[llvm] d7762a3 - [AMDGPU] Increase instruction cache line size to 128 bytes for GFX11

[Jay Foad via llvm-commits]

Author: Jay Foad
Date: 2022-06-20T14:25:10+01:00
New Revision: d7762a3b369ec2be8ccb5f585aa6a96026caaa33

URL: https://github.com/llvm/llvm-project/commit/d7762a3b369ec2be8ccb5f585aa6a96026caaa33
DIFF: https://github.com/llvm/llvm-project/commit/d7762a3b369ec2be8ccb5f585aa6a96026caaa33.diff

LOG: [AMDGPU] Increase instruction cache line size to 128 bytes for GFX11

Differential Revision: https://reviews.llvm.org/D128189

Added: 
    

Modified: 
    llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
    llvm/test/CodeGen/AMDGPU/s_code_end.ll

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
index bcb10b8b95222..0781334695491 100644
--- a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
+++ b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
@@ -301,7 +301,7 @@ bool AMDGPUTargetAsmStreamer::EmitCodeEnd(const MCSubtargetInfo &STI) {
   uint32_t Encoded_pad = Encoded_s_code_end;
 
   // Instruction cache line size in bytes.
-  const unsigned Log2CacheLineSize = 6;
+  const unsigned Log2CacheLineSize = AMDGPU::isGFX11Plus(STI) ? 7 : 6;
   const unsigned CacheLineSize = 1u << Log2CacheLineSize;
 
   // Extra padding amount in bytes to support prefetch mode 3.
@@ -824,7 +824,7 @@ bool AMDGPUTargetELFStreamer::EmitCodeEnd(const MCSubtargetInfo &STI) {
   uint32_t Encoded_pad = Encoded_s_code_end;
 
   // Instruction cache line size in bytes.
-  const unsigned Log2CacheLineSize = 6;
+  const unsigned Log2CacheLineSize = AMDGPU::isGFX11Plus(STI) ? 7 : 6;
   const unsigned CacheLineSize = 1u << Log2CacheLineSize;
 
   // Extra padding amount in bytes to support prefetch mode 3.

diff  --git a/llvm/test/CodeGen/AMDGPU/s_code_end.ll b/llvm/test/CodeGen/AMDGPU/s_code_end.ll
index 51197f326f2ec..ad7d8a14babc1 100644
--- a/llvm/test/CodeGen/AMDGPU/s_code_end.ll
+++ b/llvm/test/CodeGen/AMDGPU/s_code_end.ll
@@ -41,10 +41,10 @@ define amdgpu_kernel void @a_kernel2() #0 {
 ; GCN-ASM-NEXT:   [[END_LABEL3:\.Lfunc_end.*]]:
 ; GCN-ASM-NEXT:           .size   a_function, [[END_LABEL3]]-a_function
 ; GFX10END-ASM:           .p2alignl 6, 3214868480
-; GFX11END-ASM:           .p2alignl 6, 3214868480
+; GFX11END-ASM:           .p2alignl 7, 3214868480
 ; GFX90AEND-ASM:          .p2alignl 6, 3212836864
 ; GFX10END-ASM-NEXT:      .fill 48, 4, 3214868480
-; GFX11END-ASM-NEXT:      .fill 48, 4, 3214868480
+; GFX11END-ASM-NEXT:      .fill 96, 4, 3214868480
 ; GFX90AEND-ASM-NEXT:     .fill 256, 4, 3212836864
 ; GFX10NOEND-NOT:         .fill
 ; GFX11NOEND-NOT:         .fill