Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-03 16:45:01 -0800 (Mon, 03 Mar 2003)
Revision: 4151
Log message:
- Defined subset, subStructure (which can be used for semigroup, monoid,
group, etc.), cyclic group (including the power operation for a group
and cyclic subgroups); and proved corresponding theorems.
- Adopted Alexei's new syntax.
- Removed "itt_abelian_group" module, and moved "isCommutative" to "itt_
grouplikeobj".
Note: The subset module is rather incomplete. It only includes what I
currently will use.
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-03 16:48:43 -0800 (Mon, 03 Mar 2003)
Revision: 4152
Log message:
Added itt_cyclic_group and removed itt_abelian_group for documentation.
Changes | Path |
+1 -1 | metaprl/doc/latex/theories/itt/print.ml |
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-03 18:39:29 -0800 (Mon, 03 Mar 2003)
Revision: 4153
Log message:
- Moved "csemigroup"(the set of commutative semigroup) and "cmonoid" to
"itt_grouplikeobj" and "abelg" to "itt_group"
- Added the intro and elim rules for them. For example:
'g in group[i:l] & isCommutative{'g} <=> 'g in abelg[i:l]
- Corrected a typo in the display form of semigroup
I'm not quite sure whether these elimination rules are better or the ones
that completely unfold, say "abelg" are better. I put the second option
in comments for now.
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-03 18:43:45 -0800 (Mon, 03 Mar 2003)
Revision: 4154
Log message:
I've moved "commutative" to "itt_grouplikeobj" and "itt_group".
Changes | Path |
Deleted | metaprl/theories/itt/itt_abelian_group.ml |
Deleted | metaprl/theories/itt/itt_abelian_group.mli |
Deleted | metaprl/theories/itt/itt_abelian_group.prla |
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-03 21:56:02 -0800 (Mon, 03 Mar 2003)
Revision: 4155
Log message:
- Redefined "csemigroup", "cmonoid", and "abelg" following Alexei's advice.
- Reproved corresponding rules.
- Removed the intro resources.
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-03 23:05:03 -0800 (Mon, 03 Mar 2003)
Revision: 4156
Log message:
Can anyone take a look at the rule "/itt_group/subgroup_isect"? I can't
prove it. It states "the intersection (group) of two subgroups is again
a subgroup".
Changes | Path |
+0 -6 | metaprl/theories/itt/itt_group.ml |
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-04 18:45:52 -0800 (Tue, 04 Mar 2003)
Revision: 4157
Log message:
- Added some properties and reduction rules of the group power operation.
- **Partially proved "Every subgroup of a cyclic group is cyclic" with 2 assumptions:
1. there is a smallest positive number k such that (a^k)_g in s^car;
2. exst q,c in int. m=q*k+r & 0<=r<k.
The first one should be provable with induction.
The second is a common theorem in mathematics.
(I just found that for 2nd assumption, I used r>0 instead of r>=0, but
this is minor.)
Besides, there are two problems:
1. Why can't I apply reduceC for "m * 0" in "/itt_cyclic_group/group_power_power_reduce/2" or "/itt_cyclic_group/group_power_power_reduce/2/1"
and for "mul_uni_Assoc" in "/itt_cyclic_group/group_power_power_reduce/3/1/1/1/1/1/2".
2. I'll try writing rules like "group_power_0" as reduce resources.
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-06 02:45:27 -0800 (Thu, 06 Mar 2003)
Revision: 4158
Log message:
- Proved that for if there is a positive number x such that P[x], then there
is a smallest positive number k such that P[k], as long as P[y] is well-formed
and decidable for any integer y. (rule "smallest_positive")
- Stated as a separate rule that for any integer k>0 and m, there exists
integer r>=0 and q such that m=k*q+r. Not proved yet. (rule "int_div_rem")
- Now "Every non-trivial subgroup of a cyclic group is cyclic" "should" be
provable other than "type"{(y^a)_g in s.car} is needed for y in g.car and
a in int, which is not true. Have I made any mistake along the way or is
this theorem just incorrect here?! (see "/itt_cyclic_group/subg_cycGroup/1/1"
for example)
I'll move "If S is a non-trivial subgroup of a cyclic group G generated by b,
there exists a positive number k such that (b^k)_g in S" as a separate rule.
Changes | Path |
+80 -51 | metaprl/theories/itt/itt_cyclic_group.ml |
+2 -0 | metaprl/theories/itt/itt_cyclic_group.mli |
+6389 -2330 | metaprl/theories/itt/itt_cyclic_group.prla |
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-08 01:47:03 -0800 (Sat, 08 Mar 2003)
Revision: 4159
Log message:
- Added reduce resource for group_power;
- Moved "positiveRule1T", "positiveRule2T", and "int_div_rem" to "itt_nat".
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-08 03:43:05 -0800 (Sat, 08 Mar 2003)
Revision: 4160
Log message:
- First try of defining group homomorphism.
I define it as a set of all group homomorphisms from G to H. Any comment?
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-10 02:03:15 -0800 (Mon, 10 Mar 2003)
Revision: 4161
Log message:
- Proved some properties of group homomorphism.
"Membership is a type iff it is true" also causes problems in proving
"If f:A->B is a homomorphism and T is is a subgroup of B, then the
inverse image of T under f is a subgroup of A", as in
"/itt_group/groupHom_subg2/2" and "/itt_group/groupHom_subg2/3".
- Uncommented some reduce resources in itt_int_ext.
- Added several theorems about grouplikeobj and groups.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-11 00:41:03 -0800 (Tue, 11 Mar 2003)
Revision: 4162
Log message:
Aleksey & Xin:
- Added "AutoMustComplete" to a couple of things
- Added fnExtensionalT and fnExtenT tactics.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-11 13:56:41 -0800 (Tue, 11 Mar 2003)
Revision: 4163
Log message:
Xin & Aleksey:
- proved the subgroup intersection theorem (very straightforward,
but long proof with a lot of subcases)
- for ... >- f a in A, use intro in autoT
for ... >- f a = f' a' in A, only use intro with AutoMustComplete
- added type inference for group components
- added term contructors/descructors for the field term
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-11 22:30:11 -0800 (Tue, 11 Mar 2003)
Revision: 4164
Log message:
- Working on cleaning up and documenting the term addresses interfaces
(unfortunatelly, there are a lot of bugs in both std/gen and ds implementations
of those and I need to figure out what the expected behaviour is before I can start
fixing things).
- Minor typo in itt_record documentation.
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-12 09:09:27 -0800 (Wed, 12 Mar 2003)
Revision: 4165
Log message:
1. itt_bool - equivalence of "x=false in bool" and "assert(not x)".
2. thenLocalMT, thenLocalAT, onAllLocalMHypsT added, they respect
local labels instead of inheritted labels.
3. itt_list2/nth_wf completed. Two issues are done manually:
3.a. Use of (1)
3.b. use of splitIntT for transition from "a<>b" to "a<b or a>b"
4. several bugs fixed in arithT and normalizeC
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-13 01:19:57 -0800 (Thu, 13 Mar 2003)
Revision: 4166
Log message:
- Messed with the line_buffer and overflow code to get it do actually produce
the "..." at the end oif the string in case of overflow.
Changes | Path |
Added | metaprl/editor/ml/tests/seq_addrs_test.ml |
Properties | metaprl/editor/ml/tests/seq_addrs_test.ml |
+33 -27 | metaprl/refiner/reflib/rformat.ml |
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-13 03:04:00 -0800 (Thu, 13 Mar 2003)
Revision: 4167
Log message:
- Added intro/elim rules for all concepts in itt_grouplikeobj.ml and added
elim rule for subset.
- Changed the organization of itt_grouplikeobj.ml. Now each section reflects
a concept.
- ...
Changes by: Alexei Kopylov (kopylov at cs.caltech.edu)
Date: 2003-03-13 17:11:36 -0800 (Thu, 13 Mar 2003)
Revision: 4168
Log message:
Partially fix an old bug with parser. Now you can write
squash{'A & 'B}.
(Previously you had to write squash{('A & 'B)} or squash{.'A & 'B}).
