maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/CodeGen/AMDGPU/cc-update.ll

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

423 lines
16 KiB
LLVM
Raw Normal View History

[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx803 < %s | FileCheck --check-prefix=GFX803 %s
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 < %s | FileCheck --check-prefix=GFX900 %s
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 < %s | FileCheck --check-prefix=GFX1010 %s
define amdgpu_kernel void @test_kern_empty() local_unnamed_addr #0 {
; GFX803-LABEL: test_kern_empty:
; GFX803: ; %bb.0: ; %entry
; GFX803-NEXT: s_endpgm
;
; GFX900-LABEL: test_kern_empty:
; GFX900: ; %bb.0: ; %entry
; GFX900-NEXT: s_endpgm
;
; GFX1010-LABEL: test_kern_empty:
; GFX1010: ; %bb.0: ; %entry
; GFX1010-NEXT: s_endpgm
entry:
ret void
}
define amdgpu_kernel void @test_kern_stack() local_unnamed_addr #0 {
; GFX803-LABEL: test_kern_stack:
; GFX803: ; %bb.0: ; %entry
; GFX803-NEXT: s_add_u32 s4, s4, s7
; GFX803-NEXT: s_lshr_b32 flat_scratch_hi, s4, 8
; GFX803-NEXT: s_add_u32 s0, s0, s7
; GFX803-NEXT: s_addc_u32 s1, s1, 0
; GFX803-NEXT: v_mov_b32_e32 v0, 0
; GFX803-NEXT: s_mov_b32 flat_scratch_lo, s5
; GFX803-NEXT: buffer_store_dword v0, off, s[0:3], 0 offset:4
; GFX803-NEXT: s_endpgm
;
; GFX900-LABEL: test_kern_stack:
; GFX900: ; %bb.0: ; %entry
; GFX900-NEXT: s_add_u32 flat_scratch_lo, s4, s7
; GFX900-NEXT: s_addc_u32 flat_scratch_hi, s5, 0
; GFX900-NEXT: s_add_u32 s0, s0, s7
; GFX900-NEXT: s_addc_u32 s1, s1, 0
; GFX900-NEXT: v_mov_b32_e32 v0, 0
; GFX900-NEXT: buffer_store_dword v0, off, s[0:3], 0 offset:4
; GFX900-NEXT: s_endpgm
;
; GFX1010-LABEL: test_kern_stack:
; GFX1010: ; %bb.0: ; %entry
; GFX1010-NEXT: s_add_u32 s4, s4, s7
; GFX1010-NEXT: s_addc_u32 s5, s5, 0
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s4
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s5
[AMDGPU] More accurate gfx10 latencies Differential Revision: https://reviews.llvm.org/D81012
2020-03-11 19:33:30 +08:00
; GFX1010-NEXT: v_mov_b32_e32 v0, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: s_add_u32 s0, s0, s7
; GFX1010-NEXT: s_addc_u32 s1, s1, 0
; GFX1010-NEXT: ; implicit-def: $vcc_hi
; GFX1010-NEXT: buffer_store_dword v0, off, s[0:3], 0 offset:4
; GFX1010-NEXT: s_endpgm
entry:
%x = alloca i32, align 4, addrspace(5)
store volatile i32 0, i32 addrspace(5)* %x, align 4
ret void
}
define amdgpu_kernel void @test_kern_call() local_unnamed_addr #0 {
; GFX803-LABEL: test_kern_call:
; GFX803: ; %bb.0: ; %entry
; GFX803-NEXT: s_add_u32 s4, s4, s7
; GFX803-NEXT: s_lshr_b32 flat_scratch_hi, s4, 8
; GFX803-NEXT: s_add_u32 s0, s0, s7
; GFX803-NEXT: s_addc_u32 s1, s1, 0
; GFX803-NEXT: s_mov_b32 flat_scratch_lo, s5
; GFX803-NEXT: s_getpc_b64 s[4:5]
; GFX803-NEXT: s_add_u32 s4, s4, ex@rel32@lo+4
[AMDGPU] Fix offset for REL32_HI relocs The addend in a REL32 reloc needs to be adjusted to account for the offset from the PC value returned by the s_getpc instruction to the point where the reloc is applied. This was being done correctly for (GOTPC)REL32_LO but not for (GOTPC)REL32_HI. This will only make a difference if the target symbol happens to get loaded almost exactly a multiple of 4G away from the relocated instructions. Differential Revision: https://reviews.llvm.org/D86938
