Leonardo de Moura
af682a0981
feat(library/compiler/specialize): refine candidate selection heuristics
2018-10-16 15:50:42 -07:00
Leonardo de Moura
f80ea590ba
fix(library/compiler/specialize): typo
2018-10-16 15:50:42 -07:00
Sebastian Ullrich
f29d886c29
fix(library/init/lean/parser/string_literal): letter escapes
2018-10-16 22:43:30 +02:00
Leonardo de Moura
88de81077c
feat(library/compiler/specialize): do not specialize instances unless they are marked with [specialize]
...
@kha: I changed the specialization candidate selection procedure.
Now, instances are not considered for specializations unless we mark
them with `[specialize]`. The idea is that an instance application is
morally just creating the "dictionary" for invoking a polymorphic
function.
2018-10-15 18:29:29 -07:00
Leonardo de Moura
4cfe44bed0
feat(library/compiler/specialize): add nospecialize attribute
...
@kha I had to add this attribute because the specializer was generated
many specialization candidates for functions that take `[has_tokens ...]`
as an argument. Moreover, these candidates had a lot of
dependencies. I am trying to workaround this issue by marking the
instances with the new attribute `[nospecialize]`.
I did not mark instances created by `[derive]`. It is quite tedious to
do it.
BTW, when I was investigating the problem I stumbled at `node.view`.
Its type is:
```
node.view :
Π {α : Type} {m : Type → Type} [_inst_1 : monad m] [_inst_2 : monad_except (parsec.message syntax) m]
[_inst_3 : monad_parsec syntax m] [_inst_4 : alternative m] (k : syntax_node_kind) (rs : list (m syntax))
[i : @has_view syntax_node_kind k α], @has_view (m syntax) (@node m _inst_1 _inst_2 _inst_3 _inst_4 k rs) α
```
This looks wrong: the view depends on `[monad_parsec syntax m]`
We should also make sure definitions do not have unnecessary type
class instances. Otherwise, we will put additional stress on the code
specializer. One option is to change the frontend and filter unused
instances.
2018-10-15 17:44:26 -07:00
Leonardo de Moura
3f2c5c8b75
chore(library/compiler/csimp): fix assertion
2018-10-15 17:39:44 -07:00
Leonardo de Moura
ba55ecf063
chore(library/compiler/specialize): fix debug build
2018-10-15 17:19:28 -07:00
Leonardo de Moura
777119ceab
feat(library/compiler/specialize): collect dependencies
2018-10-15 17:15:42 -07:00
Leonardo de Moura
4f73cb18bb
feat(library/compiler/specialize): more detailed spec_info
2018-10-15 15:48:02 -07:00
Leonardo de Moura
9bd09de9e2
feat(library/compiler/specialize): find candidates for specialization
2018-10-15 15:29:38 -07:00
Leonardo de Moura
effccf9a6d
chore(library/init): mark a few combinators with [specialize]
2018-10-15 13:47:26 -07:00
Leonardo de Moura
42c056862d
feat(library/compiler/compiler): cache stage1 result before specialization
2018-10-15 13:08:48 -07:00
Leonardo de Moura
8f76df4823
feat(library/compiler): add [specialize] attribute
2018-10-15 13:02:11 -07:00
Leonardo de Moura
9ca4c362ae
feat(library/compiler/specialize): add spec_info
...
Store which arguments can be specialized.
2018-10-15 12:54:34 -07:00
Leonardo de Moura
10b99a678c
chore(util/list_ref): add objects a list of objects
2018-10-15 11:57:12 -07:00
Leonardo de Moura
8837217a2e
chore(kernel/expr): dead decls
2018-10-15 11:10:11 -07:00
Leonardo de Moura
611af5c186
feat(library/compiler/csimp): improve heuristic for deciding whether we should apply float_cases_on or not.
2018-10-15 09:50:02 -07:00
Sebastian Ullrich
db6b9db07e
fix(tests/lean/parser1): fix reprinting (`eoi node was missing)
2018-10-15 12:47:32 +02:00
Sebastian Ullrich
a820b9955f
perf(library/init/lean/parser/term): index term parsers by leading token
...
66% speedup on core.lean
2018-10-15 10:21:08 +02:00
Sebastian Ullrich
2df885d9a3
fix(library/init/lean/parser/token): broken end of comment
2018-10-13 12:01:48 -07:00
Sebastian Ullrich
eff43fab08
chore(library/init/lean/expander): comment
2018-10-13 08:11:36 -07:00
Sebastian Ullrich
8a88d4d5e2
feat(library/init/lean/parser/token): string literals
...
hard-coded for now because we do not have general support for variable-length
tokens yet
2018-10-13 08:11:36 -07:00
Sebastian Ullrich
758d258210
fix(library/init/lean/parser/term): rename term.ident to ident_univs to remove confusion with ident
...
