summaryrefslogtreecommitdiffstats
path: root/llvm/lib
Commit message (Collapse)AuthorAgeFilesLines
* Back out previous commit, it isn't safe.Nate Begeman2006-02-051-6/+0
| | | | llvm-svn: 26006
* fold c1 << (x + c2) into (c1 << c2) << x. fix a warning.Nate Begeman2006-02-051-1/+7
| | | | llvm-svn: 26005
* some stuff is doneChris Lattner2006-02-051-4/+0
| | | | llvm-svn: 26004
* Turn A % (C << N), where C is 2^k, into A & ((C << N)-1) [urem only].Chris Lattner2006-02-051-0/+35
| | | | | | | | Turn A / (C1 << N), where C1 is "1<<C2" into A >> (N+C2) [udiv only]. Tested with: rem.ll:test5, div.ll:test10 llvm-svn: 26003
* Handle urem by shifted powers of 2.Nate Begeman2006-02-051-4/+15
| | | | llvm-svn: 26001
* handle combining A / (B << N) into A >>u (log2(B)+N) when B is a power of 2Nate Begeman2006-02-051-2/+13
| | | | llvm-svn: 26000
* Use SelectRoot() as the entry to any tblgen based isel.Evan Cheng2006-02-051-1/+2
| | | | llvm-svn: 25998
* Use SelectRoot() as entry of any tblgen based isel.Evan Cheng2006-02-054-4/+8
| | | | llvm-svn: 25997
* Remove the SparcV8 backend. It has been renamed to be the Sparc backend.Chris Lattner2006-02-0519-3687/+0
| | | | llvm-svn: 25992
* remove V8 referenceChris Lattner2006-02-051-1/+1
| | | | llvm-svn: 25991
* * Added SDNode::isOnlyUse().Evan Cheng2006-02-051-3/+18
| | | | | | * Fix hasNUsesOfValue(), it should be const. llvm-svn: 25990
* Rename SPARC V8 target to be the LLVM SPARC target.Chris Lattner2006-02-0520-0/+3682
| | | | llvm-svn: 25985
* add a noteChris Lattner2006-02-051-0/+11
| | | | llvm-svn: 25984
* Re-commit the last bit of change that was backed out.Evan Cheng2006-02-051-13/+2
| | | | llvm-svn: 25983
* make sure that global doubles are aligned to 8 bytesChris Lattner2006-02-051-5/+10
| | | | llvm-svn: 25981
* Use getPreferredAlignmentLog.Chris Lattner2006-02-051-1/+1
| | | | llvm-svn: 25980
* Use the asmprinter to find out what the preferred alignment of a global is.Chris Lattner2006-02-051-1/+1
| | | | | | | This patch speeds up 172.mgrid from 31.81s to 11.39s on darwin/ppc. Many many thanks to Nate for tracking down the root cause of the issue. llvm-svn: 25979
* Implement the AsmPrinter::getPreferredAlignmentLog method.Chris Lattner2006-02-051-0/+16
| | | | llvm-svn: 25978
* linkage fix for weak functionsAndrew Lenharth2006-02-041-3/+16
| | | | llvm-svn: 25976
* Fix VC++ warning.Jeff Cohen2006-02-041-1/+0
| | | | llvm-svn: 25975
* Use SCEVExpander::InsertCastOfTo instead of our own code. This reducesChris Lattner2006-02-041-18/+1
| | | | | | #LLVM LOC, and auto-cse's cast instructions. llvm-svn: 25974
* Pull the InsertCastOfTo out of the header, implement CSE'ing of arguments.Chris Lattner2006-02-041-0/+51
| | | | llvm-svn: 25973
* Temporarily revert this patch, which probably breaks with theChris Lattner2006-02-041-7/+13
| | | | | | tblgen patch reverted. llvm-svn: 25971
* Value# select instructions, allowing -gcse to remove duplicatesChris Lattner2006-02-041-1/+36
| | | | llvm-svn: 25969
* Complex pattern's custom matcher should not call Select() on any operands.Evan Cheng2006-02-041-13/+7
| | | | | | Select them afterwards if it returns true. llvm-svn: 25968
* Custom lower VAARG for the case when we are doing vaarg(double). In thisChris Lattner2006-02-041-3/+35
| | | | | | | case, the double being loaded may not be 8-byte aligned, so we have to use our standard bit_convert game. llvm-svn: 25967
* Fix a nasty typo that broke functions with big stack frames.Chris Lattner2006-02-041-1/+1
| | | | llvm-svn: 25966
* fix a bug in my last checkinChris Lattner2006-02-041-2/+2
| | | | llvm-svn: 25965
* Fix two significant bugs in LSR:Chris Lattner2006-02-041-14/+75
