| Commit message (Collapse) | Author | Age | Files | Lines | ||
|---|---|---|---|---|---|---|
| ... | ||||||
| * | Thread an MCAsmInfo pointer through the various MC printing APIs, | Chris Lattner | 2009-09-03 | 6 | -32/+41 | |
| | | | | | | | | and fix a few things using << on MCSymbols to use ->print(). No functionality change other than unbreaking my previous patch. llvm-svn: 80890 | |||||
| * | just use dump() | Chris Lattner | 2009-09-03 | 1 | -1/+1 | |
| | | | | | llvm-svn: 80889 | |||||
| * | inline insertion operators. | Chris Lattner | 2009-09-03 | 1 | -15/+9 | |
| | | | | | llvm-svn: 80888 | |||||
| * | In C++, code is not allowed to call main. In C it is, this | Chris Lattner | 2009-09-03 | 1 | -57/+4 | |
| | | | | | | | | | | | simplifylibcalls optimization is thus valid for C++ but not C. It's not important enough to worry about for C++ apps, so just remove it. rdar://7191924 llvm-svn: 80887 | |||||
| * | merge globaladdress symbol processing stuff into other stuff. Now | Chris Lattner | 2009-09-03 | 2 | -76/+65 | |
| | | | | | | | | all global variable operand flag processing stuff is shared between different operand types. llvm-svn: 80886 | |||||
| * | Unbreak x86_64 build. | Evan Cheng | 2009-09-03 | 1 | -2/+6 | |
| | | | | | llvm-svn: 80885 | |||||
| * | Split the "operand -> symbol" logic from the "get offset and other munging | Chris Lattner | 2009-09-03 | 2 | -36/+34 | |
| | | | | | | | from operand" logic. GlobalAddress still todo. llvm-svn: 80884 | |||||
| * | implement lowering support for constant pool index operands, this gets a ↵ | Chris Lattner | 2009-09-03 | 2 | -13/+52 | |
| | | | | | | | | | bunch more olden programs working. llvm-svn: 80881 | |||||
| * | X86JITInfo::getLazyResolverFunction() should not read cpu id to determine ↵ | Evan Cheng | 2009-09-03 | 4 | -33/+25 | |
| | | | | | | | | | whether sse is available. Just use consult subtarget. No functionality changes. llvm-svn: 80880 | |||||
| * | output alignment value in hex so that we get: | Chris Lattner | 2009-09-03 | 1 | -1/+2 | |
| | | | | | | | | | | | .align 3, 0x90 instead of, .align 3, 144 suggested by eric. llvm-svn: 80875 | |||||
| * | simplify this by using SmallString::str(), much nicer! | Chris Lattner | 2009-09-03 | 1 | -1/+1 | |
| | | | | | llvm-svn: 80874 | |||||
| * | Fixed a test that ensures the LocalRewriter does not attempt to | Lang Hames | 2009-09-03 | 2 | -2/+2 | |
| | | | | | | | | | | avoid reloads by reusing clobbered registers. This was causing issues in 256.bzip2 when compiled with PIC for a while (starting at r78217), though the problem has since been masked. llvm-svn: 80872 | |||||
| * | back out my recent commit (r80858), it seems to break self-hosting ↵ | Gabor Greif | 2009-09-03 | 10 | -28/+20 | |
| | | | | | | | buildbot's stage 2 configure llvm-svn: 80871 | |||||
| * | Now Bitcode reader bug is fixed. Reapply 80839. | Devang Patel | 2009-09-03 | 5 | -35/+113 | |
| | | | | | | | | | Use CallbackVH, instead of WeakVH, to hold MDNode elements. Use FoldingSetNode to unique MDNodes in a context. Use CallbackVH hooks to update context's MDNodeSet appropriately. llvm-svn: 80868 | |||||
| * | Add new value for given index in MDValuePtrs. | Devang Patel | 2009-09-03 | 1 | -0/+3 | |
| | | | | | llvm-svn: 80867 | |||||
| * | Improve llvm::getHostTriple for some cases where the LLVM_HOSTTRIPLE is not | Daniel Dunbar | 2009-09-03 | 1 | -4/+42 | |
