summaryrefslogtreecommitdiffstats
path: root/llvm/lib/CodeGen
Commit message (Collapse)AuthorAgeFilesLines
* Preliminary soft float support.Evan Cheng2006-12-093-19/+74
| | | | llvm-svn: 32394
* Move findTiedToSrcOperand to TargetInstrDescriptor.Evan Cheng2006-12-082-4/+3
| | | | llvm-svn: 32366
* Proper fix for PR1037: to determine is a VR is a modref, check 1) whether it isEvan Cheng2006-12-081-2/+4
| | | | | | | tied to another oeprand, 2) whether is is being tied to by another operand. So the destination operand of a two-address MI can be correctly identified. llvm-svn: 32354
* findTiedToSrcOperand() changed.Evan Cheng2006-12-081-2/+2
| | | | llvm-svn: 32353
* Fix CodeGen/PowerPC/2006-12-07-SelectCrash.ll on PPC64Chris Lattner2006-12-071-1/+4
| | | | llvm-svn: 32336
* Removed more <iostream> includesBill Wendling2006-12-078-40/+31
| | | | llvm-svn: 32321
* Removing even more <iostream> includes.Bill Wendling2006-12-0711-255/+242
| | | | llvm-svn: 32320
* Revision 1.83 causes PR1037.Reid Spencer2006-12-071-2/+1
| | | | | | Reverted. llvm-svn: 32305
* Initialize {Min|Max}CSFrameIndex properly.Evan Cheng2006-12-071-2/+4
| | | | llvm-svn: 32299
* Changed llvm_ostream et all to OStream. llvm_cerr, llvm_cout, llvm_null, areBill Wendling2006-12-079-57/+51
| | | | | | now cerr, cout, and NullStream resp. llvm-svn: 32298
* MI keeps a ptr of TargetInstrDescriptor, use it.Evan Cheng2006-12-072-3/+5
| | | | llvm-svn: 32296
* getOperandConstraint returns -1 if the operand does have the specific ↵Evan Cheng2006-12-071-1/+1
| | | | | | constraint. This bug was causing excessive spills. llvm-svn: 32295
* Detemplatize the Statistic class. The only type it is instantiated withChris Lattner2006-12-0610-31/+31
| | | | | | is 'unsigned'. llvm-svn: 32279
* Some addresssed should be 64-bit and some shouldn't.Jim Laskey2006-12-061-19/+25
| | | | llvm-svn: 32278
* Move copyKillDeadInfo out-of-line. Add findRegisterUseOperand().Evan Cheng2006-12-061-0/+31
| | | | llvm-svn: 32273
* eliminate fp statisticChris Lattner2006-12-061-3/+3
| | | | llvm-svn: 32251
* Unbreak VC++ build.Jeff Cohen2006-12-023-0/+3
| | | | llvm-svn: 32113
* Match TargetInstrInfo changes.Evan Cheng2006-12-015-30/+17
| | | | llvm-svn: 32098
* better check for versionJim Laskey2006-11-301-2/+14
| | | | llvm-svn: 32045
* Change global descriptor names to provide name, full name and linkage name.Jim Laskey2006-11-302-36/+35
| | | | llvm-svn: 32036
* Temp fix to deal gdb issue.Jim Laskey2006-11-301-1/+6
| | | | llvm-svn: 32035
* - MachineInstr now keeps a ptr to TargetInstrDescriptor instead of Opcode.Evan Cheng2006-11-301-42/+35
| | | | | | - Remove the ugly TargetInstrDescriptors hack. llvm-svn: 32032
* Avoid inifinite looping if READCYCLECOUNTER isn't custom lowered.Evan Cheng2006-11-291-3/+8
| | | | llvm-svn: 32022
* Eliminate unnecessary scopes.Jim Laskey2006-11-291-21/+27
| | | | llvm-svn: 32020
* Allow target to custom lower READCYCLECOUNTER (when it doesn't have to be ↵Evan Cheng2006-11-291-2/+10
| | | | | | expanded). llvm-svn: 32016
* Fix for PR1023 by Dan Gohman.Evan Cheng2006-11-291-1/+1
| | | | llvm-svn: 32003
* Fix for PR1022 (folding loads of static initializers) by Dan Gohman.Evan Cheng2006-11-291-1/+1
| | | | llvm-svn: 32000
* add a hook to allow targets to hack on inline asms to lower them to llvmChris Lattner2006-11-291-1/+13
| | | | | | when they want to. llvm-svn: 31997
* Converted to using llvm streams instead of <iostream>sBill Wendling2006-11-295-87/+101
