| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | Fix the MASM asmprinter's lies. It does not want to emit code to .text/.data | Chris Lattner | 2006-05-09 | 4 | -7/+17 |
| | | | | | | | it wants it emitted to _text/_data. llvm-svn: 28185 | ||||
| * | Split SwitchSection into SwitchTo{Text|Data}Section methods. | Chris Lattner | 2006-05-09 | 7 | -47/+50 |
| | | | | | llvm-svn: 28184 | ||||
| * | Some notes and thoughts to myself | Chris Lattner | 2006-05-09 | 1 | -7/+7 |
| | | | | | llvm-svn: 28182 | ||||
| * | Another bad case I noticed | Chris Lattner | 2006-05-08 | 1 | -0/+37 |
| | | | | | llvm-svn: 28177 | ||||
| * | add a note | Chris Lattner | 2006-05-08 | 1 | -0/+9 |
| | | | | | llvm-svn: 28176 | ||||
| * | Yet more readme updating | Nate Begeman | 2006-05-08 | 1 | -36/+22 |
| | | | | | llvm-svn: 28172 | ||||
| * | New note about something bad happening in target independent optimizers | Nate Begeman | 2006-05-08 | 1 | -2/+18 |
| | | | | | llvm-svn: 28170 | ||||
| * | Proving once again that I am not as smart as the compiler | Nate Begeman | 2006-05-08 | 1 | -2/+1 |
| | | | | | llvm-svn: 28169 | ||||
| * | Fold more shifts into inserts, and update the README | Nate Begeman | 2006-05-08 | 2 | -20/+34 |
| | | | | | llvm-svn: 28168 | ||||
| * | When tracking demanded bits, if any bits from the sext of an SRA are demanded, | Chris Lattner | 2006-05-08 | 1 | -2/+8 |
| | | | | | | | then so is the input sign bit. This fixes mediabench/g721 on X86. llvm-svn: 28166 | ||||
| * | Fixing truncate. Previously we were emitting truncate from r16 to r8 as | Evan Cheng | 2006-05-08 | 7 | -25/+100 |
| | | | | | | | | | | | | | | | | | | | | | | | | | movw. That is we promote the destination operand to r16. So %CH = TRUNC_R16_R8 %BP is emitted as movw %bp, %cx. This is incorrect. If %cl is live, it would be clobbered. Ideally we want to do the opposite, that is emitted it as movb ??, %ch But this is not possible since %bp does not have a r8 sub-register. We are now defining a new register class R16_ which is a subclass of R16 containing only those 16-bit registers that have r8 sub-registers (i.e. AX - DX). We isel the truncate to two instructions, a MOV16to16_ to copy the value to the R16_ class, followed by a TRUNC_R16_R8. Due to bug 770, the register colaescer is not going to coalesce between R16 and R16_. That will be fixed later so we can eliminate the MOV16to16_. Right now, it can only be eliminated if we are lucky that source and destination registers are the same. llvm-svn: 28164 | ||||
| * | Update some stuff now that the new rlwimi code has gone in | Nate Begeman | 2006-05-08 | 2 | -70/+0 |
| | | | | | llvm-svn: 28162 | ||||
| * | Typo's | Evan Cheng | 2006-05-07 | 1 | -2/+2 |
| | | | | | llvm-svn: 28158 | ||||
| * | New rlwimi implementation, which is superior to the old one. There are | Nate Begeman | 2006-05-07 | 1 | -86/+45 |
| | | | | | | | | | still a couple missed optimizations, but we now generate all the possible rlwimis for multiple inserts into the same bitfield. More regression tests to come. llvm-svn: 28156 | ||||
| * | Use ComputeMaskedBits to determine # sign bits as a fallback. This allows us | Chris Lattner | 2006-05-06 | 1 | -2/+23 |