| | | | | | | | reliable. llvm-svn: 80863 | |||||
| * | re-commit r66920 (which has been backed out in r66953) I may have more luck ↵ | Gabor Greif | 2009-09-03 | 10 | -20/+28 | |
| | | | | | | | this time. I'll back out if needed... llvm-svn: 80858 | |||||
| * | Added opaque 32-, 48-, and 80-bit memory operand types to the X86 | Sean Callanan | 2009-09-03 | 4 | -0/+31 | |
| | | | | | | | | | | instruction tables to support segmented addressing (and other objects of obscure type). Modified the X86 assembly printers to handle these new operand types. Added JMP and CALL instructions that use segmented addresses. llvm-svn: 80857 | |||||
| * | Show derived host triple in --version. | Daniel Dunbar | 2009-09-02 | 1 | -0/+2 | |
| | | | | | llvm-svn: 80855 | |||||
| * | Revert 80839 for now. It causes test failures. | Devang Patel | 2009-09-02 | 5 | -113/+35 | |
| | | | | | llvm-svn: 80841 | |||||
| * | Use CallbackVH, instead of WeakVH, to hold MDNode elements. | Devang Patel | 2009-09-02 | 5 | -35/+113 | |
| | | | | | | | | Use FoldingSetNode to unique MDNodes in a context. Use CallbackVH hooks to update context's MDNodeSet appropriately. llvm-svn: 80839 | |||||
| * | More missed vdup patterns | Anton Korobeynikov | 2009-09-02 | 1 | -0/+12 | |
| | | | | | llvm-svn: 80838 | |||||
| * | switch from std::string to SmallString + raw_svector_ostream. | Chris Lattner | 2009-09-02 | 1 | -7/+5 | |
| | | | | | llvm-svn: 80807 | |||||
| * | split mcinst lowering stuff out to its own file. | Chris Lattner | 2009-09-02 | 4 | -262/+293 | |
| | | | | | llvm-svn: 80806 | |||||
| * | Refactor common code from ParseAssemblyString and ParseAssemblyFile, | Dan Gohman | 2009-09-02 | 1 | -19/+21 | |
| | | | | | | | to expose a low-level interface for parsing from an existing MemoryBuffer. llvm-svn: 80803 | |||||
| * | Rearrange code to eliminate redundancy and avoid gotos. | Bob Wilson | 2009-09-02 | 1 | -64/+40 | |
| | | | | | llvm-svn: 80798 | |||||
| * | plug another leak in LLParser::PerFunctionState::SetInstName() | Nuno Lopes | 2009-09-02 | 1 | -0/+1 | |
| | | | | | llvm-svn: 80792 | |||||
| * | plug memory leak in LLParser::PerFunctionState::SetInstName() by deleting a ↵ | Nuno Lopes | 2009-09-02 | 1 | -0/+1 | |
| | | | | | | | value after replacing it llvm-svn: 80790 | |||||
| * | Sort edges in MaximumSpanningTree more stable in case of equal weight. | Andreas Neustifter | 2009-09-02 | 1 | -2/+6 | |
| | | | | | | | (See http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090824/085890.html) llvm-svn: 80789 | |||||
| * | Changed set of BlocksToInstrument to set of InsertedBlocks that do not have to | Andreas Neustifter | 2009-09-02 | 1 | -9/+9 | |
| | | | | | | | be instrumented. llvm-svn: 80788 | |||||
| * | Code cleanups and added comments. | Andreas Neustifter | 2009-09-02 | 1 | -17/+45 | |
| | | | | | llvm-svn: 80781 | |||||
| * | Opaque types didn't work if llvm_is_multithreaded(). | Torok Edwin | 2009-09-02 | 1 | -1/+1 | |
| | | | | | | | | AlwaysOpaqueTy is always NULL at this point, and it causes an assertion failure. Fix it by using the just constructed tmp instead. llvm-svn: 80780 | |||||
| * | Fix DbgStopPointInst->getFileName/getDirectory, broken by the MDNodification in | Torok Edwin | 2009-09-02 | 2 | -8/+7 | |
| | | | | | | | r80406, and readd a -print-dbginfo test. llvm-svn: 80778 | |||||
