Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Now working, tests show a decrease in spillage | nicolas.nardino | 2021-06-16 | 1 | -14/+14 |
| | | | | | Should still find a proper way to treat the case mentioned in earlier commits | ||||
* | kinda fixed | nicolas.nardino | 2021-06-15 | 1 | -3/+17 |
| | | | | | | | | Spills are definitely reduced, but lots of arbitrary in there: See previous commit: need to determine what to do if pressure is too high but no schedulable instruction can reduce it. For now, advance time for at most 5 cycles, if still no suitable instruction, go back to CSP | ||||
* | fixing | nicolas.nardino | 2021-06-15 | 1 | -12/+35 |
| | | | | | Still need to find what to do when pressure is high but there are no instructions available that decrease it | ||||
* | was very wrong, fixing | nicolas.nardino | 2021-06-14 | 1 | -9/+65 |
| | |||||
* | Work on new schedluer | nicolas.nardino | 2021-06-10 | 1 | -9/+17 |
| | | | | | | Renamed a test file, wrote function to compute pressure deltas, Still need to pass the info in some way; beginning of the actual scheduler function | ||||
* | Fix RTLpathScheduleraux.get_live_regs_entry | nicolas.nardino | 2021-06-08 | 1 | -2/+3 |
| | |||||
* | Add some tests | nicolas.nardino | 2021-06-08 | 1 | -1/+1 |
| | |||||
* | Adding debug info | nicolas.nardino | 2021-06-06 | 1 | -42/+53 |
| | |||||
* | Fixing scope error, and non-exhaustive pattern matching | nicolas.nardino | 2021-06-06 | 1 | -5/+6 |
| | |||||
* | function written, now needs testing | nicolas.nardino | 2021-06-05 | 1 | -10/+40 |
| | |||||
* | computing live regs at sb entry from its live output regs | nicolas.nardino | 2021-06-04 | 1 | -3/+21 |
| | |||||
* | Passing info of live regs to scheduler: beginning | nicolas.nardino | 2021-06-04 | 1 | -1/+8 |
| | |||||
* | Remove flags | Léo Gourdin | 2021-04-09 | 1 | -3/+1 |
| | |||||
* | Important commit on expansions' mini CSE, and a draft for addptrofs | Léo Gourdin | 2021-04-06 | 1 | -4/+2 |
| | |||||
* | Remove first nop when doing expansion | Léo Gourdin | 2021-03-21 | 1 | -0/+1 |
| | |||||
* | Merge remote-tracking branch 'origin/riscv-work' into riscv-work-fpinit-stillexp | Léo Gourdin | 2021-03-06 | 1 | -4/+4 |
|\ | |||||
| * | Adding a mini CSE pass in the expansion oracle | Léo Gourdin | 2021-03-06 | 1 | -3/+4 |
| | | |||||
* | | [Admitted checker] Oracle expansion for float/float32 constant init | Léo Gourdin | 2021-03-02 | 1 | -1/+1 |
|/ | |||||
* | fix ci ? | Léo Gourdin | 2021-03-02 | 1 | -1/+1 |
| | |||||
* | Try to save values in virtual registers during expansion | Léo Gourdin | 2021-03-01 | 1 | -1/+1 |
| | |||||
* | Adding a compiler option -fexpanse-rtlcond | Léo Gourdin | 2021-02-16 | 1 | -1/+3 |
| | |||||
* | [Admitted checker] Checker expansion for reg Ocmp (without scratch) | Léo Gourdin | 2021-02-10 | 1 | -2/+3 |
| | |||||
* | Adding pathmap psize modification during expansion oracle | Léo Gourdin | 2021-02-08 | 1 | -5/+5 |
| | |||||
* | cond and branches expanded | Léo Gourdin | 2021-02-06 | 1 | -2/+3 |
| | |||||
* | Expansion of Ccompimm in RTL [Admitted checker] | Léo Gourdin | 2021-02-02 | 1 | -19/+15 |
| | |||||
* | Uniformizing a couple of debug print functions | Cyril SIX | 2020-12-17 | 1 | -42/+15 |
| | |||||
* | Fixing too many loads being NOTRAP | Cyril SIX | 2020-12-17 | 1 | -3/+9 |
| | |||||
* | Partially fixing turning loads into non trap | Cyril SIX | 2020-12-16 | 1 | -22/+56 |
| | | | | There are now too many loads turned into non trap. To be investigated | ||||
* | Cleanup | Cyril SIX | 2020-12-16 | 1 | -83/+0 |
| | |||||
* | Turning loads into non-trapping when necessary | Cyril SIX | 2020-12-15 | 1 | -1/+34 |
| | |||||
* | disable debug printing in scheduler | David Monniaux | 2020-11-04 | 1 | -4/+4 |
| | |||||
* | just missing OpWeights for AARCH64 | David Monniaux | 2020-09-16 | 1 | -0/+368 |