aboutsummaryrefslogtreecommitdiffstats
path: root/lib/Target/X86/X86ISelDAGToDAG.cpp
Commit message (Collapse)AuthorAgeFilesLines
* Optimize atomic lock or that doesn't use the result value.Eric Christopher2011-05-101-0/+81
| | | | | | | | | Next up: xor and and. Part of rdar://8470697 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131171 91177308-0d34-0410-b5e6-96231b3b80d8
* Silence an overzealous uninitialized variable warning from GCC.Benjamin Kramer2011-04-231-1/+1
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@130053 91177308-0d34-0410-b5e6-96231b3b80d8
* X86: Try to use a smaller encoding by transforming (X << C1) & C2 into (X & ↵Benjamin Kramer2011-04-221-0/+75
| | | | | | | | | | | | | | | | | | | | | | | (C2 >> C1)) & C1. (Part of PR5039) This tends to happen a lot with bitfield code generated by clang. A simple example for x86_64 is uint64_t foo(uint64_t x) { return (x&1) << 42; } which used to compile into bloated code: shlq $42, %rdi ## encoding: [0x48,0xc1,0xe7,0x2a] movabsq $4398046511104, %rax ## encoding: [0x48,0xb8,0x00,0x00,0x00,0x00,0x00,0x04,0x00,0x00] andq %rdi, %rax ## encoding: [0x48,0x21,0xf8] ret ## encoding: [0xc3] with this patch we can fold the immediate into the and: andq $1, %rdi ## encoding: [0x48,0x83,0xe7,0x01] movq %rdi, %rax ## encoding: [0x48,0x89,0xf8] shlq $42, %rax ## encoding: [0x48,0xc1,0xe0,0x2a] ret ## encoding: [0xc3] It's possible to save another byte by using 'andl' instead of 'andq' but I currently see no way of doing that without making this code even more complicated. See the TODOs in the code. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@129990 91177308-0d34-0410-b5e6-96231b3b80d8
* Swap VT and DebugLoc operands of getExtLoad() for consistency withStuart Hastings2011-02-161-1/+1
| | | | | | | other getNode() methods. Radar 9002173. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@125665 91177308-0d34-0410-b5e6-96231b3b80d8
* Enhance ComputeMaskedBits to know that aligned frameindexesChris Lattner2011-02-131-21/+2
| | | | | | | | | | | | | | | | | | | have their low bits set to zero. This allows us to optimize out explicit stack alignment code like in stack-align.ll:test4 when it is redundant. Doing this causes the code generator to start turning FI+cst into FI|cst all over the place, which is general goodness (that is the canonical form) except that various pieces of the code generator don't handle OR aggressively. Fix this by introducing a new SelectionDAG::isBaseWithConstantOffset predicate, and using it in places that are looking for ADD(X,CST). The ARM backend in particular was missing a lot of addressing mode folding opportunities around OR. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@125470 91177308-0d34-0410-b5e6-96231b3b80d8
* lib/Target/X86/X86ISelDAGToDAG.cpp: __main should be WINCALL64 on Win64.NAKAMURA Takumi2011-01-271-1/+1
| | | | | | CALL64 marks %xmm* as dead. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@124354 91177308-0d34-0410-b5e6-96231b3b80d8
* fix PR8514, a bug where the "heroic" transformation of shift/and Chris Lattner2011-01-161-13/+9
| | | | | | | | | | into and/shift would cause nodes to move around and a dangling pointer to happen. The code tried to avoid this with a HandleSDNode, but got the details wrong. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123578 91177308-0d34-0410-b5e6-96231b3b80d8
* 'HiReg' is written but never read. Nuke itsTed Kremenek2011-01-141-5/+5
| | | | | | | | declaration and its assignments. Found by clang static analyzer. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123486 91177308-0d34-0410-b5e6-96231b3b80d8
* PR8918 - When used with MinGW64, LLVM generates a "calll __main" at theBill Wendling2011-01-061-2/+5
| | | | | | | | | beginning of the "main" function. The assembler complains about the invalid suffix for the 'call' instruction. The right instruction is "callq __main". Patch by KS Sreeram! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122933 91177308-0d34-0410-b5e6-96231b3b80d8
* rename MVT::Flag to MVT::Glue. "Flag" is a terrible name forChris Lattner2010-12-211-5/+5
| | | | | | | | | something that just glues two nodes together, even if it is sometimes used for flags. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122310 91177308-0d34-0410-b5e6-96231b3b80d8
