summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* misched: ILP scheduler for experimental heuristics.Andrew Trick2012-10-155-20/+312
| | | | llvm-svn: 165950
* Add an entry in CREDITS.TXTArgyrios Kyrtzidis2012-10-151-0/+4
| | | | llvm-svn: 165949
* [ms-inline asm] Update the end loc for ParseIntelMemOperand.Chad Rosier2012-10-151-0/+1
| | | | llvm-svn: 165947
* [ms-inline asm] Add a few new APIs to the AsmParser class in support of MS-StyleChad Rosier2012-10-152-13/+54
| | | | | | | | | inline assembly. For the time being, these will be called directly by clang. However, in the near future I expect these to be sunk back into the MC layer and more basic APIs (e.g., getClobbers(), getConstraints(), etc.) will be called by clang. llvm-svn: 165946
* [ms-inline asm] Use incoming argument rather than hard coding to false.Chad Rosier2012-10-151-1/+1
| | | | llvm-svn: 165945
* Fix a typo in bitcode docs, from 165814.Jan Wen Voung2012-10-151-1/+1
| | | | llvm-svn: 165944
* Teach TargetInfo to hold on to the TargetOptions with which it wasDouglas Gregor2012-10-156-16/+31
| | | | | | created. llvm-svn: 165943
* Resubmit the changes to llvm core to update the functions to support ↵Micah Villmow2012-10-1540-109/+190
| | | | | | different pointer sizes on a per address space basis. llvm-svn: 165941
* PowerPC: add EmitTCEntry class for TOC creationAdhemerval Zanella2012-10-155-2/+28
| | | | | | | This patch replaces the EmitRawText by a EmitTCEntry class (specialized for each Streamer) in PowerPC64 TOC entry creation. llvm-svn: 165940
* Implement internal_memmove.Alexander Potapenko2012-10-153-0/+29
| | | | | | | | Use internal_memmove() and internal_memcpy() in the memcpy() and memmove() wrappers when building the dynamic runtime (OS X only), to work around a bug in resolver functions wrapping. See also http://code.google.com/p/address-sanitizer/issues/detail?id=116 llvm-svn: 165939
* [asan] fix a testKostya Serebryany2012-10-151-1/+1
| | | | llvm-svn: 165938
* [asan] make AddressSanitizer to be a FunctionPass instead of ModulePass. ↵Kostya Serebryany2012-10-151-1/+1
| | | | | | clang part: for FunctionPass we need to run asan at a different point, otherwise it will run before inlining llvm-svn: 165937
* [asan] make AddressSanitizer to be a FunctionPass instead of ModulePass. ↵Kostya Serebryany2012-10-152-64/+53
| | | | | | This will simplify chaining other FunctionPasses with asan. Also some minor cleanup llvm-svn: 165936
* [asan] fix more tests Kostya Serebryany2012-10-151-9/+9
| | | | llvm-svn: 165934
* Only report first error when no compilation database is found.Daniel Jasper2012-10-151-2/+7
| | | | | Review: http://llvm-reviews.chandlerc.com/D62 llvm-svn: 165933
* [asan] change the asan output slightly to avoid user confusion: a) add ':' ↵Kostya Serebryany2012-10-1511-17/+17
| | | | | | after 'AddressSanitizer' and b) changed 'crashed' to 'SEGV' llvm-svn: 165932
* Update the memcpy rewriting to fully support widened int rewriting. ThisChandler Carruth2012-10-152-9/+31
| | | | | | | | includes extracting ints for copying elsewhere and inserting ints when copying into the alloca. This should fix the CanSROA assertion coming out of Clang's regression test suite. llvm-svn: 165931
* Follow-up fix to r165928: handle memset rewriting for widened integers,Chandler Carruth2012-10-152-30/+38
| | | | | | | and generally clean up the memset handling. It had rotted a bit as the other rewriting logic got polished more. llvm-svn: 165930
* Fixed PR13938: the ARM backend was crashing because it couldn't select a ↵Silviu Baranga2012-10-152-2/+55
| | | | | | VDUPLANE node with the vector input size different from the output size. This was bacause the BUILD_VECTOR lowering code didn't check that the size of the input vector was correct for using VDUPLANE. llvm-svn: 165929
* First major step toward addressing PR14059. This teaches SROA to handleChandler Carruth2012-10-153-76/+150
