summaryrefslogtreecommitdiffstats
path: root/llvm/lib/Bytecode/Reader/Reader.cpp
Commit message (Collapse)AuthorAgeFilesLines
...
* Significantly clean up parsing of instructions. This exceptionizes andChris Lattner2003-10-091-4/+1
| | | | | | | simplifies the control flow a bit. This provides a small (~3%) speedup, but it's primarily a cleanup exercise. llvm-svn: 8983
* Eliminate the instruction placeholder. Simplify a bunch of code.Chris Lattner2003-10-091-1/+1
| | | | | | This results in no significant speedup, but does provide simpler code llvm-svn: 8980
* Eliminate the old LateResolveValues data structure, replacing it with aChris Lattner2003-10-091-23/+25
| | | | | | | | | | | | | new, simpler, ForwardReferences data structure. This is just the first simple replacement, subsequent changes will improve the code more. This simple change improves the performance of loading a file from HDF5 (contributed by Bill) from 2.36s to 1.93s, a 22% improvement. This presumably has to do with the fact that we only create ONE placeholder for a particular forward referenced values, and also may be because the data structure is much simpler. llvm-svn: 8979
* Remove a dead methodChris Lattner2003-10-091-11/+0
| | | | llvm-svn: 8978
* This patch substantially simplifies and cleans up handling of basic blocksChris Lattner2003-10-081-22/+57
| | | | | | | | | | | | | in the bytecode parser. Before we tried to shoehorn basic blocks into the "getValue" code path with other types of values. For a variety of reasons this was a bad idea, so this patch separates it out into its own data structure. This simplifies the code, makes it fit in 80 columns, and is also much faster. In a testcase provided by Bill, which has lots of PHI nodes, this patch speeds up bytecode parsing from taking 6.9s to taking 2.32s. More speedups to follow later. llvm-svn: 8977
* Inline the postResolveValues method. It was poorly named anywayChris Lattner2003-10-081-29/+24
| | | | llvm-svn: 8976
* Various cleanups and simplifications. This speeds up reading a bytecode fileChris Lattner2003-10-081-43/+38
| | | | | | Bill gave me from 8.69s to 6.90s. llvm-svn: 8971
* Transform two methods to return pointers directly instead of returning themChris Lattner2003-10-041-26/+18
| | | | | | as 'by reference' arguments. llvm-svn: 8849
* Use V for values, not D.Chris Lattner2003-10-041-16/+15
| | | | llvm-svn: 8848
* Just rethrow previous exception instead of making a new oneChris Lattner2003-09-261-1/+1
| | | | llvm-svn: 8718
* Suggestions made by Chris:Misha Brukman2003-09-231-10/+14
| | | | | | | | | * Instead of a #define, use inline function * Fix the name on the #define, errr... now inline function to be more logical: it doesn't CHECK the alignment, it PERFORMS the alignment * To get string name of a Type*, use getDescription(), not getName() llvm-svn: 8683
* * Cleaned up code:Misha Brukman2003-09-221-262/+161
| | | | | | | | | - no more passing around a string pointer to set errors - no more returning booleans and checking for errors, we use C++ exceptions * Broke functionality into 2 new classes, one reads from file, one from a stream * Implemented lazy function streaming - the parser can read in a function at-a-time llvm-svn: 8671
* Fixed spelling and grammar.Misha Brukman2003-09-111-1/+1
| | | | llvm-svn: 8489
* Remove a gross hack that was there to support bytecode files that are over a ↵Chris Lattner2003-09-081-1/+1
| | | | | | | | year old. If you still have these suckers laying around, you have GOT to rebuild them. geeze. llvm-svn: 8395
* Allow modules to have 'any' pointer size and endianness. Luckily, we hadChris Lattner2003-08-241-11/+20
| | | | | | some space for extra flags, so we don't need to bump the revision number. llvm-svn: 8118
* Spell `necessary' correctly.Misha Brukman2003-08-181-1/+1
| | | | llvm-svn: 7944
* Merged in autoconf branch. This provides configuration via the autoconfJohn Criswell2003-06-301-5/+5
| | | | | | system. llvm-svn: 7014
* Do not assert fail or crash when a bytecode parse fails!Chris Lattner2003-05-221-0/+1
| | | | llvm-svn: 6284
* Eliminate the uchar typedef, use unsigned char explicitlyChris Lattner2003-05-221-11/+19
| | | | llvm-svn: 6283
* Add support for reading and writing pointersize/endianness to and from bytecodeChris Lattner2003-04-221-0/+5
| | | | llvm-svn: 5839
* Preserve module source information in the ModuleIDChris Lattner2003-04-221-6/+7
| | | | llvm-svn: 5836
* Add support for reading unaligned bytecode buffersChris Lattner2003-04-191-0/+11
| | | | llvm-svn: 5819
* Fix a big bad bug I checked in yesterday :(Chris Lattner2003-04-181-1/+1
| | | | llvm-svn: 5807
* Add support to the bytecode reader/writer for the new linkage typesChris Lattner2003-04-161-14/+41
| | | | llvm-svn: 5790
* Add new linkage types to support a real frontendChris Lattner2003-04-161-3/+10
| | | | llvm-svn: 5786
* * Bug fixes:Chris Lattner2003-03-191-139/+133
