvt_f32_u32_e32 v8, v0
; GFX9-NEXT: v_lshlrev_b64 v[5:6], 1, v[0:1]
; GFX9-NEXT: v_add_u16_e32 v4, 1, v4
; GFX9-NEXT: v_mov_b32_e32 v7, s5
; GFX9-NEXT: v_mul_f32_e32 v9, v8, v3
; GFX9-NEXT: v_trunc_f32_e32 v9, v9
; GFX9-NEXT: v_cvt_u32_f32_e32 v10, v9
; GFX9-NEXT: v_mad_f32 v8, -v9, v2, v8
; GFX9-NEXT: v_cmp_ge_f32_e64 s[2:3], |v8|, v2
; GFX9-NEXT: v_cmp_eq_u16_e32 vcc, s7, v4
; GFX9-NEXT: v_addc_co_u32_e64 v8, s[2:3], 0, v10, s[2:3]
; GFX9-NEXT: v_mul_lo_u32 v8, v8, s6
; GFX9-NEXT: v_add_co_u32_e64 v5, s[0:1], s4, v5
; GFX9-NEXT: v_addc_co_u32_e64 v6, s[0:1], v7, v6, s[0:1]
; GFX9-NEXT: v_sub_u32_e32 v0, v0, v8
; GFX9-NEXT: global_store_short v[5:6], v0, off
; GFX9-NEXT: s_cbranch_vccz .LBB5_1
; GFX9-NEXT: ; %bb.2: ; %bb2
; GFX9-NEXT: s_endpgm
;
; GFX10-LABEL: urem16_invariant_denom:
; GFX10: ; %bb.0: ; %bb
; GFX10-NEXT: s_clause 0x1
; GFX10-NEXT: s_load_dword s4, s[0:1], 0x2c
; GFX10-NEXT: s_load_dwordx2 s[2:3], s[0:1], 0x24
; GFX10-NEXT: v_mov_b32_e32 v1, 0
; GFX10-NEXT: v_mov_b32_e32 v4, 0
; GFX10-NEXT: s_waitcnt lgkmcnt(0)
; GFX10-NEXT: s_and_b32 s1, 0xffff, s4
; GFX10-NEXT: v_cvt_f32_u32_e32 v2, s1
; GFX10-NEXT: v_rcp_iflag_f32_e32 v3, v2
; GFX10-NEXT: .LBB5_1: ; %bb3
; GFX10-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX10-NEXT: v_and_b32_e32 v0, 0xffff, v4
; GFX10-NEXT: v_add_nc_u16 v4, v4, 1
; GFX10-NEXT: v_cvt_f32_u32_e32 v7, v0
; GFX10-NEXT: v_lshlrev_b64 v[5:6], 1, v[0:1]
; GFX10-NEXT: v_mul_f32_e32 v8, v7, v3
; GFX10-NEXT: v_add_co_u32 v5, s0, s2, v5
; GFX10-NEXT: v_add_co_ci_u32_e64 v6, s0, s3, v6, s0
; GFX10-NEXT: v_trunc_f32_e32 v8, v8
; GFX10-NEXT: v_mad_f32 v7, -v8, v2, v7
; GFX10-NEXT: v_cvt_u32_f32_e32 v8, v8
; GFX10-NEXT: v_cmp_ge_f32_e64 vcc_lo, |v7|, v2
; GFX10-NEXT: v_add_co_ci_u32_e32 v7, vcc_lo, 0, v8, vcc_lo
; GFX10-NEXT: v_cmp_eq_u16_e32 vcc_lo, 0x400, v4
; GFX10-NEXT: v_mul_lo_u32 v7, v7, s1
; GFX10-NEXT: v_sub_nc_u32_e32 v0, v0, v7
; GFX10-NEXT: global_store_short v[5:6], v0, off
; GFX10-NEXT: s_cbranch_vccz .LBB5_1
; GFX10-NEXT: ; %bb.2: ; %bb2
; GFX10-NEXT: s_endpgm
bb:
br label %bb3
bb2: ; preds = %bb3
ret void
bb3: ; preds = %bb3, %bb
%tmp = phi i16 [ 0, %bb ], [ %tmp7, %bb3 ]
%tmp4 = urem i16 %tmp, %arg1
%tmp5 = zext i16 %tmp to i64
%tmp6 = getelementptr inbounds i16, i16 addrspace(1)* %arg, i64 %tmp5
store i16 %tmp4, i16 addrspace(1)* %tmp6, align 2
