[AMDGPU] Emit correct kernel descriptor on big-endian hosts

Previously we wrote multi-byte values out as-is from host memory. Use
the `emitIntN` helpers in `MCStreamer` to produce a valid descriptor
irrespective of the host endianness.

Reviewed By: arsenm, rochauha

Differential Revision: https://reviews.llvm.org/D88858

diff --git a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
index 344f22d..9d662ec 100644 (file)
--- a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
+++ b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
@@ -656,9 +656,10 @@ void AMDGPUTargetELFStreamer::EmitAmdhsaKernelDescriptor(
     KernelCodeSymbol->setVisibility(ELF::STV_PROTECTED);
 
   Streamer.emitLabel(KernelDescriptorSymbol);
-  Streamer.emitBytes(StringRef(
-      (const char*)&(KernelDescriptor),
-      offsetof(amdhsa::kernel_descriptor_t, kernel_code_entry_byte_offset)));
+  Streamer.emitInt32(KernelDescriptor.group_segment_fixed_size);
+  Streamer.emitInt32(KernelDescriptor.private_segment_fixed_size);
+  for (uint8_t Res : KernelDescriptor.reserved0)
+    Streamer.emitInt8(Res);
   // FIXME: Remove the use of VK_AMDGPU_REL64 in the expression below. The
   // expression being created is:
   //   (start of kernel code) - (start of kernel descriptor)
@@ -670,11 +671,12 @@ void AMDGPUTargetELFStreamer::EmitAmdhsaKernelDescriptor(
           KernelDescriptorSymbol, MCSymbolRefExpr::VK_None, Context),
       Context),
       sizeof(KernelDescriptor.kernel_code_entry_byte_offset));
-  Streamer.emitBytes(StringRef(
-      (const char*)&(KernelDescriptor) +
-          offsetof(amdhsa::kernel_descriptor_t, kernel_code_entry_byte_offset) +
-          sizeof(KernelDescriptor.kernel_code_entry_byte_offset),
-      sizeof(KernelDescriptor) -
-          offsetof(amdhsa::kernel_descriptor_t, kernel_code_entry_byte_offset) -
-          sizeof(KernelDescriptor.kernel_code_entry_byte_offset)));
+  for (uint8_t Res : KernelDescriptor.reserved1)
+    Streamer.emitInt8(Res);
+  Streamer.emitInt32(KernelDescriptor.compute_pgm_rsrc3);
+  Streamer.emitInt32(KernelDescriptor.compute_pgm_rsrc1);
+  Streamer.emitInt32(KernelDescriptor.compute_pgm_rsrc2);
+  Streamer.emitInt16(KernelDescriptor.kernel_code_properties);
+  for (uint8_t Res : KernelDescriptor.reserved2)
+    Streamer.emitInt8(Res);
 }

diff --git a/llvm/test/MC/AMDGPU/hsa-sgpr-init-bug-v3.s b/llvm/test/MC/AMDGPU/hsa-sgpr-init-bug-v3.s
index 644d957..35237c4 100644 (file)
--- a/llvm/test/MC/AMDGPU/hsa-sgpr-init-bug-v3.s
+++ b/llvm/test/MC/AMDGPU/hsa-sgpr-init-bug-v3.s
@@ -1,9 +1,6 @@
 // RUN: llvm-mc -mattr=+code-object-v3 -triple amdgcn-amd-amdhsa -mcpu=gfx802 -filetype=obj < %s > %t
 // RUN: llvm-objdump -s -j .rodata %t | FileCheck --check-prefix=OBJDUMP %s
 
-// big endian not supported
-// XFAIL: host-byteorder-big-endian
-
 // Check that SGPR init bug on gfx803 is corrected by the assembler, setting
 // GRANULATED_WAVEFRONT_SGPR_COUNT to 11.
 

diff --git a/llvm/test/MC/AMDGPU/hsa-v3.s b/llvm/test/MC/AMDGPU/hsa-v3.s
index 3c2ddec..a995b1b 100644 (file)
--- a/llvm/test/MC/AMDGPU/hsa-v3.s
+++ b/llvm/test/MC/AMDGPU/hsa-v3.s
@@ -3,9 +3,6 @@
 // RUN: llvm-readelf -sections -symbols -relocations %t | FileCheck --check-prefix=READOBJ %s
 // RUN: llvm-objdump -s -j .rodata %t | FileCheck --check-prefix=OBJDUMP %s
 
-// big endian not supported
-// XFAIL: host-byteorder-big-endian
-
 // READOBJ: Section Headers
 // READOBJ: .text   PROGBITS {{[0-9a-f]+}} {{[0-9a-f]+}} {{[0-9a-f]+}} {{[0-9]+}} AX {{[0-9]+}} {{[0-9]+}} 256
 // READOBJ: .rodata PROGBITS {{[0-9a-f]+}} {{[0-9a-f]+}}        000100 {{[0-9]+}}  A {{[0-9]+}} {{[0-9]+}} 64