Unfortunately, this fix is only partial:
squash{iff{'A;'B} & 'C} does not work, you need to write
squash{(iff{'A;'B} & 'C)}
Changes | Path |
+1 -1 | metaprl/filter/base/term_grammar.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-14 03:27:05 -0800 (Fri, 14 Mar 2003)
Revision: 4169
Log message:
- This is a _major change_ in how sequent contexts are now handles in rule
arguments. Now arguments corresponding to sequent contexts are passed in as _integers_.
To tell the rewriter that the corresponding context should match n hypotheses,
pass n + 1 as an argument. Also, if a context is at the _end_ of the hypotheses list,
it should _not_ be passed as an argument - instead, rewriter will just put all the remaining
hyps into it.
To summarize, a typical intro-style rule now does not need any context args at all. A typical
elim-style rule will now need only one context arg ('H) and that argument will be a straight
integer. That integer will have the same meaning as in traditional tactics - e.g. just
the hypothesis number. The only difference is that the arguments to primitive rules _have_
to be positive. Use Sequent.get_pos_hyp_num or pos_hypT to make numbers positive.
- I changed the semantics of the int arg to assertAtT and second int arg to copyHypT. Now
that int means the hyp number the new hyp will have in the resulting subgoal (whether
input is positive or negative).
- I fixed some of the proofs Yegor accidentally broke, but not all. Xin, you might have to
fix some of the cyclic_group proofs yourself, sorry.
- Cleaned up the address part of the TermAddm and TermMan interfaced, killed some
unnecessary functionality.
- More changes to line_buffer in Rformat - hopefully this time I finally got it right.
Changes by: Alexei Kopylov (kopylov at cs.caltech.edu)
Date: 2003-03-16 12:36:44 -0800 (Sun, 16 Mar 2003)
Revision: 4170
Log message:
I simplify definition of the subset relation.
And define membership: <<'x in 'A subset 'B>>.
But I don't think we will need it, if we are going to redefine usual equality.
Also add two theories:
itt_singleton -- defines singleton
itt_subset2 -- contains some theorems about subset, in particular
that intersection of two subsets is subset.
I have not finished some proofs. Most unfinished proofs should be done by autoT if autoT would be more clever. (I'm working on it).
Changes by: Alexei Kopylov (kopylov at cs.caltech.edu)
Date: 2003-03-16 19:59:48 -0800 (Sun, 16 Mar 2003)
Revision: 4171
Log message:
1. Fix broken proofs.
Some proofs became much shorter. E.g.,
-Status: `lcoset_subset' is a derived object with a complete proof (8 rule boxes, 3294 primitive steps)
+Status: `lcoset_subset' is a derived object with a complete proof (2 rule boxes, 314 primitive steps)
I have not fix incomplete proofd.
Now instead of elimination rule for subset relation you can use rule use_subset, use_superset*.
2. Change display form for subtype. Now subtype is \sqsubseteq.
My intuition is this: all types operators are sharp and square: +, \times, Type, \sqsubseteq.
And all set operations are nice and circle: \cap, \cup, Set, \subseteq. ;)
3. Add new theories to documentation.
Changes by: Alexei Kopylov (kopylov at cs.caltech.edu)
Date: 2003-03-16 21:16:16 -0800 (Sun, 16 Mar 2003)
Revision: 4172
Log message:
More on display forms and documentation
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-17 15:59:17 -0800 (Mon, 17 Mar 2003)
Revision: 4173
Log message:
Updated the list of people so that all sys descr. authors are also listed here.
Changes | Path |
+55 -29 | metaprl/doc/htmlman/mp-people.html |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-17 16:20:59 -0800 (Mon, 17 Mar 2003)
Revision: 4174
Log message:
Small no-op change in Term_ds address interface (needed to simplify adding
some additional debugging functionality to it).
Changes | Path |
+1 -2 | metaprl/refiner/refiner/refiner_ds.ml |
+0 -4 | metaprl/refiner/term_ds/term_addr_ds.ml |
+2 -3 | metaprl/refiner/term_ds/term_addr_ds.mli |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-17 16:37:46 -0800 (Mon, 17 Mar 2003)
Revision: 4175
Log message:
This is the IR for the new letrec.
Changes by: Alexei Kopylov (kopylov at cs.caltech.edu)
Date: 2003-03-17 17:33:08 -0800 (Mon, 17 Mar 2003)
Revision: 4176
Log message:
I commit my current status of definition of red-black trees.
This work in progress, everything subject to change.
I'll commit prla files later.
itt_relation_str defines algebraic structure for ordered sets, sets with decidable equality
itt_set_str defines data structure for sets.
Also change some display forms.
Changes by: Alexei Kopylov (kopylov at cs.caltech.edu)
Date: 2003-03-17 17:51:38 -0800 (Mon, 17 Mar 2003)
Revision: 4177
Log message:
Added new theories in print.ml
Changes | Path |
+7 -0 | metaprl/doc/latex/theories/itt/print.ml |
Changes by: Alexei Kopylov (kopylov at cs.caltech.edu)
Date: 2003-03-17 19:39:35 -0800 (Mon, 17 Mar 2003)
Revision: 4178
Log message:
1. Now parser never generate singleton record, but only empty record.
2. Update record_exm to use new syntax.
Still to do:
1. Allow labels to be meta-parameters and arbitrary term.
token_or_sting = constant label (as now)
@token_or_sting = meta-parameter
anything else = term
Aleksey, do you agree?
2. Rename rcrd -> record, record -> Record.
As far as field, I think probably it's not a bad name.
At least in FIR it's called field. I think we want consistency here.
Changes | Path |
+3 -4 | metaprl/filter/base/term_grammar.ml |
+2 -0 | metaprl/theories/itt/itt_record.ml |
+69 -52 | metaprl/theories/itt/itt_record_exm.ml |
+1 -1 | metaprl/theories/itt/itt_tunion.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-17 19:51:47 -0800 (Mon, 17 Mar 2003)
Revision: 4179
Log message:
Fixed some (hopefull all!) bugs where rewrite mechanism was looking (counting
hyps, taking subterms out, passing to nth*addr, etc) at the goal, while rewrite
was being applied to an assumption.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-17 20:07:14 -0800 (Mon, 17 Mar 2003)
Revision: 4180
Log message:
Updated CPS.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-18 16:24:07 -0800 (Tue, 18 Mar 2003)
Revision: 4181
Log message:
Finally all our proofs replay with TERMS=std!
Changes | Path |
+20 -24 | metaprl/refiner/term_gen/term_addr_gen.ml |
+2 -2 | metaprl/refiner/term_std/term_base_std.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-18 16:29:41 -0800 (Tue, 18 Mar 2003)
Revision: 4182
Log message:
Updated closure conversion/dead code elimination/inlining to use the
new representation of functions.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-18 18:23:51 -0800 (Tue, 18 Mar 2003)
Revision: 4183
Log message:
- The rewriter now correctly makes sure that the SO var args have
to be distinct for it to be a potential "poattern" source (and not
just an instance). This fixed BUGS 4.11 and allows eliminating the
"HACK" terms in itt_*isect
- free_vars_terms now returns StringSet.t, not string list
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-18 21:25:05 -0800 (Tue, 18 Mar 2003)
Revision: 4184
Log message:
Generate initial assembly for the M language.
This is before register allocation; that's next.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-18 22:36:02 -0800 (Tue, 18 Mar 2003)
Revision: 4185
Log message:
- I moved the sequent-related *_addr functions from TermMan to TermAddr
- I got rid of the TermSimleSig signature, use TermSig instead
- The Lib_term module now uses the Term_man_gen for converting
from/to term representation of sequents instead of reimplementing it.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-19 14:14:00 -0800 (Wed, 19 Mar 2003)
Revision: 4186
Log message:
Changed instruction format to use string parameters, like this:
Inst2["mov":s]{'dst; 'src; 'rest}
Added code to manage built-in registers.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-19 19:33:42 -0800 (Wed, 19 Mar 2003)
Revision: 4187
Log message:
Added code to split a live range.
Changes by: ( at unknown.email)
Date: 2003-03-19 19:33:42 -0800 (Wed, 19 Mar 2003)
Revision: 4188
Log message:
This commit was manufactured by cvs2svn to create branch 'lm_libmojave'.
