String Compression for ARM and ARM64
Changes on intrinsics and Code Generation on ARM and ARM64
for string compression feature. Currently the feature is off.
The size of boot.oat and boot.art for ARM before and after the
changes (feature OFF) are still. When the feature ON,
boot.oat increased by 0.60% and boot.art decreased by 9.38%.
Meanwhile for ARM64, size of boot.oat and boot.art before and
after changes (feature OFF) are still. When the feature ON,
boot.oat increased by 0.48% and boot.art decreased by 6.58%.
Turn feature on: runtime/mirror/string.h (kUseStringCompression = true)
runtime/asm_support.h (STRING_COMPRESSION_FEATURE 1)
Test: m -j31 test-art-target
All tests passed both when the mirror::kUseStringCompression
is ON and OFF.
Bug: 31040547
Change-Id: I24e86b99391df33ba27df747779b648c5a820649
diff --git a/compiler/optimizing/instruction_simplifier_arm64.cc b/compiler/optimizing/instruction_simplifier_arm64.cc
index 6d107d5..d0dd650 100644
--- a/compiler/optimizing/instruction_simplifier_arm64.cc
+++ b/compiler/optimizing/instruction_simplifier_arm64.cc
@@ -140,6 +140,13 @@
void InstructionSimplifierArm64Visitor::VisitArrayGet(HArrayGet* instruction) {
size_t data_offset = CodeGenerator::GetArrayDataOffset(instruction);
+ // Don't move the array pointer if it is charAt because we need to take the count first.
+ // TODO: Implement reading (length + compression) for String compression feature from
+ // negative offset (count_offset - data_offset) using LDP and clobbering an extra temporary.
+ // Note that "LDR (Immediate)" does not have a "signed offset" encoding.
+ if (mirror::kUseStringCompression && instruction->IsStringCharAt()) {
+ return;
+ }
if (TryExtractArrayAccessAddress(instruction,
instruction->GetArray(),
instruction->GetIndex(),