Commit message (Collapse) | Author | Age | Files | Lines | ||
---|---|---|---|---|---|---|
... | ||||||
* | Expand arith on machines without carry flags | Andrew Lenharth | 2008-10-07 | 1 | -9/+33 | |
| | | | | llvm-svn: 57243 | |||||
* | Note that ADDC and company don't actually expand yet (missing in legalize | Andrew Lenharth | 2008-10-07 | 2 | -2/+45 | |
| | | | | llvm-svn: 57226 | |||||
* | Mark negative-zero-to-int conversion as Inexact, | Dale Johannesen | 2008-10-07 | 1 | -1/+2 | |
| | | | | | | | | | since -0.0 can't be represented as an int. This prevents llvm from reducing -0.0 to a load-and-convert of int 0. Fixes ieee.exp/mzero[2356].c in gcc testsuite. llvm-svn: 57224 | |||||
* | Correctly handle calls with no return values. This fixes | Dan Gohman | 2008-10-07 | 1 | -1/+1 | |
| | | | | | | 2006-01-23-UnionInit on x86-64 when inlining is not enabled. llvm-svn: 57223 | |||||
* | Fix typo, fix PR 2865. | Devang Patel | 2008-10-06 | 1 | -2/+2 | |
| | | | | llvm-svn: 57221 | |||||
* | Be more precise about which conversions of NaNs | Dale Johannesen | 2008-10-06 | 1 | -3/+18 | |
| | | | | | | | | are Inexact. (These are not Inexact as defined by IEEE754, but that seems like a reasonable way to abstract what happens: information is lost.) llvm-svn: 57218 | |||||
* | Fix PR2850 and PR2863. Only generate movddup for 128-bit SSE vector shuffles. | Evan Cheng | 2008-10-06 | 1 | -0/+1 | |
| | | | | llvm-svn: 57210 | |||||
* | Temporarily revert functionality change from my previous patch; | Dale Johannesen | 2008-10-06 | 1 | -2/+1 | |
| | | | | | | it is too aggressive. llvm-svn: 57203 | |||||
* | Remove interfaces implemented by dead pass from the list of available passes. | Devang Patel | 2008-10-06 | 1 | -7/+17 | |
| | | | | | | Patch By Matthijs Kooijman. llvm-svn: 57202 | |||||
* | Mark shortening NaN conversions as Inexact. PR 2856. | Dale Johannesen | 2008-10-06 | 1 | -6/+9 | |
| | | | | | | Improve description of unsupported formats. llvm-svn: 57185 | |||||
* | It is possible that all functions in one module are not being | Devang Patel | 2008-10-06 | 1 | -5/+2 | |
| | | | | | | optimized for size. Set OptForSize for each function separately. llvm-svn: 57182 | |||||
* | Don't dereference the end() iterator. Thanks to | Dan Gohman | 2008-10-06 | 1 | -1/+1 | |
| | | | | | | ENABLE_EXPENSIVE_CHECKS for finding this. llvm-svn: 57181 | |||||
* | Remove unncessary isDeclaration() checks. | Devang Patel | 2008-10-06 | 3 | -3/+3 | |
| | | | | llvm-svn: 57179 | |||||
* | Allow scalarrepl to treat an all-zero GEP just as bitcast. | Matthijs Kooijman | 2008-10-06 | 1 | -14/+28 | |
| | | | | | | | This includes not marking a GEP involving a vector as unsafe, but only when it has all zero indices. This allows scalarrepl to work in a few more cases. llvm-svn: 57177 | |||||
* | reorder #include order, patch by Kenneth Boyd! | Chris Lattner | 2008-10-06 | 1 | -4/+2 | |
| | | | | llvm-svn: 57148 | |||||
* | Add #include to get alloca, patch by Kenneth Boyd! | Chris Lattner | 2008-10-06 | 1 | -0/+1 | |
| | | | | llvm-svn: 57147 | |||||
* | fix an incorrect and extremely confusing error message | Chris Lattner | 2008-10-05 | 1 | -1/+1 | |
