summaryrefslogtreecommitdiffstats
path: root/llvm/lib
Commit message (Collapse)AuthorAgeFilesLines
* Add -sse[,2,3] arguments to LLCChris Lattner2004-08-242-0/+20
| | | | llvm-svn: 16018
* Nuke commented out stuffChris Lattner2004-08-241-5/+0
| | | | llvm-svn: 16017
* This code is dodgy, but the guaranteed assertion failure doesn't help anything.Brian Gaeke2004-08-241-1/+1
| | | | llvm-svn: 16014
* Fix bug in PhyRegAlloc::setCallInterferences() handling call through aBrian Gaeke2004-08-241-2/+3
| | | | | | null pointer. llvm-svn: 16013
* Revise head-of-file comment.Brian Gaeke2004-08-241-31/+28
| | | | | | | Eliminate some excess whitespace. Fix bug in CallArgsDescriptor::get() handling call through a null pointer. llvm-svn: 16012
* Do not use .xword and friends to emit zeros on V9. Apparently there are issuesChris Lattner2004-08-241-6/+0
| | | | | | | | with emitting .xwords when not on an 8-byte boundary (.xword 0 is not the same as 8 .byte 0's). Because we do not know when or when we are not aligned, just emit bytes like the old V9 asmprinter did. llvm-svn: 16006
* Kill a majority of unnecessary sign extensions for byte loadsNate Begeman2004-08-221-12/+30
| | | | llvm-svn: 15991
* Don't hard code the offset of the saved R31 in functions with frame pointersNate Begeman2004-08-221-1/+2
| | | | llvm-svn: 15990
* Initial checkin of a pass to lower packed operations to scalars operations.Reid Spencer2004-08-211-0/+359
| | | | | | | | | This also registers the pass with opt with a -lower-packed command line option. Patch contributed by Brad Jones. llvm-svn: 15987
* Bytecode Analyzer Cleanup:Reid Spencer2004-08-211-162/+266
| | | | | | | | | | | - Rearrange output order to make more sense - Make only the function level output count as "detailed" - Output dump output directly to stream, don't buffer it. - Fix counting of block sizes - Implement new handlers for number of types, dependent libs, target triple - Compute the size of block headers. llvm-svn: 15985
* Get rid of an extraneous local variable.Reid Spencer2004-08-211-1/+0
| | | | llvm-svn: 15984
* Two Changes:Reid Spencer2004-08-211-14/+18
| | | | | | | | - Pass the output stream to the analyzer so it can write its output there directly instead of buffering it. - Don't pass a boolean to ParseBytecode because its not needed any more. llvm-svn: 15983
* Add boolean file format flags in preparation for version 5 bytecode.Reid Spencer2004-08-211-3/+18
| | | | | | | | Remove the "processFunctions" boolean from ParseBytecode as it is no longer needed. This is part of avoiding double reading of functions when analyzing bytecode. llvm-svn: 15982
* Bytecode Reader Cleanup:Reid Spencer2004-08-211-38/+66
| | | | | | | | | | | | | | | - provide the correct conversion for ModuleBlockID in read_block (a potential bug but not actually exploited because reading module block ids doesn't use read_block). - install support for handleTypeList handler - install support for handleDependentLibrary handler - install support for handleTargetTriple handler - clean up comments, output strings, - ensure that processing function arguments doesn't SIGSEGV if one of the arguments is a null pointer (yeah, it happened .. weird) - prepare for version 5 bytecode by documenting what will change. llvm-svn: 15981
* Back out branchless SetCC code. While it helped a lot in some cases, itNate Begeman2004-08-211-188/+25
| | | | | | | | hurt a lot in others. Instead, improve branching version of SetCC and Select instructions. The old code will be in CVS should we ever need to dig it up again. llvm-svn: 15979
* Switch from bytes to bits for alignment.Chris Lattner2004-08-212-6/+6
| | | | | | Also, change GPRC for PPC32 to align on 32-bit boundary instead of 64-bit llvm-svn: 15975
* Switch from bytes to bits for alignment for consistencyChris Lattner2004-08-211-10/+5
| | | | llvm-svn: 15974
* Reduce uses of getRegClassChris Lattner2004-08-211-14/+11
| | | | llvm-svn: 15973
* Convert regclass alignment from bytes to bitesChris Lattner2004-08-211-1/+1
| | | | llvm-svn: 15972
* Convert bytes to bits in alignmentChris Lattner2004-08-211-3/+3
| | | | llvm-svn: 15971
* Register info alignment is in bits, frame object alignment is (currently) inChris Lattner2004-08-211-1/+1
