summaryrefslogtreecommitdiffstats
path: root/llvm/lib
Commit message (Collapse)AuthorAgeFilesLines
* Implemented the dialect decision logic for the X86Sean Callanan2010-01-251-2/+20
| | | | | | | | TargetAsmLexer. Dialect-specific lexing code will be placed in the functions LexTokenATT() and LexTokenIntel(). llvm-svn: 94456
* Fixed the order of GPR RegisterClass regs to be: ..., R10, R11, R12, ...Johnny Chen2010-01-251-1/+1
| | | | llvm-svn: 94455
* Fix the bitcode reader to deserialize nuw/nsw/etc. bits properly in the caseDan Gohman2010-01-251-3/+3
| | | | | | of a forward-reference, which doesn't use an "abbrev" encoding. llvm-svn: 94454
* mcstreamerize gprel32 emission.Chris Lattner2010-01-254-14/+17
| | | | llvm-svn: 94452
* Remove check for an impossible condition: the condition of the while loop hasBob Wilson2010-01-251-2/+0
| | | | | | already checked that TmpBB->getSinglePredecessor() is non-null. llvm-svn: 94451
* mcize the non-gprel cases of AsmPrinter::printPICJumpTableEntryChris Lattner2010-01-251-13/+13
| | | | llvm-svn: 94450
* handle the _set_ symbol with an MCSymbol.Chris Lattner2010-01-252-7/+13
| | | | llvm-svn: 94449
* rename MAI::PICJumpTableDirective to MAI::GPRel32Directive toChris Lattner2010-01-254-12/+10
| | | | | | make it clear what it is, instead of how it is used. llvm-svn: 94448
* pull the non-pic jump table case out of printPICJumpTableEntryChris Lattner2010-01-252-11/+15
| | | | | | | and MCize the non-pic case. Now printPICJumpTableEntry really is just about printing PIC entries. llvm-svn: 94446
* remove JumpTableDirective, it is always null.Chris Lattner2010-01-253-3/+3
| | | | llvm-svn: 94445
* mcize jump table symbol manipulation.Chris Lattner2010-01-251-33/+38
| | | | llvm-svn: 94441
* sink an arm specific method out of asmprinter into the ARMAsmPrinter andChris Lattner2010-01-253-18/+17
| | | | | | rename it to avoid shadowing. llvm-svn: 94440
* remove dead code: the x86 target never sets usesGlobalOffsetTable, Chris Lattner2010-01-251-2/+0
| | | | | | even on x86-32/elf which uses a GOT. llvm-svn: 94439
* wirte up .file and .file to the mc asmparser.Chris Lattner2010-01-251-3/+7
| | | | llvm-svn: 94438
* mcstreamerize .file and .file. This also fixes an issue where theChris Lattner2010-01-256-79/+78
| | | | | | | normal form of .file would fail if the filename had a weird character in it. llvm-svn: 94437
* emit ELF .type directives through MCStreamer instead of doing it textually.Chris Lattner2010-01-253-9/+4
| | | | llvm-svn: 94436
* add symbol attribute support for the ELF .type directive.Chris Lattner2010-01-252-1/+29
| | | | llvm-svn: 94435
* Change Value::getUnderlyingObject to have the MaxLookup value specified as aBob Wilson2010-01-252-7/+6
| | | | | | | | | | | | parameter with a default value, instead of just hardcoding it in the implementation. The limit of MaxLookup = 6 was introduced in r69151 to fix a performance problem with O(n^2) behavior in instcombine, but the scalarrepl pass is relying on getUnderlyingObject to go all the way back to an AllocaInst. Making the limit part of the method signature makes it clear that by default the result is limited and should help avoid similar problems in the future. This fixes pr6126. llvm-svn: 94433
* Revert r94260 until findDbgDeclare() is made more efficientVictor Hernandez2010-01-251-38/+3
| | | | llvm-svn: 94432
* add a commentChris Lattner2010-01-251-0/+1
| | | | llvm-svn: 94417
* emit the .size directive for global variables on ELF throughChris Lattner2010-01-254-2/+11
| | | | | | mcstreamer. llvm-svn: 94416
* all supported target now have aligned common support.Chris Lattner2010-01-252-2/+1
| | | | llvm-svn: 94413
* Remove support for i386 tiger tools for aligned common symbols.Chris Lattner2010-01-251-3/+0
| | | | | | | | | Previously we would just silently miscompile code that used aligned common's, now at least you'll get a build error. tiger-ppc already triggered the build error because it didn't have a version of this logic. llvm-svn: 94412
* don't set value to its default.Chris Lattner2010-01-252-2/+0
| | | | llvm-svn: 94411
* coff targets support alignment on .commChris Lattner2010-01-251-1/+0
| | | | llvm-svn: 94410
