Inflate contended lock word by suspending owner.
Bug 6961405.
Don't inflate monitors for Notify and NotifyAll.
Tidy lock word, handle recursive lock case alongside unlocked case and move
assembly out of line (except for ARM quick). Also handle null in out-of-line
assembly as the test is quick and the enter/exit code is already a safepoint.
To gain ownership of a monitor on behalf of another thread, monitor contenders
must not hold the monitor_lock_, so they wait on a condition variable.
Reduce size of per mutex contention log.
Be consistent in calling thin lock thread ids just thread ids.
Fix potential thread death races caused by the use of FindThreadByThreadId,
make it invariant that returned threads are either self or suspended now.
Code size reduction on ARM boot.oat 0.2%.
Old nexus 7 speedup 0.25%, new nexus 7 speedup 1.4%, nexus 10 speedup 2.24%,
nexus 4 speedup 2.09% on DeltaBlue.
Change-Id: Id52558b914f160d9c8578fdd7fc8199a9598576a
diff --git a/compiler/Android.mk b/compiler/Android.mk
index 66ff461..fc2f02b 100644
--- a/compiler/Android.mk
+++ b/compiler/Android.mk
@@ -74,7 +74,6 @@
llvm/md_builder.cc \
llvm/runtime_support_builder.cc \
llvm/runtime_support_builder_arm.cc \
- llvm/runtime_support_builder_thumb2.cc \
llvm/runtime_support_builder_x86.cc \
trampolines/trampoline_compiler.cc \
utils/arm/assembler_arm.cc \
diff --git a/compiler/dex/quick/arm/call_arm.cc b/compiler/dex/quick/arm/call_arm.cc
index bba2ec5..401da2a 100644
--- a/compiler/dex/quick/arm/call_arm.cc
+++ b/compiler/dex/quick/arm/call_arm.cc
@@ -440,88 +440,120 @@
}
/*
- * Handle simple case (thin lock) inline. If it's complicated, bail
- * out to the heavyweight lock/unlock routines. We'll use dedicated
- * registers here in order to be in the right position in case we
- * to bail to oat[Lock/Unlock]Object(self, object)
- *
- * r0 -> self pointer [arg0 for oat[Lock/Unlock]Object
- * r1 -> object [arg1 for oat[Lock/Unlock]Object
- * r2 -> intial contents of object->lock, later result of strex
- * r3 -> self->thread_id
- * r12 -> allow to be used by utilities as general temp
- *
- * The result of the strex is 0 if we acquire the lock.
- *
- * See comments in monitor.cc for the layout of the lock word.
- * Of particular interest to this code is the test for the
- * simple case - which we handle inline. For monitor enter, the
- * simple case is thin lock, held by no-one. For monitor exit,
- * the simple case is thin lock, held by the unlocking thread with
- * a recurse count of 0.
- *
- * A minor complication is that there is a field in the lock word
- * unrelated to locking: the hash state. This field must be ignored, but
- * preserved.
- *
+ * Handle unlocked -> thin locked transition inline or else call out to quick entrypoint. For more
+ * details see monitor.cc.
*/
void ArmMir2Lir::GenMonitorEnter(int opt_flags, RegLocation rl_src) {
FlushAllRegs();
- DCHECK_EQ(LW_SHAPE_THIN, 0);
LoadValueDirectFixed(rl_src, r0); // Get obj
LockCallTemps(); // Prepare for explicit register usage
- GenNullCheck(rl_src.s_reg_low, r0, opt_flags);
- LoadWordDisp(rARM_SELF, Thread::ThinLockIdOffset().Int32Value(), r2);
- NewLIR3(kThumb2Ldrex, r1, r0,
- mirror::Object::MonitorOffset().Int32Value() >> 2); // Get object->lock
- // Align owner
- OpRegImm(kOpLsl, r2, LW_LOCK_OWNER_SHIFT);
- // Is lock unheld on lock or held by us (==thread_id) on unlock?
- NewLIR4(kThumb2Bfi, r2, r1, 0, LW_LOCK_OWNER_SHIFT - 1);
- NewLIR3(kThumb2Bfc, r1, LW_HASH_STATE_SHIFT, LW_LOCK_OWNER_SHIFT - 1);
- OpRegImm(kOpCmp, r1, 0);
- OpIT(kCondEq, "");
- NewLIR4(kThumb2Strex, r1, r2, r0,
- mirror::Object::MonitorOffset().Int32Value() >> 2);
- OpRegImm(kOpCmp, r1, 0);
- OpIT(kCondNe, "T");
- // Go expensive route - artLockObjectFromCode(self, obj);
- LoadWordDisp(rARM_SELF, QUICK_ENTRYPOINT_OFFSET(pLockObject).Int32Value(), rARM_LR);
- ClobberCalleeSave();
- LIR* call_inst = OpReg(kOpBlx, rARM_LR);
- MarkSafepointPC(call_inst);
- GenMemBarrier(kLoadLoad);
+ constexpr bool kArchVariantHasGoodBranchPredictor = false; // TODO: true if cortex-A15.