`protected` didn't do anything here
2018-10-13 08:11:36 -07:00
Sebastian Ullrich
fdc06b39cd
test(tests/lean/parser1): fix reprint check
2018-10-13 08:01:39 -07:00
Leonardo de Moura
5810418ddd
perf(library/init/lean/ir/extract_cpp): use derive to speedup type class resolution
...
This modification reduced processing time for this file to 1.49 secs
from 2.30 secs on my office desktop.
2018-10-12 12:27:08 -07:00
Leonardo de Moura
17d6ef3abe
chore(library/init/lean/ir/parser): minimize <|> inlining
2018-10-12 11:57:42 -07:00
Leonardo de Moura
4788c8c52c
chore(library/type_context): remove dead code
2018-10-12 11:57:41 -07:00
Leonardo de Moura
c0b76704a0
chore(library/init/lean/parser/level): typo
2018-10-12 11:56:34 -07:00
Leonardo de Moura
15a25d5aa9
chore(library/init/lean/parser): add a few comments
2018-10-11 15:54:57 -07:00
Leonardo de Moura
aff6d58659
chore(tests/lean/run/new_compiler): fix test
2018-10-10 18:38:18 -07:00
Leonardo de Moura
c74f4c16ca
feat(library/kernel,library/compiler/csimp): make sure nat.rec and nat.cases_on reduce when major premise is a nat literal
2018-10-10 18:35:15 -07:00
Leonardo de Moura
338038a05e
feat(library/init/core): add inline identity function
2018-10-10 18:17:29 -07:00
Leonardo de Moura
be0b4c998f
fix(library/compiler/lcnf): avoid unnecessary let-decl
2018-10-10 17:33:57 -07:00
Leonardo de Moura
bbb45bfc95
fix(library/compiler/csimp): another join-point management bug
2018-10-10 13:42:32 -07:00
Leonardo de Moura
18093f3f24
feat(library/compiler/csimp): eta-expand lambda-expressions during simplification
2018-10-09 19:16:41 -07:00
Leonardo de Moura
e958f20874
fix(library/compiler/csimp): beta_reduce was not preserving the "join-point" invariant
2018-10-09 19:11:16 -07:00
Leonardo de Moura
256112be5b
chore(library/compiler/lcnf): disable lambda eta-expansion during LCNF conversion
...
We should do it at `csimp`
2018-10-09 15:25:57 -07:00
Leonardo de Moura
27e6f7f424
feat(library/compiler): invoke specialize skeleton
2018-10-09 15:23:42 -07:00
Leonardo de Moura
5d726eb210
feat(library/compiler/compiler): switch to new compiler frontend
...
We also rename `vm_compiler` module to `emit_bytecode`.
We will eventually replace this module with the new IR emitter.
2018-10-08 17:38:17 -07:00
Leonardo de Moura
662c0ebb31
feat(library/compiler/compiler): cache stage1
2018-10-08 17:06:45 -07:00
Leonardo de Moura
124b4d37fe
feat(library/compiler): port simp_inductive to the new compiler stack
...
This commit also fixes a bug in the old `simp_inductive` module, and
removes now obsolete files (`compiler_step_visitor` and `old_util`).
2018-10-08 16:58:43 -07:00
Leonardo de Moura
d2fcb7af39
chore(library/compiler/compiler): add tracing to new compiler frontend
2018-10-08 15:22:36 -07:00
Leonardo de Moura
787d1ecb93
feat(library/compiler/lambda_lifting): new lambda lifting
2018-10-08 15:04:59 -07:00
Leonardo de Moura
e75d8eacb6
fix(frontends/lean/elaborator): compilation erros with g++ 4.9
2018-10-08 11:54:28 -07:00
Sebastian Ullrich
305984bab5
feat(frontends/lean/elaborator): show context of unassigned mvars
2018-10-08 09:32:41 -07:00
Sebastian Ullrich
8e0778e54a
fix(library/init/lean/parser/token): parse whitespace after raw_ident
2018-10-08 09:30:48 -07:00
Sebastian Ullrich
50e6b42f8c
fix(frontends/lean/elaborator): ensure_no_unassigned_metavars: only check mvars in parameter
...
Had forgotten to re-check the standard lib...
2018-10-07 21:11:02 -07:00
Sebastian Ullrich
aea86eb828
feat(frontends/lean/elaborator): simple, accurate mvar error position tracking
...
The accuracy of `m_last_pos` still needs to be improved at some point
2018-10-05 21:25:39 -07:00
Leonardo de Moura
eb51b9ef26
chore(library/init/core): avoid . as end-of-command and empty equations
2018-10-05 17:30:27 -07:00
Leonardo de Moura
fc62e8f3a4
chore(library/compiler): cdecl ==> comp_decl
2018-10-05 17:30:27 -07:00
