Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Protect R31's frame offset from being used by callee-saved registers, when R31 | Nicolas Geoffray | 2007-03-21 | 1 | -0/+20 |
| | | | | | | is the frame pointer. llvm-svn: 35233 | ||||
* | Added MRegisterInfo hook to re-materialize an instruction. | Evan Cheng | 2007-03-20 | 1 | -0/+9 |
| | | | | llvm-svn: 35205 | ||||
* | Minor interface change. | Evan Cheng | 2007-03-06 | 1 | -1/+2 |
| | | | | llvm-svn: 34967 | ||||
* | PEI now passes a RegScavenger ptr to eliminateFrameIndex. | Evan Cheng | 2007-02-28 | 1 | -2/+2 |
| | | | | llvm-svn: 34707 | ||||
* | Duplicate use of LR, take 2. | Jim Laskey | 2007-02-27 | 1 | -26/+34 |
| | | | | llvm-svn: 34666 | ||||
* | Backing out Jim's LR spill changes. This was causing llvm-gcc bootstrapping | Evan Cheng | 2007-02-27 | 1 | -11/+2 |
| | | | | | | | | | to infinite loop: PPCMachineFunctionInfo.h updated: 1.2 -> 1.3 PPCRegisterInfo.cpp updated: 1.110 -> 1.111 PPCRegisterInfo.h updated: 1.28 -> 1.29 llvm-svn: 34652 | ||||
* | implement support for the linux/ppc function call ABI. Patch by | Chris Lattner | 2007-02-25 | 1 | -11/+116 |
| | | | | | | Nicolas Geoffray! llvm-svn: 34574 | ||||
* | Don't spill LR as a callee saved register. | Jim Laskey | 2007-02-23 | 1 | -2/+11 |
| | | | | llvm-svn: 34533 | ||||
* | By default, spills kills the register being stored. | Evan Cheng | 2007-02-23 | 1 | -15/+15 |
| | | | | llvm-svn: 34515 | ||||
* | Support to provide exception and selector registers. | Jim Laskey | 2007-02-21 | 1 | -1/+8 |
| | | | | llvm-svn: 34482 | ||||
* | Re-apply my liveintervalanalysis changes. Now with PR1207 fixes. | Evan Cheng | 2007-02-19 | 1 | -9/+30 |
| | | | | llvm-svn: 34428 | ||||
* | For PR1207: | Reid Spencer | 2007-02-19 | 1 | -30/+9 |
| | | | | | | | Revert patches that caused the problem. Evan, please investigate and reapply when you've discovered the problem. llvm-svn: 34399 | ||||
* | Added getReservedRegs(). | Evan Cheng | 2007-02-17 | 1 | -9/+30 |
| | | | | llvm-svn: 34376 | ||||
* | Support for non-landing pad exception handling. | Jim Laskey | 2007-02-01 | 1 | -0/+4 |
| | | | | llvm-svn: 33755 | ||||
* | Fix CodeGen/PowerPC/2007-01-31-InlineAsmAddrMode.ll | Chris Lattner | 2007-02-01 | 1 | -0/+3 |
| | | | | llvm-svn: 33732 | ||||
* | Only gather frame info if debug or eh. | Jim Laskey | 2007-01-29 | 1 | -2/+2 |
| | | | | llvm-svn: 33639 | ||||
* | Landing pad-less eh for PPC. | Jim Laskey | 2007-01-29 | 1 | -4/+3 |
| | | | | llvm-svn: 33622 | ||||
* | Change the MachineDebugInfo to MachineModuleInfo to better reflect usage | Jim Laskey | 2007-01-26 | 1 | -8/+8 |
| | | | | | | for debugging and exception handling. llvm-svn: 33550 | ||||
* | Make LABEL a builtin opcode. | Jim Laskey | 2007-01-26 | 1 | -2/+2 |
| | | | | llvm-svn: 33537 | ||||
* | Fix comment. | Evan Cheng | 2007-01-25 | 1 | -2/+1 |
| | | | | llvm-svn: 33508 | ||||
* | - Tell PEI that PPC will handle stack frame rounding itself. | Evan Cheng | 2007-01-25 | 1 | -7/+2 |
