aboutsummaryrefslogtreecommitdiff
path: root/lib/CodeGen/SelectionDAG/ScheduleDAGList.cpp
AgeCommit message (Collapse)Author
2008-06-21Remove ScheduleDAG's SUnitMap altogether. Instead, use SDNode's NodeIdDan Gohman
field, which is otherwise unused after instruction selection, as an index into the SUnit array. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52583 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-21Change ScheduleDAG's SUnitMap from DenseMap<SDNode*, vector<SUnit*> >Dan Gohman
to DenseMap<SDNode*, SUnit*>, and adjust the way cloned SUnit nodes are handled so that only the original node needs to be in the map. This speeds up llc on 447.dealII.llvm.bc by about 2%. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52576 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-15Treat EntryToken nodes as "passive" so that they aren't added to theDan Gohman
ScheduleDAG; they don't correspond to any actual instructions so they don't need to be scheduled. This fixes a bug where the EntryToken was being scheduled multiple times in some cases, though it ended up not causing any trouble because EntryToken doesn't expand into anything. With this fixed the schedulers reliably schedule the expected number of units, so we can check this with an assertion. This requires a tweak to test/CodeGen/X86/loop-hoist.ll because it ends up getting scheduled differently in a trivial way, though it was enough to fool the prcontext+grep that the test does. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49701 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-10Rename MRegisterInfo to TargetRegisterInfo.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46930 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-29Use empty() instead of comparing size() with zero.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46514 91177308-0d34-0410-b5e6-96231b3b80d8
2007-12-31Rename SSARegMap -> MachineRegisterInfo in keeping with the idea Chris Lattner
that "machine" classes are used to represent the current state of the code being compiled. Given this expanded name, we can start moving other stuff into it. For now, move the UsedPhysRegs and LiveIn/LoveOuts vectors from MachineFunction into it. Update all the clients to match. This also reduces some needless #includes, such as MachineModuleInfo from MachineFunction. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@45467 91177308-0d34-0410-b5e6-96231b3b80d8
2007-12-29Remove attribution from file headers, per discussion on llvmdev.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@45418 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15Make CalcLatency() non-recursive.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@43017 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-28Trim some unneeded fields.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42442 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-25Added major new capabilities to scheduler (only BURR for now) to support ↵Evan Cheng
physical register dependency. The BURR scheduler can now backtrace and duplicate instructions in order to avoid "expensive / impossible to copy" values (e.g. status flag EFLAGS for x86) from being clobbered. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42284 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-19Use struct SDep instead of std::pair for SUnit pred and succ lists. First stepEvan Cheng
in tracking physical register output dependencies. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42125 91177308-0d34-0410-b5e6-96231b3b80d8
2007-07-13Skeleton of post-RA scheduler; doesn't do anything yet.Dale Johannesen
Change name of -sched option and DEBUG_TYPE to pre-RA-sched; adjust testcases. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@39816 91177308-0d34-0410-b5e6-96231b3b80d8
2007-02-03switch the sched unit map over to use a DenseMap instead of std::map. ThisChris Lattner
speeds up isel as a whole time by 2.6%. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@33810 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-19Eliminate static ctors from StatisticsChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32698 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-07Removing even more <iostream> includes.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32320 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-06Detemplatize the Statistic class. The only type it is instantiated withChris Lattner
is 'unsigned'. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32279 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-04Changes to use operand constraints to process two-address instructions.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31453 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-27s|llvm/Support/Visibility.h|llvm/Support/Compiler.h|Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29911 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-17switch the SUnit pred/succ sets from being std::sets to being smallvectors.Chris Lattner
This reduces selectiondag time on kc++ from 5.43s to 4.98s (9%). More significantly, this speeds up the default ppc scheduler from ~1571ms to 1063ms, a 33% speedup. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29743 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-02Final polish on machine pass registries.Jim Laskey
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29471 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-011. Change use of "Cache" to "Default".Jim Laskey
2. Added argument to instruction scheduler creators so the creators can do special things. 3. Repaired target hazard code. 4. Misc. More to follow. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29450 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-01Introducing plugable register allocators and instruction schedulers.Jim Laskey
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29434 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-20Reduce number of exported symbolsAndrew Lenharth
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29220 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-28Shave another 27K off libllvmgcc.dylib with visibility hiddenChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28973 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-30When a priority_queue is empty, the behavior of top() operator isEvan Cheng
non-deterministic. Returns NULL when it's empty! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28560 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-12Refactor a bunch of includes so that TargetMachine.h doesn't have to includeOwen Anderson
TargetData.h. This should make recompiles a bit faster with my current TargetData tinkering. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28238 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-11Refactor scheduler code. Move register-reduction list scheduler to aEvan Cheng
