-
Erwan Jahier authored
Also provide a default version of those macros using memcpy.
Erwan Jahier authoredAlso provide a default version of those macros using memcpy.
todo.org_archive 40.16 KiB
# -*- mode: org -*-
Archived entries from file /home/jahier/lus2lic/todo.org
* TODO Mettre à jour le tout nouveau file:README.org wrt src
SCHEDULED: <2012-12-10 Mon>
- State "TODO" from "" [2012-12-10 Mon 17:02]
:PROPERTIES:
:ARCHIVE_TIME: 2012-12-13 Thu 15:37
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Urgent
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO Renommer tous les modules et faire du ménage
SCHEDULED: <2012-12-10 Mon>
- State "TODO" from "" [2012-12-10 Mon 16:55]
:PROPERTIES:
:ARCHIVE_TIME: 2012-12-13 Thu 15:37
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Urgent
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO rebrancher le nodeExpand.ml et structArrayExpand.ml
SCHEDULED: <2012-12-20 Thu>
- State "TODO" from "" [2012-12-10 Mon 16:55]
:PROPERTIES:
:ARCHIVE_TIME: 2012-12-20 Thu 17:23
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Urgent
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
file:src/l2lExpandNodes.mli
file:src/l2lExpandArrays.mli
que Pascal les a débranché lors de son ménage d'été.
* TODO Pascal a shunté mon LicName dans split. Avait-il (une bonne) raison ?
SCHEDULED: <2012-12-20 Thu>
- State "TODO" from "" [2012-12-17 Mon 16:37]
:PROPERTIES:
:ARCHIVE_TIME: 2012-12-20 Thu 17:23
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Urgent
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
par ex, file:~/lus2lic/src/l2lExpandArrays.ml::50
dois-je faire comme lui ou comme avant ?
* TODO le mode ec genere tous les noeuds (ce que n'aime pas ec2c)
- State "TODO" from "" [2013-01-11 Fri 09:33]
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-11 Fri 18:15
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Regressions introduites par Pascal
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO il ne faut pas générer l'alias de type en ec
- State "TODO" from "" [2013-01-11 Fri 09:41]
:PROPERTIES: :ARCHIVE_TIME: 2013-01-11 Fri 18:15
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Regressions introduites par Pascal
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
file:/tmp/t0.ec
* TODO y'a un List.fold_left2 qui plante
- State "TODO" from "" [2013-01-11 Fri 10:08]
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-16 Wed 16:30
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Regressions introduites par Pascal
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
oops: lus2lic internal error
File "objlinux/l2lExpandNodes.ml", line 131, column 4
when compiling lustre program should_work/test.lus
file:test/should_work/test.lus
file:~/lus2lic/src/l2lExpandNodes.ml::131
* TODO Ya un probleme avec ce fichier lustre (compilait avant)
- State "TODO" from "" [2013-01-07 Mon 16:20]
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-17 Thu 10:57
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Urgent
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
../objlinux/lus2lic should_work/lionel/triSel.lus -en -ei -o /tmp/xxx.lus
* TODO le -o ne fonctionne plus
SCHEDULED: <2013-01-07 Mon>
- State "TODO" from "" [2013-01-07 Mon 14:28]
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-17 Thu 10:57
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Urgent
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO pb dans la verif de définition unique de variable en mode -en
- State "TODO" from "" [2013-01-11 Fri 09:49]
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-17 Thu 10:58
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Regressions introduites par Pascal
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
file:~/lus2lic/testsuite/should_work/heater_control.lus::39
* TODO Regression dans le mode -ec
- State "TODO" from "" [2013-01-11 Fri 09:33]
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-17 Thu 10:59
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Regressions introduites par Pascal
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
undefined new var :
file:testsuite/should_work/Gyroscope2.lus
file:/home/jahier/lus2lic/testsuite/should_work/deSimone.lus::47
var defined twice:
file:/home/jahier/lus2lic/testsuite/should_work/heater_control.lus
Peut-etre est-ce lié au pb précédent ? plus précisement, peut-etre
que le bug existait mais qu'on ne le voyait pas car tous les noeuds
n'étaient pas générés ?
Quoiqu'il en soit, c'est interessant de tester tous les noeuds ce que
ne font pas mes scripts tcl en l'état.
* TODO internal error in l2lExpandArrays.ml with -esa
- State "TODO" from "" [2013-01-18 Fri 10:48]
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-18 Fri 18:01
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Regressions introduites par Pascal
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
./lus2lic -esa -o /tmp/ply02.ec should_work/broken/ply02.lus
oops: lus2lic internal error
File "objlinux/l2lExpandArrays.ml", line 213, column 14
when compiling lustre program file:test/should_work/broken/ply02.lus
nb : le mode -en ne marche pas bas bien non plus...
