Commit message (Collapse) | Author | Age | Files | Lines | ||
---|---|---|---|---|---|---|
... | ||||||
* | make irbuilder use the new optimized debug info accessors. | Chris Lattner | 2010-03-31 | 1 | -13/+0 | |
| | | | | llvm-svn: 99984 | |||||
* | add new apis for getting/setting !dbg metadata on | Chris Lattner | 2010-03-31 | 1 | -0/+4 | |
| | | | | | | | | | instructions. In addition to being a convenience, they are faster than the old apis, particularly when not going from an MDKindID like people should be doing. llvm-svn: 99982 | |||||
* | Fix a major source of compile-time slowness at -O0 -g by optimizing | Chris Lattner | 2010-03-30 | 3 | -14/+43 | |
| | | | | | | | | | | | | | | the storage of !dbg metadata kinds in the instruction themselves. The on-the-side hash table works great for metadata that not-all instructions get, or for metadata that only exists when optimizing. But when compile-time is everything, it isn't great. I'm not super thrilled with the fact that this plops a TrackingVH in Instruction, because it grows it by 3 words. I'm investigating alternatives, but this should be a step in the right direction in any case. llvm-svn: 99957 | |||||
* | Revert Mon Ping's change 99928, since it broke all the llvm-gcc buildbots. | Bob Wilson | 2010-03-30 | 1 | -70/+0 | |
| | | | | llvm-svn: 99948 | |||||
* | Added support for address spaces and added a isVolatile field to memcpy, ↵ | Mon P Wang | 2010-03-30 | 1 | -0/+70 | |
| | | | | | | | | | memmove, and memset, e.g., llvm.memcpy.i32(i8*, i8*, i32, i32) -> llvm.memcpy.p0i8.p0i8.i32(i8*, i8*, i32, i32, i1) A update of langref will occur in a subsequent checkin. llvm-svn: 99928 | |||||
* | move some method definitions to files that make sense. | Chris Lattner | 2010-03-30 | 3 | -55/+54 | |
| | | | | llvm-svn: 99927 | |||||
* | Remove the pmulld intrinsic and autoupdate it as a vector multiply. | Eric Christopher | 2010-03-30 | 1 | -0/+17 | |
| | | | | | | | Rewrite the pmulld patterns, and make sure that they fold in loads of arguments into the instruction. llvm-svn: 99910 | |||||
* | reapply my timer rewrite with a change for PassManager to store | Chris Lattner | 2010-03-30 | 1 | -33/+21 | |
| | | | | | | timers by pointer instead of by-value. llvm-svn: 99871 | |||||
* | revert r99862 which is causing FNT failures. | Chris Lattner | 2010-03-30 | 1 | -17/+31 | |
| | | | | llvm-svn: 99870 | |||||
* | fairly major rewrite of various timing related stuff. | Chris Lattner | 2010-03-30 | 1 | -31/+17 | |
| | | | | llvm-svn: 99862 | |||||
* | add support for zero initialized unions, patch by Tim Northover! | Chris Lattner | 2010-03-29 | 1 | -1/+3 | |
| | | | | llvm-svn: 99818 | |||||
* | Oops, r98447 dropped the reference to ForwardType in the wrong place in | Jeffrey Yasskin | 2010-03-28 | 1 | -5/+5 | |
| | | | | | | | | Type::destroy(), so it got skipped for FunctionTypes, StructTypes, and UnionTypes. This fixes the resulting leaks in test/Feature/opaquetypes.ll and test/Integer/opaquetypes_bt.ll. llvm-svn: 99732 | |||||
* | rename pred_const_iterator to const_pred_iterator for consistency's sake | Gabor Greif | 2010-03-25 | 1 | -1/+1 | |
| | | | | llvm-svn: 99567 | |||||
* | rename use_const_iterator to const_use_iterator for consistency's sake | Gabor Greif | 2010-03-25 | 5 | -8/+8 | |
| | | | | llvm-svn: 99564 | |||||
* | Include isFunctionLocal while calculating folding node set profile for a MDNode. | Devang Patel | 2010-03-25 | 1 | -13/+14 | |
| | | | | llvm-svn: 99490 | |||||
* | Include isFunctionLocal while calculating folding node set provide for a MDNode. | Devang Patel | 2010-03-25 | 1 | -0/+1 | |
