jeltz/heptagon
1
0
Fork 0
Code Issues Pull requests Releases Activity
633 commits 2 branches 0 tags 6.9 MiB
Commit graph

216 commits

Author SHA1 Message Date
Adrien Guatto
2283fcdfaa Properly handle fused outputs. 2011-07-08 11:56:38 +02:00
Adrien Guatto
3657018861 Tomato: finer equivalence classes for Eapp and 
Eiterator when a reset variable is present.
 2011-07-08 11:56:38 +02:00
Adrien Guatto
3c5bb4e8b7 Tomato working with clocks and when. 2011-07-08 11:56:37 +02:00
Adrien Guatto
891174d73c All-new and fresh tomato! 
Not working ATM:
  * tuples
  * when
 2011-07-08 11:56:37 +02:00
Léonard Gérard
0518ecafe6 Merge branch 'multidimensional' into decade 
Conflicts:
	compiler/obc/c/cgen.ml
	compiler/obc/java/java_main.ml
 2011-07-08 11:36:46 +02:00
Brice Gelineau
42c2936040 Bugfix for the clocking analysis of iterators 2011-07-06 15:56:56 +02:00
Adrien Guatto
cedcabfbc4 Fix memory normalization: the correct clock was 
missing.
 2011-07-06 11:50:02 +02:00
Léonard Gérard
5837f3906f Multidimensional iterators in java. 2011-06-28 17:50:42 +02:00
Brice Gelineau
c70d34ec06 Preliminary version of multidimensional iterators. 2011-06-28 17:50:42 +02:00
Léonard Gérard
83b0182874 Merge branch 'clocked_inputs' into decade 
Conflicts:
	compiler/heptagon/analysis/typing.ml
	compiler/heptagon/parsing/hept_scoping.ml
	compiler/heptagon/parsing/hept_static_scoping.ml
	compiler/main/mls2obc.ml
	compiler/obc/c/cmain.ml
 2011-06-28 14:46:43 +02:00
Léonard Gérard
108981c0eb Static constraints in the source. Equal removed 
( = ) in pervasives is a stub, it will be typed in a polymorphic way.
This is necessary to have a simple way to transform exp into a static_exp
even when there is the = operator.
 2011-06-09 14:53:40 +02:00
Léonard Gérard
46ed39b7b3 Remove some warnings. 2011-06-09 14:40:31 +02:00
Brice Gelineau
938e8897b2 n-ary Sarray_power Earray_fill 2011-05-26 18:43:30 +02:00
Léonard Gérard
fe74030f78 Fix clocking of iterators. 2011-05-26 16:08:59 +02:00
Léonard Gérard
b73e6502a6 resets are now without constraint. 2011-05-26 15:56:59 +02:00
Léonard Gérard
d7d7552be4 Fix printing and parsing of clocks 2011-05-26 15:56:59 +02:00
Léonard Gérard
fd00f099f5 correct reset and level clock 2011-05-26 15:56:59 +02:00
Léonard Gérard
2a2b363bf7 Real asynchronous resets. 2011-05-26 15:56:59 +02:00
Léonard Gérard
fbfa6eda1a Add Ewhen to the minils ast, And clocking of iterators. 
Ewhen is now the only case of possible recursion for minils exps.
This add was motivated by equations like :
(y,z) = f(x) when c
This equation to be correctly normalized in minils before needed :
y',z' = f(x)
y = y' when c
z = z' when c
But this new variables where needless since the final translation of when c
is the identity.
 2011-05-26 15:56:59 +02:00
Léonard Gérard
87dc76f113 Code formating fixes (typos, tabs, spaces at endofline,...) 2011-05-26 15:56:59 +02:00
Léonard Gérard
de49540a76 clean signature printing. 2011-05-26 15:56:59 +02:00
Léonard Gérard
74cc367a0e correct base_clock and signature. 2011-05-26 15:56:59 +02:00
Léonard Gérard
9a17648516 good clocks in signatures 2011-05-26 15:56:59 +02:00
Leonard Gerard
88497a2da3 parsing clocks. 2011-05-26 15:56:58 +02:00
Leonard Gerard
a8215c8083 compile ! nearly all tests passed. 2011-05-26 15:56:58 +02:00
Gwenal Delaval
bf4d80c5a3 Correction of Mls_compare for new AST 2011-05-11 14:25:45 +02:00
Cédric Pasteur
f110727568 Better fix for Normalize_mem 
We have to modify the var_dec and arg as they may
have other fields that need to be kept
 2011-05-02 11:20:37 +02:00
Léonard Gérard
9fa8e7e6ff mk_equation in hept computes the stateful field. 2011-04-29 15:04:47 +02:00
Léonard Gérard
b8ee46d58f fix normalize_mem 2011-04-27 14:29:21 +02:00
Léonard Gérard
6b86b86e9d really fix normalize_mem. 
in the old fix was missing the renaming, and there was wrong equations
order :
f () returns x
[here rename x by mem_x]
mem_x = fby ...
x = mem_x