%tmp7 = add nuw nsw i16 %tmp, 1
%tmp8 = icmp eq i16 %tmp7, 1024
br i1 %tmp8, label %bb2, label %bb3
}
define amdgpu_kernel void @sdiv16_invariant_denom(i16 addrspace(1)* nocapture %arg, i16 %arg1) {
; GFX9-LABEL: sdiv16_invariant_denom:
; GFX9: ; %bb.0: ; %bb
; GFX9-NEXT: s_load_dword s2, s[0:1], 0x2c
; GFX9-NEXT: v_mov_b32_e32 v1, 0
; GFX9-NEXT: s_movk_i32 s5, 0x400
; GFX9-NEXT: v_mov_b32_e32 v4, 0
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: s_sext_i32_i16 s4, s2
; GFX9-NEXT: v_cvt_f32_i32_e32 v2, s4
; GFX9-NEXT: s_load_dwordx2 s[2:3], s[0:1], 0x24
; GFX9-NEXT: v_rcp_iflag_f32_e32 v3, v2
; GFX9-NEXT: .LBB6_1: ; %bb3
; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT: v_bfe_i32 v5, v4, 0, 16
; GFX9-NEXT: v_and_b32_e32 v0, 0xffff, v4
; GFX9-NEXT: v_cvt_f32_i32_e32 v9, v5
; GFX9-NEXT: v_xor_b32_e32 v8, s4, v5
; GFX9-NEXT: v_lshlrev_b64 v[5:6], 1, v[0:1]
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: v_mov_b32_e32 v7, s3
; GFX9-NEXT: v_add_co_u32_e64 v5, s[0:1], s2, v5
; GFX9-NEXT: v_addc_co_u32_e64 v6, s[0:1], v7, v6, s[0:1]
; GFX9-NEXT: v_mul_f32_e32 v7, v9, v3
; GFX9-NEXT: v_trunc_f32_e32 v7, v7
; GFX9-NEXT: v_ashrrev_i32_e32 v0, 30, v8
; GFX9-NEXT: v_cvt_i32_f32_e32 v8, v7
; GFX9-NEXT: v_mad_f32 v7, -v7, v2, v9
; GFX9-NEXT: v_add_u16_e32 v4, 1, v4
; GFX9-NEXT: v_or_b32_e32 v0, 1, v0
; GFX9-NEXT: v_cmp_ge_f32_e64 s[0:1], |v7|, |v2|
; GFX9-NEXT: v_cmp_eq_u16_e32 vcc, s5, v4
; GFX9-NEXT: v_cndmask_b32_e64 v0, 0, v0, s[0:1]
; GFX9-NEXT: v_add_u32_e32 v0, v8, v0
; GFX9-NEXT: global_store_short v[5:6], v0, off
; GFX9-NEXT: s_cbranch_vccz .LBB6_1
; GFX9-NEXT: ; %bb.2: ; %bb2
; GFX9-NEXT: s_endpgm
;
; GFX10-LABEL: sdiv16_invariant_denom:
; GFX10: ; %bb.0: ; %bb
; GFX10-NEXT: s_clause 0x1
; GFX10-NEXT: s_load_dword s4, s[0:1], 0x2c
; GFX10-NEXT: s_load_dwordx2 s[2:3], s[0:1], 0x24
; GFX10-NEXT: v_mov_b32_e32 v1, 0
; GFX10-NEXT: v_mov_b32_e32 v4, 0
; GFX10-NEXT: s_waitcnt lgkmcnt(0)
; GFX10-NEXT: s_sext_i32_i16 s4, s4
; GFX10-NEXT: v_cvt_f32_i32_e32 v2, s4
; GFX10-NEXT: v_rcp_iflag_f32_e32 v3, v2
; GFX10-NEXT: .LBB6_1: ; %bb3
; GFX10-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX10-NEXT: v_bfe_i32 v5, v4, 0, 16
; GFX10-NEXT: v_and_b32_e32 v0, 0xffff, v4
; GFX10-NEXT: v_add_nc_u16 v4, v4, 1
; GFX10-NEXT: v_cvt_f32_i32_e32 v7, v5
; GFX10-NEXT: v_xor_b32_e32 v8, s4, v5
