brintos

brintos / llvm-project-archived public Read only

0
0
Text · 5.4 KiB · e3c383c Raw
127 lines · plain
1 2; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=kaveri | FileCheck --check-prefix=HSA %s3; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=kaveri -mattr=-flat-for-global | FileCheck --check-prefix=HSA-CI %s4; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=carrizo | FileCheck --check-prefix=HSA %s5; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=carrizo -mattr=-flat-for-global | FileCheck --check-prefix=HSA-VI %s6; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=kaveri -filetype=obj | llvm-readobj -S --sd --syms - | FileCheck --check-prefix=ELF %s7; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=kaveri | llvm-mc -filetype=obj -triple amdgcn--amdhsa -mcpu=kaveri --amdhsa-code-object-version=4 | llvm-readobj -S --sd --syms - | FileCheck %s --check-prefix=ELF8; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx1010 -mattr=+wavefrontsize32 | FileCheck --check-prefix=GFX10 --check-prefix=GFX10-W32 %s9; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx1010 -mattr=+wavefrontsize64 | FileCheck --check-prefix=GFX10 --check-prefix=GFX10-W64 %s10; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx1100 -mattr=+wavefrontsize32 | FileCheck --check-prefix=GFX10 --check-prefix=GFX10-W32 %s11; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx1100 -mattr=+wavefrontsize64 | FileCheck --check-prefix=GFX10 --check-prefix=GFX10-W64 %s12 13; The SHT_NOTE section contains the output from the .hsa_code_object_*14; directives.15 16; ELF: Section {17; ELF: Name: .text18; ELF: Type: SHT_PROGBITS (0x1)19; ELF: Flags [ (0x6)20; ELF: SHF_ALLOC (0x2)21; ELF: SHF_EXECINSTR (0x4)22; ELF: }23 24; ELF: SHT_NOTE25; ELF: Flags [ (0x2)26; ELF: SHF_ALLOC (0x2)27; ELF: ]28; ELF: SectionData (29; ELF:   0000: 07000000 A8020000 20000000 414D444730; ELF:   0010: 50550000 83AE616D 64687361 2E6B657231; ELF:   0020: 6E656C73 928DA52E 61726773 9185AE2E32; ELF:   0030: 61646472 6573735F 73706163 65A6676C33; ELF:   0040: 6F62616C A52E6E61 6D65A36F 7574A72E34; ELF:   0050: 6F666673 657400A5 2E73697A 6508AB2E35; ELF:   0060: 76616C75 655F6B69 6E64AD67 6C6F626136; ELF:   0070: 6C5F6275 66666572 B92E6772 6F75705F37; ELF:   0080: 7365676D 656E745F 66697865 645F736938; ELF:   0090: 7A6500B6 2E6B6572 6E617267 5F73656739; ELF:   00A0: 6D656E74 5F616C69 676E08B5 2E6B657240; ELF:   00B0: 6E617267 5F736567 6D656E74 5F73697A41; ELF:   00C0: 6508B82E 6D61785F 666C6174 5F776F7242; ELF:   00D0: 6B67726F 75705F73 697A65CD 0400A52E43; ELF:   00E0: 6E616D65 A673696D 706C65BB 2E70726944; ELF:   00F0: 76617465 5F736567 6D656E74 5F66697845; ELF:   0100: 65645F73 697A6500 AB2E7367 70725F6346; ELF:   0110: 6F756E74 10B12E73 6770725F 7370696C47; ELF:   0120: 6C5F636F 756E7400 A72E7379 6D626F6C48; ELF:   0130: A973696D 706C652E 6B64AB2E 7667707249; ELF:   0140: 5F636F75 6E7403B1 2E766770 725F737050; ELF:   0150: 696C6C5F 