aboutsummaryrefslogtreecommitdiff
path: root/lib/CodeGen
AgeCommit message (Collapse)Author
2010-06-24Remove the now unused LiveIntervals::getVNInfoSourceReg().Jakob Stoklund Olesen
This method was always a bit too simplistic for the real world. It didn't really deal with subregisters and such. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106781 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Teach AdjustCopiesBackFrom to also use CoalescerPair to identify compatible ↵Jakob Stoklund Olesen
copies. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106780 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Remove the -fast-spill option.Jakob Stoklund Olesen
This code path has never really been used, and we are going to be handling spilling through the Spiller interface in the future. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106777 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Loosen up the requirements in the Horrible Hack(tm) to include all selectorsBill Wendling
which don't have a catch-all associated with them not just clean-ups. This fixes the SingleSource/Benchmarks/Shootout-C++/except.cpp testcase that broke because of my change r105902. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106772 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Replace a big gob of old coalescer logic with the new CoalescerPair class.Jakob Stoklund Olesen
CoalescerPair can determine if a copy can be coalesced, and which register gets merged away. The old logic in SimpleRegisterCoalescing had evolved into something a bit too convoluted. This second attempt fixes some crashes that only occurred Linux. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106769 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Print the LSBs of a SlotIndex symbolically using letters referring to theJakob Stoklund Olesen
[L]oad, [u]se, [d]ef, or [S]tore slots. This makes it easier to see if two indices refer to the same instruction, avoiding mental mod 4 calculations. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106766 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Simplify this code; switch lowering shouldn't produce casesDan Gohman
which trivially fold away. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106765 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Be more strict about subreg-to-subreg copies in CoalescerPair.Jakob Stoklund Olesen
Also keep track of the original DstREg before subregister adjustments. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106753 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Verify that VNI kills are pointing to existing instructions.Jakob Stoklund Olesen
In this case it is essential that the kill is real because the spiller will decide to omit a spill if it thinks there is a later kill. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106751 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Eliminate the other half of the BRCOND optimization, and updateDan Gohman
as many tests as possible. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106749 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Eliminate the first have of the optimization which eliminates BRCONDDan Gohman
when the condition is constant. This optimization shouldn't be necessary, because codegen shouldn't be able to find dead control paths that the IR-level optimizer can't find. And it's undesirable, because it encourages bugpoint to leave "br i1 false" branches in its output. And it wasn't updating the CFG. I updated all the tests I could, but some tests are too reduced and I wasn't able to meaningfully preserve them. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106748 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Reapply r106634, now that the bug it exposed is fixed.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106746 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Optimize the "bit test" code path for switch lowering in theDan Gohman
case where the bit mask has exactly one bit. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106716 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Revert "Replace a big gob of old coalescer logic with the new CoalescerPair ↵Jakob Stoklund Olesen
class." Whiny buildbots. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106710 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24Replace a big gob of old coalescer logic with the new CoalescerPair class.Jakob Stoklund Olesen
CoalescerPair can determine if a copy can be coalesced, and which register gets merged away. The old logic in SimpleRegisterCoalescing had evolved into something a bit too convoluted. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106701 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-23MorphNodeTo doesn't preserve the memory operands. Because we're morphing a nodeBill Wendling
into the same node, but with different non-memory operands, we need to replace the memory operands after it's finished morphing. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106643 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-23Revert r106263, "Fold the ShrinkDemandedOps pass into the regular ↵Daniel Dunbar
DAGCombiner pass,"... it was causing both 'file' (with clang) and 176.gcc (with llvm-gcc) to be miscompiled. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106634 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-23Some targets don't require the fencing MEMBARRIER instructions surroundingJim Grosbach
atomic intrinsics, either because the use locking instructions for the atomics, or because they perform the locking directly. Add support in the DAG combiner to fold away the fences. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106630 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-23Add a few VNInfo data structure checks.Jakob Stoklund Olesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106627 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-23Revert r106066, "Create a more targeted fix for not sinking instructions ↵Daniel Dunbar
into a range where it"... it causes bzip2 to be miscompiled by Clang. Conflicts: lib/CodeGen/MachineSink.cpp git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106614 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-22Also convert SUBREG_TO_REG to a KILL when relevant, like the other subregJakob Stoklund Olesen
instructions. This does not affect codegen much because SUBREG_TO_REG is only used by X86 and X86 does not use the register scavenger, but it prevents verifier errors. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106583 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-22Move PHIElimination's SplitCriticalEdge for MachineBasicBlocks outDan Gohman
into a utility routine, teach it how to update MachineLoopInfo, and make use of it in MachineLICM to split critical edges on demand. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106555 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-22Remove the SimpleJoin optimization from SimpleRegisterCoalescing.Jakob Stoklund Olesen
Measurements show that it does not speed up coalescing, so there is no reason the keep the added complexity around. Also clean out some unused methods and static functions. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106548 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-22Use pre-increment instead of post-increment when the result is not used.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106542 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-22When unfolding a load, avoid assuming which instruction thatDan Gohman
kill and dead flags will end up on. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106520 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-22Use single interface, using twine, to get named metadata.Devang Patel