* it turns out that when ".with.overflow" intrinsics were added to the X86Chris Lattner2010-12-051-4/+30
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | backend that they were all implemented except umul. This one fell back to the default implementation that did a hi/lo multiply and compared the top. Fix this to check the overflow flag that the 'mul' instruction sets, so we can avoid an explicit test. Now we compile: void *func(long count) { return new int[count]; } into: __Z4funcl: ## @_Z4funcl movl $4, %ecx ## encoding: [0xb9,0x04,0x00,0x00,0x00] movq %rdi, %rax ## encoding: [0x48,0x89,0xf8] mulq %rcx ## encoding: [0x48,0xf7,0xe1] seto %cl ## encoding: [0x0f,0x90,0xc1] testb %cl, %cl ## encoding: [0x84,0xc9] movq $-1, %rdi ## encoding: [0x48,0xc7,0xc7,0xff,0xff,0xff,0xff] cmoveq %rax, %rdi ## encoding: [0x48,0x0f,0x44,0xf8] jmp __Znam ## TAILCALL instead of: __Z4funcl: ## @_Z4funcl movl $4, %ecx ## encoding: [0xb9,0x04,0x00,0x00,0x00] movq %rdi, %rax ## encoding: [0x48,0x89,0xf8] mulq %rcx ## encoding: [0x48,0xf7,0xe1] testq %rdx, %rdx ## encoding: [0x48,0x85,0xd2] movq $-1, %rdi ## encoding: [0x48,0xc7,0xc7,0xff,0xff,0xff,0xff] cmoveq %rax, %rdi ## encoding: [0x48,0x0f,0x44,0xf8] jmp __Znam ## TAILCALL Other than the silly seto+test, this is using the o bit directly, so it's going in the right direction. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120935 91177308-0d34-0410-b5e6-96231b3b80d8
* Use a MemIntrinsicSDNode for ISD::PREFETCH, which touchesDale Johannesen2010-10-261-1/+0
| | | | | | | | | | memory, so a MachineMemOperand is useful (not propagated into the MachineInstr yet). No functional change except for dump output. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@117413 91177308-0d34-0410-b5e6-96231b3b80d8
* Use #NAME# to have the CMOV multiclass define things with the same names as ↵Chris Lattner2010-10-051-3/+3
| | | | | | | | | before (e.g. CMOVBE16rr instead of CMOVBErr16). git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@115705 91177308-0d34-0410-b5e6-96231b3b80d8
* switch CMOVBE to the multipattern:Chris Lattner2010-10-051-3/+3
| | | | | | | | | 21 insertions(+), 53 deletions(-) Moar change coming before I switch the rest. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@115697 91177308-0d34-0410-b5e6-96231b3b80d8
* Temporarily work around new address lowering while I figure out whatEric Christopher2010-09-221-1/+2
| | | | | | | needs to happen for darwin. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114577 91177308-0d34-0410-b5e6-96231b3b80d8
* reimplement elf TLS support in terms of addressing modes, eliminating ↵Chris Lattner2010-09-221-43/+35
| | | | | | SegmentBaseAddress. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114529 91177308-0d34-0410-b5e6-96231b3b80d8
* convert the last 4 X86ISD nodes that should have memoperands to have them.Chris Lattner2010-09-221-5/+1
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114523 91177308-0d34-0410-b5e6-96231b3b80d8
* give X86ISD::FNSTCW16m a memoperand, since it touches memory. It onlyChris Lattner2010-09-221-1/+0
| | | | | | | can access the stack due to how it is generated though. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114522 91177308-0d34-0410-b5e6-96231b3b80d8
* give FP_TO_INT16_IN_MEM and friends a memoperand. They are onlyChris Lattner2010-09-221-3/+1
| | | | | | | used with stack slots, but hey, lets be safe. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114521 91177308-0d34-0410-b5e6-96231b3b80d8
* give VZEXT_LOAD a memory operand, it now works with segment registers.Chris Lattner2010-09-221-1/+0
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114515 91177308-0d34-0410-b5e6-96231b3b80d8
* revert r114386 now that address modes work correctly, we get a niceChris Lattner2010-09-221-4/+0
| | | | | | | call through gs-relative memory now. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114510 91177308-0d34-0410-b5e6-96231b3b80d8
* give LCMPXCHG_DAG[8] a memory operand, allowing it to work with addrspace ↵Chris Lattner2010-09-211-1/+0
| | | | | | 256/257 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114508 91177308-0d34-0410-b5e6-96231b3b80d8