| | | | | | | | - Do not round up to max. alignment of stack object if it is > stack alignment. It will have to be handled with dynamic aligning code. llvm-svn: 33505 | ||||
* | Repair debug frames as a prelude to eh_frames. Switched to using MachineMoves | Jim Laskey | 2007-01-24 | 1 | -15/+43 |
| | | | | | | by value so that clean up is less confusing (these vectors tend to be small.) llvm-svn: 33488 | ||||
* | hasFP() is now a virtual method of MRegisterInfo. | Evan Cheng | 2007-01-23 | 1 | -1/+1 |
| | | | | llvm-svn: 33455 | ||||
* | Fix naming inconsistency. | Evan Cheng | 2007-01-02 | 1 | -10/+10 |
| | | | | llvm-svn: 32823 | ||||
* | What should be the last unnecessary <iostream>s in the library. | Bill Wendling | 2006-12-07 | 1 | -3/+2 |
| | | | | llvm-svn: 32333 | ||||
* | fix CodeGen/PowerPC/2006-12-07-LargeAlloca.ll on ppc64 | Chris Lattner | 2006-12-07 | 1 | -0/+1 |
| | | | | llvm-svn: 32331 | ||||
* | Make it easier for gdb to find the return address. | Jim Laskey | 2006-12-06 | 1 | -45/+89 |
| | | | | llvm-svn: 32277 | ||||
* | MachineInstr::setOpcode -> MachineInstr::setInstrDescriptor | Evan Cheng | 2006-11-30 | 1 | -1/+1 |
| | | | | llvm-svn: 32034 | ||||
* | Change MachineInstr ctor's to take a TargetInstrDescriptor reference instead | Evan Cheng | 2006-11-27 | 1 | -84/+95 |
| | | | | | | of opcode and number of operands. llvm-svn: 31947 | ||||
* | r13 is the thread pointer on darwin/ppc64, don't allocate it. | Chris Lattner | 2006-11-20 | 1 | -2/+2 |
| | | | | llvm-svn: 31882 | ||||
* | Assert unhandled case. | Jim Laskey | 2006-11-17 | 1 | -0/+2 |
| | | | | llvm-svn: 31828 | ||||
* | 1. Ignore the -disable-fp-elim when the routine is a leaf. | Jim Laskey | 2006-11-17 | 1 | -2/+2 |
| | | | | | | 2. Offsets on 64-bit stores are still in bytes. llvm-svn: 31824 | ||||
* | This is a general clean up of the PowerPC ABI. Address several problems and | Jim Laskey | 2006-11-16 | 1 | -132/+244 |
| | | | | | | | | | bugs including making sure that the TOS links back to the previous frame, that the maximum call frame size is not included twice when using frame pointers, no longer growing the frame on calls, double storing of SP and a cleaner/faster dynamic alloca. llvm-svn: 31792 | ||||
* | Fix ppc64 epilog bug. | Chris Lattner | 2006-11-15 | 1 | -1/+1 |
| | | | | llvm-svn: 31771 | ||||
* | Properly transfer kill / dead info. | Evan Cheng | 2006-11-15 | 1 | -13/+17 |
| | | | | llvm-svn: 31765 | ||||
* | Fix the PPC regressions last night | Chris Lattner | 2006-11-15 | 1 | -6/+6 |
| | | | | llvm-svn: 31752 | ||||
* | Rework PPC64 calls. Now we have a LR8/CTR8 register which the PPC64 calls | Chris Lattner | 2006-11-14 | 1 | -31/+51 |
| | | | | | | | | clobber. This allows LR8 to be save/restored correctly as a 64-bit quantity, instead of handling it as a 32-bit quantity. This unbreaks ppc64 codegen when the code is actually located above the 4G boundary. llvm-svn: 31734 | ||||
* | Matches MachineInstr changes. | Evan Cheng | 2006-11-13 | 1 | -10/+11 |
