aboutsummaryrefslogtreecommitdiff
path: root/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
AgeCommit message (Collapse)Author
2007-07-09Preserve volatililty and alignment information when lowering orDan Gohman
simplifying loads and stores. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@38473 91177308-0d34-0410-b5e6-96231b3b80d8
2007-07-09Fix this warning:Chris Lattner
DAGCombiner.cpp: In member function 'llvm::SDOperand<unnamed>::DAGCombiner::visitOR(llvm::SDNode*)': DAGCombiner.cpp:1608: warning: passing negative value '-0x00000000000000001' for argument 1 to 'llvm::SDOperand llvm::SelectionDAG::getConstant(uint64_t, llvm::MVT::ValueType, bool)' oiy. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@38458 91177308-0d34-0410-b5e6-96231b3b80d8
2007-07-03Fix several over-aggressive folds for undef nodes in dagcombine, toDan Gohman
follow the rules for undef used in instcombine. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37851 91177308-0d34-0410-b5e6-96231b3b80d8
2007-07-02Teach GetNegatedExpression to negate 0-B to B in UnsafeFPMath mode, andDan Gohman
visitFSUB to fold 0-B to -B in UnsafeFPMath mode. Also change visitFNEG to use isNegatibleForFree/GetNegatedExpression instead of doing a subset of the same thing manually. This fixes test/CodeGen/X86/negative-sin.ll. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37842 91177308-0d34-0410-b5e6-96231b3b80d8
2007-06-25Generalize MVT::ValueType and associated functions to be able to representDan Gohman
extended vector types. Remove the special SDNode opcodes used for pre-legalize vector operations, and the special MVT::Vector type used with them. Adjust lowering and legalize to work with the normal SDNode kinds instead, and to use the normal MVT functions to work with vector types instead of using the two special operands that the pre-legalize nodes held. This allows pre-legalize and post-legalize DAGs, and the code that operates on them, to be more consistent. Pre-legalize vector operators can be handled more consistently with scalar operators. And, -view-dag-combine1-dags and -view-legalize-dags now look prettier for vector code. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37719 91177308-0d34-0410-b5e6-96231b3b80d8
2007-06-22Move ComputeMaskedBits, MaskedValueIsZero, and ComputeNumSignBits fromDan Gohman
TargetLowering to SelectionDAG so that they have more convenient access to the current DAG, in preparation for the ValueType routines being changed from standalone functions to members of SelectionDAG for the pre-legalize vector type changes. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37704 91177308-0d34-0410-b5e6-96231b3b80d8
2007-06-21Xforms:Evan Cheng
(add (select cc, 0, c), x) -> (select cc, x, (add, x, c)) (sub x, (select cc, 0, c)) -> (select cc, x, (sub, x, c)) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37685 91177308-0d34-0410-b5e6-96231b3b80d8
2007-06-19Pass a SelectionDAG into SDNode::dump everywhere it's used, in preprationDan Gohman
for needing the DAG node to print pre-legalize extended value types, and to get better debug messages with target-specific nodes. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37656 91177308-0d34-0410-b5e6-96231b3b80d8
2007-06-14Rename MVT::getVectorBaseType to MVT::getVectorElementType.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37579 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-25tighten up recursion depth againChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37330 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-24Fix a typo that caused combiner to create mal-formed pre-indexed store where ↵Evan Cheng
value store is the same as the base pointer. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37318 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-23prevent exponential recursion in isNegatibleForFreeChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37310 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-18Qualify calls to getTypeForValueType with MVT:: too.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37233 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-16Don't fold bitconvert(load) for preinc/postdec loads. Likewise stores.Dale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37130 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-16Use a ptr set instead of a linear search to unique TokenFactor operands.Chris Lattner
This fixes PR1423 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37102 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-16Bug fix: should check ABI alignment, not pref. alignment.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37094 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-15Fix an infinite recursion in GetNegatedExpression.Lauro Ramos Venancio