./lus2lic -en -o /tmp/ply02.ec should_work/broken/ply02.lus
Error in file "/home/jahier/lus2lic/test/should_work/broken/ply02.lus", line 8, col 2 to 2, token 'o':
Variable o is already defined.
ok : it was dut to a bug in l2lRmPoly.ml: var types were not susbtituted in "Lic.left".
* TODO Refaire marcher les tests de non-reg qui sont cassés
SCHEDULED: <2013-01-07 lun.>
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-21 lun. 09:17
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Regressions introduites par Pascal
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
suites aux modifs de Pascal de l'été 2012
- State "TODO" from "" [2012-10-26 Fri 14:59]
En fait, ce qui est cassé, c'est surtout le -ec puisqu'il n'a pas
rebranché l'expansion de tableau et de noeud.
il a fait
- mv tests _tests
- mkdir tests
- cp _tests/Makefile tests
- créer des tests qu'il a mis dans tests...
* DONE compat pb with array type def
- State "DONE" from "WAITING" [2013-01-21 lun. 09:28]
- State "STARTED" from "WAITING" [2013-01-21 lun. 09:27]
- State "TODO" from "" [2013-01-19 Sat 09:40]
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-21 lun. 09:29
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Regressions introduites par Pascal
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: DONE
:END:
file:test/should_work/broken/pfs.lus::43./lus2lic -ec -o /tmp/pfs.ec should_work/broken/pfs.lus
Error in file "/home/jahier/lus2lic/test/should_work/broken/pfs.lus", line 43, col 22 to 22, token '[':
syntax error
It's not clear it should be valid v6: i put it into the to_sort dir.
* TODO autoriser les , dans les structures ?
- State "TODO" from "" [2013-01-19 Sat 09:56]
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-21 lun. 09:30
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Regressions introduites par Pascal
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
./lus2lic -o /tmp/Gyro.out should_work/broken/Gyro.lus
Error in file "/home/jahier/lus2lic/test/should_work/broken/Gyro.lus", line 11, col 42 to 42, token ',':
* TODO mixing v6 packages and v4 implicit package does not work. Should it?
- State "TODO" from "" [2013-01-19 Sat 09:36]
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-21 lun. 09:32
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Regressions introduites par Pascal
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
./lus2lic -o /tmp/old_style_and_pack.out should_work/broken/old_style_and_pack.lus
Error in file "/home/jahier/lus2lic/test/should_work/broken/old_style_and_pack.lus",
line 17, col 1 to 4, token 'node':
file:test/should_work/broken/old_style_and_pack.lus
-> to_sort
* TODO Mettre le sha dans l'affichage de --version
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-21 lun. 09:40
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Pas dur
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
car c'est plus facile dans git pour retrouver ses petits
- State "TODO" from "" [2012-10-26 Fri 14:59]
* TODO Changer la facon de numeroter les versions dans lus2lic (faire comme dans lurette)
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-21 lun. 09:43
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Vieux trucs (?)
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
[2010-11-08 Mon]
* TODO Verifier les boucles combinatoires
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-24 Thu 09:13
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Languages issues
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
quand on genere du ec (car ec2c ne fait pas la verif !)
[2010-06-07 Mon]
On pourrait peut-etre le faire en modifiant celui la : file:src/l2lCheckOutputs.ml ?
* TODO pb d'instance de package ??? - State "TODO" from "" [2013-01-19 Sat 09:43]
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-25 Fri 11:25
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Packages, modeles, etc.
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
./lus2lic -o /tmp/packages.out should_work/broken/packages.lus
./lus2lic -ec -o /tmp/packages.ec should_work/broken/packages.lus
Error in file "/home/jahier/lus2lic/test/should_work/broken/packages.lus", line 22, col 27 to 29, token 'int':
syntax error
file:test/should_work/broken/packages.lus::22
ok : c'etait juste un pb de syntaxe (depuis toujours vraisemblablement...)
* TODO instanciation de noeuds ne marche pas ?
- State "TODO" from "" [2013-01-19 Sat 10:02]
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-29 Tue 09:37
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Packages, modeles, etc.
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
./lus2lic -o /tmp/multipar.out should_work/broken/multipar.lus
file:test/should_work/broken/multipar.lus::20
Error in file "/home/jahier/lus2lic/test/should_work/broken/multipar.lus", line 20, col 12 to 12, token 'g':
unknown node: g
known nodes are: sil, bok, gup, lis
FAIL: without any option: ./lus2lic { -o /tmp/multipar.out should_work/broken/multipar.lus}
* TODO Condact is not working
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-29 Tue 16:31
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Languages issues
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
file:test/should_work/broken/cond01.lus
- State "TODO" from "" [2013-01-18 Fri 23:18]
./lus2lic -ec -o /tmp/cond01.ec should_work/broken/cond01.lus
oops: lus2lic internal error
File "objlinux/l2lExpandMetaOp.ml", line 310, column 4
when compiling lustre program should_work/broken/cond01.lus
* TODO Merge is not working
:PROPERTIES:
:ARCHIVE_TIME: 2013-01-31 Thu 17:52
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Languages issues
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
./lus2lic should_work/broken/merge.lus
file:test/should_work/broken/merge.lus::7
* TODO fix unresolved tests (timeout -> performance bugs)
- State "TODO" from "" [2013-01-11 Fri 11:04]
:PROPERTIES:
:ARCHIVE_TIME: 2013-02-04 Mon 21:33
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Testing process
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
par ex, file:./test/should_work/ec.lus prend un temps infini alors
qu'il n'est pas si gros (y'a 500 variables, mais bon). bon, en fait
il prend 26 secondes, ce qui n'est pas infini, mais bien long tout de
meme.