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-19 22:45:00 -0800 (Wed, 19 Mar 2003)
Revision: 4189
Log message:
Fixed a typo.
Changes | Path |
+1 -1 | metaprl/theories/czf/czf_itt_comment.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-20 00:13:58 -0800 (Thu, 20 Mar 2003)
Revision: 4190
Log message:
- Finished the proofs (and cleaned up) in ctt_markov
- Cleaned up czf_itt_bool to get rid of the error messages those
"saved on the side" subproofs generate.
Changes | Path |
+4031 -10529 | metaprl/theories/czf/czf_itt_bool.prla |
+708 -1179 | metaprl/theories/itt/ctt_markov.prla |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-20 00:54:48 -0800 (Thu, 20 Mar 2003)
Revision: 4191
Log message:
Killed some unused code.
Changes | Path |
+0 -37 | metaprl/refiner/reflib/unify_mm.ml |
+1 -49 | metaprl/refiner/term_std/term_subst_std.ml |
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-20 02:06:15 -0800 (Thu, 20 Mar 2003)
Revision: 4192
Log message:
Changed some rule names for consistence.
Changes | Path |
+10 -10 | metaprl/theories/itt/itt_grouplikeobj.ml |
+2 -0 | metaprl/theories/itt/itt_grouplikeobj.mli |
+3900 -3938 | metaprl/theories/itt/itt_grouplikeobj.prla |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-20 14:19:08 -0800 (Thu, 20 Mar 2003)
Revision: 4194
Log message:
This is a branch commit.
Added lm_libmojave to MetaPRL.
Changes | Path |
+16 -3 | metaprl-branches/lm_libmojave/Makefile |
+11 -7 | metaprl-branches/lm_libmojave/mk/rules |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-20 15:33:50 -0800 (Thu, 20 Mar 2003)
Revision: 4195
Log message:
Added register allocator code. So far unused.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-20 17:11:36 -0800 (Thu, 20 Mar 2003)
Revision: 4196
Log message:
Added x86 backend description.
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-20 17:16:26 -0800 (Thu, 20 Mar 2003)
Revision: 4197
Log message:
- Fixed some broken proofs.
- Updated structure of itt_group.
- subStructure and subgroup need redefining and anything related with
subgroup needs to be updated.
Changes | Path |
+1264 -1081 | metaprl/theories/itt/itt_cyclic_group.prla |
+242 -92 | metaprl/theories/itt/itt_group.ml |
+8 -0 | metaprl/theories/itt/itt_group.mli |
+6150 -5246 | metaprl/theories/itt/itt_group.prla |
Changes by: Alexei Kopylov (kopylov at cs.caltech.edu)
Date: 2003-03-20 17:35:40 -0800 (Thu, 20 Mar 2003)
Revision: 4198
Log message:
Commited some proofs.
Most of the proofs are incomplete (if not all of them).
If they broke, it would be ok, I am going to change them anyway.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-20 17:49:30 -0800 (Thu, 20 Mar 2003)
Revision: 4199
Log message:
Added main call to register allocator.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-20 19:07:23 -0800 (Thu, 20 Mar 2003)
Revision: 4200
Log message:
Changed the srripts to include the module name in each "status" line
along with the rule name - this makes the logs much easier to read
(no need to be guessing where the rule came from).
Changes | Path |
+1 -1 | metaprl/util/check-status.sh |
+1 -1 | metaprl/util/status-all.sh |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-20 22:02:32 -0800 (Thu, 20 Mar 2003)
Revision: 4201
Log message:
Register allocation runs and produces an assignment, but we
fail alpha-equality. Hard to debug...
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-20 22:50:59 -0800 (Thu, 20 Mar 2003)
Revision: 4202
Log message:
The problem with alpha-equality was because some MOVs _must_
be coalesced. Specifically, Let{Register{v}, v. e['v]} must
be coalesced.
This makes the first successful register allocation of the fib
program.
What's left:
1. Functions must use a standard calling convention, rather
than having the allocator choose locations of arguments
at random.
2. The LetReg[reg:s]{'src; v. 'e[v]} requires that v be the
specific register mentioned in the parameter. The register
allocator currently ignores them.
Once that's done, we need to pull the runtime from the class version
of fjava, and try compiling for real.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-21 01:16:52 -0800 (Fri, 21 Mar 2003)
Revision: 4203
Log message:
Proved couple of rules that used to be prim.
Changes | Path |
+15 -17 | metaprl/theories/itt/itt_set.ml |
Added | metaprl/theories/itt/itt_set.prla |
Properties | metaprl/theories/itt/itt_set.prla |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-21 01:49:42 -0800 (Fri, 21 Mar 2003)
Revision: 4204
Log message:
Added a set membership elimination rule.
Alexei, what's your take on membership elimination rules in general - how should
we manage those? Proving a separate mem. elim for each elim seems somewhat silly...
Alos, what's a good place for this rule? In can not go into itt_set since it requires
itt_logic, but itt_struct2 does not seem right either.
Changes | Path |
+9 -0 | metaprl/theories/itt/itt_struct2.ml |
+4285 -4430 | metaprl/theories/itt/itt_struct2.prla |
Changes by: ( at unknown.email)
Date: 2003-03-21 01:49:42 -0800 (Fri, 21 Mar 2003)
Revision: 4205
Log message:
This commit was manufactured by cvs2svn to create tag
'meta-prl-0_7_2'.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-21 04:25:13 -0800 (Fri, 21 Mar 2003)
Revision: 4206
Log message:
*** WARNING ***
*** This check-in breaks .prlb compatibility ***
*** Export any unsaved proofs to .prla ***
*** before running "cvs update" ***
- I have added support for Hypothesis that do not introduce a binding variable.
I do not yet have much I/O support for them - there is no mechanism yet for
inputing such hypotheses explicitly and the display format for them is not
ideal. I will add this later.
Note:
1) A hypothesis with an unused binding and a hypothesis without a binding are
considered equivalent (e.g. in alpha equality comparisons).
2) All the sequents on _rule box_ level are normalized. When a sequent is normalized,
all the binding hyps that introduce an unused variable a changed to non-binding ones
(ideally - for some reason it does yet not always work correctly).
3) If you want to give a name to a no-name variable, use nameHypT or nameHypsT tactic.
Note that you have to use that name right away in the _same_ rulebox, otherwise
normalization (see note 2) will remove it again!
I plan to extend nameHypT/nameHypsT to be able to rename existing bindings as well,
but have not done that yet.
- I have changed the way variable names are handled in the rewriter - now it
will watch them much more carefully, most of the bugs related to the variable
names are fixed (but a few still remain).
- I fixed all the proofs these changes broke (there were quite a few), except for
/itt_record/recordEliminationL
- vnewname: now "v_1" would be renamed into "v_2", not "v_1_1"
- I bumped MetaPRL version number to 0.8
- I bumped the ASCII version number to 1.0.1 (it will still read the old 1.0.0,
but will correctly mark all new .prla as 1.0.1 which will prevent the old MetaPRL
from thinking it can read them)
- I changed the magic numbers for .prlb (or at least I hope I did it correctly), so
hopefully it will ignore any old .prlb you might have and not crash on them.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-21 05:24:23 -0800 (Fri, 21 Mar 2003)
Revision: 4207
Log message:
Print a meaningful error message (as opposed to uncaught Not_found) when
a magic number does not match.
Changes | Path |
+17 -2 | metaprl/filter/base/filter_exn.ml |
+2 -2 | metaprl/mllib/file_base.ml |
+9 -3 | metaprl/mllib/file_type_base.ml |
+3 -2 | metaprl/mllib/file_type_base.mli |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-21 09:45:08 -0800 (Fri, 21 Mar 2003)
Revision: 4208
Log message:
Restored the TESTS=yes compile.