| | | | | | | | | to handle all kinds of stuff, including silly things like: sextinreg(setcc,i16) -> setcc. llvm-svn: 28155 | ||||
| * | Add some more sign propagation cases | Chris Lattner | 2006-05-06 | 1 | -10/+77 |
| | | | | | llvm-svn: 28154 | ||||
| * | Add some more simple sign bit propagation cases. | Chris Lattner | 2006-05-06 | 1 | -27/+67 |
| | | | | | llvm-svn: 28149 | ||||
| * | Fix some loose ends in MASM support. | Jeff Cohen | 2006-05-06 | 2 | -60/+72 |
| | | | | | llvm-svn: 28148 | ||||
| * | Add some really really simple code for computing sign-bit propagation. | Chris Lattner | 2006-05-06 | 1 | -0/+95 |
| | | | | | | | This will certainly be enhanced in the future. llvm-svn: 28145 | ||||
| * | Teach the X86 backend about non-i32 inline asm register classes. | Chris Lattner | 2006-05-06 | 1 | -5/+25 |
| | | | | | llvm-svn: 28139 | ||||
| * | Fold (trunc (srl x, c)) -> (srl (trunc x), c) | Chris Lattner | 2006-05-06 | 1 | -0/+32 |
| | | | | | llvm-svn: 28138 | ||||
| * | Implement ComputeMaskedBits/SimplifyDemandedBits for ISD::TRUNCATE | Chris Lattner | 2006-05-05 | 1 | -0/+18 |
| | | | | | llvm-svn: 28135 | ||||
| * | Print a grouping around inline asm blocks so that we can tell when we are | Chris Lattner | 2006-05-05 | 1 | -1/+2 |
| | | | | | | | using them. llvm-svn: 28134 | ||||
| * | Print *some* grouping around inline asm blocks so we know where they are. | Chris Lattner | 2006-05-05 | 1 | -1/+2 |
| | | | | | llvm-svn: 28133 | ||||
| * | Teach the code generator to use cvtss2sd as extload f32 -> f64 | Chris Lattner | 2006-05-05 | 2 | -5/+1 |
| | | | | | llvm-svn: 28131 | ||||
| * | Need extload patterns after Chris' DAG combiner changes | Evan Cheng | 2006-05-05 | 1 | -1/+11 |
| | | | | | llvm-svn: 28127 | ||||
| * | Better implementation of truncate. ISel matches it to a pseudo instruction | Evan Cheng | 2006-05-05 | 6 | -240/+162 |
| | | | | | | | | | that gets emitted as movl (for r32 to i16, i8) or a movw (for r16 to i8). And if the destination gets allocated a subregister of the source operand, then the instruction will not be emitted at all. llvm-svn: 28119 | ||||
| * | New note, Nate, please check to see if I'm full of it :) | Chris Lattner | 2006-05-05 | 1 | -0/+33 |
| | | | | | llvm-svn: 28118 | ||||
| * | Remove and simplify some more machineinstr/machineoperand stuff. | Chris Lattner | 2006-05-04 | 2 | -15/+15 |
| | | | | | llvm-svn: 28105 | ||||
| * | Rename MO_VirtualRegister -> MO_Register. Clean up immediate handling. | Chris Lattner | 2006-05-04 | 7 | -13/+13 |
| | | | | | llvm-svn: 28104 | ||||
| * | Move some methods out of MachineInstr into MachineOperand | Chris Lattner | 2006-05-04 | 8 | -25/+23 |
| | | | | | llvm-svn: 28102 | ||||
| * | There shalt be only one "immediate" operand type! | Chris Lattner | 2006-05-04 | 15 | -53/+45 |
| | | | | | llvm-svn: 28099 | ||||
| * | Revert Nate's CR patch from last night, which caused many regressions (e.g. ↵ | Chris Lattner | 2006-05-04 | 2 | -26/+9 |
| | | | | | | | | | | fhourstones). Loading and storing off R0 isn't what we wanted. Also, taking some CR's out of CRRC seems to cause failures as well. Further investigation is required. llvm-svn: 28097 | ||||
| * | Make external globals public; other minor cleanup. | Jeff Cohen | 2006-05-04 | 1 | -15/+17 |