nb: c'était deja le cas avant les changements de Pascal.
cf file:test/lus2lic.gprof
69% du temps est passé dans unify clock !!!!!
J'ai l'impression que c'est lié au fait que ce programme ne definit
que des contantes. Or les constantes sont potentiellement sur
n'importe quelle horloge, ce qui fait que l'algo manipule un gros paquet
de 'clock_var of int' et que l'on passe beaucoup de temps à faire des
apply_substs2
cf file:test/perf/ contenant les resultats de gprof et ocamlprof sur ec.lus
-----> ca y est, j'ai trouvé : encore une histoire de Verbose pas lazy !!!
* TODO Certains programmes dans should_fail n'echouent qu'en ec (ou bien c'est ec2c) ou l'inverse
- State "TODO" from "STARTED" [2013-01-23 Wed 18:26]
:PROPERTIES:
:ARCHIVE_TIME: 2013-02-06 Wed 16:09
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Testing process
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
du coup les stats sont un peu fausses. a revoir.
* TODO on devrait se passer de 'static_arg list' pour le champ =PREDEF_CALL=
:PROPERTIES:
:ARCHIVE_TIME: 2013-02-06 Wed 17:06
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Aesthetes issues
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
(c'est censé marcher)
cf file:./src/lic.ml::206
- State "TODO" from "" [2012-10-26 Fri 14:59]
* TODO le Eff.WITH (aka if statique) n'a pas lieu d'être !
:PROPERTIES:
:ARCHIVE_TIME: 2013-02-06 Wed 17:57
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Aesthetes issues
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
y virer !!
- State "TODO" from "" [2012-10-26 Fri 14:59]
* TODO const_to_val_eff n'a vraiment rien à faire dans UnifyClock !!!
- State "TODO" from "" [2013-01-29 Tue 14:26]
:PROPERTIES:
:ARCHIVE_TIME: 2013-02-07 Thu 09:42
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Aesthetes issues
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
file:src/unifyClock.ml::271
mouarf, pourquoi pas en fait.
* TODO aucune fonction dans Lic.*_to_string ne devrait dependre des options de compil
- State "TODO" from "" [2013-02-01 Fri 17:54]
:PROPERTIES:
:ARCHIVE_TIME: 2013-02-07 Thu 09:54
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Aesthetes issues
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
cf le XXX file:src/lic.ml::655
* TODO Rajouter le with à la caml pour les structures
- State "TODO" from "" [2012-10-26 Fri 14:59]
:PROPERTIES:
:ARCHIVE_TIME: 2013-02-07 Thu 16:10
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Languages issues
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
main peut-on l'appeller with ?
file:test/should_work/struct_with.lus
* TODO Refaire une passe pour virer une fois pour toute cette histoire d'idref dans Eff.
:PROPERTIES:
:ARCHIVE_TIME: 2013-02-13 Wed 09:22
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Aesthetes issues
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
cf file:/~/lus2lic/src/lic.ml::189
QU: Pascal l'a fait ?
En fait, il y a 3 types d'idents
- long :
+ pack+nom
+ Sert à dénoter les items (node/type/const) externes au pack courant
- t :
+ juste le nom
+ Sert à dénoter les items locaux au pack courant + les variables
- idref :
+ le pack est en option
+ utile dans les premieres phases de la compil
+ actuellement utilisé comme machin pour ranger des trucs dont on
ne sait pas trop s'il s'agit d'Ident.t ou Ident.long et qui
permet d'entretenir une espece de flou désagréable dans le code
et mon esprit...
les idref sont sensé disparaitre une fois que tous les problemes
d'identificateurs ont été résolu (ce qui n'est pas le cas
actuellement).
Le Lic.id_solver est là pour résoudre les ident.t ; bizzarement, il
prend en entrée ces fameux idref que je voudrais voir disparaitre de
Lic.
Le truc, c'est que cet id_solver permet aussi de récuperer des infos
sur les long, d'ou le idref qui est plus versatile...
sol 1 : dupliquer id2const, id2type, etc et créer des long2const, etc.
du coup, plus d'idref !