getNamedMetadata(). git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106518 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-22Tail merging pass shall not break up IT blocks. rdar://8115404Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106517 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-22Discard special LLVM prefix from linkage name.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106516 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-22Do not rely on Twine temporaries to survive.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106515 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-22Fix the new load-unfolding code to update LiveVariable's dead flags,Dan Gohman
in addition to the kill flags. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106512 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-21Teach two-address lowering how to unfold a load to open up commutingDan Gohman
opportunities. For example, this lets it emit this: movq (%rax), %rcx addq %rdx, %rcx instead of this: movq %rdx, %rcx addq (%rax), %rcx in the case where %rdx has subsequent uses. It's the same number of instructions, and usually the same encoding size on x86, but it appears faster, and in general, it may allow better scheduling for the load. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106493 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-21Use A.append(...) instead of A.insert(A.end(), ...) when A is aDan Gohman
SmallVector, and other SmallVector simplifications. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106452 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-21Revert r106422, which is breaking the non-fast-isel path.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106423 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-21More changes for non-top-down fast-isel.Dan Gohman
Split the code for materializing a value out of SelectionDAGBuilder::getValue into a helper function, so that it can be used in other ways. Add a new getNonRegisterValue function which uses it, for use in code which doesn't want a CopyFromReg even when FuncMap.ValueMap already has an entry for it. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106422 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-21Do one lookup instead of two.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106415 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-21Generalize this to look in the regular ValueMap in addition toDan Gohman
the LocalValueMap, to make it more flexible when fast-isel isn't proceding straight top-down. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106414 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-19Tidy.Bob Wilson
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106383 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18Teach regular and fast isel to set dead flags on unused implicit defsDan Gohman
on calls and similar instructions. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106353 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18Only run CoalesceExtSubRegs when we can expect LiveIntervalAnalysis to clean upJakob Stoklund Olesen
the inserted INSERT_SUBREGs after us. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106345 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18Allow ARM if-converter to be run after post allocation scheduling.Evan Cheng
- This fixed a number of bugs in if-converter, tail merging, and post-allocation scheduler. If-converter now runs branch folding / tail merging first to maximize if-conversion opportunities. - Also changed the t2IT instruction slightly. It now defines the ITSTATE register which is read by instructions in the IT block. - Added Thumb2 specific hazard recognizer to ensure the scheduler doesn't change the instruction ordering in the IT block (since IT mask has been finalized). It also ensures no other instructions can be scheduled between instructions in the IT block. This is not yet enabled. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106344 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18back-end libcall handling for ATOMIC_SWAP (__sync_lock_test_and_set)Jim Grosbach
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106342 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18TwoAddressInstructionPass::CoalesceExtSubRegs can insert INSERT_SUBREGJakob Stoklund Olesen
instructions, but it doesn't really understand live ranges, so the first INSERT_SUBREG uses an implicitly defined register. Fix it in LiveVariableAnalysis by adding the <undef> flag. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106333 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18Fix an inverted condition.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106330 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18Fix cross initialization compilation error.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106324 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18Teach iff-converter to properly count # of dups. It was not skipping over ↵Evan Cheng
dbg_value's which resulted in non-duplicated instructions being deleted. rdar://8104384. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106323 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18Add Expand-to-libcall support for additional atomics. This covers the usualJim Grosbach
entries used by llvm-gcc. *_[U]MIN and such can be added later if needed. This enables the front ends to simplify handling of the atomic intrinsics by removing the target-specific decision about which targets can handle the intrinsics. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106321 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18Don't leak RegClass2VRegMap, which is now a new[] array instead of aDan Gohman
std::vector. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106298 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18Start TargetRegisterClass indices at 0 instead of 1, so thatDan Gohman
MachineRegisterInfo doesn't have to confusingly allocate an extra entry. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106296 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18Fix PR7372: Conditional branches (at least on ARM) are treated as predicated,Bob Wilson
so when IfConverter::CopyAndPredicateBlock checks to see if it should ignore an instruction because it is a branch, it should not check if the branch is predicated. This case (when IgnoreBr is true) is only relevant from IfConvertTriangle, where new branches are inserted after the block has been copied and predicated. If the original branch is not removed, we end up with multiple conditional branches (possibly conflicting) at the end of the block. Aside from any immediate errors resulting from that, this confuses the AnalyzeBranch functions so that the branches are not analyzable. That in turn causes the IfConverter to think that the "Simple" pattern can be applied, and things go downhill fast because the "Simple" pattern does _not_ apply if the block can fall through. This is pretty fragile. If there are other degenerate cases where AnalyzeBranch fails, but where the block may still fall through, the IfConverter should not perform its "Simple" if-conversion. But, I don't know how to do that with the current AnalyzeBranch interface, so for now, the best thing seems to be to avoid creating branches that AnalyzeBranch cannot handle. Evan, please review! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106291 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18Don't bother calling releaseMemory before destroying the DominatorTreeBase.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106287 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2025-10-01 01:45:38 +0000