

@1506

9 years 
mulligan 
changes to costs proof over weekend



@1503

9 years 
mulligan 
inductive type complete



@1502

9 years 
mulligan 
changes to inductive defn



@1501

9 years 
sacerdot 
We must take in account the labelled_p predicate.



@1500

9 years 
sacerdot 
Proof sketch for one of the two main proofs.



@1499

9 years 
mulligan 
part way through main statement transcription



@1498

9 years 
mulligan 
added new file for proof that costs are preserved at asm level



@1497

9 years 
mulligan 
a bit of tidying up, removing dead code, etc.



@1496

9 years 
mulligan 
problem fixed with tactics missing a semicolon to stop greedy parsing



@1495

9 years 
mulligan 
proper calculation of costs



@1494

9 years 
mulligan 
changes to get everything compiling again



@1493

9 years 
mulligan 
finished well labeled check, up to injectivity of the label map



@1489

9 years 
campbell 
Fix up a couple of lemmas affected by the change to add_with_carries.



@1487

9 years 
mulligan 
committing some code for well labelling



@1486

9 years 
mulligan 
finished asm costs



@1485

9 years 
sacerdot 
Less nice definitiion of add_with_carries that avoids a quadratic …



@1484

9 years 
sacerdot 
…



@1482

9 years 
sacerdot 
1. very long standing conflict committed (but don't ask me what the …



@1481

9 years 
sacerdot 
Proof fixed. The new standard library does not index any longer the …



@1480

9 years 
sacerdot 
Proof changed (to use new automation).
BUG FOUND: automation fails if …



@1479

9 years 
boender 
 added insert_lookup_opt
 assembly compiles now



@1478

9 years 
sacerdot 
Almost completed (up to is_finals).



@1476

9 years 
sacerdot 
…



@1475

9 years 
sacerdot 
Towards the two fullexec transition systems that represent …



@1474

9 years 
mulligan 
adding missing asmcosts file for computing the costs of an assembly …



@1472

9 years 
mulligan 
moved proof utils to erasure.ma



@1471

9 years 
mulligan 
finished erasure and generalised so as to work on arbitrary joint programs



@1470

9 years 
mulligan 
finished, pretty ugly though as matita's disambiguation is a …



@1469

9 years 
mulligan 
finished new relabelling for graphs subject to one axiom closed



@1467

9 years 
mulligan 
small change, adding entry and exit labels into the internal function, …



@1466

9 years 
mulligan 
erasure for graph based joint languages almost complete



@1465

9 years 
sacerdot 
Dead code removed.



@1464

9 years 
campbell 
Use unification hints to simplify the graph monotonicity proofs.



@1463

9 years 
mulligan 
added erasure for lin



@1461

9 years 
mulligan 
rewrote erasure for assembly programs



@1460

9 years 
mulligan 
most of cost label erasure for assembly language complete, with one …



@1459

9 years 
boender 
 moved stronger occurs_exactly_once lemma to its proper place in …



@1458

9 years 
mulligan 
added skeleton file for erasure function for joint languages



@1457

9 years 
sacerdot 
Bug fixed: when calling an internal function, the pc block is now set …



@1453

9 years 
sacerdot 
…



@1452

9 years 
sacerdot 
Bug fixed: labels MUST be represented as pointers whose block is the …



@1451

9 years 
sacerdot 
1. All axioms in LIN/semantics.ma closed
2. succ_pc and …



@1450

9 years 
sacerdot 
Disambiguation problem avoided.



@1431

9 years 
sacerdot 
Backend todo (in Italian only, sorry...)



@1430

9 years 
sacerdot 
Bug fixed: push/pop must work on the isp (now added).
Note: the sp is …



@1429

9 years 
sacerdot 
Useless and removed.



@1426

9 years 
boender 
removed axiom



@1425

9 years 
mulligan 
changes to the fixpoint calculation in ertl



@1424

9 years 
sacerdot 
1. fold function over BitVectorTries? moved from ERTLToLTL to …



@1423

9 years 
sacerdot 
 spill no longer used
 BUG IN Interference: generating the destruct …



@1419

9 years 
sacerdot 
All axioms closed.