* TODO Définir les fonctions de UglyStuff proprement - State "TODO" from "" [2012-12-10 Mon 16:38]
:PROPERTIES:
:ARCHIVE_TIME: 2013-02-13 Wed 09:23
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Aesthetes issues
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
file:~/lus2lic/src/uglyStuff.ml
bon j'y ai mis dans IdSolver ; c'est pas si choquant maintenant que
ces fonctions ne sont plus dans Lic.
* TODO definir un Lic.STRUCT_WITH plutot que de passer par un type option
- State "TODO" from "" [2013-02-12 Tue 18:32]
:PROPERTIES:
:ARCHIVE_TIME: 2013-02-13 Wed 09:31
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Aesthetes issues
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO Mettre tout les val_exp dans les operand dans Lic
- State "TODO" from "" [2013-02-12 Tue 18:30]
:PROPERTIES:
:ARCHIVE_TIME: 2013-02-13 Wed 11:16
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Aesthetes issues
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
XXX essayer de virer le constructeur Oper qui n'a pas l'air de servir à grand chose
* STARTED Intégrer le résultat de mly2bnf dans le manuel
:PROPERTIES:
:ARCHIVE_TIME: 2013-02-13 Wed 14:15
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Divers
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: STARTED
:END:
* TODO tableaux (hat, array, etc)
- State "TODO" from "" [2013-03-19 Tue 10:33]
:PROPERTIES:
:ARCHIVE_TIME: 2013-03-22 Fri 10:25
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO diese et nor
- State "TODO" from "" [2013-03-30 Sat 14:31]
:PROPERTIES:
:ARCHIVE_TIME: 2013-04-02 Tue 10:21
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO meta operateurs
- State "TODO" from "" [2013-03-19 Tue 10:33]
:PROPERTIES:
:ARCHIVE_TIME: 2013-04-02 Tue 10:37
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO when, current
- State "TODO" from "" [2013-03-19 Tue 10:33] :PROPERTIES:
:ARCHIVE_TIME: 2013-04-03 Wed 08:50
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO merge
- State "TODO" from "" [2013-03-19 Tue 10:33]
:PROPERTIES:
:ARCHIVE_TIME: 2013-04-05 Fri 15:13
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO fby
- State "TODO" from "" [2013-04-04 Thu 17:10]
:PROPERTIES:
:ARCHIVE_TIME: 2013-04-05 Fri 15:13
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO condact
- State "TODO" from "" [2013-03-19 Tue 10:33]
:PROPERTIES:
:ARCHIVE_TIME: 2013-04-05 Fri 17:23
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO Traiter la fleche plus proprement.
- State "TODO" from "" [2013-04-02 Tue 08:33]
:PROPERTIES:
:ARCHIVE_TIME: 2013-04-08 Mon 13:07
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
En fait il me suffirait de m'inspirer de ce que j'ai fait dans le condact
avec la variable "first_instant" !!!
* TODO Translate the fby properly into a soc
- State "TODO" from "" [2013-04-04 Thu 17:10]
:PROPERTIES:
:ARCHIVE_TIME: 2013-04-08 Mon 13:07
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
vs "-> pre" as it is done actually in file:~/lus2lic/src/ast2lic.ml::468
* TODO slice
- State "TODO" from "" [2013-04-05 Fri 11:18]
:PROPERTIES:
:ARCHIVE_TIME: 2013-04-08 Mon 13:27
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO :END:
* TODO Unifier les modules Data de Lustre V6 et Lutin
- State "TODO" from "" [2013-04-02 Tue 08:33]
:PROPERTIES:
:ARCHIVE_TIME: 2013-04-17 Wed 10:52
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
ie. rajouter les tableaux, struct, et enum en lutin quitte à ne pas s'en servir
tout de suite.
Ca va facilité l'utilisation sans duplic de code du module rif_base
* TODO revoir l'intégration à rif_base et genlex
- State "TODO" from "" [2013-03-19 Tue 10:25]
:PROPERTIES:
:ARCHIVE_TIME: 2013-04-17 Wed 10:52
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO Integrer ce mode -exec dans lurette, à la facon dont je l'ai fait pour Lutin
- State "TODO" from "" [2013-03-30 Sat 14:35]
:PROPERTIES:
:ARCHIVE_TIME: 2013-04-17 Wed 10:53
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
Question : comment j'integre ? via un lus2lic.a ?
file:~/lurette/source/common/lustreRun.mli
* TODO Trouver un moyen d'automatiser des tests
- State "TODO" from "" [2013-03-19 Tue 10:35]
:PROPERTIES:
:ARCHIVE_TIME: 2013-04-23 Tue 16:54
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
via lurette ?
