summaryrefslogtreecommitdiffstats
path: root/llvm/lib
Commit message (Collapse)AuthorAgeFilesLines
...
* Add debugtype, make output marginally more niceChris Lattner2003-10-191-3/+4
| | | | llvm-svn: 9251
* Fix PR #47Chris Lattner2003-10-191-3/+3
| | | | llvm-svn: 9250
* Fix bug: Jello/2003-10-18-PHINode-ConstantExpr-CondCode-Failure.llxChris Lattner2003-10-191-9/+21
| | | | | | This also fixes miscompilation of 176.gcc. llvm-svn: 9249
* Without this option, the -run-llc mode does not work with shared objects atChris Lattner2003-10-181-0/+1
| | | | | | | all, making it pointless for use with the code generator debugger. With it, it works like a charm. llvm-svn: 9245
* Fix warningChris Lattner2003-10-181-0/+1
| | | | llvm-svn: 9236
* add support for new linkage typesChris Lattner2003-10-182-8/+18
| | | | llvm-svn: 9228
* Fix PR#50Chris Lattner2003-10-181-6/+6
| | | | llvm-svn: 9227
* Add support for the new varargs intrinsics and instructionsChris Lattner2003-10-184-20/+38
| | | | llvm-svn: 9226
* ADd support for the new varargs instructionsChris Lattner2003-10-181-1/+2
| | | | llvm-svn: 9225
* Add support for the new varargs intrinsicsChris Lattner2003-10-181-26/+36
| | | | llvm-svn: 9224
* Update the sparc backend to at least compile correctly with the new varargs ↵Chris Lattner2003-10-182-4/+16
| | | | | | stuff even if it's not all implemented yet. llvm-svn: 9223
* Interpret the new varargs intrinsics correctlyChris Lattner2003-10-183-28/+13
| | | | llvm-svn: 9222
* Updated to emit the final 1.0 bytecode format. This supports weak linkage,Chris Lattner2003-10-182-22/+40
| | | | | | more efficient encoding of varargs calls, and the new varargs intrinsics. llvm-svn: 9221
* * New revised variable argument handling supportChris Lattner2003-10-185-149/+290
| | | | | | | * More dense bytecode encoding for varargs calls (like printf) * Eliminated the extremely old bytecode format. rev #0 is now 1.0 llvm-svn: 9220
* New revised variable argument handling supportChris Lattner2003-10-182-5/+108
| | | | llvm-svn: 9219
* Chris says it's better to assert that TheModule is valid than silently return 0.Misha Brukman2003-10-171-2/+1
| | | | llvm-svn: 9212
* You can't just blat the address into memory, you have to blat itsBrian Gaeke2003-10-171-4/+7
| | | | | | displacement. llvm-svn: 9210
* Don't release the Module, as that invalidates the Module* within theMisha Brukman2003-10-171-2/+2
| | | | | | | ModuleProvider, which has bad consequences in lli::callAsMain() which tries to access that same Module*. llvm-svn: 9205
* Implement replaceMachineCodeForFunction() for x86.Brian Gaeke2003-10-171-0/+7
| | | | llvm-svn: 9204
* Refactor jump insertion code from CompilationCallback() into insertJumpAtAddr().Brian Gaeke2003-10-171-27/+33
| | | | | | | | Make insertFarJumpAtAddr() return void, because nothing uses its return value. Remove some commented-out code. Implement replaceMachineCodeForFunction() for SPARC. llvm-svn: 9203
* Add prototype for replaceMachineCodeForFunction().Brian Gaeke2003-10-172-0/+3
| | | | llvm-svn: 9202
* Return the Module that we just materialized.Misha Brukman2003-10-171-2/+5
| | | | llvm-svn: 9201
* Tidy up doxygen comment for getPointerToFunction().Brian Gaeke2003-10-171-0/+10
| | | | | | Add prototypes for recompileAndRelinkFunction() and runJITOnFunction(). llvm-svn: 9200
* Refactor running the JIT passes on a single function into the new method,Brian Gaeke2003-10-171-7/+35
| | | | | | | | | runJITOnFunction(). Add new method for recompiling and patching in new versions of functions, recompileAndRelinkFunction(). llvm-svn: 9199
* Remove extra blank line.Brian Gaeke2003-10-171-1/+0
| | | | llvm-svn: 9196
* Do not crash on empty structuresChris Lattner2003-10-171-0/+5
| | | | llvm-svn: 9195
* Tighten up handling of checks for shift instructionsChris Lattner2003-10-171-2/+4
| | | | llvm-svn: 9191
* Eliminate some extraneous code in SlotCalculator::insertVal().Alkis Evlogimenos2003-10-176-64/+62