| | | | | | bytes. llvm-svn: 15970
* Reduce uses of getRegClassChris Lattner2004-08-212-8/+6
| | | | llvm-svn: 15968
* Now that we have per-register spill size/alignment info, remove more usesChris Lattner2004-08-211-4/+5
| | | | | | of getRegClass llvm-svn: 15967
* Fix warningChris Lattner2004-08-211-1/+1
| | | | llvm-svn: 15964
* Examine the type code in the setcc class of instructions and if itReid Spencer2004-08-211-0/+4
| | | | | | | | is a PackedType, throw an error. Temporary solution. Patch contributed by Brad Jones. llvm-svn: 15963
* Move XForm instructions over to the auto-generated asm writerNate Begeman2004-08-213-71/+133
| | | | llvm-svn: 15962
* Add two valuesChris Lattner2004-08-211-0/+11
| | | | llvm-svn: 15959
* Fix a nasty bug that Nate ran into. In particular, instead of emitting a 2+MBChris Lattner2004-08-211-6/+13
| | | | | | | | | | | | error message, print out: llvm-link: error linking in 'g.2.rbc': Global Variable Collision on ' %struct.D_Reduction* %d_reduction_10_gram' - External linkage globals have different initializers That's a bit more concise, huh? llvm-svn: 15958
* remove some things from the todo list.Nate Begeman2004-08-201-6/+0
| | | | llvm-svn: 15956
* Do not register ppc64 yet, as it breaks the SparcV9 backendChris Lattner2004-08-201-0/+3
| | | | llvm-svn: 15955
* quish a warningChris Lattner2004-08-201-1/+1
| | | | llvm-svn: 15954
* Fix a bug found exposed by: Regression/Other/2004-08-20-PackedControlFlow.llReid Spencer2004-08-201-1/+5
| | | | | | | | Packed types need to be allowed in type statements too. Patch provided by Brad Jones. llvm-svn: 15953
* Implement code to convert SetCC into straight line code where appropriate. ↵Nate Begeman2004-08-202-22/+173
| | | | | | Add necessary instructions for this transformation to the .td file. llvm-svn: 15952
* Packed types, brought to you by Brad JonesBrian Gaeke2004-08-209-6/+336
| | | | llvm-svn: 15938
* If we are linking two global variables and they have the same size, do notChris Lattner2004-08-201-8/+9
| | | | | | spew warnings, even if the types don't match. llvm-svn: 15933
* Fix opcodes being printed in caps (the more general fix may be `AsmWriter')Misha Brukman2004-08-191-1/+1
| | | | llvm-svn: 15932
* Stack space for argument passing is 32 regardless of 32- vs. 64-bit arch.Misha Brukman2004-08-191-6/+6
| | | | | | Thanks to Nate Begeman for pointing this out. llvm-svn: 15930
* LR needs to be saved at 16-byte offset on a 64-bit archMisha Brukman2004-08-192-6/+6
| | | | llvm-svn: 15929
* On 64-bit PowerPC, pointers are 8 bytes, so parameter area offset is 48, not 24Misha Brukman2004-08-191-7/+10
| | | | llvm-svn: 15928
* This PHI has 4 additional operands, not 2.Misha Brukman2004-08-191-1/+1
| | | | llvm-svn: 15926
* Use the appropriate 64-bit register description file.Misha Brukman2004-08-191-1/+1
| | | | llvm-svn: 15922
* Fix more remaining 32-bit vestiges of PowerPCMisha Brukman2004-08-191-6/+2
| | | | llvm-svn: 15919
* Fix another vestige of the 32-bit PowerPC backend.Misha Brukman2004-08-191-3/+1
| | | | llvm-svn: 15918
* Correct character prepended to global symbols ('.'), use Mangler consistentlyMisha Brukman2004-08-191-2/+2
| | | | llvm-svn: 15917
* * Eliminate global base register, r2 is used for that on AIX/PowerPCMisha Brukman2004-08-191-45/+6
| | | | | | * Fix bug from 32-bit PowerPC days of 2-register long split llvm-svn: 15916
* Wrap long lines.Misha Brukman2004-08-191-4/+5
| | | | llvm-svn: 15915
* Convert casts that will have no effect into move instructions.Nate Begeman2004-08-191-4/+22
| | | | llvm-svn: 15914
* Clean up floating point instruction selection.Nate Begeman2004-08-194-110/+30
| | | | | | | | Change int->float cast code to put conversion constants in constant pool. Shorten code sequence for constant pool fp loads. Remove LOADLoDirect/LOADLoIndirect psuedo instructions and tweak asmwriter llvm-svn: 15913
* M_DUMMY_PHI_FLAG is no longer used to distinguish V9::PHI. Get rid of it andBrian Gaeke2004-08-181-2/+1
| | | | | | its TargetInstrInfo accessor. llvm-svn: 15907
* Instead of using isDummyPhiInstr, we just compare the opcode with V9::PHI.Brian Gaeke2004-08-184-9/+8
| | | | llvm-svn: 15906
OpenPOWER on IntegriCloud