en comparant le mode -exec avec le résultat obtenu via ec puis ec2c
faudrait rajouter une option dans lurette qui, en cas de variables
manquantes, genere le programme lutin qui va bien (loop true) plutot
que de lancer luciole
Un truc de ce gout là semble fonctionner :
./lus2lic _trivial3_condact_oracle.lus -n trivial3_condact_oracle -ec > trivial3_condact_oracl
./lus2lic trivial3.lus -n trivial3_condact --gen-autotest
./lus2lic trivial3.lus -n trivial3_condact -ec >> _trivial3_condact_oracle.lus
./lus2lic _trivial3_condact_oracle.lus -n trivial3_condact_oracle -ec > trivial3_condact_oracle.ec
~/lurette/bin/lurettetop -rp "sut:v6:trivial3.lus:trivial3_condact" -rp "env:lutin:_trivial3_condact_env.lut" -rp "oracle:ec:trivial3_condact_oracle.ec:" -go
Reste à l'integrer dans file:test/lus2lic.tests/non-reg.exp
* TODO assert violation ~/lus2lic/src/socExecValue.ml::239
- State "TODO" from "" [2013-05-10 Fri 17:38]
:PROPERTIES: :ARCHIVE_TIME: 2013-05-14 Tue 16:38
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
../utils/test_lus2lic_no_node should_work/X3.lus
* TODO assert violation ~/lus2lic/src/socExecValue.ml::267
- State "TODO" from "" [2013-05-10 Fri 17:38]
:PROPERTIES:
:ARCHIVE_TIME: 2013-05-16 Thu 09:06
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
../utils/test_lus2lic_no_node should_work/is_stable.lus
* TODO oops: lus2lic internal error
- State "TODO" from "" [2013-05-14 Tue 14:35]
:PROPERTIES:
:ARCHIVE_TIME: 2013-05-16 Thu 13:50
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
File "objlinux/lic2soc.ml", line 648, column 14
./lus2lic should_work/noeudsIndependants.lus -n noeudsIndependants --gen-autotest -np
C'est parce que il y a un clash avec Lustre::gt qui est polymorphe,
et donc l'alias vers gt ne genere pas de code...
Il faudrait vraiment que je trouve un moyen de ne pas remplir les tables avec
le contenu du module Lustre de facon strict (ie, le faire lazy !!)
cf file:src/licTab.ml::68
Bon, finalement, j'oblige les gens a ecrire Lustre::gt et puis ca marre.
* TODO oops: lus2lic internal error
- State "TODO" from "" [2013-05-10 Fri 17:46]
:PROPERTIES:
:ARCHIVE_TIME: 2013-05-21 Tue 09:31
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
File "objlinux/lic2soc.ml", line 680, column 18
when compiling lustre program should_work/morel3.lus
-> It was due to a bad handling in slices apprearinf at the lhs od equations.
* TODO pb avec -en sur programme avec des when
- State "TODO" from "" [2013-05-21 Tue 11:10]
:PROPERTIES:
:ARCHIVE_TIME: 2013-05-21 Tue 17:30
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Divers
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
./lus2lic _ck5_oracle.lus -n ck5_oracle -en
* oops: lus2lic internal error
:PROPERTIES:
:ARCHIVE_TIME: 2013-05-31 Fri 13:55
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec :ARCHIVE_CATEGORY: lv6
:END:
File "objlinux/data.ml", line 114, column 15
when compiling lustre program should_work/left.lus
./lus2lic should_work/left.lus -exec
* oops: lus2lic internal error
:PROPERTIES:
:ARCHIVE_TIME: 2013-05-31 Fri 13:56
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:END:
File "objlinux/data.ml", line 114, column 15
when compiling lustre program should_work/left.lus
./lus2lic should_work/left.lus -exec
* TODO oops: lus2lic internal error
- State "TODO" from "" [2013-05-10 Fri 18:05]
:PROPERTIES:
:ARCHIVE_TIME: 2013-05-31 Fri 13:56
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
File "objlinux/lic2soc.ml", line 680, column 18
when compiling lustre program should_work/left.lus
* TODO internal error
- State "TODO" from "" [2013-05-10 Fri 17:19]
:PROPERTIES:
:ARCHIVE_TIME: 2013-05-31 Fri 15:52
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
File "objlinux/unifyType.ml", line 106, column 3
../utils/test_lus2lic_no_node should_work/t1.lus
file:src/unifyType.ml::110
* TODO Fix a bug occurring using -en and -esa together.
- State "TODO" from "" [2013-04-24 Wed 09:22]
:PROPERTIES:
:ARCHIVE_TIME: 2013-05-31 Fri 16:10
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Testing process
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
#+begin_src sh
./lus2lic _assertion_oracle.lus -n assertion_oracle -en -esa
#+end_src
** Error in file "/home/jahier/lus2lic/test/_assertion_oracle.lus", line 38, col 3 to 7, token 'v02_1':
**
** Variable _0v02_1_1 is already defined.
#+end_src
-> on verra ca quand les autres test passeront ; je ferai alors une passe
complete qur le nommage des variables fraiches.