| | | | | | | | | | | | | | | | | | | | | - Fix problems where the constant table would not get updated when resolving constants causes other constants to change. Changes to the V2 bytecode format - Null values are implicitly encoded instead of explicitly, this makes things more compact! - More compactly represent ConstantPointerRefs - Bytecode files are represented as: Header|GlobalTypes|GlobalVars/Function Protos|Constants|Functions|SymTab instead of Header|GlobalTypes|Constants|GlobalVars/Function Protos|Functions|SymTab which makes a lot of things simpler. Changes to the reader: - Function loading code is much simpler. We now no longer make function PlaceHolderHelper objects to be replaced with real functions. llvm-svn: 5748
* Infrastructure for more compact bytecode files and REAL support for versioningChris Lattner2003-03-061-6/+40
| | | | llvm-svn: 5716
* s/Method/Function in variable and method namesChris Lattner2003-03-061-11/+11
| | | | llvm-svn: 5715
* Continue simplifying error handling, s/method/functionChris Lattner2003-03-061-57/+45
| | | | llvm-svn: 5714
* Cleanup error handling constructsChris Lattner2003-03-061-37/+47
| | | | llvm-svn: 5713
* Pull common code outChris Lattner2003-03-061-16/+14
| | | | llvm-svn: 5712
* Use the std namespace explicitlyChris Lattner2003-03-061-12/+10
| | | | llvm-svn: 5708
* - Eliminated the deferred symbol table stuff in Module & Function, it reallyChris Lattner2002-11-201-2/+2
| | | | | | wasn't an optimization and it was causing lots of bugs. llvm-svn: 4779
* - Fix a really nasty bug in the bytecode reader that caused it to failChris Lattner2002-10-231-10/+11
| | | | | | | | reading bytecode files with > 255 types in them, but only when optimization is enabled. This was caused by GCC shrinking an enum to a single byte instead of a whole word. llvm-svn: 4266
* Wow, for some reason, when reading bytecode from stdin, we were only readingChris Lattner2002-10-221-1/+1
| | | | | | 4 bytes at a time. Change this to read blocks of 4k llvm-svn: 4265
* There is no way to guarantee that constants are not forward referenced.Chris Lattner2002-10-141-0/+28
| | | | | | | | Handle forward referenced constants in a general way. This fixes bug: Assembler/2002-10-13-ConstantEncodingProblem.llx and allows the SPEC 197.parser benchmark to be built llvm-svn: 4161
* - Change Function's so that their argument list is populated when they areChris Lattner2002-10-131-8/+3
| | | | | | | constructed. Before, external functions would have an empty argument list, now a Function ALWAYS has a populated argument list. llvm-svn: 4149
* Changed uchar to unsigned char in function ParseBytecodeBuffer, because ↵Anand Shukla2002-09-201-1/+1
| | | | | | seemingly it affects opt in uncertain ways :) llvm-svn: 3866
* Emit an obnoxious warning message for bytecode that includes load/storeChris Lattner2002-08-211-1/+2
| | | | | | | instructions that use indexing. Convert them transparently into a pair of instructions. llvm-svn: 3431
* Turn off a debugging optimization that just slows bytecode reading downChris Lattner2002-08-181-1/+1
| | | | llvm-svn: 3377
* - Finally nailed: test/Regression/Assembler/2002-08-16-ConstExprInlined.llxChris Lattner2002-08-171-194/+111
| | | | | | | | | | | | | | - ParseConstantPool was resolving reference to value using the function slot # instead of the global slot #. - Bytecode reader changes: - Remove the failure<> template from Bytecode Reader - Remove extraneous #includes - s/method/function/ a bit - Eliminate the fwdRefs class that just added abstraction where it was not needed, making things more complex. - Use a vector instead of a list for function signatures. llvm-svn: 3366
* added std:: to pairAnand Shukla2002-07-161-0/+1
| | | | llvm-svn: 2917
* Add support for reading ConstantExpr nodes.Vikram S. Adve2002-07-141-32/+111
| | | | | | | Add class ConstantFwdRefs to resolve forward references to constants and to globals (unified old code for globals into this). llvm-svn: 2895
* MEGAPATCH checkin.Chris Lattner2002-06-251-3/+2
| | | | | | For details, See: docs/2002-06-25-MegaPatchInfo.txt llvm-svn: 2778
* Replace all usages of Type::isPointerType with isa<PointerType>Chris Lattner2002-05-061-1/+1
| | | | llvm-svn: 2486
* Eliminate duplicate or unneccesary #include'sChris Lattner2002-04-291-1/+0
| | | | llvm-svn: 2397
* Split ConstantVals.h into Constant.h and Constants.hChris Lattner2002-04-281-1/+1
| | | | llvm-svn: 2378
* Move FunctionArgument out of iOther.h into Argument.h and rename class toChris Lattner2002-04-091-1/+2
| | | | | | be 'Argument' instead of FunctionArgument. llvm-svn: 2216
* s/Method/FunctionChris Lattner2002-03-291-16/+16
| | | | llvm-svn: 2035
* Change references from Method to FunctionChris Lattner2002-03-261-4/+4
| | | | | | change references from MethodARgument to FunctionArgument llvm-svn: 1991
OpenPOWER on IntegriCloud