636F756E 7400AF2E 7761766551; ELF:   0160: 66726F6E 745F7369 7A65408D A52E617252; ELF:   0170: 677390B9 2E67726F 75705F73 65676D6553; ELF:   0180: 6E745F66 69786564 5F73697A 6500B62E54; ELF:   0190: 6B65726E 6172675F 7365676D 656E745F55; ELF:   01A0: 616C6967 6E04B52E 6B65726E 6172675F56; ELF:   01B0: 7365676D 656E745F 73697A65 00B82E6D57; ELF:   01C0: 61785F66 6C61745F 776F726B 67726F7558; ELF:   01D0: 705F7369 7A65CD04 00A52E6E 616D65B259; ELF:   01E0: 73696D70 6C655F6E 6F5F6B65 726E617260; ELF:   01F0: 6773BB2E 70726976 6174655F 7365676D61; ELF:   0200: 656E745F 66697865 645F7369 7A6500AB62; ELF:   0210: 2E736770 725F636F 756E740E B12E736763; ELF:   0220: 70725F73 70696C6C 5F636F75 6E7400A764; ELF:   0230: 2E73796D 626F6CB5 73696D70 6C655F6E65; ELF:   0240: 6F5F6B65 726E6172 67732E6B 64AB2E7666; ELF:   0250: 6770725F 636F756E 7402B12E 7667707267; ELF:   0260: 5F737069 6C6C5F63 6F756E74 00AF2E7768; ELF:   0270: 61766566 726F6E74 5F73697A 6540AD6169; ELF:   0280: 6D646873 612E7461 72676574 BD616D6470; ELF:   0290: 67636E2D 756E6B6E 6F776E2D 616D646871; ELF:   02A0: 73612D2D 67667837 3030AE61 6D64687372; ELF:   02B0: 612E7665 7273696F 6E92010173; ELF: )74 75; ELF: Symbol {76; ELF: Name: simple77; ELF: Size: 3278; ELF: }79 80; HSA-NOT: .AMDGPU.config81; HSA: .text82; HSA-CI: .amdgcn_target "amdgcn-unknown-amdhsa--gfx700"83; HSA-VI: .amdgcn_target "amdgcn-unknown-amdhsa--gfx801"84 85; HSA-LABEL: {{^}}simple:86 87; PRE-GFX10: wavefront_size = 688 89; HSA: s_load_{{dwordx2|b64}} s[{{[0-9]+:[0-9]+}}], s[4:5], 0x090 91; Make sure we are setting the ATC bit:92; HSA-CI: s_mov_b32 s[[HI:[0-9]]], 0x100f00093; On VI+ we also need to set MTYPE = 294; HSA-VI: s_mov_b32 s[[HI:[0-9]]], 0x1100f00095; Make sure we generate flat store for HSA96; PRE-GFX10: flat_store_dword v{{\[[0-9]+:[0-9]+\]}}, v{{[0-9]+}}97; GFX10: global_store_{{dword|b32}} v{{\[[0-9]+:[0-9]+\]}}, v{{[0-9]+}}, off98 99; HSA: .amdhsa_user_sgpr_private_segment_buffer 1100; HSA: .amdhsa_user_sgpr_kernarg_segment_ptr 1101 102; PRE-GFX10-NOT: .amdhsa_wavefront_size32103; GFX10-W32: .amdhsa_wavefront_size32 1104; GFX10-W64: .amdhsa_wavefront_size32 0105 106; HSA: .Lfunc_end0:107; HSA: .size   simple, .Lfunc_end0-simple108 109define amdgpu_kernel void @simple(ptr addrspace(1) %out) #0 {110entry:111  store i32 0, ptr addrspace(1) %out112  ret void113}114 115; HSA-LABEL: {{^}}simple_no_kernargs:116; HSA: .amdhsa_user_sgpr_kernarg_segment_ptr 0117define amdgpu_kernel void @simple_no_kernargs() #0 {118entry:119  store volatile i32 0, ptr addrspace(1) poison120  ret void121}122 123attributes #0 = { "amdgpu-no-dispatch-id" "amdgpu-no-dispatch-ptr" "amdgpu-no-implicitarg-ptr" "amdgpu-no-lds-kernel-id" "amdgpu-no-queue-ptr" "amdgpu-no-workgroup-id-x" "amdgpu-no-workgroup-id-y" "amdgpu-no-workgroup-id-z" "amdgpu-no-workitem-id-x" "amdgpu-no-workitem-id-y" "amdgpu-no-workitem-id-z" }124 125!llvm.module.flags = !{!0}126!0 = !{i32 1, !"amdhsa_code_object_version", i32 400}127