* Performances :PROPERTIES:
:ARCHIVE_TIME: 2013-05-31 Fri 16:24
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_CATEGORY: lv6
:END:
Ceux là sont particulierement long (dont certains vraiment sans raison)
en fait, quand ecexe se plante (oracle), il bloque tout le monde...
# of unresolved testcases 12
../utils/test_lus2lic_no_node should_work/assertion.lus
../utils/test_lus2lic_no_node should_work/normal.lus
../utils/test_lus2lic_no_node should_work/Gyroscope2.lus
../utils/test_lus2lic_no_node should_work/ELMU.lus
../utils/test_lus2lic_no_node should_work/aux.lus
../utils/test_lus2lic_no_node should_work/X1.lus
../utils/test_lus2lic_no_node should_work/alarme.lus
../utils/test_lus2lic_no_node should_work/onlyroll2.lus
../utils/test_lus2lic_no_node should_work/integrator.lus
../utils/test_lus2lic_no_node should_work/over2.lus
* WAITING Ya un probleme avec ce fichier lustre (compilait avant)
- State "WAITING" from "STARTED" [2013-01-17 Thu 10:48]
- State "STARTED" from "TODO" [2013-01-17 Thu 10:48]
:PROPERTIES:
:ARCHIVE_TIME: 2013-05-31 Fri 17:05
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Pas clair
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: WAITING
:END:
file:test/should_work/call/call04.lus
il semble y avoir une confusion entre parametre et arguments lors
d'appels de noeuds définis via des itérateurs de tableau
idem pour
file:test/should_work/fab_test/morel2.lus
et pleins d'autres. hm, y'aurait pas d'itérateurs dans celui la.
-> ok : c'était à de l2lAliasType (que j'ai déranché du coup) qui
faisait que les types lic n'étaient plus uniques et du coup les
substitutions dans l'expansion des noeuds ne se faisaient plus
correctement.
je met en attente en attendant de savoir ce qu'on fait de ce module.
moi j'ai bien envie de virer AbstractType de Lic.type_. En effet, j'avais
fait attention à tous les virer pour éviter les soucis, mais le plus
propre c'est d'y virer vraiment.
Pascal lui, s'en est servi pour faire des types alias, alors que ca
n'est pas fait pour. Cela dit, si on créé des types alias, on risque
d'avoir le meme genre de soucis. A quoi ca sert d'avoir de tels types
? pour moi le role de la compil ca serait plutot de les virer que de
les rajouter, mais bon. A discuter. cf point d'apres
* TODO Les assertions qui echouent font planter les tests : reflechir a une parade
:PROPERTIES:
:ARCHIVE_TIME: 2013-05-31 Fri 17:21
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Divers
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
car apres tout, c'est voulu dans le prog, et ca n'est donc pas un pb du compilo
- State "TODO" from "" [2013-05-31 Fri 16:30]
../utils/test_lus2lic_no_node should_work/assertion.lus
../utils/test_lus2lic_no_node should_work/drapfab.lus
../utils/test_lus2lic_no_node should_work/s.lus
../utils/test_lus2lic_no_node should_work/eq1.lus
../utils/test_lus2lic_no_node should_work/lecteur.lus
../utils/test_lus2lic_no_node should_work/alarme.lus
* TODO pb gen-autotest
- State "TODO" from "" [2013-06-03 Mon 09:51]
:PROPERTIES:
:ARCHIVE_TIME: 2013-06-03 Mon 09:57
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
../utils/test_lus2lic_no_node should_work/plus.lus
+ ./lus2lic _plus_oracle.lus -n plus_oracle -lv4 -eei -en --no-prefix -o lv4_plus_oracle.lus
Error in file "/home/jahier/lus2lic/test/_plus_oracle.lus", line 6, col 28 to 28, token '=':
tuple size error:
the tuple size is
4 for the left-hand-side, and
1 for the right-hand-side (in c_bis,d_bis,e_bis,f_bis = a + b)
et aussi :
../utils/test_lus2lic_no_node should_work/bred.lus
../utils/test_lus2lic_no_node should_work/alias.lus
../utils/test_lus2lic_no_node should_work/bred_lv4.lus
../utils/test_lus2lic_no_node should_work/minus.lus
* TODO il reste des pbs de portée de nom
- State "TODO" from "" [2013-06-03 Mon 09:56]
:PROPERTIES:
:ARCHIVE_TIME: 2013-06-03 Mon 10:21
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
./lus2lic should_work/myplus.lus -n plus -ec
../utils/test_lus2lic_no_node should_work/minus.lus
* TODO Lurette trouve un mismatch sur ce prog au step 0
- State "TODO" from "" [2013-05-10 Fri 17:08]
:PROPERTIES:
:ARCHIVE_TIME: 2013-06-04 Tue 10:15
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
../utils/test_lus2lic_no_node should_work/mapiter.lus
../utils/test_lus2lic_no_node should_work/arrays.lus
../utils/test_lus2lic_no_node should_work/nested.lus
../utils/test_lus2lic_no_node should_work/matrice.lus
../utils/test_lus2lic_no_node should_work/mapdeRed.lus
../utils/test_lus2lic_no_node should_work/over2.lus
