| Commit message (Collapse) | Author | Age | Files | Lines | ||
|---|---|---|---|---|---|---|
| ... | ||||||
| * | Reapply 79977. | Devang Patel | 2009-08-28 | 2 | -4/+4 | |
| | | | | | | | Use MDNodes to encode debug info in llvm IR. llvm-svn: 80406 | |||||
| * | unbreak the build, yay for symlinks + makefiles. :( | Chris Lattner | 2009-08-27 | 1 | -1/+1 | |
| | | | | | llvm-svn: 80205 | |||||
| * | enhance InlineFunction to be able to optionally return | Chris Lattner | 2009-08-27 | 1 | -6/+15 | |
| | | | | | | | a the list of static allocas that it inlined. llvm-svn: 80203 | |||||
| * | smallvectorize the list of returns built by CloneAndPruneFunctionInto. | Chris Lattner | 2009-08-27 | 3 | -8/+8 | |
| | | | | | llvm-svn: 80202 | |||||
| * | remove CloneTrace, which appears to be dead since 2004. | Chris Lattner | 2009-08-27 | 2 | -120/+0 | |
| | | | | | llvm-svn: 80201 | |||||
| * | reduce inlining factor some stuff out to a static helper function, | Chris Lattner | 2009-08-27 | 1 | -126/+155 | |
| | | | | | | | and other code cleanups. No functionality change. llvm-svn: 80199 | |||||
| * | Revert 79977. It causes llvm-gcc bootstrap failures on some platforms. | Devang Patel | 2009-08-26 | 2 | -4/+4 | |
| | | | | | llvm-svn: 80073 | |||||
| * | Eliminate the unused Context argument on one of the ICmpInst and FCmpInst | Dan Gohman | 2009-08-25 | 1 | -1/+1 | |
| | | | | | | | constructors. llvm-svn: 80049 | |||||
| * | Rename Instruction::isIdenticalTo to Instruction::isIdenticalToWhenDefined, | Dan Gohman | 2009-08-25 | 2 | -4/+10 | |
| | | | | | | | | | | | | | | | | | and introduce a new Instruction::isIdenticalTo which tests for full identity, including the SubclassOptionalData flags. Also, fix the Instruction::clone implementations to preserve the SubclassOptionalData flags. Finally, teach several optimizations how to handle SubclassOptionalData correctly, given these changes. This fixes the counterintuitive behavior of isIdenticalTo not comparing the full value, and clone not returning an identical clone, as well as some subtle bugs that could be caused by these. Thanks to Nick Lewycky for reporting this, and for an initial patch! llvm-svn: 80038 | |||||
| * | Comment-ify. | Owen Anderson | 2009-08-25 | 1 | -0/+3 | |
| | | | | | llvm-svn: 80009 | |||||
| * | Switch to SmallVector. | Owen Anderson | 2009-08-25 | 1 | -2/+2 | |
| | | | | | llvm-svn: 80007 | |||||
| * | Pull out this predicate loop into a helper function. | Owen Anderson | 2009-08-25 | 1 | -11/+15 | |
| | | | | | llvm-svn: 80006 | |||||
| * | Update DebugInfo interface to use metadata, instead of special named ↵ | Devang Patel | 2009-08-25 | 2 | -4/+4 | |
| | | | | | | | | | llvm.dbg.... global variables, to encode debugging information in llvm IR. This is mostly a mechanical change that tests metadata support very well. This change speeds up llvm-gcc by more then 6% at "-O0 -g" (measured by compiling InstructionCombining.cpp!) llvm-svn: 79977 | |||||
| * | Handle a corner case when extracing code regions where one of the immediate ↵ | Owen Anderson | 2009-08-25 | 1 | -2/+20 | |
| | | | | | | | | | | | successor of an extracted block contains a PHI using a value defined in the extracted region. With this patch, the partial inliner now passes MultiSource/Applications. llvm-svn: 79963 | |||||
| * | When extracting SEME regions of code, the extractor needs to update the ↵ | Owen Anderson | 2009-08-24 | 1 | -2/+18 | |