2020-09-01 21:52:01 +08:00
; GFX803-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX803-NEXT: s_mov_b32 s32, 0
; GFX803-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX803-NEXT: s_endpgm
;
; GFX900-LABEL: test_kern_call:
; GFX900: ; %bb.0: ; %entry
; GFX900-NEXT: s_add_u32 flat_scratch_lo, s4, s7
; GFX900-NEXT: s_addc_u32 flat_scratch_hi, s5, 0
; GFX900-NEXT: s_add_u32 s0, s0, s7
; GFX900-NEXT: s_addc_u32 s1, s1, 0
; GFX900-NEXT: s_getpc_b64 s[4:5]
; GFX900-NEXT: s_add_u32 s4, s4, ex@rel32@lo+4
[AMDGPU] Fix offset for REL32_HI relocs The addend in a REL32 reloc needs to be adjusted to account for the offset from the PC value returned by the s_getpc instruction to the point where the reloc is applied. This was being done correctly for (GOTPC)REL32_LO but not for (GOTPC)REL32_HI. This will only make a difference if the target symbol happens to get loaded almost exactly a multiple of 4G away from the relocated instructions. Differential Revision: https://reviews.llvm.org/D86938
2020-09-01 21:52:01 +08:00
; GFX900-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX900-NEXT: s_mov_b32 s32, 0
; GFX900-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX900-NEXT: s_endpgm
;
; GFX1010-LABEL: test_kern_call:
; GFX1010: ; %bb.0: ; %entry
; GFX1010-NEXT: s_add_u32 s4, s4, s7
; GFX1010-NEXT: s_mov_b32 s32, 0
; GFX1010-NEXT: s_addc_u32 s5, s5, 0
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s4
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s5
; GFX1010-NEXT: s_add_u32 s0, s0, s7
; GFX1010-NEXT: s_addc_u32 s1, s1, 0
; GFX1010-NEXT: s_getpc_b64 s[4:5]
; GFX1010-NEXT: s_add_u32 s4, s4, ex@rel32@lo+4
[AMDGPU] Fix offset for REL32_HI relocs The addend in a REL32 reloc needs to be adjusted to account for the offset from the PC value returned by the s_getpc instruction to the point where the reloc is applied. This was being done correctly for (GOTPC)REL32_LO but not for (GOTPC)REL32_HI. This will only make a difference if the target symbol happens to get loaded almost exactly a multiple of 4G away from the relocated instructions. Differential Revision: https://reviews.llvm.org/D86938
2020-09-01 21:52:01 +08:00
; GFX1010-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: ; implicit-def: $vcc_hi
; GFX1010-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX1010-NEXT: s_endpgm
entry:
tail call void @ex() #0
ret void
}
define amdgpu_kernel void @test_kern_stack_and_call() local_unnamed_addr #0 {
; GFX803-LABEL: test_kern_stack_and_call:
; GFX803: ; %bb.0: ; %entry
; GFX803-NEXT: s_add_u32 s4, s4, s7
; GFX803-NEXT: s_lshr_b32 flat_scratch_hi, s4, 8
; GFX803-NEXT: s_add_u32 s0, s0, s7
; GFX803-NEXT: s_addc_u32 s1, s1, 0
; GFX803-NEXT: v_mov_b32_e32 v0, 0
; GFX803-NEXT: s_mov_b32 flat_scratch_lo, s5
; GFX803-NEXT: s_getpc_b64 s[4:5]
; GFX803-NEXT: s_add_u32 s4, s4, ex@rel32@lo+4
[AMDGPU] Fix offset for REL32_HI relocs The addend in a REL32 reloc needs to be adjusted to account for the offset from the PC value returned by the s_getpc instruction to the point where the reloc is applied. This was being done correctly for (GOTPC)REL32_LO but not for (GOTPC)REL32_HI. This will only make a difference if the target symbol happens to get loaded almost exactly a multiple of 4G away from the relocated instructions. Differential Revision: https://reviews.llvm.org/D86938