* reimplement support for GS and FS relative address space matchingChris Lattner2010-09-211-8/+38
| | | | | | | | | | | | | | | | | | by having X86DAGToDAGISel::SelectAddr get passed in the parent node of the operand match (the load/store/atomic op) and having it get the address space from that, instead of having special FS/GS addr mode operations that require duplicating the entire instruction set to support. This makes FS and GS relative accesses *far* more predictable and work much better. It also simplifies the X86 backend a bit, more to come. There is still a pending issue with nodes like ISD::PREFETCH and X86ISD::FLD, which really should be MemSDNode's but aren't. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114491 91177308-0d34-0410-b5e6-96231b3b80d8
* fix a long standing wart: all the ComplexPattern's were beingChris Lattner2010-09-211-12/+12
| | | | | | | | | | | | passed the root of the match, even though only a few patterns actually needed this (one in X86, several in ARM [which should be refactored anyway], and some in CellSPU that I don't feel like detangling). Instead of requiring all ComplexPatterns to take the dead root, have targets opt into getting the root by putting SDNPWantRoot on the ComplexPattern. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114471 91177308-0d34-0410-b5e6-96231b3b80d8
* even though I'm about to rip it out, simplify the address mode stuffChris Lattner2010-09-211-4/+2
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114468 91177308-0d34-0410-b5e6-96231b3b80d8
* propagate MachinePointerInfo through various uses of the oldChris Lattner2010-09-211-2/+3
| | | | | | | SelectionDAG::getExtLoad overload, and eliminate it. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114446 91177308-0d34-0410-b5e6-96231b3b80d8
* fix rdar://8453210, a crash handling a call through a GS relative load.Chris Lattner2010-09-211-1/+7
| | | | | | | For now, just disable folding the load into the call. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114386 91177308-0d34-0410-b5e6-96231b3b80d8
* zap dead code.Chris Lattner2010-09-041-6/+0
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@113073 91177308-0d34-0410-b5e6-96231b3b80d8
* Don't call Predicate_* from X86 target.Jakob Stoklund Olesen2010-09-031-8/+19
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@112921 91177308-0d34-0410-b5e6-96231b3b80d8
* Remove dead recursive function. Yay for clang -Wunused-function.Benjamin Kramer2010-08-251-7/+0
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@112060 91177308-0d34-0410-b5e6-96231b3b80d8
* PR7814: Truncates cannot be ignored for signed comparisons.Eli Friedman2010-08-041-1/+2
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@110268 91177308-0d34-0410-b5e6-96231b3b80d8
* Change LEA to have 5 operands for its memory operand, justChris Lattner2010-07-081-6/+7
| | | | | | | | | | | | like all other instructions, even though a segment is not allowed. This resolves a bunch of gross hacks in the encoder and makes LEA more consistent with the rest of the instruction set. No functionality change. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@107934 91177308-0d34-0410-b5e6-96231b3b80d8
* Move getExtLoad() and (some) getLoad() DebugLoc argument after EVT argument ↵Evan Cheng2010-07-071-1/+1
| | | | | | for consistency sake. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@107820 91177308-0d34-0410-b5e6-96231b3b80d8
* Propagate debug loc.Devang Patel2010-07-061-1/+2
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@107710 91177308-0d34-0410-b5e6-96231b3b80d8
* When creating X86 MUL8 and DIV8 instructions, make sure we don't produceJakob Stoklund Olesen2010-06-261-37/+49
| | | | | | | | | | | CopyFromReg nodes for aliasing registers (AX and AL). This confuses the fast register allocator. Instead of CopyFromReg(AL), use ExtractSubReg(CopyFromReg(AX), sub_8bit). This fixes PR7312. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106934 91177308-0d34-0410-b5e6-96231b3b80d8
* Change UpdateNodeOperands' operand and return value from SDValue toDan Gohman2010-06-181-3/+3
| | | | | | | SDNode *, since it doesn't care about the ResNo value. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106282 91177308-0d34-0410-b5e6-96231b3b80d8
