Commit message (Collapse) | Author | Age | Files | Lines | ||
---|---|---|---|---|---|---|
... | ||||||
| * | python: Infrastructure for generic arch Python API | David Shah | 2019-04-02 | 2 | -0/+79 | |
| | | | | | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | | timing_opt: Add locks to optimise() | Alyssa Milburn | 2019-04-16 | 1 | -0/+2 | |
| | | | | | | | | Signed-off-by: Alyssa Milburn <amilburn@zall.org> | |||||
* | | clangformat | David Shah | 2019-04-08 | 1 | -6/+9 | |
| | | | | | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | | placer1: Check correctness of incremental updates with --debug | David Shah | 2019-04-08 | 1 | -0/+18 | |
| | | | | | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | | placer1: Improve incremental bounding box updates | David Shah | 2019-04-07 | 1 | -38/+213 | |
| | | | | | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | | common: avoid std::ofstream copy | Janos Farkas | 2019-04-02 | 1 | -2/+2 | |
|/ | | | | | Using a copy constructor to set the logfile is the only thing that stops compilation with the libstdc++ shipping with gcc 4.8 (maybe 4.7) | |||||
* | HeAP: Don't call Eigen if system is empty | David Shah | 2019-04-01 | 1 | -1/+2 | |
| | | | | | | Fixes #259 Signed-off-by: David Shah <dave@ds0.me> | |||||
* | Fix broken header guard | whitequark | 2019-03-25 | 1 | -1/+1 | |
| | ||||||
* | HeAP: Make log output more consistent | David Shah | 2019-03-25 | 2 | -5/+9 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Add missing newline | David Shah | 2019-03-25 | 1 | -1/+1 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | placer1: Restore old weighting in budget-based mode | David Shah | 2019-03-24 | 1 | -1/+1 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Per-iteration output all on one line | David Shah | 2019-03-24 | 1 | -7/+5 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | Add --placer option and refactor placer selection | David Shah | 2019-03-24 | 2 | -8/+38 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | placer1: Fix regression moving chained cells pre-legalise | David Shah | 2019-03-23 | 1 | -2/+3 | |
| | | | | Signed-off-by: David Shah <davey1576@gmail.com> | |||||
* | clangformat | David Shah | 2019-03-22 | 2 | -4/+3 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | placer1: Only get criticalities when in timing-driven mode | David Shah | 2019-03-22 | 1 | -1/+1 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | common: Add early return path to getNetinfoRouteDelay for fully unrouted nets | David Shah | 2019-03-22 | 1 | -0/+3 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Add PlacerHeapCfg | David Shah | 2019-03-22 | 3 | -9/+37 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Make HeAP placer optional | David Shah | 2019-03-22 | 1 | -0/+16 | |
| | | | | | | | | | | | | | A CMake option 'BUILD_HEAP' (default on) configures building of the HeAP placer and the associated Eigen3 dependency. Default for the iCE40 is SA placer, with --heap-placer to use HeAP Default for the ECP5 is HeAP placer, as SA placer can take 1hr+ for large ECP5 designs and HeAP tends to give better QoR. --sa-placer can be used to use SA instead, and auto-fallback to SA if HeAP not built. Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Switching from TAUCS to Eigen | David Shah | 2019-03-22 | 3 | -123/+28 | |
| | | | | Signed-off-by: David Shah <davey1576@gmail.com> | |||||
* | HeAP: Fix regression | David Shah | 2019-03-22 | 1 | -0/+2 | |
| | | | | Signed-off-by: David Shah <davey1576@gmail.com> | |||||
* | HeAP: Fix occupancy count | David Shah | 2019-03-22 | 1 | -2/+3 | |
| | | | | Signed-off-by: David Shah <davey1576@gmail.com> | |||||
* | HeAP: Legaliser fixes | David Shah | 2019-03-22 | 1 | -2/+2 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Avoid getting stuck in legaliser ripup | David Shah | 2019-03-22 | 1 | -3/+26 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Chain support | David Shah | 2019-03-22 | 2 | -8/+103 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: tidying up | David Shah | 2019-03-22 | 2 | -115/+85 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Make strict legalisation wirelength driven where needed | David Shah | 2019-03-22 | 1 | -6/+62 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Use for ECP5 as well as iCE40 | David Shah | 2019-03-22 | 1 | -9/+20 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Implement 'all+rotate' HeAP strategy | David Shah | 2019-03-22 | 3 | -30/+56 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Weight arcs by criticality for timing-driven placement | David Shah | 2019-03-22 | 1 | -6/+18 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Add multithreading | David Shah | 2019-03-22 | 1 | -27/+24 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: fine tuning | David Shah | 2019-03-22 | 1 | -28/+100 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Add performance counters | David Shah | 2019-03-22 | 1 | -3/+27 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Add SA-based iterative refinement after AP | David Shah | 2019-03-22 | 3 | -87/+143 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Spreading working acceptably | David Shah | 2019-03-22 | 1 | -64/+51 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Make cut-based spreading recursive | David Shah | 2019-03-22 | 1 | -17/+56 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Cut peturbation, binning and intra-bin linear spreading | David Shah | 2019-03-22 | 1 | -19/+132 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Cut finder for spreading | David Shah | 2019-03-22 | 1 | -18/+130 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Region finder for spreading and strict legaliser | David Shah | 2019-03-22 | 1 | -36/+454 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Fix how initial placement handles chains | David Shah | 2019-03-22 | 1 | -9/+67 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Successful solver convergance | David Shah | 2019-03-22 | 3 | -27/+115 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Add TAUCS wrapper and integration | David Shah | 2019-03-22 | 4 | -1/+166 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | HeAP: Initial infrastructure | David Shah | 2019-03-22 | 1 | -0/+366 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | placer1: Legalise after reaching a diameter, not temperature | David Shah | 2019-03-22 | 1 | -6/+4 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | placer1: Add (currently-unused) option for fanout threshold for timing cost calc | David Shah | 2019-03-22 | 2 | -3/+5 | |
| | | | | Signed-off-by: David Shah <davey1576@gmail.com> | |||||
* | placer1: Implement non-timing-driven mode | David Shah | 2019-03-22 | 1 | -26/+31 | |
| | | | | Signed-off-by: David Shah <davey1576@gmail.com> | |||||
* | placer1: Improve handling of regions | David Shah | 2019-03-22 | 1 | -4/+53 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | pycontainers: Implement __contains__ | David Shah | 2019-03-22 | 1 | -0/+14 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | Add Python helper functions for floorplanning | David Shah | 2019-03-22 | 4 | -1/+67 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> | |||||
* | placer1: Consider regions during placement | David Shah | 2019-03-22 | 3 | -16/+38 | |
| | | | | Signed-off-by: David Shah <dave@ds0.me> |