| * | Retype from unsigned to CallingConv::ID accordingly. Approved by Bob Wilson. | Sandeep Patel | 2009-09-02 | 38 | -159/+181 | |
| | | | | | llvm-svn: 80773 | |||||
| * | fix PR4815: some cases where DeleteDeadInstruction can delete | Chris Lattner | 2009-09-02 | 1 | -3/+13 | |
| | | | | | | | the instruction BBI points to. llvm-svn: 80768 | |||||
| * | clean up this code a bit. | Chris Lattner | 2009-09-02 | 1 | -64/+67 | |
| | | | | | llvm-svn: 80767 | |||||
| * | eliminate VISIBILITY_HIDDEN from Transforms/Scalar. PR4861 | Chris Lattner | 2009-09-02 | 30 | -118/+82 | |
| | | | | | llvm-svn: 80766 | |||||
| * | refactor select 'sched insertion' out to its own method. | Chris Lattner | 2009-09-02 | 2 | -60/+73 | |
| | | | | | llvm-svn: 80764 | |||||
| * | Add support for modeling whether or not the processor has support for | Chris Lattner | 2009-09-02 | 3 | -10/+21 | |
| | | | | | | | | conditional moves as a subtarget feature. This is the easy part of PR4841. llvm-svn: 80763 | |||||
| * | fix PR4848 an infinite loop when indexing down through a recursive gep | Chris Lattner | 2009-09-02 | 1 | -1/+1 | |
| | | | | | | | | and we get the original pointer type. This doesn't mean that we're at the first pointer being indexed. Correct the predicate. llvm-svn: 80762 | |||||
| * | fix PR4837, some bugs folding vector compares. These | Chris Lattner | 2009-09-02 | 1 | -7/+7 | |
| | | | | | | | return a vector of i1, not i1 itself. llvm-svn: 80761 | |||||
| * | revert my patch, duncan points out what is wrong with my logic. Add | Chris Lattner | 2009-09-02 | 1 | -4/+8 | |
| | | | | | | | a comment so that I don't change this in the future :) llvm-svn: 80760 | |||||
| * | one more try at making this simpler, hopefully it won't break everything :) | Chris Lattner | 2009-09-02 | 1 | -4/+4 | |
| | | | | | llvm-svn: 80759 | |||||
| * | Complicate Chris's simplification, avoiding complaints | Duncan Sands | 2009-09-02 | 1 | -1/+5 | |
| | | | | | | | | about singular iterators when building with expensive checks turned on. llvm-svn: 80757 | |||||
| * | Fixed the asmstrings for 8-bit, 16-bit, and 32-bit ADD %rAX, imm instructions. | Sean Callanan | 2009-09-02 | 2 | -3/+28 | |
| | | | | | | | | Added a 64-bit ADD %RAX, imm32 instruction. Added all 4 forms for AND %rAX, imm and CMP %rAX, imm. llvm-svn: 80746 | |||||
| * | Fix PR4845: r77946 completely broke x86_64 Darwin (or any situation where the | Evan Cheng | 2009-09-02 | 1 | -10/+32 | |
| | | | | | | | | | desired triplet is a sub-target, e.g. thumbv7 vs. arm host). Reverting the patch isn't quite right either since the previous behavior does not allow the triplet to be overridden with -march. llvm-svn: 80742 | |||||
| * | For now disable MDNode uniquing. This fixes llvm-gcc bootstrap failure on ↵ | Devang Patel | 2009-09-01 | 3 | -19/+8 | |
| | | | | | | | certain Mac OS X 10.5. I am working on a proper fix. llvm-svn: 80738 | |||||
| * | Avoid calling removeVirtualRegisterKilled which iterates over the operands | Bob Wilson | 2009-09-01 | 1 | -1/+2 | |
| | | | | | | | to find the kill, since we already have the operand. llvm-svn: 80736 | |||||
| * | Refactor some code into separate functions. No functional changes. | Bob Wilson | 2009-09-01 | 1 | -52/+82 | |
| | | | | | llvm-svn: 80733 | |||||
| * | Move use of LV inside condition that guards for null LV. | Bob Wilson | 2009-09-01 | 1 | -5/+5 | |
| | | | | | llvm-svn: 80731 | |||||

