Commit message (Collapse) | Author | Age | Files | Lines | ||
---|---|---|---|---|---|---|
... | ||||||
* | - Added MCSubtargetInfo to capture subtarget features and scheduling | Evan Cheng | 2011-07-01 | 1 | -68/+132 | |
| | | | | | | | | | itineraries. - Refactor TargetSubtarget to be based on MCSubtargetInfo. - Change tablegen generated subtarget info to initialize MCSubtargetInfo and hide more details from targets. llvm-svn: 134257 | |||||
* | Fix the ridiculous SubtargetFeatures API where it implicitly expects CPU name to | Evan Cheng | 2011-06-30 | 1 | -6/+6 | |
| | | | | | | | | | | be the first encoded as the first feature. It then uses the CPU name to look up features / scheduling itineray even though clients know full well the CPU name being used to query these properties. The fix is to just have the clients explictly pass the CPU name! llvm-svn: 134127 | |||||
* | Sink SubtargetFeature and TargetInstrItineraries (renamed ↵ | Evan Cheng | 2011-06-29 | 1 | -2/+2 | |
| | | | | | | MCInstrItineraries) into MC. llvm-svn: 134049 | |||||
* | Increase SubtargetFeatureKV Value and Implies fields to 64 bits since some ↵ | Evan Cheng | 2011-04-15 | 1 | -6/+12 | |
| | | | | | | targets are getting very close to 32 subtarget features. Also teach tablegen to error when there are more than 64 features to guard against undefined behavior. rdar://9282332 llvm-svn: 129590 | |||||
* | Add annotations to tablegen-generated processor itineraries, or replace them ↵ | Andrew Trick | 2011-04-01 | 1 | -15/+19 | |
| | | | | | | with something meaningful. I want to be able to read and debug the generated tables. llvm-svn: 128703 | |||||
* | whitespace | Andrew Trick | 2011-04-01 | 1 | -65/+65 | |
| | | | | llvm-svn: 128701 | |||||
* | eliminate the Records global variable, patch by Garrison Venn! | Chris Lattner | 2010-12-13 | 1 | -1/+1 | |
| | | | | llvm-svn: 121659 | |||||
* | Model Cortex-a9 load to SUB, RSB, ADD, ADC, SBC, RSC, CMN, MVN, or CMP | Evan Cheng | 2010-09-29 | 1 | -16/+21 | |
| | | | | | | pipeline forwarding path. llvm-svn: 115098 | |||||
* | Add support to model pipeline bypass / forwarding. | Evan Cheng | 2010-09-28 | 1 | -1/+46 | |
| | | | | llvm-svn: 115005 | |||||
* | For each instruction itinerary class, specify the number of micro-ops each | Evan Cheng | 2010-09-09 | 1 | -18/+26 | |
| | | | | | | | | | | instruction in the class would be decoded to. Or zero if the number of uOPs must be determined dynamically. This will be used to determine the cost-effectiveness of predicating a micro-coded instruction. llvm-svn: 113513 | |||||
* | Make processor FUs unique for given itinerary. This extends the limit of 32 | Anton Korobeynikov | 2010-04-18 | 1 | -7/+28 | |
| | | | | | | | FU per CPU arch to 32 per intinerary allowing precise modelling of quite complex pipelines in the future. llvm-svn: 101754 | |||||
* | Initial support for different kinds of FU reservation. | Anton Korobeynikov | 2010-04-07 | 1 | -3/+6 | |
| | | | | llvm-svn: 100645 | |||||
* | Have TableGen emit code that uses dbgs() rather than errs(). | David Greene | 2010-01-05 | 1 | -2/+2 | |
| | | | | llvm-svn: 92738 | |||||
* | Show command-line args and features passed into backend in debug output. ↵ | Sandeep Patel | 2009-11-11 | 1 | -0/+4 | |
| | | | | | | Approved by Evan Cheng. llvm-svn: 86797 | |||||
* | Make the end-of-itinerary mark explicit. Some cleanup. | David Goodwin | 2009-09-24 | 1 | -5/+3 | |
| | | | | llvm-svn: 82709 | |||||
* | Extend the instruction itinerary model to include the ability to indicate ↵ | David Goodwin | 2009-08-17 | 1 | -39/+102 | |
| | | | | | | the def and use cycle for each operand. This additional information is optional, so existing itineraries do not need to be changed. llvm-svn: 79247 | |||||
* | Enhance the InstrStage object to enable the specification of an Itinerary ↵ | David Goodwin | 2009-08-12 | 1 | -4/+7 | |
| | | | | | | with overlapping stages. The default is to maintain the current behavior that the "next" stage immediately follows the previous one. llvm-svn: 78827 | |||||
* | Replace std::iostreams with raw_ostream in TableGen. | Daniel Dunbar | 2009-07-03 | 1 | -10/+10 | |
| | | | | | | | | - Sorry, I can't help myself. - No intended functionality change. llvm-svn: 74742 | |||||
* | Propagate CPU string out of SubtargetFeatures | Anton Korobeynikov | 2009-05-23 | 1 | -4/+5 | |
| | | | | llvm-svn: 72335 | |||||
* | fix a few spelling errors and typos | Jim Grosbach | 2009-03-26 | 1 | -1/+1 | |
| | | | | llvm-svn: 67758 | |||||
* | lib/Target/SubtargetFeature.cpp asserts that the FeatureKV[] table be sorted | Jim Grosbach | 2008-09-11 | 1 | -1/+1 | |