so the simplest was :
f () returns out_x
var x;
[ nothing to do]
out_x = x
x = fby ...

pay attention to the order of equations since we are after the
scheduling..
probably it should be done before the scheduling anyway ?
 2011-04-20 20:47:06 +02:00
Cédric Pasteur
2757e7c1bf Fixed normalized_mem 
It inverted the order of equations
 2011-04-20 16:23:44 +02:00
Léonard Gérard
d6240e1c67 100 and java scalarized 2011-04-19 18:45:56 +02:00
Cédric Pasteur
802178fb28 Added a pass to normalize memories 
This adds a copy for outputs that are also
memories.
 2011-04-19 17:19:40 +02:00
Cédric Pasteur
740114764c Fixed stupid bug in the order of args 2011-04-19 10:38:48 +02:00
Cédric Pasteur
a2303ec7e1 Other compile fixes 
The java backend is disabled temporarily
 2011-04-19 09:49:00 +02:00
Leonard Gerard
8da5ce4648 no order in declarations 2011-04-18 19:20:35 +02:00
Leonard Gerard
c0602c6df6 etuple removed 2011-04-18 16:09:07 +02:00
Cédric Pasteur
c6aa63b444 Heptc compiles again 
AFor now expects expressions instead of static exps
 2011-04-18 15:38:42 +02:00
Léonard Gérard
5d2f7dfa85 jeudi soir. still on mls2obc. 2011-04-14 20:14:21 +02:00
Cédric Pasteur
2f346f873c Control optimization as a pass on Obc 
There is now a obc_compiler.ml file, as in hept
and minils.
 2011-04-14 11:53:39 +02:00
Cédric Pasteur
6adb45c3ad Normalize in Heptagon compiles 2011-04-13 16:21:28 +02:00
Cédric Pasteur
611c94bbbd Itfusion moved to heptagon 2011-04-13 15:10:15 +02:00
Cédric Pasteur
cbf92beba2 First try at a normalization in Heptagon 
I can't see if it compiles yet
 2011-04-13 14:40:06 +02:00
Cédric Pasteur
2c7b609d2e Forme normale Minils 2011-04-12 14:07:05 +02:00
Cédric Pasteur
2fdf2855d3 Added a new truncated select operator 
a[>e<] returns the element in the array at index
e, a[0] if e < 0 and a[n-1] if e >= n
 2011-03-22 22:12:59 +01:00
Cédric Pasteur
99eeacbceb Added mapi iterator 
The last argument of the iterated function is the
index of the element in the array.
 2011-03-22 09:28:41 +01:00
Léonard Gérard
b8b16a7355 blop 2011-03-21 17:41:00 +01:00
Léonard Gérard
02730b8a0b refactoring static evaluation. 2011-03-21 17:41:00 +01:00
Cédric Pasteur
b1b8e103f2 Added partial application for iterators 
For instance:
  ... = map<<n>> (f<<se>>)((t1, t1'))(t2, t3)

is translated to:
  for(int i =...)
    ... = f(t1, t1', t2[i], t3[i])
 2011-03-21 17:22:03 +01:00
Cédric Pasteur
35775c4131 C backend ported to recent API changes in Obc 2011-03-09 09:46:00 +01:00