| | | | | | | | | Rename SlotCalculator::getValSlot() to SlotCalculator::getSlot(), SlotCalculator::insertValue() to SlotCalculator::getOrCreateSlot(), SlotCalculator::insertVal() to SlotCalculator::insertValue(), and SlotCalculator::doInsertVal() to SlotCalculator::doInsertValue(). llvm-svn: 9190
* Fix typo in comment.Brian Gaeke2003-10-161-1/+1
| | | | llvm-svn: 9187
* Fix a typo in a comment, and zap a blank line.Brian Gaeke2003-10-161-2/+1
| | | | llvm-svn: 9184
* Add separator between different types of readers.Misha Brukman2003-10-161-0/+1
| | | | llvm-svn: 9183
* JIT.cpp:Misha Brukman2003-10-162-5/+7
| | | | | | | | | | | | * #include "llvm/ModuleProvider" * alphabetize #includes * omit extraneous parens in pointer expressions VM.cpp: * #include "llvm/ModuleProvider" * alphabetize #includes llvm-svn: 9182
* * Reorder includes as per the style guideMisha Brukman2003-10-161-6/+22
| | | | | | | | * Move the constructors from .h file here * Document ExecutionEngine::create() * Catch exception possibly thrown by ModuleProvider::releaseModule() llvm-svn: 9181
* Minor cleanupsChris Lattner2003-10-161-30/+28
| | | | llvm-svn: 9177
* Add support for 'weak' linkage.Chris Lattner2003-10-165-11/+46
| | | | llvm-svn: 9171
* Add support for 'weak' linkage.Chris Lattner2003-10-162-4/+25
| | | | | | | | For now, we translate linkonce into weak linkage in the bytecode format because we don't have enough bits to represent it. We will rev the bytecode version soon anyways, so this will be fixed in the near future. llvm-svn: 9170
* Completely rewrite support for the Value::use_* list. Now, all operations onChris Lattner2003-10-161-37/+11
| | | | | | | | | | | | | | | | | | | this list (except use_size()) are constant time. Before the killUse method (used whenever something stopped using a value) was linear time, and thus very very slow for large programs. This speeds GCCAS up _substantially_ on large programs: almost 2x for 176.gcc: 176.gcc: 77.07s -> 37.38s 177.mesa: 7.59s -> 5.57s 252.eon: 21.02s -> 19.52s (*) 253.perlbmk: 11.40s -> 13.05s 254.gap: 7.25s -> 7.42s 252.eon would speed up a whole lot more, but optimization time is being dominated by the inlining pass, which needs to be fixed. llvm-svn: 9160
* This code does not require random access use_listsChris Lattner2003-10-161-7/+2
| | | | llvm-svn: 9156
* Eliminate using declarationChris Lattner2003-10-161-29/+35
| | | | | | Rewrite code to work with use_lists what are either random access or bidirectional llvm-svn: 9155
* #include vector which we will need here soonChris Lattner2003-10-151-0/+1
| | | | llvm-svn: 9144
* Fix up error message.Brian Gaeke2003-10-151-1/+2
| | | | llvm-svn: 9139
* Decrease usage of use_size()Chris Lattner2003-10-159-22/+22
| | | | llvm-svn: 9135
* CleanupChris Lattner2003-10-151-5/+2
| | | | llvm-svn: 9133
* Generalize abstract interpreter interface to allow linking in an arbitrary ↵Chris Lattner2003-10-141-18/+23
| | | | | | number of shared objects llvm-svn: 9129
* Order #includes as per the style guide.Misha Brukman2003-10-141-3/+3
| | | | llvm-svn: 9128
* Enabling incremental bytecode loading in the JIT:Misha Brukman2003-10-141-3/+10
| | | | | | | * FunctionPassManager ctor now takes in a ModuleProvider * run() materializes function before running passes on it llvm-svn: 9126
* Enabling incremental bytecode loading in the JIT:Misha Brukman2003-10-143-7/+15
| | | | | | * The VM is now constructed with a ModuleProvider llvm-svn: 9125
* Enabling incremental bytecode loading in the JIT:Misha Brukman2003-10-141-4/+5
| | | | | | * ExecutionEngine and VM can be constructed using a ModuleProvider. llvm-svn: 9124
* Substantial cleanups:Chris Lattner2003-10-141-28/+35
| | | | | | | | | | | | | * Add header comment * Remove extraneous #includes * Move the FileType enum into the GCC class * The GCC class is not virtual. * Move all of the "constructor" functions into the classes themselves * Stop using cl::list as arguments, use std::vector instead (which cl::list derives from) * Improve comments llvm-svn: 9121
* Actually return an error if something bad happens, don't just exit.Chris Lattner2003-10-141-5/+4
| | | | llvm-svn: 9119
OpenPOWER on IntegriCloud