| | | | | | | | dominator tree for split return blocks. llvm-svn: 79957 | |||||
| * | Fix -Asserts warnings. | Daniel Dunbar | 2009-08-23 | 1 | -0/+3 | |
| | | | | | llvm-svn: 79849 | |||||
| * | Change Pass::print to take a raw ostream instead of std::ostream, | Chris Lattner | 2009-08-23 | 1 | -1/+1 | |
| | | | | | | | update all code that this affects. llvm-svn: 79830 | |||||
| * | eliminate the "Value" printing methods that print to a std::ostream. | Chris Lattner | 2009-08-23 | 4 | -45/+46 | |
| | | | | | | | This required converting a bunch of stuff off DOUT and other cleanups. llvm-svn: 79819 | |||||
| * | Fix up PHI nodes correctly in the presence of unreachable BBs, part two. Also | Nick Lewycky | 2009-08-19 | 1 | -7/+30 | |
| | | | | | | | delete a newed pointer, and improve readability a little bit. llvm-svn: 79411 | |||||
| * | Be more careful when modifying PHI nodes. Patch by Andre Tavares. | Nick Lewycky | 2009-08-19 | 1 | -1/+2 | |
| | | | | | llvm-svn: 79407 | |||||
| * | Update comments to new-style syntax. | Dan Gohman | 2009-08-17 | 1 | -2/+2 | |
| | | | | | llvm-svn: 79263 | |||||
| * | Don't crash on critical edge. Patch by Andre Tavares. | Nick Lewycky | 2009-08-17 | 1 | -1/+1 | |
| | | | | | llvm-svn: 79252 | |||||
| * | Fix for PR3016: detect the tricky case, where there are | Eli Friedman | 2009-08-16 | 1 | -59/+37 | |
| | | | | | | | | | | | | | | | | | unfoldable references to a PHI node in the block being folded, and disable the transformation in that case. The correct transformation of such PHI nodes depends on whether BB dominates Succ, and dominance is expensive to compute here. (Alternatively, it's possible to check whether any uses are live, but that's also essentially a dominance calculation. Another alternative is to use reg2mem, but it probably isn't a good idea to use that in simplifycfg.) Also, remove some incorrect code from CanPropagatePredecessorsForPHIs which is made unnecessary with this patch: it didn't consider the case where a PHI node in BB has multiple uses. llvm-svn: 79174 | |||||
| * | SSI construction should just go ahead and ignore instructions in unreachable | Nick Lewycky | 2009-08-15 | 1 | -2/+6 | |
| | | | | | | | blocks. llvm-svn: 79132 | |||||
| * | Actually privatize a IntegerTypes, and fix a few bugs exposed by this. | Owen Anderson | 2009-08-13 | 1 | -1/+1 | |
| | | | | | llvm-svn: 78955 | |||||
| * | Push LLVMContexts through the IntegerType APIs. | Owen Anderson | 2009-08-13 | 13 | -100/+131 | |
| | | | | | llvm-svn: 78948 | |||||
| * | Remove a bunch more now-unnecessary Context arguments. | Dan Gohman | 2009-08-12 | 2 | -8/+5 | |
| | | | | | llvm-svn: 78809 | |||||
| * | Remove unnecessary casts. | Dan Gohman | 2009-08-11 | 1 | -1/+1 | |
| | | | | | llvm-svn: 78664 | |||||
| * | Fix a bunch of namespace pollution. | Dan Gohman | 2009-08-07 | 1 | -0/+3 | |
| | | | | | llvm-svn: 78363 | |||||
| * | Privatize the StructType table, which unfortunately involves routing ↵ | Owen Anderson | 2009-08-05 | 2 | -3/+3 | |
| | | | | | | | contexts through a number of APIs. llvm-svn: 78258 | |||||
| * | Move a few more APIs back to 2.5 forms. The only remaining ones left to ↵ | Owen Anderson | 2009-07-31 | 4 | -32/+17 | |
| | | | | | | | | | change back are metadata related, which I'm waiting on to avoid conflicting with Devang. llvm-svn: 77721 | |||||