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37086 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-14implement a simple fneg optimization/propagation thing. This compiles:Chris Lattner
CodeGen/PowerPC/fneg.ll into: _t4: fmul f0, f3, f4 fmadd f1, f1, f2, f0 blr instead of: _t4: fneg f0, f3 fmul f0, f0, f4 fmsub f1, f1, f2, f0 blr git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@37054 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-09Can't fold the bit_convert is the store is a truncating store.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@36962 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-07Forgot a check.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@36910 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-07Enable a couple of xforms:Evan Cheng
- (store (bitconvert v)) -> (store v) if resultant store does not require higher alignment - (bitconvert (load v)) -> (load (bitconvert*)v) if resultant load does not require higher alignment git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@36908 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-03Don't create indexed load / store with zero offset!Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@36716 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-01Forgot about chain result; also UNDEF cannot have multiple values.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@36622 91177308-0d34-0410-b5e6-96231b3b80d8
2007-05-01* Only turn a load to UNDEF if all of its outputs have no uses (indexed loadsEvan Cheng
produce two results.) * Do not touch volatile loads. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@36604 91177308-0d34-0410-b5e6-96231b3b80d8
2007-04-22PR400 phase 2. Propagate attributed load/store information through DAGs.Christopher Lamb
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@36356 91177308-0d34-0410-b5e6-96231b3b80d8
2007-04-21Revert Christopher Lamb's load/store alignment changes.Reid Spencer
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@36309 91177308-0d34-0410-b5e6-96231b3b80d8
2007-04-21add support for alignment attributes on load/store instructionsChristopher Lamb
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@36301 91177308-0d34-0410-b5e6-96231b3b80d8
2007-04-18allow SRL to simplify its operands, as it doesn't demand all bits as input.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@36245 91177308-0d34-0410-b5e6-96231b3b80d8
2007-04-18When replacing a node in SimplifyDemandedBits, if the old node used anyChris Lattner
single-use nodes, they will be dead soon. Make sure to remove them before processing other nodes. This implements CodeGen/X86/shl_elim.ll git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@36244 91177308-0d34-0410-b5e6-96231b3b80d8
2007-04-17SIGN_EXTEND_INREG does not demand its top bits. Give SimplifyDemandedBitsChris Lattner
a chance to hack on it. This compiles: int baz(long long a) { return (short)(((int)(a >>24)) >> 9); } into: _baz: slwi r2, r3, 8 srwi r2, r2, 9 extsh r3, r2 blr instead of: _baz: srwi r2, r4, 24 rlwimi r2, r3, 8, 0, 23 srwi r2, r2, 9 extsh r3, r2 blr This implements CodeGen/PowerPC/sign_ext_inreg1.ll git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@36212 91177308-0d34-0410-b5e6-96231b3b80d8
2007-04-11fix an infinite loop compiling ldecod, notice by JeffC.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35910 91177308-0d34-0410-b5e6-96231b3b80d8
2007-04-11Fix this harder.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35888 91177308-0d34-0410-b5e6-96231b3b80d8
2007-04-11don't create shifts by zero, fix some problems with my previous patchChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35887 91177308-0d34-0410-b5e6-96231b3b80d8
2007-04-11Teach the codegen to turn [aez]ext (setcc) -> selectcc of 1/0, which oftenChris Lattner
allows other simplifications. For example, this compiles: int isnegative(unsigned int X) { return !(X < 2147483648U); } Into this code: x86: movl 4(%esp), %eax shrl $31, %eax ret arm: mov r0, r0, lsr #31 bx lr thumb: lsr r0, r0, #31 bx lr instead of: x86: cmpl $0, 4(%esp) sets %al movzbl %al, %eax ret arm: mov r3, #0 cmp r0, #0 movlt r3, #1 mov r0, r3 bx lr thumb: mov r2, #1 mov r1, #0 cmp r0, #0 blt LBB1_2 @entry LBB1_1: @entry cpy r2, r1 LBB1_2: @entry cpy r0, r2 bx lr Testcase here: test/CodeGen/Generic/ispositive.ll git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35883 91177308-0d34-0410-b5e6-96231b3b80d8
2007-04-11Codegen integer abs more efficiently using the trick from the PPC CWG. ThisChris Lattner