| | | | llvm-svn: 31992
* Use llvm streams instead of <iostream>Bill Wendling2006-11-283-20/+18
| | | | llvm-svn: 31985
* Put the #include for a module first.Bill Wendling2006-11-281-1/+1
| | | | llvm-svn: 31958
* Clean up.Evan Cheng2006-11-281-2/+2
| | | | llvm-svn: 31957
* Changed to using llvm streams.Bill Wendling2006-11-282-15/+14
| | | | llvm-svn: 31954
* Fix PR1016Chris Lattner2006-11-281-5/+4
| | | | llvm-svn: 31950
* Change MachineInstr ctor's to take a TargetInstrDescriptor reference insteadEvan Cheng2006-11-274-24/+19
| | | | | | of opcode and number of operands. llvm-svn: 31947
* Fix the dag combiner bug corresponding to PR1014.Chris Lattner2006-11-271-8/+8
| | | | llvm-svn: 31943
* For better or worse, load from i1 is assumed to be zero extended. Do notChris Lattner2006-11-271-1/+6
| | | | | | form a load from i1 from larger loads that may not be zext'd. llvm-svn: 31933
* If a brcond condition is promoted, make sure to zero extend it, even if notChris Lattner2006-11-271-6/+6
| | | | | | expanded into BR_CC. llvm-svn: 31932
* For PR950:Reid Spencer2006-11-274-80/+235
| | | | | | | | | | The long awaited CAST patch. This introduces 12 new instructions into LLVM to replace the cast instruction. Corresponding changes throughout LLVM are provided. This passes llvm-test, llvm/test, and SPEC CPUINT2000 with the exception of 175.vpr which fails only on a slight floating point output difference. llvm-svn: 31931
* Global label not handled correctly.Jim Laskey2006-11-201-3/+11
| | | | llvm-svn: 31883
* Fix PR1011 and CodeGen/Generic/2006-11-20-DAGCombineCrash.llChris Lattner2006-11-201-2/+2
| | | | llvm-svn: 31878
* setOperand should not zap the operand list or add implicit operands to anChris Lattner2006-11-201-39/+0
| | | | | | | | | instruction. Doing so breaks the FP stackifier, the alpha branch selector the sparc fpmover. This fixes PR1012 and CodeGen/X86/fp-stack-compare.ll llvm-svn: 31876
* For PR950:Reid Spencer2006-11-201-0/+56
| | | | | | | | First in a series of patches to convert SetCondInst into ICmpInst and FCmpInst using only two opcodes and having the instructions contain their predicate value. Nothing uses these classes yet. More patches to follow. llvm-svn: 31867
* Don't transform in another bad case: if the block is empty, it should beChris Lattner2006-11-181-1/+2
| | | | | | simplified before we do this xform so that our cost model is accurate. llvm-svn: 31864
* Fix another case we *don't* want to do this xform.Chris Lattner2006-11-181-2/+26
| | | | llvm-svn: 31861
* Implement operator<< for machine basic blocks to make it easier to dump them.Chris Lattner2006-11-181-0/+4
| | | | llvm-svn: 31857
* make the previous change more aggressive, moving any block with no fallthrough.Chris Lattner2006-11-181-9/+11
| | | | | | This speeds up yacr2 by 7% on a core2. llvm-svn: 31856
* Minor code layout tweak: if we have something like this:Chris Lattner2006-11-181-0/+50
| | | | | | | | | | | | | | | if (cond) goto BB2 BB1: ... return; BB2: ... Move BB1 to the end of the function so that the code falls through in the non-return case. This has the effect of moving assert (and other no-return call) bodies and return blocks out of loops. llvm-svn: 31855
* Fixing the ENABLE_OPTIMIZED=1 DISABLE_ASSERTIONS=1 build.Jim Laskey2006-11-171-0/+1
| | | | llvm-svn: 31822
* More removal of std::cerr and DEBUG, replacing with DOUT instead.Bill Wendling2006-11-173-65/+72
| | | | llvm-svn: 31806
OpenPOWER on IntegriCloud