+ if (kArchVariantHasGoodBranchPredictor) {
+ LIR* null_check_branch;
+ if ((opt_flags & MIR_IGNORE_NULL_CHECK) && !(cu_->disable_opt & (1 << kNullCheckElimination))) {
+ null_check_branch = nullptr; // No null check.
+ } else {
+ // If the null-check fails its handled by the slow-path to reduce exception related meta-data.
+ null_check_branch = OpCmpImmBranch(kCondEq, r0, 0, NULL);
+ }
+ LoadWordDisp(rARM_SELF, Thread::ThinLockIdOffset().Int32Value(), r2);
+ NewLIR3(kThumb2Ldrex, r1, r0, mirror::Object::MonitorOffset().Int32Value() >> 2);
+ LIR* not_unlocked_branch = OpCmpImmBranch(kCondNe, r1, 0, NULL);
+ NewLIR4(kThumb2Strex, r1, r2, r0, mirror::Object::MonitorOffset().Int32Value() >> 2);
+ LIR* lock_success_branch = OpCmpImmBranch(kCondEq, r1, 0, NULL);
+
+
+ LIR* slow_path_target = NewLIR0(kPseudoTargetLabel);
+ not_unlocked_branch->target = slow_path_target;
+ if (null_check_branch != nullptr) {
+ null_check_branch->target = slow_path_target;
+ }
+ // TODO: move to a slow path.
+ // Go expensive route - artLockObjectFromCode(obj);
+ LoadWordDisp(rARM_SELF, QUICK_ENTRYPOINT_OFFSET(pLockObject).Int32Value(), rARM_LR);
+ ClobberCalleeSave();
+ LIR* call_inst = OpReg(kOpBlx, rARM_LR);
+ MarkSafepointPC(call_inst);
+
+ LIR* success_target = NewLIR0(kPseudoTargetLabel);
+ lock_success_branch->target = success_target;
+ GenMemBarrier(kLoadLoad);
+ } else {
+ // Explicit null-check as slow-path is entered using an IT.
+ GenNullCheck(rl_src.s_reg_low, r0, opt_flags);
+ LoadWordDisp(rARM_SELF, Thread::ThinLockIdOffset().Int32Value(), r2);
+ NewLIR3(kThumb2Ldrex, r1, r0, mirror::Object::MonitorOffset().Int32Value() >> 2);
+ OpRegImm(kOpCmp, r1, 0);
+ OpIT(kCondEq, "");
+ NewLIR4(kThumb2Strex/*eq*/, r1, r2, r0, mirror::Object::MonitorOffset().Int32Value() >> 2);
+ OpRegImm(kOpCmp, r1, 0);
+ OpIT(kCondNe, "T");
+ // Go expensive route - artLockObjectFromCode(self, obj);
+ LoadWordDisp/*ne*/(rARM_SELF, QUICK_ENTRYPOINT_OFFSET(pLockObject).Int32Value(), rARM_LR);
+ ClobberCalleeSave();
+ LIR* call_inst = OpReg(kOpBlx/*ne*/, rARM_LR);
+ MarkSafepointPC(call_inst);
+ GenMemBarrier(kLoadLoad);
+ }
}
/*
- * For monitor unlock, we don't have to use ldrex/strex. Once
- * we've determined that the lock is thin and that we own it with
- * a zero recursion count, it's safe to punch it back to the
- * initial, unlock thin state with a store word.
+ * Handle thin locked -> unlocked transition inline or else call out to quick entrypoint. For more
+ * details see monitor.cc. Note the code below doesn't use ldrex/strex as the code holds the lock
+ * and can only give away ownership if its suspended.
*/
void ArmMir2Lir::GenMonitorExit(int opt_flags, RegLocation rl_src) {
- DCHECK_EQ(LW_SHAPE_THIN, 0);
FlushAllRegs();
LoadValueDirectFixed(rl_src, r0); // Get obj
LockCallTemps(); // Prepare for explicit register usage
- GenNullCheck(rl_src.s_reg_low, r0, opt_flags);
- LoadWordDisp(r0, mirror::Object::MonitorOffset().Int32Value(), r1); // Get lock
+ LIR* null_check_branch;
LoadWordDisp(rARM_SELF, Thread::ThinLockIdOffset().Int32Value(), r2);
- // Is lock unheld on lock or held by us (==thread_id) on unlock?