Changes | Path |
+9 -9 | metaprl/editor/ml/tests/prop-pigeon.ml |
+4 -4 | metaprl/theories/itt/itt_test.ml |
+1 -1 | metaprl/theories/itt/jprover_tests.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-21 09:51:49 -0800 (Fri, 21 Mar 2003)
Revision: 4209
Log message:
Tailcalls take a variable number of arguments.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-21 11:08:54 -0800 (Fri, 21 Mar 2003)
Revision: 4210
Log message:
Added memory reserve statements to control GC.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-21 21:34:25 -0800 (Fri, 21 Mar 2003)
Revision: 4211
Log message:
| Register allocator now adheres to calling convention.
| We need some work on the spills:
| 1. The choice of variables to spill is very bad (this
| must be affecting us in mcc as well).
| 2. The MetaPRL spill code needs to be smarter. Given a
| variable "v" to spill, it splits the live range of that
| variable by inserting this code in a random location,
| then trying to migrate the top move up, and the bottom
| move down.
|
| MOV %v, spill
| MOV spill, %v
|
| The choice of the location is currently quite bad.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-21 21:44:50 -0800 (Fri, 21 Mar 2003)
Revision: 4212
Log message:
- Be even more strict about the variable name clashes (now that hypSubstT
in Itt_record.recordEliminationL will not be allowed).
- I reordered the args in StringSet.remove to be able to use fold_left
instead of fold_right.
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-22 08:29:26 -0800 (Sat, 22 Mar 2003)
Revision: 4213
Log message:
Prove of int_div_rem is 80% complete.
Only some local problems with my normalizeC/arithT remain.
Changes | Path |
+5703 -4288 | metaprl/theories/itt/itt_nat.prla |
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-22 09:22:43 -0800 (Sat, 22 Mar 2003)
Revision: 4214
Log message:
Forgot two important branches.
For some reason "export" raises "Not found" exception so I deleted previous
.prla file before exporting.
Changes | Path |
+9196 -8593 | metaprl/theories/itt/itt_nat.prla |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-22 14:17:51 -0800 (Sat, 22 Mar 2003)
Revision: 4215
Log message:
One very important thing I forgot in the last commit: I converted
the assembly to use three-operand register instructions. What this
really means is that all assignments to registers are binding
occurrences. In other words, registers are immutable, and we can
call this functional assembly code (an oxymoron if I ever heard one).
This allows much greater flexibility in assembly code motion.
The printer knows how to print these so that they are valid.
For example, the instruction ADD (src1, src2, dst) prints as:
add src1, dst /* if src2 = dst (dst is a register) */
or
mov src2, dst /* if src1 != dst */
add src1, dst
This commit makes a change to Jcc. They now look like regular
conditionals, so the assembly has the usual tree structure.
The printer knows how to print so that this is actually valid
assembly.
This change makes it easier to move code into the branches of
the conditional.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-22 17:15:41 -0800 (Sat, 22 Mar 2003)
Revision: 4216
Log message:
We now generate code the assembles without errors.
Added a few assembly optimizations too.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-22 19:08:38 -0800 (Sat, 22 Mar 2003)
Revision: 4217
Log message:
Implemented Alexei's suggestion for the meaning of negative seq. context
argumets in the rewriter.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-22 20:40:13 -0800 (Sat, 22 Mar 2003)
Revision: 4218
Log message:
| Yes! Our first working program is fib! It is just about as fast
| as the MCC version from what I can tell.
|
| 1. Added the runtime.
| 2. Random fixes in the other files. Especially, CPS was broken
| (and we should examine it carefully soon), and Reserve was
| computed wrong.
| 3. Integer arithmetic is now 31-bit so the GC can tell between a
| pointer and integer. Need to examine the shifting in M_x86_codegen.
| I think Shift will fail on some operands.
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-23 02:24:53 -0800 (Sun, 23 Mar 2003)
Revision: 4219
Log message:
Updated the M parser to use records for functions.
Also added the fib test case in m_test.ml.
Changes | Path |
+103 -56 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_ast.pho |
+12 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_test.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-23 07:02:37 -0800 (Sun, 23 Mar 2003)
Revision: 4220
Log message:
*** WARNING ***
*** This breaks prlb compatibility again! ***
- I got rid of the variable arguments to rules and rewrites. Now the rewriter
handles all the variable naming internally.
- All the maybe_new_vars1 - maybe_new_vars9 functions are gone. When creating
a bind term by var_subst, instead of creating a new variable for it, call
Perv.var_subst_to_bind
- To rename a variable in a hypothesis, use the nameHypT tactic - it
now can handle both binding and binding-free hyps.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-23 12:23:15 -0800 (Sun, 23 Mar 2003)
Revision: 4221
Log message:
Got rid of the signature files that were only used once.
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-23 13:41:36 -0800 (Sun, 23 Mar 2003)
Revision: 4222
Log message:
In process of completing itt_nat/int_div_rem and debugging the arith tactic.
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-23 14:42:38 -0800 (Sun, 23 Mar 2003)
Revision: 4223
Log message:
Some typo fixes to documentation. (I'm tempted to go through and
fill out all the documentation, but probably it is not worth it
just yet.) Minor tweak to the LetTuple display form.
I think I get the basics of the IR. Only one statement really
confuses me at this point: ``For now, variables are represented
as variables; we don't need separate atoms.'' Then what is
AtomVar for? I am guessing that that statement is out-of-date.
Changes | Path |
+2 -2 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_ir.ml |
+2 -2 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_ir.mli |
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-23 18:02:46 -0800 (Sun, 23 Mar 2003)
Revision: 4224
Log message:
"subgoals" debug variable added.
Changes | Path |
+14 -4 | metaprl/theories/itt/itt_int_arith.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-23 18:17:07 -0800 (Sun, 23 Mar 2003)
Revision: 4225
Log message:
Enforce the type distinctions in Ascii_io (it used to be unnecessary,
but now an entry for a binding-less hyp can look exactly the same as
an entry for a binding-less bterm!). This was what cause the "Not_found"
problems when re-exporting some .prla files.
Changes | Path |
+7 -6 | metaprl/refiner/reflib/ascii_io.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-23 18:55:45 -0800 (Sun, 23 Mar 2003)
Revision: 4226
Log message:
eflush should be passed via %t, no need to call it on a separate line.
Changes | Path |
+1 -2 | metaprl/doc/htmlman/developer-guide/debugging.html |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-23 19:46:10 -0800 (Sun, 23 Mar 2003)
Revision: 4227
Log message:
| Whew! Fixed the register allocator, and completely rewrote the spill code.
| Current allocation performance is excellent. We might try to migrate these
| ideas into mcc, though it may not be as easy to express.
|
| The main spill algorithm is this.
|
| 1. Remember that the assembly has only one def for each var.
| Remember we are pretending that we have three-operand instructions.
|
| 2. When a variable $v$ is selected for spilling:
| a. add this code after the def:
|
| add src1,src2,%v // For example
| mov %v, spill[name] // This is a binding occurrence of name
|
| b. convert all following uses of the variable
| to the operand (SpillRegister (v, name)).
|
| c. add a mov before each use
|
| mov SpillRegister (v, name), v'
|
| and daisy-chain these mov's so that there is at most two
| uses for each spilled var (one for the real use, and one
| to copy the daisy-chain).
|
| The meaning of SpillRegister (v, name) is that the value is in _both_
| the register and the spill area. The reason for the daisy-chain is so
| the register allocator can now spill only part of a live range.
|
| 3. If the register allocator decides to spill part of a daisy-chain,
| that part looks like this:
|
| mov %v, spill[name]
| ...lots of code that does not use v...
| mov SpillRegister (v, name), v'
| add 'v, dst
|
| Convert this code to:
|
| mov %v, spill[name]
| ...lots of code that does not use v...
| mov spill[name], v'
| add v', dst
|
| That is, forget that the spill was ever in the register, and
| just fetch it from the memory area.
|
| Spill variables have to be coalesced just like registers, but I
| haven't gotten to it.
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-23 20:34:05 -0800 (Sun, 23 Mar 2003)
Revision: 4228
Log message:
Minor modifications of debug and error reporting code according to Aleksey's
recommendations.
Changes | Path |
+11 -15 | metaprl/theories/itt/itt_int_arith.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-23 22:30:32 -0800 (Sun, 23 Mar 2003)
Revision: 4229
Log message:
Fixed a bug in alpha equality...
Changes | Path |
+5 -4 | metaprl/refiner/term_ds/term_subst_ds.ml |
+1703 -1601 | metaprl/theories/itt/ctt_markov.prla |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-24 11:24:39 -0800 (Mon, 24 Mar 2003)
Revision: 4230
Log message:
Killing the tab characters in the README file and adding my
perpetual nitpick about deadcode-elim and semantics. (-:
Also adding rewrites for AtomTrue and AtomFalse to CPS.
