summaryrefslogtreecommitdiffstats
path: root/llvm/tools/opt
Commit message (Collapse)AuthorAgeFilesLines
...
* split dom frontier handling stuff out to its own DominanceFrontier header,Chris Lattner2011-01-021-7/+4
| | | | | | so that Dominators.h is *just* domtree. Also prune #includes a bit. llvm-svn: 122714
* Print breakpoints for call instructions. This is used by optimized debug ↵Devang Patel2010-12-091-0/+13
| | | | | | info test harness. llvm-svn: 121432
* Add a simple breakpoint location printer. This will be used by upcoming ↵Devang Patel2010-12-071-1/+59
| | | | | | "debug info in optimized code" quality test harness to set breakpoints at "interesting" locations. llvm-svn: 121078
* Move check of command line options after command line parsing.Tobias Grosser2010-12-021-5/+5
| | | | | | | | | The check to not allow -analyze and -disable-output at the same time was done before parsing the command line flags. Therefore it never triggered, and in case both options where used opt segfaulted. Fix this by moving this check a after command line parsing. llvm-svn: 120732
* Merge System into Support.Michael J. Spencer2010-11-291-1/+1
| | | | llvm-svn: 120298
* Tweak the opt -O2 / opt -O3 inliner thresholds to be the same as llvm-gcc andJakob Stoklund Olesen2010-11-021-2/+2
| | | | | | clang are using. llvm-svn: 118118
* Add RegionPass support.Tobias Grosser2010-10-201-0/+38
| | | | | | | A RegionPass is executed like a LoopPass but on the regions detected by the RegionInfo pass instead of the loops detected by the LoopInfo pass. llvm-svn: 116905
* Get rid of static constructors for pass registration. Instead, every pass ↵Owen Anderson2010-10-191-0/+12
| | | | | | | | | | | | | | | | | exposes an initializeMyPassFunction(), which must be called in the pass's constructor. This function uses static dependency declarations to recursively initialize the pass's dependencies. Clients that only create passes through the createFooPass() APIs will require no changes. Clients that want to use the CommandLine options for passes will need to manually call the appropriate initialization functions in PassInitialization.h before parsing commandline arguments. I have tested this with all standard configurations of clang and llvm-gcc on Darwin. It is possible that there are problems with the static dependencies that will only be visible with non-standard options. If you encounter any crash in pass registration/creation, please send the testcase to me directly. llvm-svn: 116820
* Move tool_output_file into its own file.Dan Gohman2010-10-072-2/+2
| | | | llvm-svn: 115973
* Revert "CMake: Get rid of LLVMLibDeps.cmake and export the libraries normally."Michael J. Spencer2010-09-131-9/+2
| | | | | | | | | | This reverts commit r113632 Conflicts: cmake/modules/AddLLVM.cmake llvm-svn: 113819
* CMake: Get rid of LLVMLibDeps.cmake and export the libraries normally.Michael J. Spencer2010-09-101-2/+9
| | | | llvm-svn: 113632
* Execute all Pass Printers even if -quiet is set.Tobias Grosser2010-09-081-19/+16
| | | | | | | | | | | | Follow the same logic in the LoopPass, ModulePass and CallGraphSCCPass printers, as it was already used in the BasicBlockPass and FunctionPass printers. This is more consistent. The other option would have been to completely disable dumping the analysis information. However, as this information is the only information printed if the -analysis flag is set, calling opt would not do anything at all. llvm-svn: 113360
* Include original pass name in the PassPrinter's name.Tobias Grosser2010-09-081-10/+36
| | | | llvm-svn: 113359
* Make tool_output_file's raw_ostream instance a member variable insteadDan Gohman2010-09-012-12/+17
| | | | | | | | | | of a base class. This makes it possible to unregister the file from FilesToRemove when the file is done. Also, this eliminates the need for formatted_tool_output_file. llvm-svn: 112706
* Convert tools to use tool_output_file, and introduce errorDan Gohman2010-08-201-5/+11
| | | | | | checking to places which previously lacked it. llvm-svn: 111651
* Use the new tool_output_file in several tools. This fixes a varietyDan Gohman2010-08-201-12/+8
| | | | | | | | of problems with output files being left behind or output streams being left unclosed. Fix llvm-mc to respect the -o option in all modes, rather than hardcoding outs() in some cases. llvm-svn: 111603
* Make the SCC printing passes use errs() instead of outs(), as theDan Gohman2010-08-201-10/+10
| | | | | | other printing passes do, and update the documentation accordingly. llvm-svn: 111601
* Minor cleanups to follow the common convention for passDan Gohman2010-08-201-7/+7
| | | | | | registration variables. llvm-svn: 111598
* Minor cleanups to follow the common convention for passDan Gohman2010-08-202-14/+16
| | | | | | registration variables. llvm-svn: 111596
* Allow the -analyze option to follow the -o option, which defaults toDan Gohman2010-08-181-30/+35
| | | | | | standard output, instead of just hardcoding outs(). llvm-svn: 111372
* Don't translate "-" to outs() manually; raw_ostream does that automatically.Dan Gohman2010-08-181-31/+23
| | | | llvm-svn: 111371
* Reapply r110396, with fixes to appease the Linux buildbot gods.Owen Anderson2010-08-064-21/+24
| | | | llvm-svn: 110460
* Revert r110396 to fix buildbots.Owen Anderson2010-08-064-24/+21
| | | | llvm-svn: 110410