- OpRegRegImm(kOpAnd, r3, r1,
- (LW_HASH_STATE_MASK << LW_HASH_STATE_SHIFT));
- // Align owner
- OpRegImm(kOpLsl, r2, LW_LOCK_OWNER_SHIFT);
- NewLIR3(kThumb2Bfc, r1, LW_HASH_STATE_SHIFT, LW_LOCK_OWNER_SHIFT - 1);
- OpRegReg(kOpSub, r1, r2);
- OpIT(kCondEq, "EE");
- StoreWordDisp(r0, mirror::Object::MonitorOffset().Int32Value(), r3);
- // Go expensive route - UnlockObjectFromCode(obj);
- LoadWordDisp(rARM_SELF, QUICK_ENTRYPOINT_OFFSET(pUnlockObject).Int32Value(), rARM_LR);
- ClobberCalleeSave();
- LIR* call_inst = OpReg(kOpBlx, rARM_LR);
- MarkSafepointPC(call_inst);
- GenMemBarrier(kStoreLoad);
+ constexpr bool kArchVariantHasGoodBranchPredictor = false; // TODO: true if cortex-A15.
+ if (kArchVariantHasGoodBranchPredictor) {
+ if ((opt_flags & MIR_IGNORE_NULL_CHECK) && !(cu_->disable_opt & (1 << kNullCheckElimination))) {
+ null_check_branch = nullptr; // No null check.
+ } else {
+ // If the null-check fails its handled by the slow-path to reduce exception related meta-data.
+ null_check_branch = OpCmpImmBranch(kCondEq, r0, 0, NULL);
+ }
+ LoadWordDisp(r0, mirror::Object::MonitorOffset().Int32Value(), r1);
+ LoadConstantNoClobber(r3, 0);
+ LIR* slow_unlock_branch = OpCmpBranch(kCondNe, r1, r2, NULL);
+ StoreWordDisp(r0, mirror::Object::MonitorOffset().Int32Value(), r3);
+ LIR* unlock_success_branch = OpUnconditionalBranch(NULL);
+
+ LIR* slow_path_target = NewLIR0(kPseudoTargetLabel);
+ slow_unlock_branch->target = slow_path_target;
+ if (null_check_branch != nullptr) {
+ null_check_branch->target = slow_path_target;
+ }
+ // TODO: move to a slow path.
+ // Go expensive route - artUnlockObjectFromCode(obj);
+ LoadWordDisp(rARM_SELF, QUICK_ENTRYPOINT_OFFSET(pUnlockObject).Int32Value(), rARM_LR);
+ ClobberCalleeSave();
+ LIR* call_inst = OpReg(kOpBlx, rARM_LR);
+ MarkSafepointPC(call_inst);
+
+ LIR* success_target = NewLIR0(kPseudoTargetLabel);
+ unlock_success_branch->target = success_target;
+ GenMemBarrier(kStoreLoad);
+ } else {
+ // Explicit null-check as slow-path is entered using an IT.
+ GenNullCheck(rl_src.s_reg_low, r0, opt_flags);
+ LoadWordDisp(r0, mirror::Object::MonitorOffset().Int32Value(), r1); // Get lock
+ LoadWordDisp(rARM_SELF, Thread::ThinLockIdOffset().Int32Value(), r2);
+ LoadConstantNoClobber(r3, 0);
+ // Is lock unheld on lock or held by us (==thread_id) on unlock?
+ OpRegReg(kOpCmp, r1, r2);
+ OpIT(kCondEq, "EE");
+ StoreWordDisp/*eq*/(r0, mirror::Object::MonitorOffset().Int32Value(), r3);
+ // Go expensive route - UnlockObjectFromCode(obj);
+ LoadWordDisp/*ne*/(rARM_SELF, QUICK_ENTRYPOINT_OFFSET(pUnlockObject).Int32Value(), rARM_LR);
+ ClobberCalleeSave();
+ LIR* call_inst = OpReg(kOpBlx/*ne*/, rARM_LR);
+ MarkSafepointPC(call_inst);
+ GenMemBarrier(kStoreLoad);
+ }
}
void ArmMir2Lir::GenMoveException(RegLocation rl_dest) {
diff --git a/compiler/dex/quick/arm/int_arm.cc b/compiler/dex/quick/arm/int_arm.cc
index 07782d9..4fa0387 100644
--- a/compiler/dex/quick/arm/int_arm.cc
+++ b/compiler/dex/quick/arm/int_arm.cc
@@ -24,8 +24,7 @@
namespace art {
-LIR* ArmMir2Lir::OpCmpBranch(ConditionCode cond, int src1,
- int src2, LIR* target) {
+LIR* ArmMir2Lir::OpCmpBranch(ConditionCode cond, int src1, int src2, LIR* target) {
OpRegReg(kOpCmp, src1, src2);
return OpCondBranch(cond, target);
}
diff --git a/compiler/dex/quick/gen_common.cc b/compiler/dex/quick/gen_common.cc
index 4dd55d7..f38225a 100644
--- a/compiler/dex/quick/gen_common.cc
+++ b/compiler/dex/quick/gen_common.cc
@@ -1761,4 +1761,16 @@
suspend_launchpads_.Insert(launch_pad);
}
+/* Call out to helper assembly routine that will null check obj and then lock it. */
+void Mir2Lir::GenMonitorEnter(int opt_flags, RegLocation rl_src) {
