mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-11-26 04:32:44 +01:00
f7eb7bd43b
Summary: Original patch by: Fabian Wahlster <razor@singul4rity.com> Change-Id: I148f692a88432541fad468963f58da9ddf79fac5 Reviewers: arsenm, rampitec Subscribers: kzhuravl, jvesely, wdng, yaxunl, dstuttard, tpr, t-tye, b-sumner, llvm-commits Differential Revision: https://reviews.llvm.org/D51995 llvm-svn: 348488
95 lines
3.0 KiB
LLVM
95 lines
3.0 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc < %s -mtriple=amdgcn-amd-mesa3d -mcpu=fiji -verify-machineinstrs | FileCheck -check-prefix=VI %s
|
|
; RUN: llc < %s -mtriple=amdgcn-amd-mesa3d -mcpu=gfx900 -verify-machineinstrs | FileCheck -check-prefix=GFX9 %s
|
|
|
|
; ===================================================================================
|
|
; V_LSHL_ADD_U32
|
|
; ===================================================================================
|
|
|
|
define amdgpu_ps float @shl_add(i32 %a, i32 %b, i32 %c) {
|
|
; VI-LABEL: shl_add:
|
|
; VI: ; %bb.0:
|
|
; VI-NEXT: v_lshlrev_b32_e32 v0, v1, v0
|
|
; VI-NEXT: v_add_u32_e32 v0, vcc, v0, v2
|
|
; VI-NEXT: ; return to shader part epilog
|
|
;
|
|
; GFX9-LABEL: shl_add:
|
|
; GFX9: ; %bb.0:
|
|
; GFX9-NEXT: v_lshl_add_u32 v0, v0, v1, v2
|
|
; GFX9-NEXT: ; return to shader part epilog
|
|
%x = shl i32 %a, %b
|
|
%result = add i32 %x, %c
|
|
%bc = bitcast i32 %result to float
|
|
ret float %bc
|
|
}
|
|
|
|
; ThreeOp instruction variant not used due to Constant Bus Limitations
|
|
define amdgpu_ps float @shl_add_vgpr_a(i32 %a, i32 inreg %b, i32 inreg %c) {
|
|
; VI-LABEL: shl_add_vgpr_a:
|
|
; VI: ; %bb.0:
|
|
; VI-NEXT: v_lshlrev_b32_e32 v0, s2, v0
|
|
; VI-NEXT: v_add_u32_e32 v0, vcc, s3, v0
|
|
; VI-NEXT: ; return to shader part epilog
|
|
;
|
|
; GFX9-LABEL: shl_add_vgpr_a:
|
|
; GFX9: ; %bb.0:
|
|
; GFX9-NEXT: v_lshlrev_b32_e32 v0, s2, v0
|
|
; GFX9-NEXT: v_add_u32_e32 v0, s3, v0
|
|
; GFX9-NEXT: ; return to shader part epilog
|
|
%x = shl i32 %a, %b
|
|
%result = add i32 %x, %c
|
|
%bc = bitcast i32 %result to float
|
|
ret float %bc
|
|
}
|
|
|
|
define amdgpu_ps float @shl_add_vgpr_all(i32 %a, i32 %b, i32 %c) {
|
|
; VI-LABEL: shl_add_vgpr_all:
|
|
; VI: ; %bb.0:
|
|
; VI-NEXT: v_lshlrev_b32_e32 v0, v1, v0
|
|
; VI-NEXT: v_add_u32_e32 v0, vcc, v0, v2
|
|
; VI-NEXT: ; return to shader part epilog
|
|
;
|
|
; GFX9-LABEL: shl_add_vgpr_all:
|
|
; GFX9: ; %bb.0:
|
|
; GFX9-NEXT: v_lshl_add_u32 v0, v0, v1, v2
|
|
; GFX9-NEXT: ; return to shader part epilog
|
|
%x = shl i32 %a, %b
|
|
%result = add i32 %x, %c
|
|
%bc = bitcast i32 %result to float
|
|
ret float %bc
|
|
}
|
|
|
|
define amdgpu_ps float @shl_add_vgpr_ab(i32 %a, i32 %b, i32 inreg %c) {
|
|
; VI-LABEL: shl_add_vgpr_ab:
|
|
; VI: ; %bb.0:
|
|
; VI-NEXT: v_lshlrev_b32_e32 v0, v1, v0
|
|
; VI-NEXT: v_add_u32_e32 v0, vcc, s2, v0
|
|
; VI-NEXT: ; return to shader part epilog
|
|
;
|
|
; GFX9-LABEL: shl_add_vgpr_ab:
|
|
; GFX9: ; %bb.0:
|
|
; GFX9-NEXT: v_lshl_add_u32 v0, v0, v1, s2
|
|
; GFX9-NEXT: ; return to shader part epilog
|
|
%x = shl i32 %a, %b
|
|
%result = add i32 %x, %c
|
|
%bc = bitcast i32 %result to float
|
|
ret float %bc
|
|
}
|
|
|
|
define amdgpu_ps float @shl_add_vgpr_const(i32 %a, i32 %b) {
|
|
; VI-LABEL: shl_add_vgpr_const:
|
|
; VI: ; %bb.0:
|
|
; VI-NEXT: v_lshlrev_b32_e32 v0, 3, v0
|
|
; VI-NEXT: v_add_u32_e32 v0, vcc, v0, v1
|
|
; VI-NEXT: ; return to shader part epilog
|
|
;
|
|
; GFX9-LABEL: shl_add_vgpr_const:
|
|
; GFX9: ; %bb.0:
|
|
; GFX9-NEXT: v_lshl_add_u32 v0, v0, 3, v1
|
|
; GFX9-NEXT: ; return to shader part epilog
|
|
%x = shl i32 %a, 3
|
|
%result = add i32 %x, %b
|
|
%bc = bitcast i32 %result to float
|
|
ret float %bc
|
|
}
|