summaryrefslogtreecommitdiffstats
path: root/llvm/lib
Commit message (Collapse)AuthorAgeFilesLines
* Machine LICM will check that operands are defined outside of the loop. AlsoBill Wendling2008-01-022-40/+1
| | | | | | check that register isn't 0 before going further. llvm-svn: 45498
* Remove dead code.Bill Wendling2008-01-021-35/+0
| | | | llvm-svn: 45496
* darwin9 and above support aligned common symbols.Chris Lattner2008-01-023-12/+36
| | | | llvm-svn: 45494
* leopard and above support alignment for common symbols.Chris Lattner2008-01-023-11/+32
| | | | llvm-svn: 45493
* Use the new architecture to get the containing machine basic block for a machineBill Wendling2008-01-021-28/+28
| | | | | | | instruction. Also, use "splice" to move the new instruction instead of remove/insert (where it was leaking memory anyway). llvm-svn: 45492
* Don't be rude, emit debugging info where asked to.Nick Lewycky2008-01-021-5/+5
| | | | llvm-svn: 45485
* Move some more instruction creation methods from RegisterInfo into InstrInfo.Owen Anderson2008-01-0138-1198/+1276
| | | | llvm-svn: 45484
* Make MachineRegisterInfo::getVRegDef more efficient by aiming the keep the ↵Chris Lattner2008-01-011-4/+10
| | | | | | def of the vreg at the start of the list, so the list doesn't need to be traversed. llvm-svn: 45483
* switch the register iterator to act more like hte LLVM value iterator: ↵Chris Lattner2008-01-012-3/+17
| | | | | | | | | | | dereferencing it now returns the machineinstr of the use. To get the operand, use I.getOperand(). Add a new MachineRegisterInfo::replaceRegWith, which is basically like Value::replaceAllUsesWith. llvm-svn: 45482
* Adding C bindings for SwitchInst::addCase.Gordon Henriksen2008-01-011-0/+5
| | | | | | Patch by Bryan O'Sullivan! llvm-svn: 45481
* Add a trivial but handy function to efficiently return the machine Chris Lattner2008-01-011-0/+15
| | | | | | instruction that defines the specified vreg. Crazy. llvm-svn: 45480
* Implement automatically updated def/use lists for all MachineInstr register Chris Lattner2008-01-013-8/+331
| | | | | | | operands. The lists are currently kept in MachineRegisterInfo, but it does not yet provide an iterator interface to them. llvm-svn: 45477
* Fix a bug in my previous patch: refer to the impl not the pure virtual ↵Chris Lattner2008-01-012-2/+2
| | | | | | version. It's unclear why gcc would ever compile this... llvm-svn: 45476
* Fix a problem where lib/Target/TargetInstrInfo.h would include and useChris Lattner2008-01-0118-60/+74
| | | | | | | | | | a header file from libcodegen. This violates a layering order: codegen depends on target, not the other way around. The fix to this is to split TII into two classes, TII and TargetInstrInfoImpl, which defines stuff that depends on libcodegen. It is defined in libcodegen, where the base is not. llvm-svn: 45475
* Fix PR1833 - eh.exception and eh.selector return twoDuncan Sands2007-12-311-6/+28
| | | | | | | | values, which means doing extra legalization work. It would be easier to get this kind of thing right if there was some documentation... llvm-svn: 45472
* Move copyRegToReg from MRegisterInfo to TargetInstrInfo. This is part of theOwen Anderson2007-12-3129-261/+312
| | | | | | Machine-level API cleanup instigated by Chris. llvm-svn: 45470
* properly encapsulate the parent field of MBB and MI with get/set accessors.Chris Lattner2007-12-313-31/+35
| | | | llvm-svn: 45469
* update a couple of references to SSARegMap.Chris Lattner2007-12-312-3/+3
| | | | llvm-svn: 45468
* Rename SSARegMap -> MachineRegisterInfo in keeping with the idea Chris Lattner2007-12-3145-396/+413
| | | | | | | | | | | | | | 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. llvm-svn: 45467
* Add new shorter predicates for testing machine operands for various types: Chris Lattner2007-12-3039-179/+172
| | | | | | | | | | | | e.g. MO.isMBB() instead of MO.isMachineBasicBlock(). I don't plan on switching everything over, so new clients should just start using the shorter names. Remove old long accessors, switching everything over to use the short accessor: getMachineBasicBlock() -> getMBB(), getConstantPoolIndex() -> getIndex(), setMachineBasicBlock -> setMBB(), etc. llvm-svn: 45464
* More cleanups for MachineOperand:Chris Lattner2007-12-308-139/+131
| | | | | | | | | | - Eliminate the static "print" method for operands, moving it into MachineOperand::print. - Change various set* methods for register flags to take a bool for the value to set it to. Remove unset* methods. - Group methods more logically by operand flavor in MachineOperand.h llvm-svn: 45461
* MachineOperand:Chris Lattner2007-12-301-57/+38
| | | | | | | | | | | | | | | | | - Add getParent() accessors. - Move SubReg out of the AuxInfo union, to make way for future changes. - Remove the getImmedValue/setImmedValue methods. - in some MachineOperand::Create* methods, stop initializing fields that are dead. MachineInstr: - Delete one copy of the MachineInstr printing code, now there is only one dump format and one copy of the code. - Make MachineOperand use the parent field to get info about preg register names if no target info is otherwise available. - Move def/use/kill/dead flag printing to the machineoperand printer, so they are always printed for an operand. llvm-svn: 45460