+ FlushAllRegs();
+ CallRuntimeHelperRegLocation(QUICK_ENTRYPOINT_OFFSET(pLockObject), rl_src, true);
+}
+
+/* Call out to helper assembly routine that will null check obj and then unlock it. */
+void Mir2Lir::GenMonitorExit(int opt_flags, RegLocation rl_src) {
+ FlushAllRegs();
+ CallRuntimeHelperRegLocation(QUICK_ENTRYPOINT_OFFSET(pUnlockObject), rl_src, true);
+}
+
} // namespace art
diff --git a/compiler/dex/quick/mips/call_mips.cc b/compiler/dex/quick/mips/call_mips.cc
index d53c012..9a5ca2c 100644
--- a/compiler/dex/quick/mips/call_mips.cc
+++ b/compiler/dex/quick/mips/call_mips.cc
@@ -261,36 +261,6 @@
MarkSafepointPC(call_inst);
}
-/*
- * TODO: implement fast path to short-circuit thin-lock case
- */
-void MipsMir2Lir::GenMonitorEnter(int opt_flags, RegLocation rl_src) {
- FlushAllRegs();
- LoadValueDirectFixed(rl_src, rMIPS_ARG0); // Get obj
- LockCallTemps(); // Prepare for explicit register usage
- GenNullCheck(rl_src.s_reg_low, rMIPS_ARG0, opt_flags);
- // Go expensive route - artLockObjectFromCode(self, obj);
- int r_tgt = LoadHelper(QUICK_ENTRYPOINT_OFFSET(pLockObject));
- ClobberCalleeSave();
- LIR* call_inst = OpReg(kOpBlx, r_tgt);
- MarkSafepointPC(call_inst);
-}
-
-/*
- * TODO: implement fast path to short-circuit thin-lock case
- */
-void MipsMir2Lir::GenMonitorExit(int opt_flags, RegLocation rl_src) {
- FlushAllRegs();
- LoadValueDirectFixed(rl_src, rMIPS_ARG0); // Get obj
- LockCallTemps(); // Prepare for explicit register usage
- GenNullCheck(rl_src.s_reg_low, rMIPS_ARG0, opt_flags);
- // Go expensive route - UnlockObjectFromCode(obj);
- int r_tgt = LoadHelper(QUICK_ENTRYPOINT_OFFSET(pUnlockObject));
- ClobberCalleeSave();
- LIR* call_inst = OpReg(kOpBlx, r_tgt);
- MarkSafepointPC(call_inst);
-}
-
void MipsMir2Lir::GenMoveException(RegLocation rl_dest) {
int ex_offset = Thread::ExceptionOffset().Int32Value();
RegLocation rl_result = EvalLoc(rl_dest, kCoreReg, true);
@@ -318,6 +288,7 @@
FreeTemp(reg_card_base);
FreeTemp(reg_card_no);
}
+
void MipsMir2Lir::GenEntrySequence(RegLocation* ArgLocs, RegLocation rl_method) {
int spill_count = num_core_spills_ + num_fp_spills_;
/*
diff --git a/compiler/dex/quick/mips/codegen_mips.h b/compiler/dex/quick/mips/codegen_mips.h
index 8d0b347..95b2e77 100644
--- a/compiler/dex/quick/mips/codegen_mips.h
+++ b/compiler/dex/quick/mips/codegen_mips.h
@@ -123,8 +123,6 @@
void GenFusedLongCmpBranch(BasicBlock* bb, MIR* mir);
void GenSelect(BasicBlock* bb, MIR* mir);
void GenMemBarrier(MemBarrierKind barrier_kind);
- void GenMonitorEnter(int opt_flags, RegLocation rl_src);
- void GenMonitorExit(int opt_flags, RegLocation rl_src);
void GenMoveException(RegLocation rl_dest);
void GenMultiplyByTwoBitMultiplier(RegLocation rl_src, RegLocation rl_result, int lit,
int first_bit, int second_bit);
diff --git a/compiler/dex/quick/mir_to_lir.h b/compiler/dex/quick/mir_to_lir.h
index 7d6f968..71b68fe 100644
--- a/compiler/dex/quick/mir_to_lir.h
+++ b/compiler/dex/quick/mir_to_lir.h
@@ -629,8 +629,6 @@
virtual void GenFusedLongCmpBranch(BasicBlock* bb, MIR* mir) = 0;
virtual void GenSelect(BasicBlock* bb, MIR* mir) = 0;
virtual void GenMemBarrier(MemBarrierKind barrier_kind) = 0;
- virtual void GenMonitorEnter(int opt_flags, RegLocation rl_src) = 0;
- virtual void GenMonitorExit(int opt_flags, RegLocation rl_src) = 0;
virtual void GenMoveException(RegLocation rl_dest) = 0;
virtual void GenMultiplyByTwoBitMultiplier(RegLocation rl_src,
RegLocation rl_result, int lit, int first_bit,
@@ -689,6 +687,10 @@
virtual bool InexpensiveConstantLong(int64_t value) = 0;
virtual bool InexpensiveConstantDouble(int64_t value) = 0;
+ // May be optimized by targets.
