Searched refs:NSA (Results 1 – 16 of 16) sorted by relevance
| /llvm-project-15.0.7/llvm/test/CodeGen/AMDGPU/ |
| H A D | llvm.amdgcn.image.nsa.ll | 2 …=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,NSA,GFX1010-NSA %s 3 …=amdgcn -mcpu=gfx1030 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,NSA,GFX1030-NSA %s 5 …ch=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,NSA,GFX11-NSA %s 9 ; TODO: use NSA here 22 ; NSA: image_sample v[0:3], [v1, v2, v0], 75 ; GFX1010-NSA: image_sample v9, v[5:7], 77 ; GFX1030-NSA: image_sample v1, v[5:7], 79 ; GFX11-NSA: image_sample v1, v[5:7], 91 ; GFX1010-NSA: image_sample v9, v[5:7], 93 ; GFX1030-NSA: image_sample v1, v[5:7], [all …]
|
| H A D | sgpr-regalloc-flags.ll | 24 ; DEFAULT-NEXT: GCN NSA Reassign 52 ; BASIC-DEFAULT-NEXT: GCN NSA Reassign 64 ; DEFAULT-BASIC-NEXT: GCN NSA Reassign 82 ; BASIC-BASIC-NEXT: GCN NSA Reassign
|
| H A D | nsa-vmem-hazard.mir | 39 # Non-NSA
|
| H A D | llc-pipeline.ll | 346 ; GCN-O1-NEXT: GCN NSA Reassign 636 ; GCN-O1-OPTS-NEXT: GCN NSA Reassign 928 ; GCN-O2-NEXT: GCN NSA Reassign 1233 ; GCN-O3-NEXT: GCN NSA Reassign
|
| H A D | llvm.amdgcn.intersect_ray.ll | 161 ; TODO: NSA reassign is very limited and cannot work with VGPR tuples and subregs.
|
| /llvm-project-15.0.7/llvm/docs/AMDGPU/ |
| H A D | gfx1013_vaddr_cdc744.rst | 15 … using either :ref:`standard VGPR syntax<amdgpu_synid_v>` or special :ref:`NSA VGPR syntax<amdgpu_… 19 * If specified using :ref:`NSA VGPR syntax<amdgpu_synid_nsa>`, the size is 1-13 dwords.
|
| H A D | gfx1030_vaddr_cdc744.rst | 15 … using either :ref:`standard VGPR syntax<amdgpu_synid_v>` or special :ref:`NSA VGPR syntax<amdgpu_… 19 * If specified using :ref:`NSA VGPR syntax<amdgpu_synid_nsa>`, the size is 1-13 dwords.
|
| H A D | gfx10_vaddr_cdc744.rst | 15 … using either :ref:`standard VGPR syntax<amdgpu_synid_v>` or special :ref:`NSA VGPR syntax<amdgpu_… 19 * If specified using :ref:`NSA VGPR syntax<amdgpu_synid_nsa>`, the size is 1-13 dwords.
|
| H A D | gfx1013_vaddr_49d53a.rst | 15 … using either :ref:`standard VGPR syntax<amdgpu_synid_v>` or special :ref:`NSA VGPR syntax<amdgpu_… 19 * If specified using :ref:`NSA VGPR syntax<amdgpu_synid_nsa>`, the size is 8-12 dwords.
|
| H A D | gfx1030_vaddr_49d53a.rst | 15 … using either :ref:`standard VGPR syntax<amdgpu_synid_v>` or special :ref:`NSA VGPR syntax<amdgpu_… 19 * If specified using :ref:`NSA VGPR syntax<amdgpu_synid_nsa>`, the size is 8-12 dwords.
|
| /llvm-project-15.0.7/llvm/lib/Target/AMDGPU/ |
| H A D | GCNNSAReassign.cpp | 179 bool NSA = false; in CheckNSA() local 232 NSA = true; in CheckNSA() 235 return NSA ? NSA_Status::NON_CONTIGUOUS : NSA_Status::CONTIGUOUS; in CheckNSA()
|
| H A D | MIMGInstructions.td | 15 // - MIMGEncGfx10Default: gfx10 default (non-NSA) encoding 16 // - MIMGEncGfx10NSA: gfx10 NSA encoding 17 // - MIMGEncGfx11Default: gfx11 default (non-NSA) encoding 18 // - MIMGEncGfx11NSA: gfx11 NSA encoding 248 int NSA = !if(!le(num_addrs, 1), ?, 277 // Base class of all non-NSA gfx10 MIMG instructions. 290 // Base class for all NSA MIMG instructions. 305 let nsa = nsah.NSA; 308 // Base class of all non-NSA gfx11 MIMG instructions. 321 // Base class for all NSA MIMG instructions. [all …]
|
| H A D | AMDGPU.td | 237 "MIMG-NSA followed by VMEM fail if EXEC_LO or EXEC_HI equals zero" 243 "MIMG-NSA in a hard clause has unpredictable results on GFX10.1" 488 "Support NSA encoding for image instructions"
|
| /llvm-project-15.0.7/clang/test/CXX/drs/ |
| H A D | dr1xx.cpp | 414 namespace NSA { namespace
|
| /llvm-project-15.0.7/llvm/docs/ |
| H A D | AMDGPUOperandSyntax.rst | 85 GFX10 *Image* instructions may use special *NSA* (Non-Sequential Address) syntax for *image address… 95 in *NSA* sequence are not required to have
|
| /llvm-project-15.0.7/llvm/test/MC/Disassembler/AMDGPU/ |
| H A D | gfx10_mimg.txt | 309 # MIMG, NSA address 312 # NOTE: Contents of unused NSA bytes are NOT preserved.
|