| | | | | llvm-svn: 99484 | |||||
* | Reapply r99451 with a fix to move the NoInline check to the cost functions | Eric Christopher | 2010-03-25 | 1 | -2/+13 | |
| | | | | | | instead of InlineFunction. llvm-svn: 99483 | |||||
* | Temporarily revert this, it's causing an issue with an internal project. | Eric Christopher | 2010-03-24 | 1 | -13/+2 | |
| | | | | llvm-svn: 99451 | |||||
* | Finally land the InvokeInst operand reordering. | Gabor Greif | 2010-03-24 | 4 | -20/+28 | |
| | | | | | | | | | | | | I have audited all getOperandNo calls now, fixing hidden assumptions. CallSite related uglyness will be eliminated successively. Note this patch has a long and griveous history, for all the back-and-forths have a look at CallSite.h's log. llvm-svn: 99399 | |||||
* | add some accessors to callsite/callinst/invokeinst to check | Chris Lattner | 2010-03-23 | 1 | -2/+13 | |
| | | | | | | | | for the noinline attribute, and make the inliner refuse to inline a call site when the call site is marked noinline even if the callee isn't. This fixes PR6682. llvm-svn: 99341 | |||||
* | fix an infinite loop in Module::getEndianness, PR6684 | Chris Lattner | 2010-03-23 | 1 | -1/+1 | |
| | | | | | | patch by Alex Mac! llvm-svn: 99330 | |||||
* | add assert in argpromotion, which cannot trigger | Gabor Greif | 2010-03-23 | 1 | -2/+2 | |
| | | | | | | | | if Function::hasAddressTaken works as advertised also included some cosmetic cleanups llvm-svn: 99276 | |||||
* | word-o | Gabor Greif | 2010-03-23 | 1 | -1/+1 | |
| | | | | llvm-svn: 99275 | |||||
* | Change intrinsic result type for void to store it as an empty list | Chris Lattner | 2010-03-22 | 1 | -34/+41 | |
| | | | | | | | instead of as a single element list with VoidTy. Now with a fix for the verifier. llvm-svn: 99206 | |||||
* | backing out r99170 because it still fails on clang-x86_64-darwin10-fnt | Gabor Greif | 2010-03-22 | 3 | -17/+9 | |
| | | | | llvm-svn: 99171 | |||||
* | Now that hopefully all direct accesses to InvokeInst operands are fixed | Gabor Greif | 2010-03-22 | 3 | -9/+17 | |
| | | | | | | we can reapply the InvokeInst operand reordering patch. (see r98957). llvm-svn: 99170 | |||||
* | Free all Constants in ~LLVMConstantImpl. We avoid assertion failures | Jeffrey Yasskin | 2010-03-22 | 2 | -6/+27 | |
| | | | | | | | | | by dropping all references from all constants that can use other constants before trying to destroy any of them. I also had to free bugpoint's Module in ~BugDriver(). llvm-svn: 99160 | |||||
* | Re-add the Metadata.h include to LLVMContextImpl.h so that MDNode is complete | Jeffrey Yasskin | 2010-03-21 | 2 | -3/+1 | |
| | | | | | | | where FoldingSet<MDNode> is instantiated. Clang and MSVC complain; gcc doesn't. llvm-svn: 99147 | |||||
* | Move the LLVMContextImpl implementation into a .cpp file. | Jeffrey Yasskin | 2010-03-21 | 3 | -69/+87 | |
| | | | | llvm-svn: 99146 | |||||
* | Memoize InlineAsms into the LLVMContext and delete them on shutdown. | Jeffrey Yasskin | 2010-03-21 | 3 | -11/+68 | |
| | | | | | | Fixes PR803. llvm-svn: 99143 | |||||
* | call void @llvm.dbg.declare(metadata !{i32* null}, metadata !1) | Devang Patel | 2010-03-19 | 1 | -4/+0 | |
| | | | | | | is valid, but not useful, when variable identified by !1 is optimized away by the optimizer. llvm-svn: 98986 | |||||
* | back out r98957, it broke ↵ | Gabor Greif | 2010-03-19 | 3 | -17/+9 | |
| | | | | | | http://smooshlab.apple.com:8010/builders/clang-x86_64-darwin10-fnt/builds/703 in the nightly test suite llvm-svn: 98958 | |||||
* | Recommit r80858 again (which has been backed out in r80871). | Gabor Greif | 2010-03-19 | 3 | -9/+17 | |