2020-09-01 21:52:01 +08:00
; GFX803-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX803-NEXT: s_movk_i32 s32, 0x400
; GFX803-NEXT: buffer_store_dword v0, off, s[0:3], 0 offset:4
; GFX803-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX803-NEXT: s_endpgm
;
; GFX900-LABEL: test_kern_stack_and_call:
; GFX900: ; %bb.0: ; %entry
; GFX900-NEXT: s_add_u32 flat_scratch_lo, s4, s7
; GFX900-NEXT: s_addc_u32 flat_scratch_hi, s5, 0
; GFX900-NEXT: s_add_u32 s0, s0, s7
; GFX900-NEXT: s_addc_u32 s1, s1, 0
; GFX900-NEXT: v_mov_b32_e32 v0, 0
; GFX900-NEXT: s_getpc_b64 s[4:5]
; GFX900-NEXT: s_add_u32 s4, s4, ex@rel32@lo+4
[AMDGPU] Fix offset for REL32_HI relocs The addend in a REL32 reloc needs to be adjusted to account for the offset from the PC value returned by the s_getpc instruction to the point where the reloc is applied. This was being done correctly for (GOTPC)REL32_LO but not for (GOTPC)REL32_HI. This will only make a difference if the target symbol happens to get loaded almost exactly a multiple of 4G away from the relocated instructions. Differential Revision: https://reviews.llvm.org/D86938
2020-09-01 21:52:01 +08:00
; GFX900-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX900-NEXT: s_movk_i32 s32, 0x400
; GFX900-NEXT: buffer_store_dword v0, off, s[0:3], 0 offset:4
; GFX900-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX900-NEXT: s_endpgm
;
; GFX1010-LABEL: test_kern_stack_and_call:
; GFX1010: ; %bb.0: ; %entry
; GFX1010-NEXT: s_add_u32 s4, s4, s7
; GFX1010-NEXT: s_movk_i32 s32, 0x200
; GFX1010-NEXT: s_addc_u32 s5, s5, 0
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s4
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s5
[AMDGPU] More accurate gfx10 latencies Differential Revision: https://reviews.llvm.org/D81012
2020-03-11 19:33:30 +08:00
; GFX1010-NEXT: v_mov_b32_e32 v0, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: s_add_u32 s0, s0, s7
; GFX1010-NEXT: s_addc_u32 s1, s1, 0
; GFX1010-NEXT: s_getpc_b64 s[4:5]
; GFX1010-NEXT: s_add_u32 s4, s4, ex@rel32@lo+4
[AMDGPU] Fix offset for REL32_HI relocs The addend in a REL32 reloc needs to be adjusted to account for the offset from the PC value returned by the s_getpc instruction to the point where the reloc is applied. This was being done correctly for (GOTPC)REL32_LO but not for (GOTPC)REL32_HI. This will only make a difference if the target symbol happens to get loaded almost exactly a multiple of 4G away from the relocated instructions. Differential Revision: https://reviews.llvm.org/D86938
2020-09-01 21:52:01 +08:00
; GFX1010-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: ; implicit-def: $vcc_hi
; GFX1010-NEXT: buffer_store_dword v0, off, s[0:3], 0 offset:4
; GFX1010-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX1010-NEXT: s_endpgm
entry:
%x = alloca i32, align 4, addrspace(5)
store volatile i32 0, i32 addrspace(5)* %x, align 4
tail call void @ex() #0
ret void
}
define amdgpu_kernel void @test_force_fp_kern_empty() local_unnamed_addr #2 {
; GFX803-LABEL: test_force_fp_kern_empty:
; GFX803: ; %bb.0: ; %entry
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX803-NEXT: s_mov_b32 s33, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX803-NEXT: s_endpgm