* Don't use PassInfo* as a type identifier for passes. Instead, use the ↵Owen Anderson2010-08-054-21/+24
| | | | | | | | address of the static ID member as the sole unique type identifier. Clean up APIs related to this change. llvm-svn: 110396
* simplify by using CallSite constructors; virtually eliminates CallSite::get ↵Gabor Greif2010-07-281-2/+2
| | | | | | from the tree llvm-svn: 109687
* Convert some tab stops into spaces.Duncan Sands2010-07-121-1/+1
| | | | llvm-svn: 108130
* Change another reference to the "indirect callgraph node" toDuncan Sands2010-06-091-1/+1
| | | | | | refer to the "external node" instead. llvm-svn: 105731
* Output "external node" rather than "Indirect CallGraph node" when printingDuncan Sands2010-06-091-1/+1
| | | | | | | | callgraph SCC's. This makes it match what the node itself would print. Also, "indirect callgraph node" doesn't make sense - it has nothing particularly to do with indirect calls. llvm-svn: 105730
* Don't special-case stdout in llvm::WriteBitcodeToFile; just considerDan Gohman2010-05-271-2/+6
| | | | | | | it to be the caller's responsibility to provide a stream in binary mode. This fixes a layering violation and avoids an outs() call. llvm-svn: 104878
* Don't create an output stream when output is disabled.Dan Gohman2010-05-271-19/+21
| | | | llvm-svn: 104875
* Avoid calling outs() and fouts() when the stream isn't really needed.Dan Gohman2010-05-271-3/+7
| | | | llvm-svn: 104873
* Use regular PassManager instead of FunctionPassManager in opt, since itDan Gohman2010-05-141-12/+8
| | | | | | isn't doing lazy streaming. This also fixes a missing doFinalization call. llvm-svn: 103774
* introduce a new CallGraphSCC class, and pass it aroundChris Lattner2010-04-161-5/+4
| | | | | | | | to CallGraphSCCPass's instead of passing around a std::vector<CallGraphNode*>. No functionality change, but now we have a much tidier interface. llvm-svn: 101558
* Remove unnecessary uses of <iostream>.Daniel Dunbar2010-04-152-5/+3
| | | | llvm-svn: 101338
* Trim #includes.Dan Gohman2010-03-241-1/+0
| | | | llvm-svn: 99416
* Avoid leaking the FunctionPassManager from opt.Jeffrey Yasskin2010-03-221-2/+2
| | | | llvm-svn: 99180
* Avoid a dangling pointer dereference, PassManager::add can delete the Pass.Benjamin Kramer2010-02-181-1/+2
| | | | llvm-svn: 96576
* Kill ModuleProvider and ghost linkage by inverting the relationship betweenJeffrey Yasskin2010-01-271-2/+1
| | | | | | | | | | | | | | | | | | | | | Modules and ModuleProviders. Because the "ModuleProvider" simply materializes GlobalValues now, and doesn't provide modules, it's renamed to "GVMaterializer". Code that used to need a ModuleProvider to materialize Functions can now materialize the Functions directly. Functions no longer use a magic linkage to record that they're materializable; they simply ask the GVMaterializer. Because the C ABI must never change, we can't remove LLVMModuleProviderRef or the functions that refer to it. Instead, because Module now exposes the same functionality ModuleProvider used to, we store a Module* in any LLVMModuleProviderRef and translate in the wrapper methods. The bindings to other languages still use the ModuleProvider concept. It would probably be worth some time to update them to follow the C++ more closely, but I don't intend to do it. Fixes http://llvm.org/PR5737 and http://llvm.org/PR5735. llvm-svn: 94686
* make -fno-rtti the default unless a directory builds with REQUIRES_RTTI.Chris Lattner2010-01-241-1/+0
| | | | llvm-svn: 94378
* Stop building RTTI information for *most* llvm libraries. NotableChris Lattner2010-01-221-0/+1
| | | | | | | | | | | missing ones are libsupport, libsystem and libvmcore. libvmcore is currently blocked on bugpoint, which uses EH. Once it stops using EH, we can switch it off. This #if 0's out 3 unit tests, because gtest requires RTTI information. Suggestions welcome on how to fix this. llvm-svn: 94164
* elimiante the dynamic_cast's from opt.Chris Lattner2010-01-221-5/+12
| | | | llvm-svn: 94160
* simplify code.Chris Lattner2010-01-221-9/+4
| | | | llvm-svn: 94159
* Make opt -O3 act more like clang -O3 etc., by making the inlining thresholdsEli Friedman2010-01-181-1/+11
| | | | | | match. llvm-svn: 93798
* Don't create a (empty) output file, and don't warn about bitcode outputDan Gohman2010-01-171-13/+18
| | | | | | | | | | | to a console, when --analyze is used. Similarly, avoid creating an empty output file when --disable-output is used. Print a warning when the -o option appears with either --analyze or --disable-output, to indicate that the option is being ignored. llvm-svn: 93685
* Enable debug buffering.David Greene2010-01-051-0/+4
| | | | llvm-svn: 92667
* when opt crashes, print its command line arguments as a pretty stack trace.Chris Lattner2009-12-091-2/+5
| | | | | | Somehow opt was missed when this was added. llvm-svn: 90912
* Remove ShortNames from getNodeLabel in DOTGraphTraitsTobias Grosser2009-11-301-2/+1
| | | | llvm-svn: 90134
* Instantiate DefaultDOTGraphTraitsTobias Grosser2009-11-301-0/+3
| | | | llvm-svn: 90133
* Make opt default to not adding a target data string and update tests that ↵Kenneth Uildriks2009-11-031-5/+1
| | | | | | depend on target data to supply it within the test llvm-svn: 85900
* nothing opt uses can throw, remove the try block and -fexceptions whenChris Lattner2009-10-223-163/+150
| | | | | | building opt. llvm-svn: 84816
OpenPOWER on IntegriCloud