; GFX10-NEXT: v_lshlrev_b64 v[5:6], 1, v[0:1]
; GFX10-NEXT: v_cmp_eq_u16_e32 vcc_lo, 0x400, v4
; GFX10-NEXT: v_mul_f32_e32 v0, v7, v3
; GFX10-NEXT: v_ashrrev_i32_e32 v8, 30, v8
; GFX10-NEXT: v_add_co_u32 v5, s0, s2, v5
; GFX10-NEXT: v_trunc_f32_e32 v0, v0
; GFX10-NEXT: v_or_b32_e32 v8, 1, v8
; GFX10-NEXT: v_add_co_ci_u32_e64 v6, s0, s3, v6, s0
; GFX10-NEXT: s_and_b32 vcc_lo, exec_lo, vcc_lo
; GFX10-NEXT: v_mad_f32 v7, -v0, v2, v7
; GFX10-NEXT: v_cvt_i32_f32_e32 v0, v0
; GFX10-NEXT: v_cmp_ge_f32_e64 s1, |v7|, |v2|
; GFX10-NEXT: v_cndmask_b32_e64 v7, 0, v8, s1
; GFX10-NEXT: v_add_nc_u32_e32 v0, v0, v7
; GFX10-NEXT: global_store_short v[5:6], v0, off
; GFX10-NEXT: s_cbranch_vccz .LBB6_1
; GFX10-NEXT: ; %bb.2: ; %bb2
; GFX10-NEXT: s_endpgm
bb:
br label %bb3
bb2: ; preds = %bb3
ret void
bb3: ; preds = %bb3, %bb
%tmp = phi i16 [ 0, %bb ], [ %tmp7, %bb3 ]
%tmp4 = sdiv i16 %tmp, %arg1
%tmp5 = zext i16 %tmp to i64
%tmp6 = getelementptr inbounds i16, i16 addrspace(1)* %arg, i64 %tmp5
store i16 %tmp4, i16 addrspace(1)* %tmp6, align 2
%tmp7 = add nuw nsw i16 %tmp, 1
%tmp8 = icmp eq i16 %tmp7, 1024
br i1 %tmp8, label %bb2, label %bb3
}
define amdgpu_kernel void @srem16_invariant_denom(i16 addrspace(1)* nocapture %arg, i16 %arg1) {
; GFX9-LABEL: srem16_invariant_denom:
; GFX9: ; %bb.0: ; %bb
; GFX9-NEXT: s_load_dword s2, s[0:1], 0x2c
; GFX9-NEXT: s_load_dwordx2 s[4:5], s[0:1], 0x24
; GFX9-NEXT: v_mov_b32_e32 v1, 0
; GFX9-NEXT: s_movk_i32 s7, 0x400
; GFX9-NEXT: v_mov_b32_e32 v4, 0
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: s_sext_i32_i16 s6, s2
; GFX9-NEXT: v_cvt_f32_i32_e32 v2, s6
; GFX9-NEXT: v_rcp_iflag_f32_e32 v3, v2
; GFX9-NEXT: .LBB7_1: ; %bb3
; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT: v_bfe_i32 v7, v4, 0, 16
; GFX9-NEXT: v_cvt_f32_i32_e32 v10, v7
; GFX9-NEXT: v_and_b32_e32 v0, 0xffff, v4
; GFX9-NEXT: v_xor_b32_e32 v9, s6, v7
; GFX9-NEXT: v_lshlrev_b64 v[5:6], 1, v[0:1]
; GFX9-NEXT: v_ashrrev_i32_e32 v0, 30, v9
; GFX9-NEXT: v_mul_f32_e32 v9, v10, v3
; GFX9-NEXT: v_trunc_f32_e32 v9, v9
; GFX9-NEXT: v_cvt_i32_f32_e32 v11, v9
; GFX9-NEXT: v_mad_f32 v9, -v9, v2, v10
; GFX9-NEXT: v_or_b32_e32 v0, 1, v0
; GFX9-NEXT: v_cmp_ge_f32_e64 s[2:3], |v9|, |v2|
; GFX9-NEXT: v_cndmask_b32_e64 v0, 0, v0, s[2:3]
; GFX9-NEXT: v_add_u32_e32 v0, v11, v0
; GFX9-NEXT: v_mul_lo_u32 v0, v0, s6