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | cases where we have partial integer loads and stores to an otherwise promotable alloca to widen[1] those loads and stores to cover the entire alloca and bitcast them into the appropriate type such that promotion can proceed. These partial loads and stores stem from an annoying confluence of ARM's calling convention and ABI lowering and the FCA pre-splitting which takes place in SROA. Clang lowers a { double, double } in-register function argument as a [4 x i32] function argument to ensure it is placed into integer 32-bit registers (a really unnerving implicit contract between Clang and the ARM backend I would add). This results in a FCA load of [4 x i32]* from the { double, double } alloca, and SROA decomposes this into a sequence of i32 loads and stores. Inlining proceeds, code gets folded, but at the end of the day, we still have i32 stores to the low and high halves of a double alloca. Widening these to be i64 operations, and bitcasting them to double prior to loading or storing allows promotion to proceed for these allocas. I looked quite a bit changing the IR which Clang produces for this case to be more friendly, but small changes seem unlikely to help. I think the best representation we could use currently would be to pass 4 i32 arguments thereby avoiding any FCAs, but that would still require this fix. It seems like it might eventually be nice to somehow encode the ABI register selection choices outside of the parameter type system so that the parameter can be a { double, double }, but the CC register annotations indicate that this should be passed via 4 integer registers. This patch does not address the second problem in PR14059, which is the reverse: when a struct alloca is loaded as a *larger* single integer. This patch also does not address some of the code quality issues with the FCA-splitting. Those don't actually impede any optimizations really, but they're on my list to clean up. [1]: Pedantic footnote: for those concerned about memory model issues here, this is safe. For the alloca to be promotable, it cannot escape or have any use of its address that could allow these loads or stores to be racing. Thus, widening is always safe. llvm-svn: 165928
* Hoist the canConvertValue predicate and the convertValue transform outChandler Carruth2012-10-151-44/+52
| | | | | | | into static helper functions. They're really quite generic and are going to be needed elsewhere shortly. llvm-svn: 165927
* Use enum values instead of magic numbers for indexing into the attribute list.Bill Wendling2012-10-154-7/+14
| | | | llvm-svn: 165925
* Add an enum for the return and function indexes into the AttrListPtr object. ↵Bill Wendling2012-10-1513-84/+130
| | | | | | This gets rid of some magic numbers. llvm-svn: 165924
* Use a ::get method to create the attribute from Attributes::AttrVals instead ↵Bill Wendling2012-10-152-20/+24
| | | | | | of a constructor. llvm-svn: 165923
* Supply a default 'operator=' method.Bill Wendling2012-10-151-0/+4
| | | | llvm-svn: 165922
* Introduce the notion of excluded headers into the module mapDouglas Gregor2012-10-158-39/+126
| | | | | | | | description. Previously, one could emulate this behavior by placing the header in an always-unavailable submodule, but Argyrios guilted me into expressing this idea properly. llvm-svn: 165921
* Move the AttributesImpl header file into the VMCore directory so that it can ↵Bill Wendling2012-10-154-5/+7
| | | | | | be opaque. llvm-svn: 165920
* Attributes RewriteBill Wendling2012-10-157-16/+28
| | | | | | | | Convert the uses of the Attributes class over to the new format. The Attributes::get method call now takes an LLVM context so that the attributes object can be uniquified and stored. llvm-svn: 165918
* Attributes RewriteBill Wendling2012-10-1521-125/+165