* fix typo duncan noticed!Chris Lattner2007-12-301-1/+1
| | | | llvm-svn: 45459
* simpilfy some register printing code.Chris Lattner2007-12-301-16/+6
| | | | llvm-svn: 45458
* eliminate a copy of the machineoperand printing stuff. Keep the copy thatChris Lattner2007-12-301-34/+4
| | | | | | knows how to print offsets. llvm-svn: 45457
* Simplify and clean up some machine operand/instr printing/dumping stuff.Chris Lattner2007-12-301-9/+9
| | | | llvm-svn: 45456
* two register machineoperands are not identical unless their subregs match.Chris Lattner2007-12-301-1/+2
| | | | llvm-svn: 45455
* MachineOperand::getImmedValue -> MachineOperand::getImmChris Lattner2007-12-304-12/+9
| | | | llvm-svn: 45454
* Use MachineOperand::getImm instead of MachineOperand::getImmedValue. ↵Chris Lattner2007-12-3031-159/+156
| | | | | | Likewise setImmedValue -> setImm llvm-svn: 45453
* Add some doxygen comments to llvm-c/Core.h.Gordon Henriksen2007-12-301-1/+2
| | | | llvm-svn: 45450
* make machine operands fatter: give each one an up-pointer to the Chris Lattner2007-12-301-1/+7
| | | | | | machineinstr that owns it. llvm-svn: 45449
* If we have a load of a global address that's not modified during theBill Wendling2007-12-302-3/+60
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | function, then go ahead and hoist it out of the loop. This is the result: $ cat a.c volatile int G; int A(int N) { for (; N > 0; --N) G++; } $ llc -o - -relocation-model=pic _A: ... LBB1_2: # bb movl L_G$non_lazy_ptr-"L1$pb"(%eax), %esi incl (%esi) incl %edx cmpl %ecx, %edx jne LBB1_2 # bb ... $ llc -o - -relocation-model=pic -machine-licm _A: ... movl L_G$non_lazy_ptr-"L1$pb"(%eax), %eax LBB1_2: # bb incl (%eax) incl %edx cmpl %ecx, %edx jne LBB1_2 # bb ... I'm limiting this to the MOV32rm x86 instruction for now. llvm-svn: 45444
* use simplified operand addition methods.Chris Lattner2007-12-302-10/+11
| | | | llvm-svn: 45437
* use simplified operand addition methods.Chris Lattner2007-12-301-17/+25
| | | | llvm-svn: 45436
* use simplified operand addition methods.Chris Lattner2007-12-301-30/+22
| | | | llvm-svn: 45435
* Shrinkify the machine operand creation method names.Chris Lattner2007-12-301-1/+1
| | | | llvm-svn: 45433
* Start using the simplified methods for adding operands.Chris Lattner2007-12-303-28/+29
| | | | llvm-svn: 45432
* simplify some code by factoring operand construction better.Chris Lattner2007-12-301-22/+4
| | | | llvm-svn: 45428
* Remove attribution from file headers, per discussion on llvmdev.Chris Lattner2007-12-29548-1292/+1451
| | | | llvm-svn: 45418
* remove attribution from lib Makefiles.Chris Lattner2007-12-2935-71/+71
| | | | llvm-svn: 45415
* this is done.Chris Lattner2007-12-291-7/+0
| | | | llvm-svn: 45408
* One readme entry is done, one is really easy (Evan, want to investigateChris Lattner2007-12-291-77/+25
| | | | | | | | eliminating the llvm.x86.sse2.loadl.pd intrinsic?), one shuffle optzn may be done (if shufps is better than pinsw, Evan, please review), and we already know about LICM of simple instructions. llvm-svn: 45407
* Fold comparisons against a constant nan, and optimize ORD/UNORD Chris Lattner2007-12-292-17/+22
| | | | | | | | | | | | | | | | | | | | | | | | comparisons with a constant. This allows us to compile isnan to: _foo: fcmpu cr7, f1, f1 mfcr r2 rlwinm r3, r2, 0, 31, 31 blr instead of: LCPI1_0: ; float .space 4 _foo: lis r2, ha16(LCPI1_0) lfs f0, lo16(LCPI1_0)(r2) fcmpu cr7, f1, f0 mfcr r2 rlwinm r3, r2, 0, 31, 31 blr llvm-svn: 45405
* this xform is implemented.Chris Lattner2007-12-291-22/+0
| | | | llvm-svn: 45404
* Disable null pointer folding transforms for non-generic address spaces. This ↵Christopher Lamb2007-12-292-4/+12
| | | | | | should probably be a target-specific predicate based on address space. That way for targets where this isn't applicable the predicate can be optimized away. llvm-svn: 45403
* make sure not to zap volatile stores, thanks a lot to Dale for noticing this!Chris Lattner2007-12-291-1/+2
| | | | llvm-svn: 45402
* Codegen:Chris Lattner2007-12-291-4/+29
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | as: _bar: pushl %esi subl $8, %esp movl 16(%esp), %esi call L_foo$stub fstps (%esi) addl $8, %esp popl %esi #FP_REG_KILL ret instead of: _bar: pushl %esi subl $8, %esp movl 16(%esp), %esi call L_foo$stub fstpl (%esi) cvtsd2ss (%esi), %xmm0 movss %xmm0, (%esi) addl $8, %esp popl %esi #FP_REG_KILL ret llvm-svn: 45401
* don't fold fp_round(fp_extend(load)) -> fp_round(extload)Chris Lattner2007-12-291-0/+4
| | | | llvm-svn: 45400
* avoid going through a stack slot to convert from fpstack to xmm regChris Lattner2007-12-292-29/+39
| | | | | | | | | if we are just going to store it back anyway. This improves things like: double foo(); void bar(double *P) { *P = foo(); } llvm-svn: 45399
* Delete a store whose input is a load from the same pointer:Chris Lattner2007-12-291-1/+12
| | | | | | | x = load p store x -> p llvm-svn: 45398
OpenPOWER on IntegriCloud