| | | | | | | | | | 1. When rewriting code in outer loops, sometimes we would insert code into inner loops that is invariant in that loop. 2. Notice that 4*(2+x) is 8+4*x and use that to simplify expressions. This is a performance neutral change. llvm-svn: 25964
* Remove some stuff that now worksNate Begeman2006-02-041-53/+0
| | | | llvm-svn: 25963
* add a noteChris Lattner2006-02-041-0/+1
| | | | llvm-svn: 25962
* Two changes:Chris Lattner2006-02-045-17/+24
| | | | | | | | | | 1. Treat FMOVD as a copy instruction, to help with coallescing in V9 mode 2. When in V9 mode, insert FMOVD instead of FpMOVD instructions, as we don't ever rewrite FpMOVD instructions into FMOVS instructions, thus we just end up with commented out copies! This should fix a bunch of failures in V9 mode on sparc. llvm-svn: 25961
* Get rid of some memory leaks identified by ValgrindEvan Cheng2006-02-042-2/+9
| | | | llvm-svn: 25960
* Let bugpoint work on sparc with v9 instructions enabled.Chris Lattner2006-02-041-0/+6
| | | | llvm-svn: 25958
* Fix VC++ warning.Jeff Cohen2006-02-041-1/+1
| | | | llvm-svn: 25957
* Add initial support for immediates. This allows us to compile this:Chris Lattner2006-02-042-6/+16
| | | | | | | | | | | | | | | | | | int %rlwnm(int %A, int %B) { %C = call int asm "rlwnm $0, $1, $2, $3, $4", "=r,r,r,n,n"(int %A, int %B, int 4, int 17) ret int %C } into: _rlwnm: or r2, r3, r3 or r3, r4, r4 rlwnm r2, r2, r3, 4, 17 ;; note the immediates :) or r3, r2, r2 blr llvm-svn: 25955
* Remove an unnecessary predicate.Evan Cheng2006-02-041-2/+1
| | | | llvm-svn: 25954
* Separate FILD and FILD_FLAG, the later is only used for SSE2. It produces aEvan Cheng2006-02-043-8/+15
| | | | | | flag so it can be flagged to a FST. llvm-svn: 25953
* Initial early support for non-register operands, like immediatesChris Lattner2006-02-041-15/+42
| | | | llvm-svn: 25952
* implementation of some methods for inlineasmChris Lattner2006-02-041-1/+41
| | | | llvm-svn: 25951
* Handle another case exposed on X86.Chris Lattner2006-02-031-0/+4
| | | | llvm-svn: 25949
* Fix a nasty problem on two-address machines in the following situation:Chris Lattner2006-02-031-9/+27
| | | | | | | | | | | | | | | | | | | | | | | | store EAX -> [ss#0] [ss#0] += 1 ... use(EAX) In this case, it is not valid to rewrite this as: store EAX -> [ss#0] EAX += 1 store EAX -> [ss#0] ;;; this would also delete the store above ... use(EAX) ... because EAX is not a dead at that point. Keep track of which registers we are allowed to clobber, and which ones we aren't, and don't clobber the ones we're not supposed to. :) This should resolve the issues on X86 last night. llvm-svn: 25948
* significantly simplify the VirtRegMap code by pulling the SpillSlotsAvailableChris Lattner2006-02-031-70/+98
| | | | | | | | | and PhysRegsAvailable maps out into a new AvailableSpills struct. No functionality change. This paves the way for a bugfix, coming up next. llvm-svn: 25947
* Implement some feedback from sabreNate Begeman2006-02-031-5/+5
| | | | llvm-svn: 25946
* Add a framework for eliminating instructions that produces undemanded bits.Nate Begeman2006-02-032-11/+87
| | | | llvm-svn: 25945
* add a noteChris Lattner2006-02-031-0/+2
| | | | llvm-svn: 25944
* another case Nate came up withChris Lattner2006-02-031-0/+3
| | | | llvm-svn: 25943
* add a noteChris Lattner2006-02-031-0/+4
| | | | llvm-svn: 25942
* remove some #ifdef'd out code, which should properly be in the dag combiner ↵Chris Lattner2006-02-031-30/+0
| | | | | | anyway. llvm-svn: 25941
* remove an old commentChris Lattner2006-02-031-3/+2
| | | | llvm-svn: 25940
OpenPOWER on IntegriCloud