44 lines · plain
1; RUN: llc < %s -mtriple=amdgcn -mcpu=verde | FileCheck --check-prefixes=GCN,PRE-GFX10,PRE-GFX12 %s2; RUN: llc < %s -mtriple=amdgcn -mcpu=gfx900 | FileCheck --check-prefixes=GCN,PRE-GFX10,PRE-GFX12 %s3; RUN: llc < %s -mtriple=amdgcn -mcpu=gfx1010 | FileCheck --check-prefixes=GCN,GFX10,PRE-GFX12 %s4; RUN: llc < %s -mtriple=amdgcn -mcpu=gfx1100 | FileCheck --check-prefixes=GCN,GFX10,PRE-GFX12 %s5; RUN: llc < %s -mtriple=amdgcn -mcpu=gfx1200 | FileCheck --check-prefixes=GCN,GFX12 %s6 7; GCN-LABEL: {{^}}getlod_1d:8; PRE-GFX10: image_get_lod v[0:3], v0, s[0:7], s[8:11] dmask:0xf{{$}}9; GFX10: image_get_lod v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D10; GFX12: image_get_lod v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D11; PRE-GFX12: s_waitcnt vmcnt(0)12; GFX12: s_wait_samplecnt 0x013define amdgpu_ps <4 x float> @getlod_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {14main_body:15 %r = call <4 x float> @llvm.amdgcn.image.getlod.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)16 ret <4 x float> %r17}18 19; GCN-LABEL: {{^}}getlod_2d:20; PRE-GFX10: image_get_lod v[0:1], v[0:1], s[0:7], s[8:11] dmask:0x3{{$}}21; GFX10: image_get_lod v[0:1], v[0:1], s[0:7], s[8:11] dmask:0x3 dim:SQ_RSRC_IMG_2D22; GFX12: image_get_lod v[0:1], [v0, v1], s[0:7], s[8:11] dmask:0x3 dim:SQ_RSRC_IMG_2D23; PRE-GFX12: s_waitcnt vmcnt(0)24; GFX12: s_wait_samplecnt 0x025define amdgpu_ps <2 x float> @getlod_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %t) {26main_body:27 %r = call <2 x float> @llvm.amdgcn.image.getlod.2d.v2f32.f32(i32 3, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)28 ret <2 x float> %r29}30 31; GCN-LABEL: {{^}}adjust_writemask_getlod_none_enabled:32; GCN-NOT: image33define amdgpu_ps <4 x float> @adjust_writemask_getlod_none_enabled(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %t) {34main_body:35 %r = call <4 x float> @llvm.amdgcn.image.getlod.2d.v4f32.f32(i32 0, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)36 ret <4 x float> %r37}38 39declare <4 x float> @llvm.amdgcn.image.getlod.1d.v4f32.f32(i32, float, <8 x i32>, <4 x i32>, i1, i32, i32) #040declare <4 x float> @llvm.amdgcn.image.getlod.2d.v4f32.f32(i32, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #041declare <2 x float> @llvm.amdgcn.image.getlod.2d.v2f32.f32(i32, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #042 43attributes #0 = { nounwind readnone }44