;
; GFX900-LABEL: test_force_fp_kern_empty:
; GFX900: ; %bb.0: ; %entry
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX900-NEXT: s_mov_b32 s33, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX900-NEXT: s_endpgm
;
; GFX1010-LABEL: test_force_fp_kern_empty:
; GFX1010: ; %bb.0: ; %entry
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX1010-NEXT: s_mov_b32 s33, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: s_endpgm
entry:
ret void
}
define amdgpu_kernel void @test_force_fp_kern_stack() local_unnamed_addr #2 {
; GFX803-LABEL: test_force_fp_kern_stack:
; GFX803: ; %bb.0: ; %entry
; GFX803-NEXT: s_add_u32 s4, s4, s7
; GFX803-NEXT: s_lshr_b32 flat_scratch_hi, s4, 8
; GFX803-NEXT: s_add_u32 s0, s0, s7
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX803-NEXT: s_mov_b32 s33, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX803-NEXT: s_addc_u32 s1, s1, 0
; GFX803-NEXT: v_mov_b32_e32 v0, 0
; GFX803-NEXT: s_mov_b32 flat_scratch_lo, s5
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX803-NEXT: buffer_store_dword v0, off, s[0:3], s33 offset:4
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX803-NEXT: s_endpgm
;
; GFX900-LABEL: test_force_fp_kern_stack:
; GFX900: ; %bb.0: ; %entry
; GFX900-NEXT: s_add_u32 flat_scratch_lo, s4, s7
; GFX900-NEXT: s_addc_u32 flat_scratch_hi, s5, 0
; GFX900-NEXT: s_add_u32 s0, s0, s7
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX900-NEXT: s_mov_b32 s33, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX900-NEXT: s_addc_u32 s1, s1, 0
; GFX900-NEXT: v_mov_b32_e32 v0, 0
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX900-NEXT: buffer_store_dword v0, off, s[0:3], s33 offset:4
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX900-NEXT: s_endpgm
;
; GFX1010-LABEL: test_force_fp_kern_stack:
; GFX1010: ; %bb.0: ; %entry
; GFX1010-NEXT: s_add_u32 s4, s4, s7
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX1010-NEXT: s_mov_b32 s33, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: s_addc_u32 s5, s5, 0
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s4
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s5
[AMDGPU] More accurate gfx10 latencies Differential Revision: https://reviews.llvm.org/D81012
2020-03-11 19:33:30 +08:00
; GFX1010-NEXT: v_mov_b32_e32 v0, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: s_add_u32 s0, s0, s7
; GFX1010-NEXT: s_addc_u32 s1, s1, 0
; GFX1010-NEXT: ; implicit-def: $vcc_hi
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX1010-NEXT: buffer_store_dword v0, off, s[0:3], s33 offset:4
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: s_endpgm
entry:
%x = alloca i32, align 4, addrspace(5)
store volatile i32 0, i32 addrspace(5)* %x, align 4
ret void
}
define amdgpu_kernel void @test_force_fp_kern_call() local_unnamed_addr #2 {
; GFX803-LABEL: test_force_fp_kern_call:
; GFX803: ; %bb.0: ; %entry
; GFX803-NEXT: s_add_u32 s4, s4, s7
; GFX803-NEXT: s_lshr_b32 flat_scratch_hi, s4, 8
; GFX803-NEXT: s_add_u32 s0, s0, s7
; GFX803-NEXT: s_addc_u32 s1, s1, 0
; GFX803-NEXT: s_mov_b32 flat_scratch_lo, s5
; GFX803-NEXT: s_getpc_b64 s[4:5]
; GFX803-NEXT: s_add_u32 s4, s4, ex@rel32@lo+4