@1417

9 years 
boender 
 proved that jumps always increase  this should make termination easy



@1416

9 years 
sacerdot 
Maps from hardware registers to beval now implemented in ASM/I8051 (in …



@1415

9 years 
sacerdot 
1. hwreg_store/retrieve no longer returns a res (but it is still …



@1412

9 years 
sacerdot 
Tailcalls (via ids or pointers) to internal functions implemented. …



@1411

9 years 
sacerdot 
1. sem_params2 splitted into sem_params1 + sem_params2 to take out the …



@1410

9 years 
campbell 
Remove a few old workarounds.



@1408

9 years 
sacerdot 
1. Added joint/BEGlobalenvs that is a modification of …



@1404

9 years 
boender 
 reworked + added
 added an axiom to arithmetic, but should be provable



@1401

9 years 
ricciott 
Changes concerning the new behavior of destruct.



@1396

9 years 
sacerdot 
Proof obligation closed.



@1395

9 years 
sacerdot 
1) New versions of pointer_of_beval/beval_of_pointer with a stricter …



@1393

9 years 
boender 
 added invariant for policy trie to assembly
 change (syntax only) …



@1390

9 years 
sacerdot 
All fetch_result implementations have been factorized out, leaving …



@1389

9 years 
sacerdot 
One more axiom closed.



@1388

9 years 
sacerdot 
fetch_result implemented for ERTL. This required a different …



@1387

9 years 
sacerdot 
Further simplification *params1 no longer used.



@1386

9 years 
sacerdot 
Structure of semantic parameters simplified.



@1385

9 years 
sacerdot 
1. fetch_result and pop_frame now takes the genv in input
2. …



@1384

9 years 
sacerdot 
* fetch_ra taken out of pop_frame again since it is used uniformly and …



@1383

9 years 
sacerdot 
Potential bug fixed and bug found: the way pointers and labels are put …



@1382

9 years 
sacerdot 
 succ_pc generalized to return a res (necessary for LIN semantics)
 …



@1381

9 years 
sacerdot 
Old commented out code removed.



@1380

9 years 
sacerdot 
LTL and LIN semantics factorized out in joint_LTL_LIN_semantics.ma. …



@1379

9 years 
sacerdot 
Invariant on LIN code removed. In Paris it was decided that a simpler …



@1378

9 years 
sacerdot 
New file LIN/joint_LTL_LIN.ma to factorize out the syntactic …



@1377

9 years 
sacerdot 
pop_frame now incorporates the fetch_result (that made sense only for …



@1376

9 years 
sacerdot 
Stack deallocation for RTL implemented in pop_frame.



@1372

9 years 
sacerdot 
save_frame now takes the stacksize to allow RTL to allocate the stack frame



@1371

9 years 
sacerdot 
save_frame changed to accept also the formal/actual argument pairs, …



@1369

9 years 
campbell 
Put type information into frontend unary ops.
Slight change to …



@1368

9 years 
sacerdot 
A bug in the clear tactic makes the previous (correct) commit wrong. …



@1367

9 years 
sacerdot 
Proof improvement, still somehow a bit slow.



@1363

9 years 
boender 
 done stuff with create_label_trie



@1359

9 years 
sacerdot 
1. more work on the RTL semantics
2. changes to joint/semantics to …



@1358

9 years 
mulligan 
got rtlabs to rtl compiling, foldi_strong needs examining



@1356

9 years 
mulligan 
deleted redundant directory. added outlines for both reports, and …



@1355

9 years 
sacerdot 
monadic fold_lefti added



@1354

9 years 
sacerdot 
One axiom closed.



@1353

9 years 
sacerdot 
This commit is made necessary by the last Matita change.
Inclusion is …



@1352

9 years 
sacerdot 
This commit is made necessary by the last Matita change.
Inclusion is …



@1351

9 years 
campbell 
Tidy up some loose ends from the invariants branch merge.



@1350

9 years 
sacerdot 
Porting to latest destruct tactic.
Note: the tactics has a few …



@1348

9 years 
sacerdot 
…



@1347

9 years 
campbell 
Remove obsolete definitions.



@1344

9 years 
sacerdot 
Ported to new destruct.



@1343

9 years 
mulligan 
fixed some bugs in the translation



@1342

9 years 
sacerdot 
The new auto is much more powerful.



@1341

9 years 
sacerdot 
Empty directory removed.



@1339

9 years 
sacerdot 
Automation is now stronger.