| * | Move more code back to 2.5 APIs. | Owen Anderson | 2009-07-30 | 8 | -21/+16 | |
| | | | | | llvm-svn: 77635 | |||||
| * | Twines: Don't allow implicit conversion from integers, this is too tricky. | Daniel Dunbar | 2009-07-30 | 1 | -1/+1 | |
| | | | | | llvm-svn: 77605 | |||||
| * | Switch obvious clients to Twine instead of utostr (when they were already using | Daniel Dunbar | 2009-07-30 | 1 | -3/+2 | |
| | | | | | | | | | | a Twine, e.g., for names). - I am a little ambivalent about this; we don't want the string conversion of utostr, but using overload '+' mixed with string and integer arguments is sketchy. On the other hand, this particular usage is something of an idiom. llvm-svn: 77579 | |||||
| * | Move types back to the 2.5 API. | Owen Anderson | 2009-07-29 | 5 | -17/+15 | |
| | | | | | llvm-svn: 77516 | |||||
| * | Remove now unused Context variables. | Benjamin Kramer | 2009-07-29 | 1 | -1/+0 | |
| | | | | | llvm-svn: 77495 | |||||
| * | Move ConstantExpr to 2.5 API. | Owen Anderson | 2009-07-29 | 4 | -8/+8 | |
| | | | | | llvm-svn: 77494 | |||||
| * | Rename MDNode.h header. It defines MDnode and other metadata classes. | Devang Patel | 2009-07-28 | 1 | -1/+1 | |
| | | | | | | | New name is Metadata.h. llvm-svn: 77370 | |||||
| * | Return ConstantVector to 2.5 API. | Owen Anderson | 2009-07-28 | 1 | -1/+1 | |
| | | | | | llvm-svn: 77366 | |||||
| * | Change ConstantArray to 2.5 API. | Owen Anderson | 2009-07-28 | 2 | -3/+3 | |
| | | | | | llvm-svn: 77347 | |||||
| * | Fix a small little typo. | Mike Stump | 2009-07-28 | 1 | -1/+1 | |
| | | | | | llvm-svn: 77289 | |||||
| * | Fix a release-asserts warning. Debug functions should be marked used, | Mike Stump | 2009-07-27 | 1 | -0/+2 | |
| | | | | | | | | if there are no other uses. If people don't need this routine anymore, if should be deleted. llvm-svn: 77274 | |||||
| * | Avoid build warnings. | Mike Stump | 2009-07-27 | 1 | -0/+1 | |
| | | | | | llvm-svn: 77271 | |||||
| * | Move ConstantStruct back to 2.5 API. | Owen Anderson | 2009-07-27 | 1 | -1/+1 | |
| | | | | | llvm-svn: 77266 | |||||
| * | Do not seed mstadata into the value map. | Devang Patel | 2009-07-27 | 1 | -25/+4 | |
| | | | | | llvm-svn: 77208 | |||||
| * | Remove Value::getName{Start,End}, the last of the old Name APIs. | Daniel Dunbar | 2009-07-26 | 1 | -12/+12 | |
| | | | | | llvm-svn: 77152 | |||||
| * | Remove Value::{isName, getNameRef}. | Daniel Dunbar | 2009-07-25 | 1 | -1/+1 | |
| | | | | | | | Also, change MDString to use a StringRef. llvm-svn: 77098 | |||||
| * | Convert a few more uses of llvm/Support/Streams.h to raw_ostream. | Dan Gohman | 2009-07-25 | 2 | -18/+11 | |
| | | | | | llvm-svn: 77033 | |||||
| * | Make AliasAnalysis and related classes use | Dan Gohman | 2009-07-25 | 1 | -1/+1 | |
| | | | | | | | getAnalysisIfAvailable<TargetData>(). llvm-svn: 77028 | |||||
| * | More migration to raw_ostream, the water has dried up around the iostream hole. | Daniel Dunbar | 2009-07-25 | 2 | -6/+8 | |
| | | | | | | | | | | | - Some clients which used DOUT have moved to DEBUG. We are deprecating the "magic" DOUT behavior which avoided calling printing functions when the statement was disabled. In addition to being unnecessary magic, it had the downside of leaving code in -Asserts builds, and of hiding potentially unnecessary computations. llvm-svn: 77019 | |||||