| | | | | | | | | | Convert the internal representation of the Attributes class into a pointer to an opaque object that's uniqued by and stored in the LLVMContext object. The Attributes class then becomes a thin wrapper around this opaque object. Eventually, the internal representation will be expanded to include attributes that represent code generation options, etc. llvm-svn: 165917
* instcombine: Migrate strcmp and strncmp optimizationsMeador Inge2012-10-1510-274/+351
| | | | | | | This patch migrates the strcmp and strncmp optimizations from the simplify-libcalls pass into the instcombine library call simplifier. llvm-svn: 165915
* At -O0, emit an @llvm.trap() call at the end of a value-returning function whichRichard Smith2012-10-152-3/+11
| | | | | | fails to return a value, to make debugging this issue easier. llvm-svn: 165914
* Provide a slightly more helpful diagnostic if -fcatch-undefined-behavior findsRichard Smith2012-10-141-0/+6
| | | | | | | a problem with __int128 arithmetic but the runtime was not built with __int128 support. llvm-svn: 165913
* Use a better test for __int128 support, which works on old GCC versions.Richard Smith2012-10-141-1/+1
| | | | llvm-svn: 165912
* Make sure temporary files are deleted when recovering from a crash when ↵Benjamin Kramer2012-10-141-0/+1
| | | | | | compiling modules. llvm-svn: 165911
* Delete temporary output files when an error occurs during PCH reading.Benjamin Kramer2012-10-142-0/+2
| | | | | | | | This reduces the spam make test leaves behind in /tmp. The assert isn't particularly useful because it's not run with -disable-free (the default when using the clang driver) but should cover all -cc1 tests. llvm-svn: 165910
* Update CMake build.Benjamin Kramer2012-10-141-0/+1
| | | | llvm-svn: 165908
* Fix a typo that made ImmutableMap::getMaxElement() useless.Benjamin Kramer2012-10-142-1/+51
| | | | | | Add a basic unit test for ImmutableMap. Found by inspection. llvm-svn: 165907
* Simplify code. No functionality change.Benjamin Kramer2012-10-141-10/+4
| | | | llvm-svn: 165905
* Simplify code. No functionality change.Benjamin Kramer2012-10-141-5/+3
| | | | llvm-svn: 165904
* Unquadratize SetVector removal loops in DSE.Benjamin Kramer2012-10-141-27/+36
| | | | | | | | | Erasing from the beginning or middle of the vector is expensive, remove_if can do it in linear time even though it's a bit ugly without lambdas. No functionality change. llvm-svn: 165903
* Remove dead methods.Bill Wendling2012-10-143-13/+0
| | | | llvm-svn: 165902
* Remove operator cast method in favor of querying with the correct method.Bill Wendling2012-10-141-3/+5
| | | | llvm-svn: 165900
* Remove operator cast method in favor of querying with the correct method.Bill Wendling2012-10-146-23/+39
| | | | llvm-svn: 165899
* Fix use after free when deleting attributes in a chained folding set.Benjamin Kramer2012-10-141-3/+5
| | | | | | Can't follow the intrusive linked list when the element is gone. llvm-svn: 165898
* Don't use the new syntax just yet.Bill Wendling2012-10-141-2/+2
| | | | llvm-svn: 165897
* Remove the bitwise AND operators from the Attributes class. Replace it with ↵Bill Wendling2012-10-145-17/+9
| | | | | | the equivalent from the builder class. llvm-svn: 165896
* Remove the bitwise assignment OR operator from the Attributes class. Replace ↵Bill Wendling2012-10-148-26/+36
| | | | | | it with the equivalent from the builder class. llvm-svn: 165895
* Remove the bitwise OR operator from the Attributes class. Replace it with ↵Bill Wendling2012-10-142-8/+10
| | | | | | the equivalent from the builder class. llvm-svn: 165894
* Remove the bitwise XOR operator from the Attributes class. Replace it with ↵Bill Wendling2012-10-144-8/+6
| | | | | | the equivalent from the builder class. llvm-svn: 165893
* Remove the bitwise NOT operator from the Attributes class. Replace it with ↵Bill Wendling2012-10-147-22/+35
| | | | | | the equivalent from the builder class. llvm-svn: 165892
OpenPOWER on IntegriCloud