| | | | | llvm-svn: 31712 | ||||
* | Make sure stack link is set in 64-bit. | Jim Laskey | 2006-11-11 | 1 | -3/+9 |
| | | | | llvm-svn: 31690 | ||||
* | implement proper PPC64 prolog/epilog codegen. | Chris Lattner | 2006-11-11 | 1 | -30/+76 |
| | | | | llvm-svn: 31684 | ||||
* | Running with frame pointers prevented debugging, external probes and | Jim Laskey | 2006-11-11 | 1 | -2/+5 |
| | | | | | | | | potentially some system calls/exception handling from working. TOS must always link to previous frame. This is a short term workaround until alloca scheme is reworked. llvm-svn: 31677 | ||||
* | For PR786: | Reid Spencer | 2006-11-02 | 1 | -1/+0 |
| | | | | | | | | | | Turn on -Wunused and -Wno-unused-parameter. Clean up most of the resulting fall out by removing unused variables. Remaining warnings have to do with unused functions (I didn't want to delete code without review) and unused variables in generated code. Maintainers should clean up the remaining issues when they see them. All changes pass DejaGnu tests and Olden. llvm-svn: 31380 | ||||
* | Completely eliminate def&use operands. Now a register operand is EITHER a | Chris Lattner | 2006-09-05 | 1 | -3/+3 |
| | | | | | | def operand or a use operand. llvm-svn: 30109 | ||||
* | Handle callee saved registers in dwarf frame info (lead up to exception | Jim Laskey | 2006-08-29 | 1 | -6/+13 |
| | | | | | | handling.) llvm-svn: 29954 | ||||
* | Fix some comments. | Jim Laskey | 2006-08-25 | 1 | -1/+4 |
| | | | | llvm-svn: 29880 | ||||
* | The PPC64 JIT needs register numbers to encode instructions. | Chris Lattner | 2006-07-11 | 1 | -35/+36 |
| | | | | llvm-svn: 29114 | ||||
* | In 64-bit mode, 64-bit GPRs are callee saved, not 32-bit ones. | Chris Lattner | 2006-07-11 | 1 | -59/+94 |
| | | | | llvm-svn: 29096 | ||||
* | Fix rewriting frame offsets with ixaddr instructions, which implicitly shift | Chris Lattner | 2006-06-27 | 1 | -8/+21 |
| | | | | | | the offset two bits to the left. llvm-svn: 28933 | ||||
* | Rename OR4 -> OR. Move some PPC64-specific stuff to the 64-bit file | Chris Lattner | 2006-06-20 | 1 | -3/+3 |
| | | | | llvm-svn: 28889 | ||||
* | Gaar! Don't use r11 for CR save/restore, use R0. R11 can be register | Chris Lattner | 2006-06-12 | 1 | -9/+11 |
| | | | | | | | | | | | | | | | | | | | allocated, thus live across the save/reload. This fixes llc-beta /MultiSource/Applications/spiff/spiff llc-beta /MultiSource/Benchmarks/sim/sim: llc-beta /MultiSource/Benchmarks/Ptrdist/bc/bc llc-beta /MultiSource/Benchmarks/McCat/12-IOtest/iotest: llc-beta /MultiSource/Benchmarks/FreeBench/fourinarow/fourinarow llc-beta /MultiSource/Benchmarks/Fhourstones-3.1/fhourstones3.1 llc-beta /MultiSource/Benchmarks/mediabench/adpcm/rawdaudio/rawdaudio llc-beta /MultiSource/Benchmarks/mediabench/adpcm/rawcaudio/rawcaudio llc-beta /MultiSource/Benchmarks/mediabench/g721/g721encode/encode llc-beta /MultiSource/Benchmarks/mediabench/jpeg/jpeg-6a/cjpeg and probably others, with -regalloc=local. llvm-svn: 28761 |