| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | Add support for llvm.setjmp and longjmp. Only 3 SingleSource/UnitTests fail ↵ | Chris Lattner | 2005-01-08 | 1 | -1/+8 |
| | | | | | | | now. llvm-svn: 19404 | ||||
| * | Tighten up assertions. | Chris Lattner | 2005-01-08 | 1 | -8/+6 |
| | | | | | llvm-svn: 19397 | ||||
| * | Silence VS warnings | Chris Lattner | 2005-01-08 | 1 | -7/+8 |
| | | | | | llvm-svn: 19388 | ||||
| * | Silence warnings from VS | Chris Lattner | 2005-01-08 | 2 | -3/+5 |
| | | | | | llvm-svn: 19386 | ||||
| * | Silence VS warnings | Chris Lattner | 2005-01-08 | 2 | -2/+2 |
| | | | | | llvm-svn: 19385 | ||||
| * | Silence VS warnings. | Chris Lattner | 2005-01-08 | 2 | -5/+4 |
| | | | | | llvm-svn: 19384 | ||||
| * | Implement handling of most long operators through libcalls. | Chris Lattner | 2005-01-08 | 1 | -5/+65 |
| | | | | | | | Fix a bug legalizing "ret (Val,Val)" llvm-svn: 19375 | ||||
| * | Adjust to changes in LowerCAllTo interfaces | Chris Lattner | 2005-01-08 | 1 | -14/+27 |
| | | | | | llvm-svn: 19374 | ||||
| * | Add support for FP->INT conversions and back. | Chris Lattner | 2005-01-08 | 3 | -17/+42 |
| | | | | | llvm-svn: 19369 | ||||
| * | Implement the 'store FPIMM, Ptr' -> 'store INTIMM, Ptr' optimization for | Chris Lattner | 2005-01-08 | 1 | -0/+25 |
| | | | | | | | all targets. llvm-svn: 19366 | ||||
| * | 1ULL << 64 is undefined, don't do it. | Chris Lattner | 2005-01-08 | 1 | -2/+3 |
| | | | | | llvm-svn: 19365 | ||||
| * | Fix a pointer invalidation problem. This fixes Generic/badarg6.ll | Chris Lattner | 2005-01-07 | 1 | -13/+7 |
| | | | | | llvm-svn: 19361 | ||||
| * | Fold conditional branches on constants away. | Chris Lattner | 2005-01-07 | 1 | -0/+6 |
| | | | | | llvm-svn: 19360 | ||||
| * | Fix a thinko in the reassociation code, fixing Generic/badlive.ll | Chris Lattner | 2005-01-07 | 1 | -1/+1 |
| | | | | | llvm-svn: 19359 | ||||
| * | Add support for truncating integer casts from long. | Chris Lattner | 2005-01-07 | 1 | -1/+13 |
| | | | | | llvm-svn: 19358 | ||||
| * | Fix a bug in load expansion legalization and ret legalization. This fixes | Chris Lattner | 2005-01-07 | 1 | -9/+17 |
| | | | | | | | CodeGen/Generic/select.ll:castconst. llvm-svn: 19357 | ||||
| * | Legalize unconditional branches too | Chris Lattner | 2005-01-07 | 1 | -0/+6 |
| | | | | | llvm-svn: 19356 | ||||
| * | Implement support for long GEP indices on 32-bit archs and support for | Chris Lattner | 2005-01-07 | 2 | -3/+15 |
| | | | | | | | int GEP indices on 64-bit archs. llvm-svn: 19354 | ||||
| * | Simplify: truncate ({zero|sign}_extend (X)) | Chris Lattner | 2005-01-07 | 1 | -0/+9 |
| | | | | | llvm-svn: 19353 | ||||
| * | implement legalization of a bunch more operators. | Chris Lattner | 2005-01-07 | 1 | -0/+6 |
| | | | | | llvm-svn: 19352 | ||||
| * | Fix another bug legalizing calls! | Chris Lattner | 2005-01-07 | 1 | -1/+1 |
| | | | | | llvm-svn: 19350 | ||||
| * | Fix handling of dead PHI nodes. | Chris Lattner | 2005-01-07 | 1 | -26/+30 |
| | | | | | llvm-svn: 19349 | ||||
| * | Fix a bug legalizing calls | Chris Lattner | 2005-01-07 | 1 | -1/+1 |
| | | | | | llvm-svn: 19348 | ||||
| * | After legalizing a DAG, delete dead nodes to save space. | Chris Lattner | 2005-01-07 | 1 | -6/+1 |
| | | | | | llvm-svn: 19346 | ||||
| * | Implement RemoveDeadNodes | Chris Lattner | 2005-01-07 | 1 | -0/+119 |
| | | | | | llvm-svn: 19345 | ||||
| * | Teach legalize how to handle condbranches | Chris Lattner | 2005-01-07 | 1 | -0/+10 |
| | | | | | llvm-svn: 19339 | ||||
| * | Initial implementation of the SelectionDAGISel class. This contains most | Chris Lattner | 2005-01-07 | 1 | -0/+853 |
| | | | | | | | of the code for lowering from LLVM code to a SelectionDAG. llvm-svn: 19331 | ||||
| * | This file is obsolete | Chris Lattner | 2005-01-07 | 1 | -278/+0 |
| | | | | | llvm-svn: 19330 | ||||
| * | Initial implementation of the DAG legalization. This still has a long way | Chris Lattner | 2005-01-07 | 1 | -0/+703 |
| | | | | | | | to go, but it does work for some non-trivial cases now. llvm-svn: 19329 | ||||
| * | Complete rewrite of the SelectionDAG class. | Chris Lattner | 2005-01-07 | 1 | -106/+758 |
| | | | | | llvm-svn: 19327 | ||||