Perhaps these were just overlooked?
Two more comments now:
(1) (minor) Why isn't M_ast / M_post_parsing compiled at all?
(2) I thought I understood AtomFunVar, but now I'm confused. It
appears at the AST level, in closure conversion, and in codegen.
Why is there no mention of it in CPS? Or perhaps, more to the
point: when and where is AtomFunVar supposed to be used?
Changes | Path |
+11 -3 | metaprl-branches/lm_libmojave/theories/experimental/compile/README |
+10 -4 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_cps.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-25 11:28:59 -0800 (Tue, 25 Mar 2003)
Revision: 4231
Log message:
I think CPS conversion is fine... fixing a typo in the section
where the rewrites are added to the reduce resource, and putting a
BUG marker next to a comment that seems to be out of date.
Also, tweaking the display form for TailCall so that there is a
space between the function being called and the list of arguments.
Changes | Path |
+4 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_cps.ml |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_ir.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-25 14:35:48 -0800 (Tue, 25 Mar 2003)
Revision: 4232
Log message:
Starting the paper.
The source text is in theories/experimental/compile/m_doc_*.ml.
To build the paper, this is what you do:
1. In the main meta-prl directory, run "make"
2. Go to doc/latex/theories, run "make compile"
3. View output in m-paper.ps
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-25 14:44:58 -0800 (Tue, 25 Mar 2003)
Revision: 4233
Log message:
Aleksey & Yegor:
- Fixed a term comparison bug in arith
- Cleaned up the arith signature
Changes | Path |
+37 -41 | metaprl/refiner/reflib/arith.ml |
+8 -112 | metaprl/refiner/reflib/arith.mli |
+11 -6 | metaprl/theories/itt/itt_int_arith.ml |
+263 -557 | metaprl/theories/itt/itt_nat.prla |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-25 16:07:58 -0800 (Tue, 25 Mar 2003)
Revision: 4234
Log message:
Wrote something about the IR.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-25 19:48:39 -0800 (Tue, 25 Mar 2003)
Revision: 4235
Log message:
Added a section on CPS conversion.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-25 21:21:28 -0800 (Tue, 25 Mar 2003)
Revision: 4236
Log message:
Added a section on closure conversion.
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-25 23:30:36 -0800 (Tue, 25 Mar 2003)
Revision: 4237
Log message:
Thought a bit more about CPS conversion. I think I get what's
going on now. I added a few comments, including a response to
JYH's bug marker.
Changes | Path |
+9 -2 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_cps.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-25 23:36:43 -0800 (Tue, 25 Mar 2003)
Revision: 4238
Log message:
Fixing an obvious mistake in the label/heading for the
closure conversion section of the paper.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_closure.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-26 11:33:30 -0800 (Wed, 26 Mar 2003)
Revision: 4240
Log message:
Make util _before_ libmojave - this is the only way to make sure
ocamldep exists when we start compiling libmojave (I want
"make clean; make [opt]" to work without needing an extra "make depend"
step).
Changes | Path |
+6 -5 | metaprl-branches/lm_libmojave/Makefile |
Changes by: Lori Lorigo (lorigo at sbcglobal.net)
Date: 2003-03-26 12:21:03 -0800 (Wed, 26 Mar 2003)
Revision: 4241
Log message:
Updated the connection to the nuprl library to reflect nuprl-side changes. Jprover now connects directly to the nuprl refiner and uses a single connection instead of 2. The connection command is a comment at the top of editor/ml/nuprl_run.mli.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-26 13:24:45 -0800 (Wed, 26 Mar 2003)
Revision: 4242
Log message:
Added initial empty section on x86 asm.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-26 13:52:35 -0800 (Wed, 26 Mar 2003)
Revision: 4243
Log message:
Have to be a lot more aggressive about using @docoff to avoid
extra white space.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-26 15:36:13 -0800 (Wed, 26 Mar 2003)
Revision: 4244
Log message:
Added a little moreon assembly.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/doc/latex/theories/m-paper.tex |
+2 -2 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_x86_asm.ml |
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-26 17:03:47 -0800 (Wed, 26 Mar 2003)
Revision: 4245
Log message:
Added a good chunk of the parsing section. I need to say something
about currying of functions after parsing.
Changes | Path |
+45 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_parsing.ml |
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-26 17:52:11 -0800 (Wed, 26 Mar 2003)
Revision: 4246
Log message:
Added ACM templates. Changed the title of the paper to 'Towards
Formal Compilers.'
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-26 18:27:40 -0800 (Wed, 26 Mar 2003)
Revision: 4247
Log message:
Adding a few links at the end of the introduction. The main one
to look at, I think, is the one about the FreshML project.
Note that the formatting in my comment is hiddious since I can't
seem to make a verbatim environment within an @comment work
properly, and I have no idea how to make a simple tilde character
show up...
Oh yeah, is it me, or does the paper no longer build anymore?
I haven't found a way of hacking doc/latex/theories/m-paper.tex
to the point where it compiles...
Changes | Path |
+11 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-26 19:09:56 -0800 (Wed, 26 Mar 2003)
Revision: 4248
Log message:
Added a section on the assembly.
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-26 20:54:04 -0800 (Wed, 26 Mar 2003)
Revision: 4249
Log message:
Cleaned out the main tex file.
Changes | Path |
+16 -22 | metaprl-branches/lm_libmojave/doc/latex/theories/m-paper.tex |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-26 22:26:32 -0800 (Wed, 26 Mar 2003)
Revision: 4250
Log message:
Added a few comments and a few citations (note - comments are agains the version
I had with me on the plane, some may no longer apply).
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-26 23:24:49 -0800 (Wed, 26 Mar 2003)
Revision: 4251
Log message:
Added a section on code generation.
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-27 01:04:41 -0800 (Thu, 27 Mar 2003)
Revision: 4252
Log message:
Added currying to the parsing section.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-27 01:13:52 -0800 (Thu, 27 Mar 2003)
Revision: 4253
Log message:
Added a section on register allocation.
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-27 06:02:09 -0800 (Thu, 27 Mar 2003)
Revision: 4254
Log message:
1.Some bugs are fixed in polynomial normalization.
2.itt_nat/int_div_rem is now complete but could be shortened as soon as arithT will be tought to deal with equality in conclusion.
3.Expanding of this prove takes noticable time on my computer - 3.8 secs, hope arithT has some potential for optimization.
Changes | Path |
+118 -103 | metaprl/theories/itt/itt_int_arith.ml |
+1 -1 | metaprl/theories/itt/itt_int_arith.mli |
+2222 -1075 | metaprl/theories/itt/itt_nat.prla |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-27 09:30:03 -0800 (Thu, 27 Mar 2003)
Revision: 4255
Log message:
Adding a comment at the end of the introduction about some
thoughts I had about why the paper doesn't ``feel right'' to me.
Changes | Path |
+7 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-27 10:25:39 -0800 (Thu, 27 Mar 2003)
Revision: 4256
Log message:
Adding another link to a paper (at the end of the introduction).
Changes | Path |
+7 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-27 14:13:21 -0800 (Thu, 27 Mar 2003)
Revision: 4257
Log message:
Added some sections on optimization.
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-27 14:57:00 -0800 (Thu, 27 Mar 2003)
Revision: 4258
Log message:
Typo fixes only.
Changes | Path |
+3 -3 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-27 15:29:03 -0800 (Thu, 27 Mar 2003)
Revision: 4259
Log message:
I'm switching the paper to the alternate LaTeX style
(sig-alternate.cls). The paper doesn't look as odd
(e.g., no weird gaps between paragraphs) anymore, and
there's less wasted space.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/doc/latex/theories/m-paper.tex |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-27 15:43:15 -0800 (Thu, 27 Mar 2003)
Revision: 4260
Log message:
A typo fix in x86_asm (bad label for the section), and adding some
comments to the Terminology section.
Changes | Path |
+6 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_x86_asm.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-27 16:01:01 -0800 (Thu, 27 Mar 2003)
Revision: 4261
Log message:
Adding some comments (mostly questions) to the section on the IR.