| | | | | | | | | | | | This time I did a self-hosted bootstrap on Linux x86-64, with no problems. Let's see how darwin 64-bit self-hosting goes. At the first sign of failure I'll back this out. Maybe the valgrind bots give me a hint of what may be wrong (it at all). llvm-svn: 98957 | |||||
* | add asmprinter suport for unions, fixing Feature/unions.ll to actually | Chris Lattner | 2010-03-16 | 1 | -0/+9 | |
| | | | | | | be doing something useful. llvm-svn: 98677 | |||||
* | fix some pointless layering violations. | Chris Lattner | 2010-03-14 | 1 | -2/+1 | |
| | | | | llvm-svn: 98460 | |||||
* | Allow types that have been forwarded to to be freed. | Jeffrey Yasskin | 2010-03-13 | 1 | -4/+11 | |
| | | | | | Tested: make check-lit && valgrind --dsymutil=yes --leak-check=full unittests/ExecutionEngine/JIT/Debug/JITTests llvm-svn: 98447 | |||||
* | Delete MDNodes when LLVMContext is destroyed. Previous attempts: r97918, ↵ | Jeffrey Yasskin | 2010-03-13 | 2 | -16/+25 | |
| | | | | | | | | | | r97788. Tested: clang debug bootstrap, llvm-gcc bootstrap, `make check-lit` after configuring with --with-llvmgccdir (and this did run the FrontendC* tests this time) llvm-svn: 98410 | |||||
* | Remove duplicated code. No functionality change. | Benjamin Kramer | 2010-03-10 | 1 | -9/+0 | |
| | | | | llvm-svn: 98156 | |||||
* | Roll back r97918 again. Just configuring against llvm-gcc wasn't enough to run | Jeffrey Yasskin | 2010-03-07 | 2 | -25/+16 | |
| | | | | | | the FrontendC* tests. :( llvm-svn: 97921 | |||||
* | Reapply r97788 to free MDNodes when the LLVMContext is destroyed. It | Jeffrey Yasskin | 2010-03-07 | 2 | -16/+25 | |
| | | | | | | bootstraps llvm-gcc this time. llvm-svn: 97918 | |||||
* | Add verification of union types. | Nick Lewycky | 2010-03-06 | 1 | -0/+9 | |
| | | | | llvm-svn: 97889 | |||||
* | Revert r97788 because it broke test/FrontendC/2010-02-16-DbgVarScope.c. | Jeffrey Yasskin | 2010-03-05 | 2 | -25/+16 | |
| | | | | llvm-svn: 97792 | |||||
* | Free MDNodes when the LLVMContext is destroyed. Leak found by Valgrind. | Jeffrey Yasskin | 2010-03-05 | 2 | -16/+25 | |
| | | | | llvm-svn: 97788 | |||||
* | Stop leaking MDStrings. | Jeffrey Yasskin | 2010-03-04 | 1 | -0/+5 | |
| | | | | llvm-svn: 97763 | |||||
* | Rewrite makefiles to explicitly reference DESTDIR to fix bug 3153. | Erick Tryzelaar | 2010-03-04 | 1 | -2/+2 | |
| | | | | | | We need this so can not bake DESTDIR into the O'Caml symlinks. llvm-svn: 97743 | |||||
* | Make the 'icmp pred trunc(ext(X)), CST --> icmp pred X, ext(trunc(CST))' | Nick Lewycky | 2010-03-04 | 1 | -24/+5 | |
| | | | | | | | | | | transformation much more careful. Truncating binary '01' to '1' sounds like it's safe until you realize that it switched from positive to negative under a signed interpretation, and that depends on the icmp predicate. Also a few miscellaneous cleanups. llvm-svn: 97721 | |||||
* | fix incorrect folding of icmp with undef, PR6481. | Chris Lattner | 2010-03-03 | 1 | -1/+1 | |
| | | | | llvm-svn: 97659 | |||||
* | Add Module functions in place of module providers. | Erick Tryzelaar | 2010-03-02 | 1 | -1/+6 | |
| | | | | llvm-svn: 97608 | |||||
* | Rename LLVMUseIteratorRef to LLVMUseRef since we don't refer to iterators in ↵ | Erick Tryzelaar | 2010-03-02 | 1 | -7/+10 | |
| | | | | | | llvm-c. llvm-svn: 97585 | |||||
* | Rather than passing "false" for InsertBefore, AddressSpace for ThreadLocal, | Duncan Sands | 2010-03-02 | 1 | -2/+2 | |
| | | | | | | | and nothing for AddressSpace, pass 0 for InsertBefore, "false" for ThreadLocal and AddressSpace for AddressSpace. Spotted by gcc-4.5. llvm-svn: 97563 |