| | | | | | llvm-svn: 28096 | ||||
| * | Make Intel syntax the default when LLVM is built with VC++. | Jeff Cohen | 2006-05-04 | 1 | -1/+6 |
| | | | | | llvm-svn: 28095 | ||||
| * | Remove a bunch more dead V9 specific stuff | Chris Lattner | 2006-05-04 | 1 | -6/+1 |
| | | | | | llvm-svn: 28094 | ||||
| * | Remove a bunch more SparcV9 specific stuff | Chris Lattner | 2006-05-04 | 10 | -42/+10 |
| | | | | | llvm-svn: 28093 | ||||
| * | Remove some more V9-specific stuff. | Chris Lattner | 2006-05-04 | 2 | -7/+1 |
| | | | | | llvm-svn: 28092 | ||||
| * | Remove some more unused stuff from MachineInstr that was leftover from V9. | Chris Lattner | 2006-05-04 | 6 | -31/+0 |
| | | | | | llvm-svn: 28091 | ||||
| * | Simplify handling of relocations | Chris Lattner | 2006-05-04 | 1 | -24/+38 |
| | | | | | llvm-svn: 28090 | ||||
| * | Use movsd to shuffle in the lowest two elements of a v4f32 / v4i32 vector when | Evan Cheng | 2006-05-03 | 1 | -0/+8 |
| | | | | | | | movlps cannot be used (e.g. when load from m64 has multiple uses). llvm-svn: 28089 | ||||
| * | Change from using MachineRelocation ctors to using static methods | Chris Lattner | 2006-05-03 | 3 | -8/+8 |
| | | | | | | | in MachineRelocation to create Relocations. llvm-svn: 28088 | ||||
| * | inline a simple method | Chris Lattner | 2006-05-03 | 1 | -10/+7 |
| | | | | | llvm-svn: 28083 | ||||
| * | Suck block address tracking out of targets into the JIT Emitter. This | Chris Lattner | 2006-05-03 | 3 | -47/+19 |
| | | | | | | | | simplifies the MachineCodeEmitter interface just a little bit and makes BasicBlocks work like constant pools and jump tables. llvm-svn: 28082 | ||||
| * | Fix a bug in Owen's checkin that broke the CBE on all non sparc v9 platforms. | Chris Lattner | 2006-05-03 | 1 | -1/+1 |
| | | | | | llvm-svn: 28081 | ||||
| * | Teach the x86 jit how to handle jump tables not directly used by a jump | Nate Begeman | 2006-05-03 | 1 | -0/+3 |
| | | | | | | | instruction. llvm-svn: 28080 | ||||
| * | Refactor TargetMachine, pushing handling of TargetData into the ↵ | Owen Anderson | 2006-05-03 | 19 | -49/+53 |
| | | | | | | | | | target-specific subclasses. This has one caller-visible change: getTargetData() now returns a pointer instead of a reference. This fixes PR 759. llvm-svn: 28074 | ||||
| * | Change the BasicBlockAddrs map to be a vector, indexed by MBB number. | Chris Lattner | 2006-05-03 | 2 | -14/+17 |
| | | | | | llvm-svn: 28069 | ||||
| * | Keep the alpha JIT similar to the PPC/X86 jits | Chris Lattner | 2006-05-03 | 1 | -4/+6 |
| | | | | | llvm-svn: 28068 | ||||
| * | Several related changes: | Chris Lattner | 2006-05-02 | 3 | -5/+0 |
| | | | | | | | | | | | | | | 1. Change several methods in the MachineCodeEmitter class to be pure virtual. 2. Suck emitConstantPool/initJumpTableInfo into startFunction, removing them from the MachineCodeEmitter interface, and reducing the amount of target- specific code. 3. Change the JITEmitter so that it allocates constantpools and jump tables *right* next to the functions that they belong to, instead of in a separate pool of memory. This makes all memory for a function be contiguous, and means the JITEmitter only tracks one block of memory now. llvm-svn: 28065 | ||||