Changes | Path |
+11 -4 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_ir.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-27 16:24:45 -0800 (Thu, 27 Mar 2003)
Revision: 4262
Log message:
Put the \thanks in the right place.
Changes | Path |
+9 -11 | metaprl-branches/lm_libmojave/doc/latex/theories/m-paper.tex |
+13 -12 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-27 16:53:53 -0800 (Thu, 27 Mar 2003)
Revision: 4263
Log message:
Files now pass the MetaPRL spell checker. Note, text in @comment{}
blocks is not spell-checked.
For now, when you add text, try compiling with the MP_DEBUG=spell
environment variable.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-27 18:14:28 -0800 (Thu, 27 Mar 2003)
Revision: 4264
Log message:
Make sure it actually compiles with MP_DEBUG=spell. The problem was that the
emacs-oriented comment in the end of each file had been the (*!...*) doc comment,
so the system tried to spell check it. I moved the @docoff's out of it (note - I
may have been too agressive with @docoff placement, do we want the tactics and
resource code to be displayed or @docoff'ed?) and made them be normal comments.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-27 19:06:19 -0800 (Thu, 27 Mar 2003)
Revision: 4265
Log message:
Revised the introduction.
Changes | Path |
+51 -25 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-27 19:21:09 -0800 (Thu, 27 Mar 2003)
Revision: 4266
Log message:
Few typos and fixes.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-27 20:16:44 -0800 (Thu, 27 Mar 2003)
Revision: 4267
Log message:
Modified more of the introduction.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-27 20:26:16 -0800 (Thu, 27 Mar 2003)
Revision: 4268
Log message:
Modified the introduction.
Changes | Path |
+8 -4 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-27 20:34:11 -0800 (Thu, 27 Mar 2003)
Revision: 4269
Log message:
Inlining didn't do anything for boolean values and conditions, so
I added the necessary rewrites. There's a comment about this in
the optimization section of the paper, as well.
Changes | Path |
+5 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_opt.ml |
+18 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_inline.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-27 21:00:55 -0800 (Thu, 27 Mar 2003)
Revision: 4270
Log message:
Modified the IR section.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
+31 -14 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_ir.ml |
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-27 21:07:07 -0800 (Thu, 27 Mar 2003)
Revision: 4271
Log message:
Added some related work.
Changes | Path |
+19 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-27 21:10:32 -0800 (Thu, 27 Mar 2003)
Revision: 4272
Log message:
- Make sure the system compiles with TESTS=yes (these fixes are already on
the trunk, but not yet on this branch).
- More MP_DEBUG=spell fixes (in lot of theories - we definitely need to start
thinking when and how we are planning to merge things back to the trunk).
- A few comment updates.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-27 21:31:01 -0800 (Thu, 27 Mar 2003)
Revision: 4273
Log message:
A few comments.
Changes | Path |
+11 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-27 21:52:52 -0800 (Thu, 27 Mar 2003)
Revision: 4274
Log message:
Changes to the CPS section.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-27 21:59:29 -0800 (Thu, 27 Mar 2003)
Revision: 4275
Log message:
Another comment.
Changes | Path |
+3 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-27 22:12:48 -0800 (Thu, 27 Mar 2003)
Revision: 4276
Log message:
Wider comments.
Changes | Path |
+2 -0 | metaprl-branches/lm_libmojave/doc/latex/theories/m-paper.tex |
+1 -3 | texinputs-branches/lm_libmojave/metaprl.tex |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-27 22:22:58 -0800 (Thu, 27 Mar 2003)
Revision: 4277
Log message:
Deleting my comments in the Terminology section since they are no
longer relevant. Also made a few typo fixes. I think this
section looks good.
Changes | Path |
+3 -9 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-27 23:01:20 -0800 (Thu, 27 Mar 2003)
Revision: 4278
Log message:
Modified closure conversion.
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-27 23:04:04 -0800 (Thu, 27 Mar 2003)
Revision: 4279
Log message:
Added M syntax table and some fixes to the parsing text.
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-27 23:05:32 -0800 (Thu, 27 Mar 2003)
Revision: 4280
Log message:
Adding a response to one of Nogin's comments in the introduction.
My suggested text is far from perfect, but it is an idea that I
had.
Changes | Path |
+13 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-27 23:19:00 -0800 (Thu, 27 Mar 2003)
Revision: 4281
Log message:
Adding my comments on the current version of the parsing section.
Changes | Path |
+17 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_parsing.ml |
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-27 23:20:58 -0800 (Thu, 27 Mar 2003)
Revision: 4282
Log message:
Added a comment on naming intermediate values.
Changes | Path |
+3 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_ir.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-27 23:27:41 -0800 (Thu, 27 Mar 2003)
Revision: 4283
Log message:
My apologies if this conflicts with anyone else's idea or work on
the conclusion. I just wanted to put a few comments about what I
thought should go into the conclusion, so I added that section to
the paper.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-28 00:11:40 -0800 (Fri, 28 Mar 2003)
Revision: 4284
Log message:
I'm off to get intermittent sleep for the night.
I've finished a pass through all the sections, but still need conclusion/summary.
Adam, I added a comment in the parsing section.
See you all tomorrow.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 01:38:03 -0800 (Fri, 28 Mar 2003)
Revision: 4285
Log message:
Added a bunch of comments.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 01:39:54 -0800 (Fri, 28 Mar 2003)
Revision: 4286
Log message:
Typo.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 01:46:17 -0800 (Fri, 28 Mar 2003)
Revision: 4287
Log message:
Some thoughs on the conclusion. Or summary? Is there a good distinction
between the two, or should we merge them into a single section?
Changes | Path |
+5 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_conclusion.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 03:32:57 -0800 (Fri, 28 Mar 2003)
Revision: 4288
Log message:
Started writing a brief cheat-list on semantics of compiler expressions.
Changes | Path |
Added | metaprl-branches/lm_libmojave/theories/experimental/compile/semantics.txt |
Properties | metaprl-branches/lm_libmojave/theories/experimental/compile/semantics.txt |
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-28 06:43:15 -0800 (Fri, 28 Mar 2003)
Revision: 4289
Log message:
Unintended change slipped into reduce-resource.
This update fixes broken "itt_int_arith/test6".
Changes | Path |
+1 -1 | metaprl/theories/itt/itt_int_ext.ml |
+3 -2 | metaprl/util/check-status |
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-28 06:45:12 -0800 (Fri, 28 Mar 2003)
Revision: 4290
Log message:
Sorry, didn't want to change util/check-status.
Changes | Path |
+1 -2 | metaprl/util/check-status |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-28 08:50:17 -0800 (Fri, 28 Mar 2003)
Revision: 4291
Log message:
Minor typo fixes. Added inlining of branches in conditionals to
the IR optimization section.
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-28 09:08:27 -0800 (Fri, 28 Mar 2003)
Revision: 4292
Log message:
Perhaps a bit late to be coming up with this, but I think I
finally figured out what bugged me about the paper yesterday.
I've added a comment in the introduction that contains an idea
for the ``theme/message'' of the paper. I'm going to scan over
the rest of the paper and see if I can come up with any other
comments.
Changes | Path |
+16 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-28 10:02:32 -0800 (Fri, 28 Mar 2003)
Revision: 4293
Log message:
Killing the list of keywords, since we don't strictly need them.
We can add them back later if needed.
Fixing a typo in M_doc_opt.
Adding more to my comments in the introduction about the theme of
the paper. I think I've identified what I think is missing from
the paper / what needs to be the paper's theme, and it would be
interesting to see what others have to say.
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-28 10:17:53 -0800 (Fri, 28 Mar 2003)
Revision: 4294
Log message:
Added a couple comments to Brian's.
Changes | Path |
+12 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-28 12:02:43 -0800 (Fri, 28 Mar 2003)
Revision: 4295
Log message:
Cut down the parsing section according to Jason's comments.
Changes | Path |
+30 -77 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_parsing.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 12:19:23 -0800 (Fri, 28 Mar 2003)
Revision: 4296
Log message:
check-status should indeed go to localhost.