| | | | | | | | | | | | by its first field, but TableGen doesn't actually enforce creating it that way. TableGen sorts the records that will be used to create it by the names of the records, not the Name field of those records. This patch corrects the sort to use the "Name" field of the record as the sort key. llvm-svn: 56106 | |||||
* | Stabilize 'getDwarfRegNumFull' output to not depend on random memory | Chris Lattner | 2008-08-26 | 1 | -18/+0 | |
| | | | | | | orders, part of PR2590 llvm-svn: 55359 | |||||
* | Fix generation of multi-stage instruction itineraries. Patch by | Chris Lattner | 2008-04-06 | 1 | -2/+5 | |
| | | | | | | giuma.cordes@gmail.com llvm-svn: 49276 | |||||
* | A quick nm audit turned up several fixed tables and objects that were | Dan Gohman | 2008-03-25 | 1 | -3/+3 | |
| | | | | | | | marked read-write. Use const so that they can be allocated in a read-only segment. llvm-svn: 48800 | |||||
* | Rewrite tblgen handling of subtarget features so | Dale Johannesen | 2008-02-14 | 1 | -2/+6 | |
| | | | | | | | | | | it follows the order of the enum, not alphabetical. The motivation is to make -mattr=+ssse3,+sse41 select SSE41 as it ought to. Added "ignored" enum values of 0 to PPC and SPU to avoid compiler warnings. llvm-svn: 47143 | |||||
* | remove attributions from utils. | Chris Lattner | 2007-12-29 | 1 | -2/+2 | |
| | | | | llvm-svn: 45419 | |||||
* | Add an "implies" field to features. This indicates that, if the current | Bill Wendling | 2007-05-04 | 1 | -46/+54 | |
| | | | | | | | | feature is set, then the features in the implied list should be set also. The opposite is also enforced: if a feature in the implied list isn't set, then the feature that owns that implies list shouldn't be set either. llvm-svn: 36756 | |||||
* | Fix generation of certain scheduler itineraries. | Christopher Lamb | 2007-04-22 | 1 | -1/+2 | |
| | | | | llvm-svn: 36338 | |||||
* | Ignore entries with blank names. | Jim Laskey | 2006-12-12 | 1 | -2/+4 | |
| | | | | llvm-svn: 32491 | |||||
* | Rollback changes to take a different tack. | Jim Laskey | 2006-12-12 | 1 | -1/+0 | |
| | | | | llvm-svn: 32488 | |||||
* | Honor the command line specification for machine type. | Jim Laskey | 2006-12-12 | 1 | -0/+1 | |
| | | | | llvm-svn: 32483 | |||||
* | Don't write out variables that are never used. | Reid Spencer | 2006-11-03 | 1 | -3/+0 | |
| | | | | llvm-svn: 31396 | |||||
* | For PR786: | Reid Spencer | 2006-11-02 | 1 | -1/+0 | |
| | | | | | | Remove unused variables. llvm-svn: 31381 | |||||
* | Don't make zero-sized static arrays | Chris Lattner | 2006-05-24 | 1 | -1/+1 | |
| | | | | llvm-svn: 28448 | |||||
* | remove out of date comment | Chris Lattner | 2006-03-03 | 1 | -2/+1 | |
| | | | | llvm-svn: 26492 | |||||
* | Subtarget feature can now set any variable to any value | Evan Cheng | 2006-01-27 | 1 | -3/+4 | |
| | | | | llvm-svn: 25678 | |||||
* | There is at least a 'noitinerary' itinerary now | Chris Lattner | 2006-01-27 | 1 | -1/+1 | |
| | | | | llvm-svn: 25671 | |||||
* | almost got the HP-UX tester up.. :) | Duraid Madina | 2005-12-30 | 1 | -6/+6 | |
| | | | | llvm-svn: 25051 | |||||
* | 1. Remove ranges from itinerary data. | Jim Laskey | 2005-11-03 | 1 | -9/+11 | |
| | | | | | | 2. Tidy up the subtarget emittined code. llvm-svn: 24172 | |||||
* | Allow itineraries to be passed through the Target Machine. | Jim Laskey | 2005-11-01 | 1 | -12/+33 | |
| | | | | llvm-svn: 24139 | |||||
* | Generate cpu to itinerary map. | Jim Laskey | 2005-10-31 | 1 | -7/+61 | |
| | | | | llvm-svn: 24121 | |||||
* | Rename Record::getValueAsListDef to getValueAsListOfDefs, to more accurately | Chris Lattner | 2005-10-28 | 1 | -4/+5 | |
| | | | | | | | reflect what it is. Convert some more code over to use it. llvm-svn: 24072 | |||||
* | Removed Mr. Smith from the code. | Jim Laskey | 2005-10-28 | 1 | -160/+136 | |
| | | | | llvm-svn: 24070 | |||||
* | Add some commentary. | Jim Laskey | 2005-10-28 | 1 | -19/+119 | |
| | | | | llvm-svn: 24055 | |||||
* | Keep Visual Studio happy. | Jeff Cohen | 2005-10-28 | 1 | -0/+1 | |
| | | | | llvm-svn: 24052 | |||||
* | Now generating instruction itineraries for scheduling. Not my best work, but... | Jim Laskey | 2005-10-27 | 1 | -31/+190 | |
| | | | | llvm-svn: 24050 | |||||
* | Simplify. | Jim Laskey | 2005-10-26 | 1 | -6/+1 | |
| | | | | llvm-svn: 24015 | |||||
* | Give full control of subtarget features over to table generated code. | Jim Laskey | 2005-10-26 | 1 | -14/+62 | |
| | | | | llvm-svn: 24013 | |||||
* | Refactored to make room for more stuff (scheduling info.) | Jim Laskey | 2005-10-25 | 1 | -77/+97 | |
| | | | | llvm-svn: 23975 | |||||
* | Add the needed #include, emit enums with the sizes of tables, remove | Chris Lattner | 2005-10-23 | 1 | -9/+12 | |
| | | | | | | definitions from the LLVM namespace, since they are all static. llvm-svn: 23907 |