+ virtual void GenMonitorEnter(int opt_flags, RegLocation rl_src);
+ virtual void GenMonitorExit(int opt_flags, RegLocation rl_src);
+
// Temp workaround
void Workaround7250540(RegLocation rl_dest, int value);
diff --git a/compiler/dex/quick/x86/call_x86.cc b/compiler/dex/quick/x86/call_x86.cc
index 2be2aa9..7fad6f0 100644
--- a/compiler/dex/quick/x86/call_x86.cc
+++ b/compiler/dex/quick/x86/call_x86.cc
@@ -150,43 +150,6 @@
rX86_ARG1, true);
}
-void X86Mir2Lir::GenMonitorEnter(int opt_flags, RegLocation rl_src) {
- FlushAllRegs();
- LoadValueDirectFixed(rl_src, rCX); // Get obj
- LockCallTemps(); // Prepare for explicit register usage
- GenNullCheck(rl_src.s_reg_low, rCX, opt_flags);
- // If lock is unheld, try to grab it quickly with compare and exchange
- // TODO: copy and clear hash state?
- NewLIR2(kX86Mov32RT, rDX, Thread::ThinLockIdOffset().Int32Value());
- NewLIR2(kX86Sal32RI, rDX, LW_LOCK_OWNER_SHIFT);
- NewLIR2(kX86Xor32RR, rAX, rAX);
- NewLIR3(kX86LockCmpxchgMR, rCX, mirror::Object::MonitorOffset().Int32Value(), rDX);
- LIR* branch = NewLIR2(kX86Jcc8, 0, kX86CondEq);
- // If lock is held, go the expensive route - artLockObjectFromCode(self, obj);
- CallRuntimeHelperReg(QUICK_ENTRYPOINT_OFFSET(pLockObject), rCX, true);
- branch->target = NewLIR0(kPseudoTargetLabel);
-}
-
-void X86Mir2Lir::GenMonitorExit(int opt_flags, RegLocation rl_src) {
- FlushAllRegs();
- LoadValueDirectFixed(rl_src, rAX); // Get obj
- LockCallTemps(); // Prepare for explicit register usage
- GenNullCheck(rl_src.s_reg_low, rAX, opt_flags);
- // If lock is held by the current thread, clear it to quickly release it
- // TODO: clear hash state?
- NewLIR2(kX86Mov32RT, rDX, Thread::ThinLockIdOffset().Int32Value());
- NewLIR2(kX86Sal32RI, rDX, LW_LOCK_OWNER_SHIFT);
- NewLIR3(kX86Mov32RM, rCX, rAX, mirror::Object::MonitorOffset().Int32Value());
- OpRegReg(kOpSub, rCX, rDX);
- LIR* branch = NewLIR2(kX86Jcc8, 0, kX86CondNe);
- NewLIR3(kX86Mov32MR, rAX, mirror::Object::MonitorOffset().Int32Value(), rCX);
- LIR* branch2 = NewLIR1(kX86Jmp8, 0);
- branch->target = NewLIR0(kPseudoTargetLabel);
- // Otherwise, go the expensive route - UnlockObjectFromCode(obj);
- CallRuntimeHelperReg(QUICK_ENTRYPOINT_OFFSET(pUnlockObject), rAX, true);
- branch2->target = NewLIR0(kPseudoTargetLabel);
-}
-
void X86Mir2Lir::GenMoveException(RegLocation rl_dest) {
int ex_offset = Thread::ExceptionOffset().Int32Value();
RegLocation rl_result = EvalLoc(rl_dest, kCoreReg, true);
diff --git a/compiler/dex/quick/x86/codegen_x86.h b/compiler/dex/quick/x86/codegen_x86.h
index 0f28110..29fb3a5 100644
--- a/compiler/dex/quick/x86/codegen_x86.h
+++ b/compiler/dex/quick/x86/codegen_x86.h
@@ -123,8 +123,6 @@
void GenFusedLongCmpBranch(BasicBlock* bb, MIR* mir);
void GenSelect(BasicBlock* bb, MIR* mir);
void GenMemBarrier(MemBarrierKind barrier_kind);
- void GenMonitorEnter(int opt_flags, RegLocation rl_src);
- void GenMonitorExit(int opt_flags, RegLocation rl_src);
void GenMoveException(RegLocation rl_dest);
void GenMultiplyByTwoBitMultiplier(RegLocation rl_src, RegLocation rl_result,
int lit, int first_bit, int second_bit);
diff --git a/compiler/image_test.cc b/compiler/image_test.cc
index 6464a4c..d4be7c0 100644
--- a/compiler/image_test.cc
+++ b/compiler/image_test.cc
@@ -23,6 +23,8 @@
#include "compiler/oat_writer.h"
#include "gc/space/image_space.h"
#include "image.h"
+#include "lock_word.h"
+#include "mirror/object-inl.h"
#include "signal_catcher.h"
#include "UniquePtr.h"
#include "utils.h"
@@ -158,7 +160,7 @@
// non image classes should be in a space after the image.