Changes | Path |
+2 -2 | metaprl/util/check-status |
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-28 12:21:43 -0800 (Fri, 28 Mar 2003)
Revision: 4297
Log message:
A bit more precise description of what happens during parsing.
Changes | Path |
+4 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_parsing.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-28 12:29:37 -0800 (Fri, 28 Mar 2003)
Revision: 4298
Log message:
Added some terminology. Please comment.
Changes | Path |
+36 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
+5 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_summary.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 12:58:15 -0800 (Fri, 28 Mar 2003)
Revision: 4299
Log message:
MP_DEBUG=spell fixes.
Changes | Path |
+14 -14 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
+2 -2 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_parsing.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 13:48:21 -0800 (Fri, 28 Mar 2003)
Revision: 4300
Log message:
New title.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/doc/latex/theories/m-paper.tex |
+0 -8 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Xin Yu (xiny at cs.caltech.edu)
Date: 2003-03-28 16:26:14 -0800 (Fri, 28 Mar 2003)
Revision: 4302
Log message:
Fixed the broken proof of positive_rule1.
Changes | Path |
+2696 -2285 | metaprl/theories/itt/itt_nat.prla |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-28 17:56:08 -0800 (Fri, 28 Mar 2003)
Revision: 4303
Log message:
Changed the intro.
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-28 18:29:46 -0800 (Fri, 28 Mar 2003)
Revision: 4304
Log message:
Fixing a display form so that the paper compiles.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_ir.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-28 18:40:46 -0800 (Fri, 28 Mar 2003)
Revision: 4305
Log message:
Addressing two of Justin's minor comments that we didn't get to
this afternoon.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_cps.ml |
+2 -2 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 20:59:47 -0800 (Fri, 28 Mar 2003)
Revision: 4306
Log message:
Some rough text on FreshML.
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-28 21:17:27 -0800 (Fri, 28 Mar 2003)
Revision: 4307
Log message:
Added two citations for the IR (taken from a table in Appel).
Jyh, I can add more, but the ones I added seemed to be the most
relevent (judging from the table).
Changes | Path |
+1 -7 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_ir.ml |
+7 -0 | texinputs-branches/lm_libmojave/rc.bib |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 21:17:35 -0800 (Fri, 28 Mar 2003)
Revision: 4308
Log message:
Another paper to cite.
Changes | Path |
+9 -14 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 21:37:07 -0800 (Fri, 28 Mar 2003)
Revision: 4309
Log message:
Added another paragraph and citation to the related work section
(the text is pretty rough, feel free to help me in polishing it).
Changes | Path |
+6 -3 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
+9 -1 | texinputs-branches/lm_libmojave/rc.bib |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 21:55:49 -0800 (Fri, 28 Mar 2003)
Revision: 4310
Log message:
Added the Chaitin citation.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_x86_regalloc.ml |
+12 -3 | texinputs-branches/lm_libmojave/rc.bib |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-28 22:09:57 -0800 (Fri, 28 Mar 2003)
Revision: 4311
Log message:
Added a summary section.
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-28 22:13:44 -0800 (Fri, 28 Mar 2003)
Revision: 4312
Log message:
Adding an alternate formulation of the paragraph on FreshML,
and a typo fix in the Necula paragraph (I am still thinking on how
I would word it).
Changes | Path |
+18 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 22:25:26 -0800 (Fri, 28 Mar 2003)
Revision: 4313
Log message:
Rewrote the FreshML par based on Brian's version.
Changes | Path |
+4 -17 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-28 22:31:58 -0800 (Fri, 28 Mar 2003)
Revision: 4314
Log message:
Deleting Aleksey's comment about the display forms for LetAtom and
LetClosure being identical, since it has been addressed. Also
adding a brief comment to the summary (I think ``grunge'' is fine,
and will do a closer read of the summary a little later).
Changes | Path |
+0 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_ir.ml |
+6 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_summary.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 22:54:28 -0800 (Fri, 28 Mar 2003)
Revision: 4315
Log message:
Typo.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_summary.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-28 23:12:47 -0800 (Fri, 28 Mar 2003)
Revision: 4316
Log message:
Added an abstract.
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-28 23:28:53 -0800 (Fri, 28 Mar 2003)
Revision: 4317
Log message:
Another random comment on the summary. I'm having trouble finding
any serious flaws in it, which is probably a good thing. (-:
Changes | Path |
+8 -7 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_summary.ml |
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-28 23:30:13 -0800 (Fri, 28 Mar 2003)
Revision: 4318
Log message:
Modified parsing section according to Jason's comments.
Jason, feel free to change things if you are not satisfied.
Changes | Path |
+26 -29 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_parsing.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-28 23:35:04 -0800 (Fri, 28 Mar 2003)
Revision: 4319
Log message:
Commented and uncommented in a few places.
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-28 23:49:01 -0800 (Fri, 28 Mar 2003)
Revision: 4320
Log message:
I have this paragraph that I want to write about work that has
been done on augmenting compilers written in general-purpose
programming languages with formal tools. I can't seem to find
a good way of stating anything, so if this is a topic that's not
worth pursuing, I'd like to know that. (-:
Changes | Path |
+9 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-28 23:57:30 -0800 (Fri, 28 Mar 2003)
Revision: 4321
Log message:
Various minor fixes.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-29 00:01:02 -0800 (Sat, 29 Mar 2003)
Revision: 4322
Log message:
Added a par on Liang's Lambda-Prolog compiler. It seems to be _very_
close to what we do, it includes parsing, CPS and assembly (Sparc).
I do not think I've said enough, so if anybody else is willing to take
a look - http://www.cs.hofstra.edu/~cscccl/hocompiler/padl02.ps.gz ,
it would probably be a good idea.
Changes | Path |
+7 -3 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
+9 -0 | texinputs-branches/lm_libmojave/rc.bib |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-29 00:01:47 -0800 (Sat, 29 Mar 2003)
Revision: 4323
Log message:
A typo fix to the abstract, and a comment on it (I must have
writer's block or something --- I am not completely happy with the
final sentence, but I can't come up with a replacement that makes
me happy).
Changes | Path |
+6 -1 | metaprl-branches/lm_libmojave/doc/latex/theories/m-paper.tex |
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-29 00:05:37 -0800 (Sat, 29 Mar 2003)
Revision: 4324
Log message:
Added heading for IR conversion section. I am working on the text.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
+1 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_ir.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-29 00:09:32 -0800 (Sat, 29 Mar 2003)
Revision: 4325
Log message:
The ``fun'' rewrite in assembly generation still seems wrong.
Changes | Path |
+1 -0 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_x86_asm.ml |
+2 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_x86_codegen.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-29 00:30:23 -0800 (Sat, 29 Mar 2003)
Revision: 4326
Log message:
Modified the parsing section.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-29 00:59:10 -0800 (Sat, 29 Mar 2003)
Revision: 4327
Log message:
YACC citation.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_parsing.ml |
+9 -0 | texinputs-branches/lm_libmojave/rc.bib |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-29 01:08:28 -0800 (Sat, 29 Mar 2003)
Revision: 4328
Log message:
Updated the l-Prolog paragraph. Not sure it is so much better though.
The related work is starting to look better, we need a citation for the
for the ASF+SDF work.
I think I will quit for the night. Main thing we are waiting on is the
section Adam is writing on translating AST to IR.
Let's meet then at 11am. This will get us 4 hours to work things out,
though I hope we can go home before 3pm.
Changes | Path |
+17 -5 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
+2 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_x86_asm.ml |
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-29 01:53:18 -0800 (Sat, 29 Mar 2003)
Revision: 4329
Log message:
Added section on AST-IR conversion. We are pushing the 12 page limit,
so I kept it relatively short, but satisfactory I believe.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_cps.ml |
+69 -7 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_ir.ml |
Changes by: Adam Granicz (granicz at cs.caltech.edu)
Date: 2003-03-29 02:04:49 -0800 (Sat, 29 Mar 2003)
Revision: 4330
Log message:
Added citation for ASF+SDF and some text that mentions the
main difference compared to our work.
Changes | Path |
+6 -2 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
+10 -0 | texinputs-branches/lm_libmojave/rc.bib |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-29 02:17:17 -0800 (Sat, 29 Mar 2003)
Revision: 4331
Log message:
Adding a bib entry for the Mojave homepage. This I think resolves
some citations. Made IR deadcode and inlining subsubsections.
