x86_64: Hard Float ABI support in QCG
This patch shows our efforts on resolving the ART limitations:
- passing "float"/"double" arguments via FPR
- passing "long" arguments via single GPR, not pair
- passing more than 3 agruments via GPR.
Work done:
- Extended SpecialTargetRegister enum with kARG4, kARG5, fARG4..fARG7.
- Created initial LoadArgRegs/GenDalvikX/FlushIns version in X86Mir2Lir.
- Unlimited number of long/double/float arguments support
- Refactored (v2)
Change-Id: I5deadd320b4341d5b2f50ba6fa4a98031abc3902
Signed-off-by: Serguei Katkov <serguei.i.katkov@intel.com>
Signed-off-by: Dmitry Petrochenko <dmitry.petrochenko@intel.com>
Signed-off-by: Chao-ying Fu <chao-ying.fu@intel.com>
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
diff --git a/compiler/dex/quick/mir_to_lir.h b/compiler/dex/quick/mir_to_lir.h
index ed94a8d..9718acd 100644
--- a/compiler/dex/quick/mir_to_lir.h
+++ b/compiler/dex/quick/mir_to_lir.h
@@ -910,13 +910,13 @@
void GenInvoke(CallInfo* info);
void GenInvokeNoInline(CallInfo* info);
virtual void FlushIns(RegLocation* ArgLocs, RegLocation rl_method);
- int GenDalvikArgsNoRange(CallInfo* info, int call_state, LIR** pcrLabel,
+ virtual int GenDalvikArgsNoRange(CallInfo* info, int call_state, LIR** pcrLabel,
NextCallInsn next_call_insn,
const MethodReference& target_method,
uint32_t vtable_idx,
uintptr_t direct_code, uintptr_t direct_method, InvokeType type,
bool skip_this);
- int GenDalvikArgsRange(CallInfo* info, int call_state, LIR** pcrLabel,
+ virtual int GenDalvikArgsRange(CallInfo* info, int call_state, LIR** pcrLabel,
NextCallInsn next_call_insn,
const MethodReference& target_method,
uint32_t vtable_idx,