[AMDGPU] Fix offset for REL32_HI relocs The addend in a REL32 reloc needs to be adjusted to account for the offset from the PC value returned by the s_getpc instruction to the point where the reloc is applied. This was being done correctly for (GOTPC)REL32_LO but not for (GOTPC)REL32_HI. This will only make a difference if the target symbol happens to get loaded almost exactly a multiple of 4G away from the relocated instructions. Differential Revision: https://reviews.llvm.org/D86938
2020-09-01 21:52:01 +08:00
; GFX803-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX803-NEXT: s_mov_b32 s32, 0
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX803-NEXT: s_mov_b32 s33, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX803-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX803-NEXT: s_endpgm
;
; GFX900-LABEL: test_force_fp_kern_call:
; GFX900: ; %bb.0: ; %entry
; GFX900-NEXT: s_add_u32 flat_scratch_lo, s4, s7
; GFX900-NEXT: s_addc_u32 flat_scratch_hi, s5, 0
; GFX900-NEXT: s_add_u32 s0, s0, s7
; GFX900-NEXT: s_addc_u32 s1, s1, 0
; GFX900-NEXT: s_getpc_b64 s[4:5]
; GFX900-NEXT: s_add_u32 s4, s4, ex@rel32@lo+4
[AMDGPU] Fix offset for REL32_HI relocs The addend in a REL32 reloc needs to be adjusted to account for the offset from the PC value returned by the s_getpc instruction to the point where the reloc is applied. This was being done correctly for (GOTPC)REL32_LO but not for (GOTPC)REL32_HI. This will only make a difference if the target symbol happens to get loaded almost exactly a multiple of 4G away from the relocated instructions. Differential Revision: https://reviews.llvm.org/D86938
2020-09-01 21:52:01 +08:00
; GFX900-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX900-NEXT: s_mov_b32 s32, 0
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX900-NEXT: s_mov_b32 s33, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX900-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX900-NEXT: s_endpgm
;
; GFX1010-LABEL: test_force_fp_kern_call:
; GFX1010: ; %bb.0: ; %entry
; GFX1010-NEXT: s_add_u32 s4, s4, s7
; GFX1010-NEXT: s_mov_b32 s32, 0
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX1010-NEXT: s_mov_b32 s33, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: s_addc_u32 s5, s5, 0
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s4
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s5
; GFX1010-NEXT: s_add_u32 s0, s0, s7
; GFX1010-NEXT: s_addc_u32 s1, s1, 0
; GFX1010-NEXT: s_getpc_b64 s[4:5]
; GFX1010-NEXT: s_add_u32 s4, s4, ex@rel32@lo+4
[AMDGPU] Fix offset for REL32_HI relocs The addend in a REL32 reloc needs to be adjusted to account for the offset from the PC value returned by the s_getpc instruction to the point where the reloc is applied. This was being done correctly for (GOTPC)REL32_LO but not for (GOTPC)REL32_HI. This will only make a difference if the target symbol happens to get loaded almost exactly a multiple of 4G away from the relocated instructions. Differential Revision: https://reviews.llvm.org/D86938
2020-09-01 21:52:01 +08:00
; GFX1010-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: ; implicit-def: $vcc_hi
; GFX1010-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX1010-NEXT: s_endpgm
entry:
tail call void @ex() #2
ret void
}