EXPECT_GT(reinterpret_cast<byte*>(klass), image_end) << descriptor;
}
- EXPECT_TRUE(Monitor::IsValidLockWord(*klass->GetRawLockWordAddress()));
+ EXPECT_TRUE(Monitor::IsValidLockWord(klass->GetLockWord()));
}
}
diff --git a/compiler/llvm/llvm_compilation_unit.cc b/compiler/llvm/llvm_compilation_unit.cc
index aa439cc..feb495e 100644
--- a/compiler/llvm/llvm_compilation_unit.cc
+++ b/compiler/llvm/llvm_compilation_unit.cc
@@ -82,7 +82,6 @@
#include "ir_builder.h"
#include "os.h"
#include "runtime_support_builder_arm.h"
-#include "runtime_support_builder_thumb2.h"
#include "runtime_support_builder_x86.h"
#include "utils_llvm.h"
@@ -118,12 +117,10 @@
default:
runtime_support_.reset(new RuntimeSupportBuilder(*context_, *module_, *irb_));
break;
+ case kThumb2:
case kArm:
runtime_support_.reset(new RuntimeSupportBuilderARM(*context_, *module_, *irb_));
break;
- case kThumb2:
- runtime_support_.reset(new RuntimeSupportBuilderThumb2(*context_, *module_, *irb_));
- break;
case kX86:
runtime_support_.reset(new RuntimeSupportBuilderX86(*context_, *module_, *irb_));
break;
diff --git a/compiler/llvm/runtime_support_builder.cc b/compiler/llvm/runtime_support_builder.cc
index 24e283d..c825fbf 100644
--- a/compiler/llvm/runtime_support_builder.cc
+++ b/compiler/llvm/runtime_support_builder.cc
@@ -164,89 +164,13 @@
/* Monitor */
void RuntimeSupportBuilder::EmitLockObject(::llvm::Value* object) {
- Value* monitor =
- irb_.LoadFromObjectOffset(object,
- mirror::Object::MonitorOffset().Int32Value(),
- irb_.getJIntTy(),
- kTBAARuntimeInfo);
-
- Value* real_monitor =
- irb_.CreateAnd(monitor, ~(LW_HASH_STATE_MASK << LW_HASH_STATE_SHIFT));
-
- // Is thin lock, unheld and not recursively acquired.
- Value* unheld = irb_.CreateICmpEQ(real_monitor, irb_.getInt32(0));
-
- Function* parent_func = irb_.GetInsertBlock()->getParent();
- BasicBlock* bb_fast = BasicBlock::Create(context_, "lock_fast", parent_func);
- BasicBlock* bb_slow = BasicBlock::Create(context_, "lock_slow", parent_func);
- BasicBlock* bb_cont = BasicBlock::Create(context_, "lock_cont", parent_func);
- irb_.CreateCondBr(unheld, bb_fast, bb_slow, kLikely);
-
- irb_.SetInsertPoint(bb_fast);
-
- // Calculate new monitor: new = old | (lock_id << LW_LOCK_OWNER_SHIFT)
- Value* lock_id =
- EmitLoadFromThreadOffset(Thread::ThinLockIdOffset().Int32Value(),
- irb_.getInt32Ty(), kTBAARuntimeInfo);
-
- Value* owner = irb_.CreateShl(lock_id, LW_LOCK_OWNER_SHIFT);
- Value* new_monitor = irb_.CreateOr(monitor, owner);
-
- // Atomically update monitor.