* Don't maintain a set of deleted nodes; instead, use a HandleSDNodeDan Gohman2010-06-181-55/+29
| | | | | | | to track a node over CSE events. This fixes PR7368. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106266 91177308-0d34-0410-b5e6-96231b3b80d8
* Add first pass at darwin tls compiler support.Eric Christopher2010-06-031-1/+1
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@105381 91177308-0d34-0410-b5e6-96231b3b80d8
* Rename X86 subregister indices to something shorter.Jakob Stoklund Olesen2010-05-241-6/+6
| | | | | | Use the tablegen-produced enums. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104493 91177308-0d34-0410-b5e6-96231b3b80d8
* Don't leave Base.FrameIndex uninitialized, so that it doesn'tDan Gohman2010-04-291-34/+33
| | | | | | | print randomly in debug output. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@102668 91177308-0d34-0410-b5e6-96231b3b80d8
* Enable i16 to i32 promotion by default.Evan Cheng2010-04-281-0/+3
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@102493 91177308-0d34-0410-b5e6-96231b3b80d8
* teach the x86 address matching stuff to handleChris Lattner2010-04-201-7/+22
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (shl (or x,c), 3) the same as (shl (add x, c), 3) when x doesn't have any bits from c set. This finishes off PR1135. Before we compiled the block to: to: LBB0_3: ## %bb cmpb $4, %dl sete %dl addb %dl, %cl movb %cl, %dl shlb $2, %dl addb %r8b, %dl shlb $2, %dl movzbl %dl, %edx movl %esi, (%rdi,%rdx,4) leaq 2(%rdx), %r9 movl %esi, (%rdi,%r9,4) leaq 1(%rdx), %r9 movl %esi, (%rdi,%r9,4) addq $3, %rdx movl %esi, (%rdi,%rdx,4) incb %r8b decb %al movb %r8b, %dl jne LBB0_1 Now we produce: LBB0_3: ## %bb cmpb $4, %dl sete %dl addb %dl, %cl movb %cl, %dl shlb $2, %dl addb %r8b, %dl shlb $2, %dl movzbl %dl, %edx movl %esi, (%rdi,%rdx,4) movl %esi, 8(%rdi,%rdx,4) movl %esi, 4(%rdi,%rdx,4) movl %esi, 12(%rdi,%rdx,4) incb %r8b decb %al movb %r8b, %dl jne LBB0_1 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101958 91177308-0d34-0410-b5e6-96231b3b80d8
* Use const qualifiers with TargetLowering. This eliminates severalDan Gohman2010-04-171-5/+4
| | | | | | | | | | | | | | const_casts, and it reinforces the design of the Target classes being immutable. SelectionDAGISel::IsLegalToFold is now a static member function, because PIC16 uses it in an unconventional way. There is more room for API cleanup here. And PIC16's AsmPrinter no longer uses TargetLowering. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101635 91177308-0d34-0410-b5e6-96231b3b80d8
* Add const qualifiers to CodeGen's use of LLVM IR constructs.Dan Gohman2010-04-151-4/+3
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101334 91177308-0d34-0410-b5e6-96231b3b80d8
* Delete unneeeded arguments.Dan Gohman2010-04-141-5/+5
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101276 91177308-0d34-0410-b5e6-96231b3b80d8
* use DebugLoc default ctor instead of DebugLoc::getUnknownLoc()Chris Lattner2010-04-021-1/+1
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@100214 91177308-0d34-0410-b5e6-96231b3b80d8
* X86 address mode matching code MatchAddressRecursively does some aggressive ↵Evan Cheng2010-03-171-14/+60
| | | | | | hack which require doing a RAUW. It may end up deleting some SDNode up stream. It should avoid referencing deleted nodes. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@98780 91177308-0d34-0410-b5e6-96231b3b80d8
* Do not force indirect tailcall through fixed registers: eax, r11. Add ↵Evan Cheng2010-03-141-15/+22
| | | | | | support to allow loads to be folded to tail call instructions. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@98465 91177308-0d34-0410-b5e6-96231b3b80d8
* add a comment.Chris Lattner2010-03-041-0/+1
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97709 91177308-0d34-0410-b5e6-96231b3b80d8
* factor the 'sign extended from 8 bit' patterns better so Chris Lattner2010-03-031-6/+6
| | | | | | | | | that they are not destination type specific. This allows tblgen to factor them and the type check is redundant with what the isel does anyway. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97629 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.1 at 2025-01-26 12:27:49 +0000