define amdgpu_kernel void @test_force_fp_kern_stack_and_call() local_unnamed_addr #2 {
; GFX803-LABEL: test_force_fp_kern_stack_and_call:
; GFX803: ; %bb.0: ; %entry
; GFX803-NEXT: s_add_u32 s4, s4, s7
; GFX803-NEXT: s_lshr_b32 flat_scratch_hi, s4, 8
; GFX803-NEXT: s_add_u32 s0, s0, s7
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX803-NEXT: s_mov_b32 s33, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX803-NEXT: s_addc_u32 s1, s1, 0
; GFX803-NEXT: v_mov_b32_e32 v0, 0
; GFX803-NEXT: s_mov_b32 flat_scratch_lo, s5
; GFX803-NEXT: s_getpc_b64 s[4:5]
; GFX803-NEXT: s_add_u32 s4, s4, ex@rel32@lo+4
[AMDGPU] Fix offset for REL32_HI relocs The addend in a REL32 reloc needs to be adjusted to account for the offset from the PC value returned by the s_getpc instruction to the point where the reloc is applied. This was being done correctly for (GOTPC)REL32_LO but not for (GOTPC)REL32_HI. This will only make a difference if the target symbol happens to get loaded almost exactly a multiple of 4G away from the relocated instructions. Differential Revision: https://reviews.llvm.org/D86938
2020-09-01 21:52:01 +08:00
; GFX803-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX803-NEXT: s_movk_i32 s32, 0x400
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX803-NEXT: buffer_store_dword v0, off, s[0:3], s33 offset:4
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX803-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX803-NEXT: s_endpgm
;
; GFX900-LABEL: test_force_fp_kern_stack_and_call:
; GFX900: ; %bb.0: ; %entry
; GFX900-NEXT: s_add_u32 flat_scratch_lo, s4, s7
; GFX900-NEXT: s_addc_u32 flat_scratch_hi, s5, 0
; GFX900-NEXT: s_add_u32 s0, s0, s7
; GFX900-NEXT: s_addc_u32 s1, s1, 0
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX900-NEXT: s_mov_b32 s33, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX900-NEXT: v_mov_b32_e32 v0, 0
; GFX900-NEXT: s_getpc_b64 s[4:5]
; GFX900-NEXT: s_add_u32 s4, s4, ex@rel32@lo+4
[AMDGPU] Fix offset for REL32_HI relocs The addend in a REL32 reloc needs to be adjusted to account for the offset from the PC value returned by the s_getpc instruction to the point where the reloc is applied. This was being done correctly for (GOTPC)REL32_LO but not for (GOTPC)REL32_HI. This will only make a difference if the target symbol happens to get loaded almost exactly a multiple of 4G away from the relocated instructions. Differential Revision: https://reviews.llvm.org/D86938
2020-09-01 21:52:01 +08:00
; GFX900-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX900-NEXT: s_movk_i32 s32, 0x400
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX900-NEXT: buffer_store_dword v0, off, s[0:3], s33 offset:4
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX900-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX900-NEXT: s_endpgm
;
; GFX1010-LABEL: test_force_fp_kern_stack_and_call:
; GFX1010: ; %bb.0: ; %entry
; GFX1010-NEXT: s_add_u32 s4, s4, s7
; GFX1010-NEXT: s_movk_i32 s32, 0x200
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX1010-NEXT: s_mov_b32 s33, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: s_addc_u32 s5, s5, 0
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s4
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s5
[AMDGPU] More accurate gfx10 latencies Differential Revision: https://reviews.llvm.org/D81012