improves codegen on many architectures. Tests committed as CodeGen/*/iabs.ll X86 Old: X86 New: _test: _test: movl 4(%esp), %ecx movl 4(%esp), %eax movl %ecx, %eax movl %eax, %ecx negl %eax sarl $31, %ecx testl %ecx, %ecx addl %ecx, %eax cmovns %ecx, %eax xorl %ecx, %eax ret ret PPC Old: PPC New: _test: _test: cmpwi cr0, r3, -1 srawi r2, r3, 31 neg r2, r3 add r3, r3, r2 bgt cr0, LBB1_2 ; xor r3, r3, r2 LBB1_1: ; blr mr r3, r2 LBB1_2: ; blr ARM Old: ARM New: _test: _test: rsb r3, r0, #0 add r3, r0, r0, asr #31 cmp r0, #0 eor r0, r3, r0, asr #31 movge r3, r0 bx lr mov r0, r3 bx lr Thumb Old: Thumb New: _test: _test: neg r2, r0 asr r2, r0, #31 cmp r0, #0 add r0, r0, r2 bge LBB1_2 eor r0, r2 LBB1_1: @ bx lr cpy r0, r2 LBB1_2: @ bx lr Sparc Old: Sparc New: test: test: save -96, %o6, %o6 save -96, %o6, %o6 sethi 0, %l0 sra %i0, 31, %l0 sub %l0, %i0, %l0 add %i0, %l0, %l1 subcc %i0, -1, %l1 xor %l1, %l0, %i0 bg .BB1_2 restore %g0, %g0, %g0 nop retl .BB1_1: nop or %g0, %l0, %i0 .BB1_2: restore %g0, %g0, %g0 retl nop It also helps alpha/ia64 :) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35881 91177308-0d34-0410-b5e6-96231b3b80d8
2007-04-021. Insert custom lowering hooks for ISD::ROTR and ISD::ROTL.Scott Michel
2. Help DAGCombiner recognize zero/sign/any-extended versions of ROTR and ROTL patterns. This was motivated by the X86/rotate.ll testcase, which should now generate code for other platforms (and soon-to-come platforms.) Rewrote code slightly to make it easier to read. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35605 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-30Fix incorrect combination of different loads. Reenable zext-over-truncateDale Johannesen
combination. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35517 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-29Disable load width reduction xform of variant (zext (truncate load x)) forEvan Cheng
big endian targets until llvm-gcc build issue has been resolved. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35449 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-26SIGN_EXTEND_INREG requires one extra operand, a ValueType node.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35350 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-24Adjust offset to compensate for big endian machines.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35293 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-23Make sure SEXTLOAD of the specific type is supported on the target.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35289 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-23Also replace uses of SRL if that's also folded during ReduceLoadWidth().Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35286 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-23A couple of bug fixes for reducing load width xform:Evan Cheng
1. Address offset is in bytes. 2. Make sure truncate node uses are replaced with new load. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35274 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-22More opportunities to reduce load size.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35254 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-21fold (truncate (srl (load x), c)) -> (smaller load (x+c/vt bits))Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35239 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-07Avoid combining indexed load further.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35005 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-04fold away addc nodes when we know there cannot be a carry-out.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@34913 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-04generalizeChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@34910 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-04canonicalize constants to the RHS of addc/adde. If nothing uses the carry ↵Chris Lattner
out of addc, turn it into add. This allows us to compile: long long test(long long A, unsigned B) { return (A + ((long long)B << 32)) & 123; } into: _test: movl $123, %eax andl 4(%esp), %eax xorl %edx, %edx ret instead of: _test: xorl %edx, %edx movl %edx, %eax addl 4(%esp), %eax ;; add of zero andl $123, %eax ret git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@34909 91177308-0d34-0410-b5e6-96231b3b80d8
2007-02-26Fold (sext (truncate x)) more aggressively, by avoiding creation of aChris Lattner
sextinreg if not needed. This is useful in two cases: before legalize, it avoids creating a sextinreg that will be trivially removed. After legalize if the target doesn't support sextinreg, the trunc/sext would not have been removed before. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@34621 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2025-09-19 14:08:27 +0000