| | | | | llvm-svn: 57123 | |||||
* | make the autoupgrade code for ret attributes dramatically simpler | Chris Lattner | 2008-10-05 | 1 | -30/+22 | |
| | | | | | | and actually work. We can now read the llvm 2.3 bc file from PR2849 llvm-svn: 57122 | |||||
* | regenerate | Nuno Lopes | 2008-10-05 | 2 | -209/+233 | |
| | | | | llvm-svn: 57114 | |||||
* | clean ArgTypeListI production: free the PATypeHolder | Nuno Lopes | 2008-10-05 | 1 | -2/+14 | |
| | | | | llvm-svn: 57113 | |||||
* | Emit type-correct constant null. Also fix a typo. | Anton Korobeynikov | 2008-10-05 | 1 | -7/+3 | |
| | | | | | | Patch by Robert G. Jakabosky! llvm-svn: 57110 | |||||
* | Fix weird think-o and unbreak build on all gcc-3.4.x-based platforms (e.g. ↵ | Anton Korobeynikov | 2008-10-05 | 3 | -98/+91 | |
| | | | | | | mingw) llvm-svn: 57106 | |||||
* | this case is matched now. | Chris Lattner | 2008-10-05 | 1 | -9/+1 | |
| | | | | llvm-svn: 57096 | |||||
* | rewrite bswap matching to be more general, allowing arbitrary | Chris Lattner | 2008-10-05 | 1 | -77/+121 | |
| | | | | | | | | shifting and masking inside a bswap expr. This allows it to handle the cases from PR2842, which involve the intermediate 'or' expressions being shifted, not just the input value. llvm-svn: 57095 | |||||
* | fix a bug where the bswap matcher could match a case involving | Chris Lattner | 2008-10-05 | 1 | -1/+4 | |
| | | | | | | ashr. It should only apply to lshr. llvm-svn: 57089 | |||||
* | wrap some long lines and expand i32 mul's to libcalls, inspired by a | Chris Lattner | 2008-10-04 | 1 | -4/+12 | |
| | | | | | | patch by Mikael Lepisto! llvm-svn: 57077 | |||||
* | CMake: Reflected changes on source file structure. New plugin support | Oscar Fuentes | 2008-10-04 | 1 | -0/+2 | |
| | | | | | | for llvmc2 incomplete. llvm-svn: 57076 | |||||
* | Ignore loads from and stores to local memory (i.e. allocas) | Duncan Sands | 2008-10-04 | 1 | -9/+26 | |
| | | | | | | | | when deciding whether to mark a function readnone/readonly. Since the pass is currently run before SROA, this may be quite helpful. Requested by Chris on IRC. llvm-svn: 57050 | |||||
* | Allow the construction of SCEVs with SCEVCouldNotCompute operands, by | Nick Lewycky | 2008-10-04 | 1 | -0/+67 | |
| | | | | | | implementing folding. Fixes PR2857. llvm-svn: 57049 | |||||
* | Revert r56675 - it breaks unwinding runtime everywhere. | Anton Korobeynikov | 2008-10-04 | 1 | -18/+8 | |
| | | | | llvm-svn: 57048 | |||||
* | Add implementations for sys::Memory::setWritable and ↵ | Argyrios Kyrtzidis | 2008-10-04 | 1 | -0/+8 | |
| | | | | | | sys::Memory::setExecutable on Win32 platform. llvm-svn: 57047 | |||||
* | Fix compilation error on MSVC. | Argyrios Kyrtzidis | 2008-10-04 | 1 | -1/+1 | |
| | | | | llvm-svn: 57046 | |||||
* | Fix fast-isel's handling of atomic instructions. They may | Dan Gohman | 2008-10-04 | 1 | -1/+4 | |
| | | | | | | | | expand to multiple basic blocks, in which case fast-isel needs to informed of which block to use as it resumes inserting instructions. llvm-svn: 57040 | |||||
* | Fix a bug in the local allocator's liveness computation where it | Dan Gohman | 2008-10-04 | 1 | -3/+5 | |