- Value* old_monitor =
- irb_.CompareExchangeObjectOffset(object,
- mirror::Object::MonitorOffset().Int32Value(),
- monitor, new_monitor, kTBAARuntimeInfo);
-
- Value* retry_slow_path = irb_.CreateICmpEQ(old_monitor, monitor);
- irb_.CreateCondBr(retry_slow_path, bb_cont, bb_slow, kLikely);
-
- irb_.SetInsertPoint(bb_slow);
Function* slow_func = GetRuntimeSupportFunction(runtime_support::LockObject);
irb_.CreateCall2(slow_func, object, EmitGetCurrentThread());
- irb_.CreateBr(bb_cont);
-
- irb_.SetInsertPoint(bb_cont);
}
void RuntimeSupportBuilder::EmitUnlockObject(::llvm::Value* object) {
- Value* lock_id =
- EmitLoadFromThreadOffset(Thread::ThinLockIdOffset().Int32Value(),
- irb_.getJIntTy(),
- kTBAARuntimeInfo);
- Value* monitor =
- irb_.LoadFromObjectOffset(object,
- mirror::Object::MonitorOffset().Int32Value(),
- irb_.getJIntTy(),
- kTBAARuntimeInfo);
-
- Value* my_monitor = irb_.CreateShl(lock_id, LW_LOCK_OWNER_SHIFT);
- Value* hash_state = irb_.CreateAnd(monitor, (LW_HASH_STATE_MASK << LW_HASH_STATE_SHIFT));
- Value* real_monitor = irb_.CreateAnd(monitor, ~(LW_HASH_STATE_MASK << LW_HASH_STATE_SHIFT));
-
- // Is thin lock, held by us and not recursively acquired
- Value* is_fast_path = irb_.CreateICmpEQ(real_monitor, my_monitor);
-
- Function* parent_func = irb_.GetInsertBlock()->getParent();
- BasicBlock* bb_fast = BasicBlock::Create(context_, "unlock_fast", parent_func);
- BasicBlock* bb_slow = BasicBlock::Create(context_, "unlock_slow", parent_func);
- BasicBlock* bb_cont = BasicBlock::Create(context_, "unlock_cont", parent_func);
- irb_.CreateCondBr(is_fast_path, bb_fast, bb_slow, kLikely);
-
- irb_.SetInsertPoint(bb_fast);
- // Set all bits to zero (except hash state)
- irb_.StoreToObjectOffset(object,
- mirror::Object::MonitorOffset().Int32Value(),
- hash_state,
- kTBAARuntimeInfo);
- irb_.CreateBr(bb_cont);
-
- irb_.SetInsertPoint(bb_slow);
Function* slow_func = GetRuntimeSupportFunction(runtime_support::UnlockObject);
irb_.CreateCall2(slow_func, object, EmitGetCurrentThread());
- irb_.CreateBr(bb_cont);
-
- irb_.SetInsertPoint(bb_cont);
}
diff --git a/compiler/llvm/runtime_support_builder.h b/compiler/llvm/runtime_support_builder.h
index e92ac0a..898611a 100644
--- a/compiler/llvm/runtime_support_builder.h
+++ b/compiler/llvm/runtime_support_builder.h
@@ -64,8 +64,8 @@
virtual void EmitTestSuspend();
/* Monitor */
- virtual void EmitLockObject(::llvm::Value* object);
- virtual void EmitUnlockObject(::llvm::Value* object);
+ void EmitLockObject(::llvm::Value* object);
+ void EmitUnlockObject(::llvm::Value* object);
/* MarkGCCard */
virtual void EmitMarkGCCard(::llvm::Value* value, ::llvm::Value* target_addr);
diff --git a/compiler/llvm/runtime_support_builder_arm.cc b/compiler/llvm/runtime_support_builder_arm.cc
index 569d825..cad4624 100644
--- a/compiler/llvm/runtime_support_builder_arm.cc
+++ b/compiler/llvm/runtime_support_builder_arm.cc
@@ -116,24 +116,5 @@
return old_thread_register;
}
-
-/* Monitor */
-
-void RuntimeSupportBuilderARM::EmitLockObject(Value* object) {
- RuntimeSupportBuilder::EmitLockObject(object);
- FunctionType* func_ty = FunctionType::get(/*Result=*/Type::getVoidTy(context_),
- /*isVarArg=*/false);
- InlineAsm* func = InlineAsm::get(func_ty, "dmb sy", "", true);
- irb_.CreateCall(func);
-}
-
-void RuntimeSupportBuilderARM::EmitUnlockObject(Value* object) {
- RuntimeSupportBuilder::EmitUnlockObject(object);
- FunctionType* func_ty = FunctionType::get(/*Result=*/Type::getVoidTy(context_),
- /*isVarArg=*/false);
- InlineAsm* func = InlineAsm::get(func_ty, "dmb sy", "", true);
- irb_.CreateCall(func);
-}
-
} // namespace llvm
} // namespace art
diff --git a/compiler/llvm/runtime_support_builder_arm.h b/compiler/llvm/runtime_support_builder_arm.h
index 5a353d7..0d01509 100644
--- a/compiler/llvm/runtime_support_builder_arm.h
+++ b/compiler/llvm/runtime_support_builder_arm.h
@@ -34,10 +34,6 @@
virtual void EmitStoreToThreadOffset(int64_t offset, ::llvm::Value* value,
TBAASpecialType s_ty);
virtual ::llvm::Value* EmitSetCurrentThread(::llvm::Value* thread);
-
- /* Monitor */
