Patch supports Intel(R) AVX/AVX2 MOV Instruction
This patch enhances the existing ART-Compiler
to generate Intel(R) AVX/AVX2 MOV Instructions for
doing SIMD Operations on Intel(R) Architecture CPUs.
It also provides the framework for AVX/AVX2 Instruction
encoding and dissassembly
BUG: 127881558
Test: run-test gtest
Change-Id: I9386aecc134941a2d907f9ec6b2d5522ec5ff8b5
diff --git a/compiler/optimizing/code_generator_x86.cc b/compiler/optimizing/code_generator_x86.cc
index 95118b0..ca1723b 100644
--- a/compiler/optimizing/code_generator_x86.cc
+++ b/compiler/optimizing/code_generator_x86.cc
@@ -8236,6 +8236,7 @@
void CodeGeneratorX86::Finalize(CodeAllocator* allocator) {
// Generate the constant area if needed.
X86Assembler* assembler = GetAssembler();
+
if (!assembler->IsConstantAreaEmpty() || !fixups_to_jump_tables_.empty()) {
// Align to 4 byte boundary to reduce cache misses, as the data is 4 and 8
// byte values.