2020-03-11 19:33:30 +08:00
; GFX1010-NEXT: v_mov_b32_e32 v0, 0
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: s_add_u32 s0, s0, s7
; GFX1010-NEXT: s_addc_u32 s1, s1, 0
; GFX1010-NEXT: s_getpc_b64 s[4:5]
; GFX1010-NEXT: s_add_u32 s4, s4, ex@rel32@lo+4
[AMDGPU] Fix offset for REL32_HI relocs The addend in a REL32 reloc needs to be adjusted to account for the offset from the PC value returned by the s_getpc instruction to the point where the reloc is applied. This was being done correctly for (GOTPC)REL32_LO but not for (GOTPC)REL32_HI. This will only make a difference if the target symbol happens to get loaded almost exactly a multiple of 4G away from the relocated instructions. Differential Revision: https://reviews.llvm.org/D86938
2020-09-01 21:52:01 +08:00
; GFX1010-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: ; implicit-def: $vcc_hi
[AMDGPU] Move frame pointer from s34 to s33 Remove the gap left between the stack pointer (s32) and frame pointer (s34) now that the scratch wave offset is no longer a part of the calling convention ABI. Update llvm/docs/AMDGPUUsage.rst to reflect the change. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75657
2020-03-04 09:39:47 +08:00
; GFX1010-NEXT: buffer_store_dword v0, off, s[0:3], s33 offset:4
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX1010-NEXT: s_endpgm
entry:
%x = alloca i32, align 4, addrspace(5)
store volatile i32 0, i32 addrspace(5)* %x, align 4
tail call void @ex() #2
ret void
}
define amdgpu_kernel void @test_sgpr_offset_kernel() #1 {
; GFX803-LABEL: test_sgpr_offset_kernel:
; GFX803: ; %bb.0: ; %entry
; GFX803-NEXT: s_add_u32 s4, s4, s7
; GFX803-NEXT: s_lshr_b32 flat_scratch_hi, s4, 8
; GFX803-NEXT: s_add_u32 s0, s0, s7
; GFX803-NEXT: s_addc_u32 s1, s1, 0
; GFX803-NEXT: buffer_load_dword v0, off, s[0:3], 0 offset:8
; GFX803-NEXT: s_mov_b32 s4, 0x40000
; GFX803-NEXT: s_mov_b32 flat_scratch_lo, s5
; GFX803-NEXT: s_waitcnt vmcnt(0)
; GFX803-NEXT: buffer_store_dword v0, off, s[0:3], s4 ; 4-byte Folded Spill
; GFX803-NEXT: ;;#ASMSTART
; GFX803-NEXT: ;;#ASMEND
; GFX803-NEXT: s_mov_b32 s4, 0x40000
; GFX803-NEXT: buffer_load_dword v0, off, s[0:3], s4 ; 4-byte Folded Reload
; GFX803-NEXT: s_waitcnt vmcnt(0)
; GFX803-NEXT: buffer_store_dword v0, off, s[0:3], 0 offset:8
; GFX803-NEXT: s_endpgm
;
; GFX900-LABEL: test_sgpr_offset_kernel:
; GFX900: ; %bb.0: ; %entry
; GFX900-NEXT: s_add_u32 flat_scratch_lo, s4, s7
; GFX900-NEXT: s_addc_u32 flat_scratch_hi, s5, 0
; GFX900-NEXT: s_add_u32 s0, s0, s7
; GFX900-NEXT: s_addc_u32 s1, s1, 0
; GFX900-NEXT: buffer_load_dword v0, off, s[0:3], 0 offset:8
; GFX900-NEXT: s_mov_b32 s6, 0x40000
; GFX900-NEXT: s_waitcnt vmcnt(0)
; GFX900-NEXT: buffer_store_dword v0, off, s[0:3], s6 ; 4-byte Folded Spill
; GFX900-NEXT: ;;#ASMSTART
; GFX900-NEXT: ;;#ASMEND
; GFX900-NEXT: s_mov_b32 s6, 0x40000
; GFX900-NEXT: buffer_load_dword v0, off, s[0:3], s6 ; 4-byte Folded Reload
; GFX900-NEXT: s_waitcnt vmcnt(0)
; GFX900-NEXT: buffer_store_dword v0, off, s[0:3], 0 offset:8
; GFX900-NEXT: s_endpgm
;
; GFX1010-LABEL: test_sgpr_offset_kernel:
; GFX1010: ; %bb.0: ; %entry
; GFX1010-NEXT: s_add_u32 s4, s4, s7
; GFX1010-NEXT: s_addc_u32 s5, s5, 0
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s4
; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s5
; GFX1010-NEXT: s_add_u32 s0, s0, s7
; GFX1010-NEXT: s_addc_u32 s1, s1, 0
; GFX1010-NEXT: s_mov_b32 s6, 0x20000
; GFX1010-NEXT: buffer_load_dword v0, off, s[0:3], 0 offset:8
[AMDGPU] More accurate gfx10 latencies Differential Revision: https://reviews.llvm.org/D81012
2020-03-11 19:33:30 +08:00
; GFX1010-NEXT: ; implicit-def: $vcc_hi
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: s_waitcnt vmcnt(0)
; GFX1010-NEXT: buffer_store_dword v0, off, s[0:3], s6 ; 4-byte Folded Spill
[AMDGPU] Update VMEM scalar write hazard mitigation sequence Using s_waitcnt_depctr 0xffe3 is potentially faster than v_nop. Reviewed By: rampitec, foad Differential Revision: https://reviews.llvm.org/D83872
2020-07-16 10:07:26 +08:00
; GFX1010-NEXT: s_waitcnt_depctr 0xffe3
[AMDGPU] Add Scratch Wave Offset to Scratch Buffer Descriptor in entry functions Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in the entry function prologue. This allows us to removes the scratch wave offset register from the calling convention ABI. As part of this change, allow the use of an inline constant zero for the SOffset of MUBUF instructions accessing the stack in entry functions when a frame pointer is not requested/required. Entry functions with calls still need to set up the calling convention ABI stack pointer register, and reference it in order to address arguments of called functions. The ABI stack pointer register remains unswizzled, but is now wave-relative instead of queue-relative. Non-entry functions also use an inline constant zero SOffset for wave-relative scratch access, but continue to use the stack and frame pointers as before. When the stack or frame pointer is converted to a swizzled offset it is now scaled directly, as the scratch wave offset no longer needs to be subtracted first. Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling convention. Tags: #llvm Differential Revision: https://reviews.llvm.org/D75138
2020-01-22 06:27:57 +08:00
; GFX1010-NEXT: s_mov_b32 s6, 0x20000
; GFX1010-NEXT: ;;#ASMSTART
; GFX1010-NEXT: ;;#ASMEND
; GFX1010-NEXT: buffer_load_dword v0, off, s[0:3], s6 ; 4-byte Folded Reload
; GFX1010-NEXT: s_waitcnt vmcnt(0)
; GFX1010-NEXT: buffer_store_dword v0, off, s[0:3], 0 offset:8
; GFX1010-NEXT: s_endpgm
entry:
; Occupy 4096 bytes of scratch, so the offset of the spill of %a does not
; fit in the instruction, and has to live in the SGPR offset.
%alloca = alloca i8, i32 4092, align 4, addrspace(5)
%buf = bitcast i8 addrspace(5)* %alloca to i32 addrspace(5)*
%aptr = getelementptr i32, i32 addrspace(5)* %buf, i32 1
; 0x40000 / 64 = 4096 (for wave64)
; CHECK: s_add_u32 s6, s7, 0x40000
; CHECK: buffer_store_dword v{{[0-9]+}}, off, s[{{[0-9]+:[0-9]+}}], s6 ; 4-byte Folded Spill
%a = load volatile i32, i32 addrspace(5)* %aptr
; Force %a to spill
call void asm sideeffect "", "~{v0},~{v1},~{v2},~{v3},~{v4},~{v5},~{v6},~{v7}" ()
%outptr = getelementptr i32, i32 addrspace(5)* %buf, i32 1
store volatile i32 %a, i32 addrspace(5)* %outptr
ret void
}
declare hidden void @ex() local_unnamed_addr #0
attributes #0 = { nounwind }
attributes #1 = { nounwind "amdgpu-num-vgpr"="8" }
attributes #2 = { nounwind "frame-pointer"="all" }