../utils/test_lus2lic_no_node should_work/predefOp.lus
file:test/should_work/predefOp.lus
* TODO Lurette trouve un mismatch sur ce prog au step 2
- State "TODO" from "" [2013-05-10 Fri 17:08]
:PROPERTIES:
:ARCHIVE_TIME: 2013-06-04 Tue 14:16
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6 :ARCHIVE_TODO: TODO
:END:
../utils/test_lus2lic_no_node should_work/cond01.lus
* TODO
- State "TODO" from "" [2014-06-16 Mon 08:59]
:PROPERTIES:
:ARCHIVE_TIME: 2014-06-16 Mon 10:50
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
lus2lic -exec should_work/map.lus -n pbmap
ne prend pas en compte la valeur du pre
* TODO Lurette trouve un mismatch sur ce prog au step 7
- State "TODO" from "" [2013-05-10 Fri 17:08]
:PROPERTIES:
:ARCHIVE_TIME: 2014-06-16 Mon 11:00
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
../utils/test_lus2lic_no_node should_work/call07.lus
file:test/should_work/call07.lus
* TODO pb en utilisant -2c et -esa en meme temps
- State "TODO" from "" [2014-07-04 Fri 16:16]
:PROPERTIES:
:ARCHIVE_TIME: 2014-07-10 Thu 10:17
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -2C
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
lus2lic should_work/modes3x2_v4.lus -n modes3x2_v4 -esa -o xxx -2c
* TODO Lurette trouve un mismatch sur ce prog au step 1
- State "TODO" from "" [2013-05-10 Fri 17:08]
:PROPERTIES:
:ARCHIVE_TIME: 2014-08-07 Thu 12:27
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
../utils/test_lus2lic_no_node should_work/pipeline.lus
file:test/should_work/pipeline.lus
* TODO Lurette trouve un mismatch sur ce prog au step 6
- State "TODO" from "" [2013-05-10 Fri 17:08]
:PROPERTIES:
:ARCHIVE_TIME: 2014-08-07 Thu 12:27
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
file:test/should_work/deSimone.lus
../utils/test_lus2lic_no_node should_work/deSimone.lus
* TODO Lurette trouve un mismatch sur ce prog au step 2
- State "TODO" from "" [2013-05-10 Fri 17:08]
:PROPERTIES:
:ARCHIVE_TIME: 2014-08-07 Thu 12:27
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
../utils/test_lus2lic_no_node should_work/test.lus
file:test/should_work/test.lus
* TODO le traitement du condact ne marche plus
- State "TODO" from "" [2014-06-27 Fri 15:26]
:PROPERTIES:
:ARCHIVE_TIME: 2014-08-07 Thu 16:15
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: lus2lic -exec
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
depuis le 2eme commit du 18-06-2014
lus2lic should_work/test_condact.lus -n test_condact -exec
once the activation condition has been true, and when it is false it
returns the default value instead of the previous one.
* TODO Use severale machine to launch tests in paralell
- State "TODO" from "" [2014-08-14 Thu 11:23]
:PROPERTIES:
:ARCHIVE_TIME: 2014-08-22 Fri 10:38
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Testing process
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO Traiter TOUS les warnings !!!
- State "TODO" from "" [2014-08-13 Wed 17:33]
:PROPERTIES:
:ARCHIVE_TIME: 2014-08-26 Tue 10:19
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Divers
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
* TODO lus2lic genere du ec mal typé !!
- State "TODO" from "" [2013-05-10 Fri 17:24]
:PROPERTIES:
:ARCHIVE_TIME: 2014-09-03 Wed 11:04
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Testing process
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
../utils/test_lus2lic_no_node should_work/predef03.lus
* TODO pour ceux la ils suffit d'expanser. Comment eviter que les tests echouent ?
:PROPERTIES:
:ARCHIVE_TIME: 2014-09-03 Wed 11:04
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Testing process
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
- lus2lic -2c should_work/activation2.lus -n activation2
- lus2lic -2c should_work/activation1.lus -n activation1- lus2lic -2c should_work/asservi.lus -n asservi
- lus2lic -2c should_work/speedcontrol.lus -n speedcontrol
* TODO pb d'horloge
- State "TODO" from "" [2014-08-13 Wed 17:33]
:PROPERTIES:
:ARCHIVE_TIME: 2014-09-03 Wed 11:07
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Divers
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
#+BEGIN_SRC lustre
node xxx(x:int;c:bool) returns (res1,res2:int);
var
y:int when c;
k:int when c;
let
y = x when c;
k = (0 fby (k+1)) when c ;-- erreur ici, alors que ca semble bon
res1 = current(y);
res2 = current(k);
tel
#+END_SRC
clock error: ' on c on base' is not a sub-clock of ' on base'
en plus le message est bizzare
D'une maniere generale, les messages d'erreur en cas de pb d'horloge sont vraiment mauvais.