separate file. Added an initial implementation of top-down register pressure reduction list scheduler. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28226 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-10Templatify RegReductionPriorityQueueEvan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28212 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-09Add pseudo dependency to force a def&use operand to be scheduled last (unlessEvan Cheng
the distance between the def and another use is much longer). This is under option control for now "-sched-lower-defnuse". git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28201 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-05Fix VC++ compilation error.Jeff Cohen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28117 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-04Initial support for register pressure aware scheduling. The register reductionEvan Cheng
scheduler can go into a "vertical mode" (i.e. traversing up the two-address chain, etc.) when the register pressure is low. This does seem to reduce the number of spills in the cases I've looked at. But with x86, it's no guarantee the performance of the code improves. It can be turned on with -sched-vertically option. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28108 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-03Bottom up register pressure reduction work: clean up some hacks and enhancedEvan Cheng
the heuristic to further reduce spills for several test cases. (Note, it may not necessarily translate to runtime win!) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28076 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-01Dis-favor stores moreEvan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28035 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-01Bottom up register-pressure reduction scheduler now pushes store operationsEvan Cheng
up the schedule. This helps code that looks like this: loads ... computations (first set) ... stores (first set) ... loads computations (seccond set) ... stores (seccond set) ... Without this change, the stores and computations are more likely to interleave: loads ... loads ... computations (first set) ... computations (second set) ... computations (first set) ... stores (first set) ... computations (second set) ... stores (stores set) ... This can increase the number of spills if we are unlucky. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28033 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-01Didn't mean ScheduleDAGList.cpp to make the last checkin.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28030 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-01Remove temp. option -spiller-check-liveout, it didn't cause any failure nor ↵Evan Cheng
performance regressions. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28029 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-12Don't advance the hazard recognizer when there are no hazards and no ↵Chris Lattner
instructions to be emitted. Don't add one to the latency of a completed instruction if the latency of the op is 0. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26718 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-12Chain operands aren't real uses: they don't require the full latency of theChris Lattner
predecessor to finish before they can start. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26717 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-12As a pending queue data structure to keep track of instructions whoseChris Lattner
operands have all issued, but whose results are not yet available. This allows us to compile: int G; int test(int A, int B, int* P) { return (G+A)*(B+1); } to: _test: lis r2, ha16(L_G$non_lazy_ptr) addi r4, r4, 1 lwz r2, lo16(L_G$non_lazy_ptr)(r2) lwz r2, 0(r2) add r2, r2, r3 mullw r3, r2, r4 blr instead of this, which has a stall between the lis/lwz: _test: lis r2, ha16(L_G$non_lazy_ptr) lwz r2, lo16(L_G$non_lazy_ptr)(r2) addi r4, r4, 1 lwz r2, 0(r2) add r2, r2, r3 mullw r3, r2, r4 blr git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26716 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-11rename priorityqueue -> availablequeue. When a node is scheduled, rememberChris Lattner
which cycle it lands on. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26714 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-11Make CurrCycle a local var instead of an instance varChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26713 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-11Move some methods around so that BU specific code is together, TD specific codeChris Lattner
is together, and direction independent code is together. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26712 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-11merge preds/chainpreds -> preds setChris Lattner
merge succs/chainsuccs -> succs set This has no functionality change, simplifies the code, and reduces the size of sunits. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26711 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-10Move some simple-sched-specific instance vars to the simple scheduler.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26690 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-10Make EmitNode take a SDNode instead of a NodeInfo*Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26687 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-10Move the VRBase field from NodeInfo to being a separate, explicit, map.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26686 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-10no need to build groups anymoreChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26684 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-10Create SUnits directly from the SelectionDAG.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26683 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-10Push PrepareNodeInfo/IdentifyGroups down the inheritance hierarchyChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26682 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-10Teach the latency scheduler some new tricks. In particular, to break ties,Chris Lattner
keep track of a sense of "mobility", i.e. how many other nodes scheduling one node will free up. For something like this: float testadd(float *X, float *Y, float *Z, float *W, float *V) { return (*X+*Y)*(*Z+*W)+*V; } For example, this makes us schedule *X then *Y, not *X then *Z. The former allows us to issue the add, the later only lets us issue other loads. This turns the above code from this: _testadd: lfs f0, 0(r3) lfs f1, 0(r6) lfs f2, 0(r4) lfs f3, 0(r5) fadds f0, f0, f2 fadds f1, f3, f1 lfs f2, 0(r7) fmadds f1, f0, f1, f2 blr into this: _testadd: lfs f0, 0(r6) lfs f1, 0(r5) fadds f0, f1, f0 lfs f1, 0(r4) lfs f2, 0(r3) fadds f1, f2, f1 lfs f2, 0(r7) fmadds f1, f1, f0, f2 blr git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26680 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.2.3-18-g5258 (git 2.32.0) at 2025-04-30 02:42:55 +0000