Killed some of my comments that have been addressed. Removed the
ACM copyright notice and replaced it with more suitable text (I
think).
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-29 03:18:14 -0800 (Sat, 29 Mar 2003)
Revision: 4332
Log message:
More random fixes, comment clean-up. I'm going to give up for now
on trying to finish the related work. I have some papers, but I
just can't seem to find a nice way of saying anything about
them...
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-29 08:59:25 -0800 (Sat, 29 Mar 2003)
Revision: 4333
Log message:
Still working on cleaning up the related work.
Just wanted to flush some work out, for the moment.
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-29 09:25:46 -0800 (Sat, 29 Mar 2003)
Revision: 4334
Log message:
A little blurb on typed assembly language that is not complete.
Changes | Path |
+15 -9 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-29 10:47:41 -0800 (Sat, 29 Mar 2003)
Revision: 4335
Log message:
Fixing more typos, and eliminating some redundant text.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-29 11:02:58 -0800 (Sat, 29 Mar 2003)
Revision: 4336
Log message:
Just rewrote Adam's section on AST->IR.
Be in soon.
Changes | Path |
+11 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
+83 -54 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_ir.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-29 11:09:30 -0800 (Sat, 29 Mar 2003)
Revision: 4337
Log message:
Removing the CVS conflict in the intro...
Changes | Path |
+0 -8 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: Brian Emre Aydemir (baydemir at cis.upenn.edu)
Date: 2003-03-29 11:29:53 -0800 (Sat, 29 Mar 2003)
Revision: 4338
Log message:
Fixing a bogus figure reference.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_summary.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-29 12:00:59 -0800 (Sat, 29 Mar 2003)
Revision: 4339
Log message:
Moved related work to the end.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-29 12:42:26 -0800 (Sat, 29 Mar 2003)
Revision: 4340
Log message:
Minor updates in AST->IR.
Changes | Path |
+45 -40 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_ir.ml |
+38 -26 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_parsing.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-29 12:44:17 -0800 (Sat, 29 Mar 2003)
Revision: 4341
Log message:
Oops, I committed something that didn't compile.
Changes | Path |
+4 -4 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_ir.ml |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-29 13:50:05 -0800 (Sat, 29 Mar 2003)
Revision: 4342
Log message:
Almost-final version.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-29 14:00:38 -0800 (Sat, 29 Mar 2003)
Revision: 4343
Log message:
Final?
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/doc/latex/theories/m-paper.tex |
+3 -2 | texinputs-branches/lm_libmojave/rc.bib |
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2003-03-29 14:23:24 -0800 (Sat, 29 Mar 2003)
Revision: 4344
Log message:
Forgot to commit deletion of HOAS.
Changes | Path |
+1 -1 | metaprl-branches/lm_libmojave/theories/experimental/compile/m_doc_intro.ml |
Changes by: ( at unknown.email)
Date: 2003-03-29 14:23:24 -0800 (Sat, 29 Mar 2003)
Revision: 4345
Log message:
This commit was manufactured by cvs2svn to create tag
'M_PAPER_ICFP_2003_SUBMISSION'.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-29 15:15:11 -0800 (Sat, 29 Mar 2003)
Revision: 4346
Log message:
Merged the lm_libmojave branch back to the trunk.
We have submitted the ICFP paper, so this should be the last in this
crazy flood of commit we had in the last several days.
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-29 16:52:27 -0800 (Sat, 29 Mar 2003)
Revision: 4347
Log message:
1.Now arithT can deal with equality in conclusion (now it support lt,gt,le,ge,eq). neq is still unsupported.
2.As much as possible of substitution reasoning is replaced with arithT in itt_nat/int_div_rem.
P.S. I know that number of steps is alarmingly high, I'll look into it now.
Changes | Path |
+9 -1 | metaprl/theories/itt/itt_int_arith.ml |
+15823 -13617 | metaprl/theories/itt/itt_int_arith.prla |
+279 -2380 | metaprl/theories/itt/itt_nat.prla |
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-29 18:51:38 -0800 (Sat, 29 Mar 2003)
Revision: 4348
Log message:
1.Bugfix for previous commit - equality "a=b in t" in conclusion should be processed only if t is int and a is not b.
2.Every "repeatC (sweepDnC conv)" is replaced with "repeatC (higherC conc)" they are equivalent if conv fails whenever it's not applicable (I had idC in several places). It noticably reduced number of proof steps. It also appears to be slightly faster (according to time printed by expand). It's actually strange - I expected that sweepDnC would be better - it should scan term less number of times.
Changes | Path |
+0 -15 | metaprl/Makefile |
+25 -18 | metaprl/theories/itt/itt_int_arith.ml |
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-29 18:53:14 -0800 (Sat, 29 Mar 2003)
Revision: 4349
Log message:
Oops, Makefile should not be changed (I don't know how to fix it correctly).
Changes | Path |
+15 -0 | metaprl/Makefile |
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-29 19:01:53 -0800 (Sat, 29 Mar 2003)
Revision: 4350
Log message:
Small bugfix/optimization.
Changes | Path |
+1 -1 | metaprl/theories/itt/itt_int_arith.ml |
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-29 21:33:51 -0800 (Sat, 29 Mar 2003)
Revision: 4351
Log message:
Small optimization (number of proof steps decreased but just a bit).
sweepUpC appears to be very efficient - my manual tail-recursive implementation of some rewrite is just a little bit faster.
Changes | Path |
+43 -48 | metaprl/theories/itt/itt_int_arith.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-30 14:05:47 -0800 (Sun, 30 Mar 2003)
Revision: 4352
Log message:
Yegor, does this fix your PDIRS issues?
Changes | Path |
+11 -3 | metaprl/Makefile |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-30 14:21:54 -0800 (Sun, 30 Mar 2003)
Revision: 4353
Log message:
Fixed "make latex".
Changes | Path |
+2 -1 | metaprl/theories/tactic/comment.ml |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-30 16:44:25 -0800 (Sun, 30 Mar 2003)
Revision: 4354
Log message:
- New syntax for the documentation! New syntax is
doc <:doc< ... >>
(and "doc" is a keyword!)
- Some files had documentation in .mli (there were especially many of those
in theories/experimental/compile). In most cases I just killed it...
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-30 17:17:25 -0800 (Sun, 30 Mar 2003)
Revision: 4355
Log message:
MetaPRL now compiles with unmodified version of Ocaml 3.04.
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-31 13:25:12 -0800 (Mon, 31 Mar 2003)
Revision: 4356
Log message:
1.GPL header was missing in arith.ml, arith.mli
2.Trying to prove itt_bool/not_assert_elim but don't know how (tried several formulations already)
3.negation in coclusion is now supported by arithT.
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-31 17:00:21 -0800 (Mon, 31 Mar 2003)
Revision: 4357
Log message:
Now MetaPRL requires ocaml 3.06 (get the RPMs at http://rpm.nogin.org/ocaml.html).
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-31 20:01:01 -0800 (Mon, 31 Mar 2003)
Revision: 4358
Log message:
Support for nequal in conclusion added.
Changes | Path |
+4 -2 | metaprl/theories/itt/itt_int_arith.ml |
+11 -6 | metaprl/theories/itt/itt_int_ext.ml |
+8 -4 | metaprl/theories/itt/itt_int_ext.mli |
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-31 22:12:38 -0800 (Mon, 31 Mar 2003)
Revision: 4359
Log message:
Working on support of negation in hyps.
Changes | Path |
+2 -2 | metaprl/theories/itt/itt_bool.ml |
+12 -0 | metaprl/theories/itt/itt_int_arith.ml |
Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 2003-03-31 22:15:12 -0800 (Mon, 31 Mar 2003)
Revision: 4360
Log message:
Forgot to export proofs.
Changes | Path |
+16340 -15356 | metaprl/theories/itt/itt_int_arith.prla |
Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 2003-03-31 22:38:09 -0800 (Mon, 31 Mar 2003)
Revision: 4361
Log message:
- Proved all the rules in itt_record_exm
- Refreshed a bunch of .prla to get rid of compilation warnings.