- virtual void EmitLockObject(::llvm::Value* object);
- virtual void EmitUnlockObject(::llvm::Value* object);
};
} // namespace llvm
diff --git a/compiler/llvm/runtime_support_builder_thumb2.cc b/compiler/llvm/runtime_support_builder_thumb2.cc
deleted file mode 100644
index eff29c8..0000000
--- a/compiler/llvm/runtime_support_builder_thumb2.cc
+++ /dev/null
@@ -1,90 +0,0 @@
-/*
- * Copyright (C) 2012 The Android Open Source Project
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- * http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-#include "runtime_support_builder_thumb2.h"
-
-#include "ir_builder.h"
-#include "mirror/object.h"
-#include "monitor.h"
-#include "thread.h"
-#include "utils_llvm.h"
-
-#include <llvm/IR/DerivedTypes.h>
-#include <llvm/IR/Function.h>
-#include <llvm/IR/InlineAsm.h>
-#include <llvm/IR/Module.h>
-#include <llvm/IR/Type.h>
-
-#include <inttypes.h>
-#include <vector>
-
-using ::llvm::BasicBlock;
-using ::llvm::Function;
-using ::llvm::FunctionType;
-using ::llvm::InlineAsm;
-using ::llvm::Type;
-using ::llvm::Value;
-
-namespace art {
-namespace llvm {
-
-
-void RuntimeSupportBuilderThumb2::EmitLockObject(Value* object) {
- FunctionType* func_ty = FunctionType::get(/*Result=*/irb_.getInt32Ty(),
- /*Params=*/irb_.getJObjectTy(),
- /*isVarArg=*/false);
- // $0: result
- // $1: object
- // $2: temp
- // $3: temp
- std::string asms;
- StringAppendF(&asms, "add $3, $1, #%" PRId32 "\n", mirror::Object::MonitorOffset().Int32Value());
- StringAppendF(&asms, "ldr $2, [r9, #%" PRId32 "]\n", Thread::ThinLockIdOffset().Int32Value());
- StringAppendF(&asms, "ldrex $0, [$3]\n");
- StringAppendF(&asms, "lsl $2, $2, %d\n", LW_LOCK_OWNER_SHIFT);
- StringAppendF(&asms, "bfi $2, $0, #0, #%d\n", LW_LOCK_OWNER_SHIFT - 1);
- StringAppendF(&asms, "bfc $0, #%d, #%d\n", LW_HASH_STATE_SHIFT, LW_LOCK_OWNER_SHIFT - 1);
- StringAppendF(&asms, "cmp $0, #0\n");
- StringAppendF(&asms, "it eq\n");
- StringAppendF(&asms, "strexeq $0, $2, [$3]\n");
-
- InlineAsm* func = InlineAsm::get(func_ty, asms, "=&l,l,~l,~l", true);
-
- Value* retry_slow_path = irb_.CreateCall(func, object);
- retry_slow_path = irb_.CreateICmpNE(retry_slow_path, irb_.getJInt(0));
-
- Function* parent_func = irb_.GetInsertBlock()->getParent();
- BasicBlock* basic_block_lock = BasicBlock::Create(context_, "lock", parent_func);
- BasicBlock* basic_block_cont = BasicBlock::Create(context_, "lock_cont", parent_func);
- irb_.CreateCondBr(retry_slow_path, basic_block_lock, basic_block_cont, kUnlikely);
-
- irb_.SetInsertPoint(basic_block_lock);
- Function* slow_func = GetRuntimeSupportFunction(runtime_support::LockObject);
- irb_.CreateCall2(slow_func, object, EmitGetCurrentThread());
- irb_.CreateBr(basic_block_cont);
-
- irb_.SetInsertPoint(basic_block_cont);
- { // Memory barrier
- FunctionType* asm_ty = FunctionType::get(/*Result=*/Type::getVoidTy(context_),
- /*isVarArg=*/false);
- InlineAsm* func = InlineAsm::get(asm_ty, "dmb sy", "", true);
- irb_.CreateCall(func);
- }
-}
-
-
-} // namespace llvm
-} // namespace art
diff --git a/compiler/llvm/runtime_support_builder_thumb2.h b/compiler/llvm/runtime_support_builder_thumb2.h
deleted file mode 100644
index c47a274..0000000
--- a/compiler/llvm/runtime_support_builder_thumb2.h
+++ /dev/null
@@ -1,37 +0,0 @@
-/*
- * Copyright (C) 2012 The Android Open Source Project
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- * http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-#ifndef ART_COMPILER_LLVM_RUNTIME_SUPPORT_BUILDER_THUMB2_H_
-#define ART_COMPILER_LLVM_RUNTIME_SUPPORT_BUILDER_THUMB2_H_
-
-#include "runtime_support_builder_arm.h"
-
-namespace art {
-namespace llvm {
-
-class RuntimeSupportBuilderThumb2 : public RuntimeSupportBuilderARM {
- public:
- RuntimeSupportBuilderThumb2(::llvm::LLVMContext& context, ::llvm::Module& module, IRBuilder& irb)
- : RuntimeSupportBuilderARM(context, module, irb) {}
-
- /* Monitor */
- virtual void EmitLockObject(::llvm::Value* object);
-};
-
-} // namespace llvm
-} // namespace art
-
-#endif // ART_COMPILER_LLVM_RUNTIME_SUPPORT_BUILDER_THUMB2_H_