| | | | | | | | | was setting kill flags on tied uses in two-address instructions. The kill flags were causing the allocator to think it could allocate the use and its tied def in different registers. llvm-svn: 57039 | |||||
* | Make atomic Swap work, 64-bit on x86-32. | Dale Johannesen | 2008-10-03 | 4 | -11/+33 | |
| | | | | | | Make it all work in non-pic mode. llvm-svn: 57034 | |||||
* | Clean up some multiple-return-value code that is no longer | Dan Gohman | 2008-10-03 | 1 | -5/+1 | |
| | | | | | | applicable. llvm-svn: 57033 | |||||
* | Fix function attribute verification check. | Devang Patel | 2008-10-03 | 1 | -6/+5 | |
| | | | | | | Thanks Duncan! llvm-svn: 57029 | |||||
* | Pass MemOperand through for 64-bit atomics on 32-bit, | Dale Johannesen | 2008-10-03 | 4 | -9/+27 | |
| | | | | | | | incidentally making the case where the memop is a pointer deref work. Fix cmp-and-swap regression. llvm-svn: 57027 | |||||
* | Nick Lewycky's patch. | Devang Patel | 2008-10-03 | 1 | -1/+1 | |
| | | | | | | While hosting instruction check PHI node. llvm-svn: 57025 | |||||
* | Use -1ULL instead of uint64_t(-1), at Anton's suggestion. | Dan Gohman | 2008-10-03 | 1 | -2/+2 | |
| | | | | llvm-svn: 57021 | |||||
* | Verify function attributes. | Devang Patel | 2008-10-03 | 1 | -1/+5 | |
| | | | | llvm-svn: 57020 | |||||
* | Fix typos pointed out by Duncan. Also untabify these files. | Evan Cheng | 2008-10-03 | 3 | -15/+15 | |
| | | | | llvm-svn: 57018 | |||||
* | Unbreak build. | Daniel Dunbar | 2008-10-03 | 1 | -18/+18 | |
| | | | | llvm-svn: 57017 | |||||
* | Avoid creating two TargetLowering objects for each target. | Dan Gohman | 2008-10-03 | 11 | -33/+22 | |
| | | | | | | | | | | | | Instead, just create one, and make sure everything that needs it can access it. Previously most of the SelectionDAGISel subclasses all had their own TargetLowering object, which was redundant with the TargetLowering object in the TargetMachine subclasses, except on Sparc, where SparcTargetMachine didn't have a TargetLowering object. Change Sparc to work more like the other targets here. llvm-svn: 57016 | |||||
* | Remove an unused field. | Dan Gohman | 2008-10-03 | 1 | -6/+1 | |
| | | | | llvm-svn: 57014 | |||||
* | On Darwin ARM, memory needs special handling to do JIT. This patch expands | Jim Grosbach | 2008-10-03 | 4 | -13/+51 | |
| | | | | | | | this handling to work properly for modifying stub functions, relocations back to entry points after JIT compilation, etc.. llvm-svn: 57013 | |||||
* | Indexing off by one resulted in errant encoding of source register for | Jim Grosbach | 2008-10-03 | 1 | -1/+1 | |
| | | | | | | reg->reg moves. llvm-svn: 57011 | |||||
* | NeedStub/DoesntNeedStub logic was reversed, leading to not using a stub | Jim Grosbach | 2008-10-03 | 1 | -4/+4 | |
| | | | | | | for global relocations that do need them (libc calls, for example). llvm-svn: 57010 | |||||
* | regenerate | Nuno Lopes | 2008-10-03 | 3 | -338/+346 | |
| | | | | llvm-svn: 57009 | |||||
* | fix more memleaks in ResolveTypeTo() and ParseGlobalVariable() | Nuno Lopes | 2008-10-03 | 1 | -0/+4 | |
| | | | | llvm-svn: 57008 |