* ARM does accept the .comm directive alignment.Jim Grosbach2010-01-251-1/+0
| | | | llvm-svn: 94408
* Fix PR6134.Rafael Espindola2010-01-251-3/+1
| | | | | | | We are not emitting alignments on Darwin for "bar". Not sure what is the correct way to do it. llvm-svn: 94400
* linux/ppc does use alignment in bytes, not pow-2. This fixes PR6129.Chris Lattner2010-01-241-1/+0
| | | | | | | | It looks like linux/arm and linux/mips have the same setting, which are probably wrong. Someone who cares about ARM and MIPS should investigate with the testcase in PR6129. llvm-svn: 94381
* don't reset the default.Chris Lattner2010-01-241-2/+0
| | | | llvm-svn: 94380
* make -fno-rtti the default unless a directory builds with REQUIRES_RTTI.Chris Lattner2010-01-2468-73/+0
| | | | llvm-svn: 94378
* mark some libraries that currently require RTTI.Chris Lattner2010-01-243-2/+3
| | | | llvm-svn: 94377
* reassociate should do this.Chris Lattner2010-01-241-1/+1
| | | | llvm-svn: 94374
* add a note.Chris Lattner2010-01-241-0/+39
| | | | llvm-svn: 94373
* fix some issues where we weren't emitting enough newlines.Chris Lattner2010-01-242-0/+3
| | | | llvm-svn: 94370
* when emitting DIEs, emit the comment on the same line as the directive.Chris Lattner2010-01-241-2/+3
| | | | | | This fixes FrontendObjC/2009-11-30-Objc-ID.m llvm-svn: 94369
* move PR5945 here.Chris Lattner2010-01-241-0/+29
| | | | llvm-svn: 94350
* fix a parsing problem on instructions like:Chris Lattner2010-01-242-6/+5
| | | | | | | | movw $8, (_cost_table_-L97$pb)+66(%eax) After the parens, we could still have a binop. llvm-svn: 94345
* It seems better to scalarize vectors of size 1 instead of widening them.Mon P Wang2010-01-243-1/+11
| | | | | | Add support to widen SETCC. llvm-svn: 94342
* change the canonical form of "cond ? -1 : 0" to be Chris Lattner2010-01-244-68/+49
| | | | | | | | "sext cond" instead of a select. This simplifies some instcombine code, matches the policy for zext (cond ? 1 : 0 -> zext), and allows us to generate better code for a testcase on ppc. llvm-svn: 94339
* Improved widening loads by adding support for wider loads ifMon P Wang2010-01-246-275/+439
| | | | | | | the alignment allows. Fixed a bug where we didn't use a vector load/store for PR5626. llvm-svn: 94338
* fix a potential overflow issue Eli pointed out.Chris Lattner2010-01-231-3/+3
| | | | llvm-svn: 94336
* Speculatively revert r94322 to see if it fixes darwin selfhost buildbot.Nick Lewycky2010-01-231-70/+23
| | | | llvm-svn: 94331
* third bug from PR6119: the xor dupe extension allowsChris Lattner2010-01-231-2/+2
| | | | | | | | for arbitrary terminators in predecessors, don't assume it is a conditional or uncond branch. The testcase shows an example where they can happen with switches. llvm-svn: 94323
* Teach DAE that even though it can't modify the function signature of anNick Lewycky2010-01-231-23/+70
| | | | | | | externally visible function, it can still find all callers of it and replace the parameters to a dead argument with undef. llvm-svn: 94322
* add an early out to ProcessBranchOnXOR to speed it up,Chris Lattner2010-01-231-0/+25
| | | | | | | | handle the case when we can infer an input to the xor from all inputs that agree, instead of going into an infinite loop. Another part of PR6199 llvm-svn: 94321
* fix a crash in jump threading, PR6119Chris Lattner2010-01-231-0/+6
| | | | llvm-svn: 94319
* implement a simple instcombine xform that has been in theChris Lattner2010-01-232-23/+29
| | | | | | readme forever. llvm-svn: 94318
* add a noteChris Lattner2010-01-231-0/+29
| | | | llvm-svn: 94317
* add some notes, making posix-memalign be nocapture would be an easy improvement.Chris Lattner2010-01-231-0/+31
| | | | llvm-svn: 94312
* Remove the '-disable-scheduling' flag and replace it with the 'source' option ofBill Wendling2010-01-235-696/+444
| | | | | | | | | the '-pre-RA-sched' flag. It actually makes more sense to do it this way. Also, keep track of the SDNode ordering by default. Eventually, we would like to make this ordering a way to break a "tie" in the scheduler. However, doing that now breaks the "CodeGen/X86/abi-isel.ll" test for 32-bit Linux. llvm-svn: 94308
OpenPOWER on IntegriCloud