| * | Make the 2-address instruction lowering pass smarter in two ways: | Chris Lattner | 2005-01-02 | 1 | -17/+61 |
| | | | | | | | | | | | | | | | | | | 1. If we are two-addressing a commutable instruction and the LHS is not the last use of the variable, see if the instruction is the last use of the RHS. If so, commute the instruction, allowing us to avoid a register-register copy in many cases for common instructions like ADD, OR, AND, etc on X86. 2. If #1 doesn't hold, and if this is an instruction that also existing in 3-address form, promote the instruction to a 3-address instruction to avoid the register-register copy. We can do this for several common instructions in X86, including ADDrr, INC, DEC, etc. This patch implements test/Regression/CodeGen/X86/commute-two-addr.ll, overlap-add.ll, and overlap-shift.ll when I check in the X86 support for it. llvm-svn: 19245 | ||||
| * | Move virtual method call out of loop | Chris Lattner | 2004-12-15 | 2 | -6/+6 |
| | | | | | llvm-svn: 18955 | ||||
| * | Move virtual function call out of loop to speed up getFreePhysReg by about | Nate Begeman | 2004-12-07 | 1 | -2/+2 |
| | | | | | | | 20%, shaving 0.1s off hbd compile time on my g5. Yay. llvm-svn: 18592 | ||||
| * | For PR387:\ | Reid Spencer | 2004-12-07 | 2 | -2/+2 |
| | | | | | | | | Make only one print method to avoid overloaded virtual warnings when \ compiled with -Woverloaded-virtual llvm-svn: 18589 | ||||
| * | Prevent accessing past the end of the intervals vector, this fixes | Chris Lattner | 2004-12-04 | 1 | -1/+3 |
| | | | | | | | Prolang-C/bison in the JIT llvm-svn: 18477 | ||||
| * | Fix SingleSource/UnitTests/2004-11-28-GlobalBoolLayout.c, and hopefully | Chris Lattner | 2004-11-28 | 1 | -1/+1 |
| | | | | | | | PR449 llvm-svn: 18306 | ||||
| * | Fix the FIXME, nuke the JIT specific forceCompilationOf method. | Chris Lattner | 2004-11-22 | 1 | -18/+0 |
| | | | | | llvm-svn: 18131 | ||||
| * | These methods are obsolete | Chris Lattner | 2004-11-22 | 1 | -8/+0 |
| | | | | | llvm-svn: 18129 | ||||
| * | Adjust to changed interfaces | Chris Lattner | 2004-11-20 | 1 | -7/+7 |
| | | | | | llvm-svn: 18064 | ||||
| * | Add getCurrentPCOffset() and addRelocation() methods. | Chris Lattner | 2004-11-20 | 1 | -0/+12 |
| | | | | | llvm-svn: 18034 | ||||
| * | Match change in MachineCodeEmitter prototype. | Chris Lattner | 2004-11-19 | 1 | -2/+2 |
| | | | | | llvm-svn: 18009 | ||||
| * | * There is no reason for SpillWeights to be an instance var | Chris Lattner | 2004-11-18 | 1 | -69/+71 |
| | | | | | | | | | * Do not put fixed registers into the unhandled set. This means they will never find their way into the inactive, active, or handled sets, so we can simplify a bunch of code. llvm-svn: 17945 | ||||
| * | There is no need to check to see if j overflowed in this loop as we're only | Chris Lattner | 2004-11-18 | 1 | -1/+3 |
| | | | | | | | incrementing i. llvm-svn: 17944 | ||||
| * | Moderate head scratching reveals that this conditional is not needed. If | Chris Lattner | 2004-11-18 | 1 | -4/+0 |
| | | | | | | | i->start == j->start, then certainly i->end > j->start. llvm-svn: 17943 | ||||
| * | Fix a couple of bugs where we considered physregs past their range as possibly | Chris Lattner | 2004-11-18 | 1 | -6/+11 |
| | | | | | | | intersecting an interval. llvm-svn: 17939 | ||||
| * | Fix typeo | Chris Lattner | 2004-11-18 | 1 | -1/+1 |
| | | | | | llvm-svn: 17938 | ||||
| * | Start using the iterators in the fixed_ intervals to avoid having to binary | Chris Lattner | 2004-11-18 | 1 | -5/+12 |
| | | | | | | | | search physreg intervals every time we access it. This takes another half second off of linscan. llvm-svn: 17937 | ||||
| * | Take another .7 seconds off of linear scan time. | Chris Lattner | 2004-11-18 | 1 | -3/+5 |
| | | | | | llvm-svn: 17936 | ||||
| * | Add a counter for the number of times linscan has to backtrack. Start using | Chris Lattner | 2004-11-18 | 1 | -22/+36 |
| | | | | | | | | the iterator hints we have to speed up overlaps(). This speeds linscan up by about .2s (out of 8.7) on 175.vpr for PPC. llvm-svn: 17935 | ||||
| * | Add ability to give hints to the overlaps routines. | Chris Lattner | 2004-11-18 | 2 | -7/+27 |
| | | | | | llvm-svn: 17934 | ||||