; GFX9-NEXT: v_add_u16_e32 v4, 1, v4
; GFX9-NEXT: v_mov_b32_e32 v8, s5
; GFX9-NEXT: v_cmp_eq_u16_e32 vcc, s7, v4
; GFX9-NEXT: v_add_co_u32_e64 v5, s[0:1], s4, v5
; GFX9-NEXT: v_addc_co_u32_e64 v6, s[0:1], v8, v6, s[0:1]
; GFX9-NEXT: v_sub_u32_e32 v0, v7, v0
; GFX9-NEXT: global_store_short v[5:6], v0, off
; GFX9-NEXT: s_cbranch_vccz .LBB7_1
; GFX9-NEXT: ; %bb.2: ; %bb2
; GFX9-NEXT: s_endpgm
;
; GFX10-LABEL: srem16_invariant_denom:
; GFX10: ; %bb.0: ; %bb
; GFX10-NEXT: s_clause 0x1
; GFX10-NEXT: s_load_dword s4, s[0:1], 0x2c
; GFX10-NEXT: s_load_dwordx2 s[2:3], s[0:1], 0x24
; GFX10-NEXT: v_mov_b32_e32 v1, 0
; GFX10-NEXT: v_mov_b32_e32 v4, 0
; GFX10-NEXT: s_waitcnt lgkmcnt(0)
; GFX10-NEXT: s_sext_i32_i16 s1, s4
; GFX10-NEXT: v_cvt_f32_i32_e32 v2, s1
; GFX10-NEXT: v_rcp_iflag_f32_e32 v3, v2
; GFX10-NEXT: .LBB7_1: ; %bb3
; GFX10-NEXT: ; =>This Inner Loop Header: Depth=1
; GFX10-NEXT: v_bfe_i32 v7, v4, 0, 16
; GFX10-NEXT: v_and_b32_e32 v0, 0xffff, v4
; GFX10-NEXT: v_add_nc_u16 v4, v4, 1
; GFX10-NEXT: v_cvt_f32_i32_e32 v5, v7
; GFX10-NEXT: v_xor_b32_e32 v6, s1, v7
; GFX10-NEXT: v_mul_f32_e32 v8, v5, v3
; GFX10-NEXT: v_ashrrev_i32_e32 v6, 30, v6
; GFX10-NEXT: v_trunc_f32_e32 v8, v8
; GFX10-NEXT: v_or_b32_e32 v6, 1, v6
; GFX10-NEXT: v_mad_f32 v5, -v8, v2, v5
; GFX10-NEXT: v_cvt_i32_f32_e32 v8, v8
; GFX10-NEXT: v_cmp_ge_f32_e64 vcc_lo, |v5|, |v2|
; GFX10-NEXT: v_cndmask_b32_e32 v9, 0, v6, vcc_lo
; GFX10-NEXT: v_lshlrev_b64 v[5:6], 1, v[0:1]
; GFX10-NEXT: v_cmp_eq_u16_e32 vcc_lo, 0x400, v4
; GFX10-NEXT: v_add_nc_u32_e32 v0, v8, v9
; GFX10-NEXT: v_add_co_u32 v5, s0, s2, v5
; GFX10-NEXT: v_add_co_ci_u32_e64 v6, s0, s3, v6, s0
; GFX10-NEXT: v_mul_lo_u32 v0, v0, s1
; GFX10-NEXT: v_sub_nc_u32_e32 v0, v7, v0
; GFX10-NEXT: global_store_short v[5:6], v0, off
; GFX10-NEXT: s_cbranch_vccz .LBB7_1
; GFX10-NEXT: ; %bb.2: ; %bb2
; GFX10-NEXT: s_endpgm
bb:
br label %bb3
bb2: ; preds = %bb3
ret void
bb3: ; preds = %bb3, %bb
%tmp = phi i16 [ 0, %bb ], [ %tmp7, %bb3 ]
%tmp4 = srem i16 %tmp, %arg1
%tmp5 = zext i16 %tmp to i64
%tmp6 = getelementptr inbounds i16, i16 addrspace(1)* %arg, i64 %tmp5
store i16 %tmp4, i16 addrspace(1)* %tmp6, align 2
%tmp7 = add nuw nsw i16 %tmp, 1
%tmp8 = icmp eq i16 %tmp7, 1024
br i1 %tmp8, label %bb2, label %bb3
}
Reference in New Issue View Git Blame Copy Permalink