-> non, en fait, le when c ne sert a rien piur definir le k.
* TODO msg d'erreur un peu mauvais sur ce programme
:PROPERTIES:
:ARCHIVE_TIME: 2014-09-03 Wed 11:11
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Divers
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
#+begin_src lustre
node argos_oracle(a:bool;b:bool;s0:bool;s1:bool;s2:bool) returns(ok:bool;s0_bis:bool;s1_bis:bool;s2_bis:bool);
let
(s0_bis,s1_bis,s2_bis) = argos(a,b);
s0=s0_bis;
s1=s1_bis;
s2=s2_bis;
tel;
#+end_src
-> ah bon ? bof.
* TODO ce noeud ne compile pas en mode -ec
- State "TODO" from "" [2013-04-17 Wed 15:51]
:PROPERTIES:
:ARCHIVE_TIME: 2014-09-03 Wed 11:12
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Divers
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
#+begin_src lustre
node trivial3(e: int ^2) returns (x: int ^6);
var
c: int ^6;
let
c = e|e|e;
x[0..2] = c[3..5];
x[3..5] = c[0..2];
tel
#+end_src
-> maintenant si
* TODO pb en mode -esa avec les itérateurs
- State "TODO" from "" [2014-06-05 Thu 10:00]
:PROPERTIES:
:ARCHIVE_TIME: 2014-09-03 Wed 11:14
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Divers
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
lus2lic should_work/minus.lus -n minus -esa
lus2lic should_work/normal.lus -n normal -esa
-> plus maintenant
* TODO divergence -exec et -2c (1)
- State "TODO" from "" [2014-07-11 Fri 12:02]
:PROPERTIES:
:ARCHIVE_TIME: 2014-09-04 Thu 14:42
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: Failures spotted by non reg tests (13)
:ARCHIVE_CATEGORY: lv6
:ARCHIVE_TODO: TODO
:END:
grep "FAIL:" lus2lic*.log | grep "exec" | grep "\-2c" | sed s/'FAIL: Try to compare lus2lic -exec and -2c:'/-/
1) ../utils/compare_exec_and_2c should_work/PCOND.lus 48319
Failure occured in lurette: The oracle first output should be a bool;
but ok has type nil
* A option to inline the simple nodes such as AND, OR, NOT etc.
:PROPERTIES:
:ARCHIVE_TIME: 2014-10-01 Wed 17:31
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: a classer/Willie's wishes
:ARCHIVE_CATEGORY: lv6
:END:
* The expansion of nodes with -efn --expand-following-nodes [easy]
:PROPERTIES:
:ARCHIVE_TIME: 2014-10-02 Thu 14:54
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: a classer/Willie's wishes
:ARCHIVE_CATEGORY: lv6
:END:
the current available options are:
-en, --expand-nodes
Expand the main node (use the first node if no one is
specified).
-den, --do_not-expand-nodes <string>
Do not expand the specified node (meaningful with -en only of
course).
It might be better if another option such as -efn --expand-following-nodes
<string>
The nodes which are not in the string will not be expanded.
* Generate a copy of memcpy and replace the sizeof operator with the actual size of the structure.
:PROPERTIES:
:ARCHIVE_TIME: 2014-10-02 Thu 17:54
:ARCHIVE_FILE: ~/lus2lic/todo.org
:ARCHIVE_OLPATH: a classer/Willie's wishes
:ARCHIVE_CATEGORY: lv6 :END:
It seems that the "assert" operator will generate "array" structure
where the number of the element depending on the number of the
arguments to the "assert". The generated C code tries to make a
copy of this structure with memcpy and giving
sizeof(the_structure_to_copy) as the argument.
Since OTAWA does not handle memcpy, could you generate a version of
memcpy (contains a for loop) for each memcpy call. (Or just to
paste the for loop to replace the memcpy call), while the iteration
count of the for loop is given (in actual value) so that ORANGE can
handle it.....
je comprend pas tout, mais ce que je retiens, c'est de remplacer les memcpy
par des appels a des fonctions (ou des macros) qui par default utiliseraient
des memcpy
genre
_assign_ip2(ctx.x, ctx.y);
#+BEGIN_SRC C
#ifndef _assign_ip2
#define _assign_ip2(dest, source) memcpy(dest, source, sizeof(dest))
#endif
#+END_SRC