

@891

9 years 
campbell 
Revise proofs affected by recent matita change.



@890

9 years 
sacerdot 
Better statement, begin of uniform proof.



@889

9 years 
sacerdot 
Minor changes because of the new, weaker (but much faster) delift.



@888

9 years 
campbell 
Use simplified conditionals in RTLabs, following the prototype.



@887

9 years 
campbell 
Start bringing RTLabs into line with the prototype compiler:
 a …



@886

9 years 
campbell 
Put types into parameter and variable lists in Cminor.
Temporarily …



@885

9 years 
sacerdot 
Proof almost finished, but rewritings are extremely slow.



@884

9 years 
sacerdot 
…



@883

9 years 
sacerdot 
Merged done well.



@882

9 years 
campbell 
Fix up fragile proofs for current version of matita.



@881

9 years 
campbell 
Sort out regions in Cminor to fix Clight to Cminor translation of Ederef.



@880

9 years 
campbell 
Add type information into Cminor.
As a result, the Clight to Cminor …



@879

9 years 
campbell 
Refine "AST" types to include size/signedness information.



@878

9 years 
campbell 
Removal of manually inserted record projections.



@877

9 years 
mulligan 
work from today



@876

9 years 
sacerdot 
…



@875

9 years 
sacerdot 
…



@874

9 years 
sacerdot 
Tactics no longer works.



@873

9 years 
sacerdot 
Script improved. Maybe this requires an svn up of Matita.



@872

9 years 
mulligan 
changes from today, need investigation of reduction machine



@871

9 years 
sacerdot 
…



@870

9 years 
mulligan 
more changes



@869

9 years 
sacerdot 
More progress.



@868

9 years 
mulligan 
more changes



@867

9 years 
mulligan 
changes to assembly proof



@866

9 years 
sacerdot 
Significant improvement in the proof.



@865

9 years 
sacerdot 
Renaming.



@864

9 years 
mulligan 
resolved conflict



@863

9 years 
sacerdot 
Yet another fix to the statement.



@862

9 years 
sacerdot 
…



@861

9 years 
sacerdot 
Statement of the lemma fixed (again!)
Some preliminary work on …



@860

9 years 
sacerdot 
Progress in the proof.



@859

9 years 
mulligan 
more added



@858

9 years 
sacerdot 
If then else notation improved.



@857

9 years 
sacerdot 
Notations.



@856

9 years 
sacerdot 
1. if_then_else is now a notation for match with (to allow Russell to …



@855

9 years 
mulligan 
changes from today



@854

9 years 
mulligan 
commit to avoid conflicts



@853

9 years 
sacerdot 
…



@852

9 years 
mulligan 
foldl_strong outer definition



@851

9 years 
mulligan 
strong foldl added



@850

9 years 
sacerdot 
More informative foldl: foldll.



@849

9 years 
sacerdot 
…



@848

9 years 
sacerdot 
…



@847

9 years 
sacerdot 
Several bugs fixed in Matita.



@846

9 years 
mulligan 
changes



@845

9 years 
sacerdot 
Nightmare…



@844

9 years 
sacerdot 
Useless code removed.



@843

9 years 
sacerdot 
Function moved from Interpret to Status.



@842

9 years 
sacerdot 
Bug fixed.



@841

9 years 
sacerdot 
Minor changes.



@840

9 years 
sacerdot 
sigma defined



@839

9 years 
sacerdot 
More experiments.



@838

9 years 
sacerdot 
Restored.



@837

9 years 
mulligan 
changes complete



@836

9 years 
mulligan 
changes to assembly functions



@835

9 years 
sacerdot 
Old experiments removed.



@834

9 years 
sacerdot 
Russell at work.



@833

9 years 
sacerdot 
Bug fixed to make the file compile.
But the type of the assembly …



@832

9 years 
mulligan 
work from today



@831

9 years 
sacerdot 
Progress in proofs.



@829

9 years 
sacerdot 
…



@828

9 years 
sacerdot 
Proof statement.



@827

9 years 
sacerdot 
The preamble is now part of the PseudoStatus?.



@826

9 years 
mulligan 
start of proof



@825

9 years 
mulligan 
lots of refactoring, finally got something to prove



@824

9 years 
campbell 
Some work on showing that casts around integer operations can be removed.



@823

9 years 
mulligan 
added new file for proof of correctness of pseudoassembly translation



@822

9 years 
mulligan 
removed all axioms



@821

9 years 
mulligan 
changes to introduce pseudostatus



@820

9 years 
mulligan 
changes to get the semantics of pseudoassembly working



@816

9 years 
campbell 
Clight to Cminor compilation, modulo switch statements, temporary …



@799

9 years 
mulligan 
more changes.



@798

9 years 
campbell 
Fix usual matita tactic mistake.



@797

9 years 
campbell 
Add error messages wherever the error monad is used.
Sticks to …



@795

9 years 
mulligan 
Changes from this morning.



@793

9 years 
mulligan 
Work from today on rtlabs > rtl pass.



@790

9 years 
campbell 
A little tidying: get rid of requirement for jmeq in Mem.ma, remove …



@789

9 years 
mulligan 
More work on rtlabs > rtl pass.



@786

9 years 
campbell 
A version of the clight matita term printer for the current prototype.



@784

9 years 
mulligan 
Added missing tailcall simplification file.



@783

9 years 
mulligan 
rtl to ertl pass complete (modulo some straightforward axioms that …



@782

9 years 
mulligan 
More work on rtlertl pass from today, plus resolved conflict.



@781

9 years 
campbell 
Implement labelling pass for Clight.



@780

9 years 
campbell 
Properly update set of registers that are used for pointers in Cminor …



@779

9 years 
campbell 
Add merging of tries and identifier sets (based on Dominic's earlier …



@778

9 years 
mulligan 
moved register set into correct place



@777

9 years 
mulligan 
Lots of work on RTL to ERTL pass from today.



@776

9 years 
campbell 
Fix up some minor null pointer issues in Clight.
Add corresponding …



@775

9 years 
campbell 
A few useful definitions for when RTLabs programs fail.



@774

9 years 
campbell 
Separate out the different forms of addition and subtraction in the …



@773

9 years 
campbell 
Report a couple of minor changes from the prototype compiler.



@772

9 years 
campbell 
Implement proper support for RTLabs addressing modes.



@771

9 years 
campbell 
Implement switch statements in Cminor > RTLabs phase



@770

9 years 
campbell 
Clight and Cminor examples for switch statement.



@768

9 years 
campbell 
Make Cminor tests test translation to RTLabs.



@767

9 years 
campbell 
Use variable shadowing as a poor man's state monad in cminor to rtlabs …



@766

9 years 
campbell 
Most of the Cminor to RTLabs stage.
Is buggy, generates inefficient …



@765

9 years 
campbell 
Remove superfluous register in RTLabs return statements.
Also fix up …



@764

9 years 
campbell 
Start Cminor to RTLabs phase.
Includes some syntax for matching …


