63753 Commits

Author SHA1 Message Date
Niko Matsakis
3da5daf425 change various uses of item_variances to variances_of 2017-05-05 14:34:42 -04:00
Niko Matsakis
79de56a14b remove pub modifier (and last use thereof) 2017-05-03 16:42:46 -04:00
Niko Matsakis
50df5f85d0 correct the new graphs resulting from various tests
(Now that variances are not part of signature.)
2017-05-03 16:42:07 -04:00
Niko Matsakis
e8f6fbb255 allow dep-graph assertions on fields 2017-05-03 16:42:07 -04:00
Niko Matsakis
3819ccda9c allow tests to refer to ItemVariances 2017-05-03 16:42:07 -04:00
Niko Matsakis
898f978c97 add back variance testing mechanism
make it work for traits etc uniformly
2017-05-03 16:42:07 -04:00
Niko Matsakis
4355169f8a kill the old visit_all_item_likes infrastructure 2017-05-03 16:42:07 -04:00
Niko Matsakis
4824a199ca factor variances into a proper query
There are now two queries: crate and item. The crate one computes the
variance of all items in the crate; it is sort of an implementation
detail, and not meant to be used. The item one reads from the crate one,
synthesizing correct deps in lieu of the red-green algorithm.

At the same time, remove the `variance_computed` flag, which was a
horrible hack used to force invariance early on (e.g. when type-checking
constants). This is only needed because of trait applications, and
traits are always invariant anyway. Therefore, we now change to take
advantage of the query system:

- When asked to compute variances for a trait, just return a vector
  saying 'all invariant'.
- Remove the corresponding "inferreds" from traits, and tweak the
  constraint generation code to understand that traits are always
  inferred.
2017-05-03 16:42:07 -04:00
Niko Matsakis
55412a201a track CurrentItem, not just Generics 2017-05-03 15:28:17 -04:00
Niko Matsakis
b175aef0c4 make transitive relation use a hash map 2017-05-03 15:28:16 -04:00
Niko Matsakis
488b2a3e7b add FIXME to Steal 2017-05-02 16:21:58 -04:00
Niko Matsakis
25be7983e8 remove mir_passes from Session and add a FIXME 2017-05-02 16:21:58 -04:00
Niko Matsakis
afc5acd84b fix librustc_driver 2017-05-02 16:21:58 -04:00
Niko Matsakis
e6793ac452 have borrowck fetch MIR, which will perform some errors 2017-05-02 16:21:58 -04:00
Niko Matsakis
15bc2f4ca0 remove temporary variable 2017-05-02 16:21:57 -04:00
Niko Matsakis
0afcfce8d8 update comment about heuristics 2017-05-02 16:21:57 -04:00
Niko Matsakis
b0092e89e0 move queries code into transform 2017-05-02 16:21:57 -04:00
Niko Matsakis
393fa4f1b7 rename from item_mir to optimized_mir 2017-05-02 16:21:57 -04:00
Niko Matsakis
c7023d1c2f run MIR borrowck on the validated, not optimized, MIR 2017-05-02 16:21:57 -04:00
Niko Matsakis
74b2783877 delete dead code 2017-05-02 16:21:56 -04:00
Niko Matsakis
2fa1ba3e7e pacify the mercilous tidy 2017-05-02 16:21:56 -04:00
Niko Matsakis
851a880f52 remove irrelevant comments 2017-05-02 16:21:56 -04:00
Niko Matsakis
1dd9c3e52a support inlining by asking for optimizer mir for callees
I tested this with it enabled 100% of the time, and we were able to run
mir-opt tests successfully.
2017-05-02 16:21:56 -04:00
Niko Matsakis
669d31683f simplify down to one query per pass suite 2017-05-02 16:21:55 -04:00
Niko Matsakis
9c154a67bf rip out everything but MirPass, move the logic into suites 2017-05-02 16:21:55 -04:00
Niko Matsakis
c253df5249 remove Pass and (temporarily) drop Inline 2017-05-02 16:21:55 -04:00
Niko Matsakis
69c8f9dd25 move build_mir into build directory 2017-05-02 16:21:55 -04:00
Niko Matsakis
c2cfdbbdf9 adjust privacy of various types in build 2017-05-02 14:01:37 -04:00
Niko Matsakis
c1ff10464d rename mir_map to queries and remove build_mir_for_crate 2017-05-02 14:01:37 -04:00
Niko Matsakis
532439faa1 add a README describing the whole design 2017-05-02 14:01:37 -04:00
Niko Matsakis
d9c8a2b54b use force to ensure const-qualif has been done, not read 2017-05-02 14:01:36 -04:00
Niko Matsakis
0d045d73a7 add comments to Steal and use bug! 2017-05-02 14:01:36 -04:00
Niko Matsakis
a26e966307 convert the inline pass to use the new multi result
This involves changing various details about that system,
though the basic shape remains the same.
2017-05-02 14:01:36 -04:00
Niko Matsakis
1d675ce0a3 adjust the macro to allow for multi modifier 2017-05-02 14:01:36 -04:00
Niko Matsakis
3d1095c5be introduce IntoKeyValues trait to prepare for multi-queries
The idea is that such queries will return `FxHashMap`
2017-05-02 14:01:36 -04:00
Niko Matsakis
ecc8ff9199 rework macro to prepare for more modifiers than just [pub] 2017-05-02 14:01:35 -04:00
Niko Matsakis
29263fdb54 introduce idea of "stealable" MIR
This is a more principled version of the `RefCell` we were using
before. We now allocate a `Steal<Mir<'tcx>>` for each intermediate MIR
pass; when the next pass steals the entry, any later attempts to use it
will panic (there is no way to *test* if MIR is stolen, you're just
supposed to *know*).
2017-05-02 14:01:34 -04:00
Niko Matsakis
e89a321dff rename MirPassSet to MirSuite
This seems like a better noun.
2017-05-02 14:01:01 -04:00
Niko Matsakis
2b32cb90c7 retool MIR passes completely
The new setup is as follows. There is a pipeline of MIR passes that each
run **per def-id** to optimize a particular function. You are intended
to request MIR at whatever stage you need it. At the moment, there is
only one stage you can request:

- `optimized_mir(def_id)`

This yields the final product. Internally, it pulls the MIR for the
given def-id through a series of steps. Right now, these are still using
an "interned ref-cell" but they are intended to "steal" from one
another:

- `mir_build` -- performs the initial construction for local MIR
- `mir_pass_set` -- performs a suite of optimizations and transformations
- `mir_pass` -- an individual optimization within a suite

So, to construct the optimized MIR, we invoke:

    mir_pass_set((MIR_OPTIMIZED, def_id))

which will build up the final MIR.
2017-05-02 14:01:01 -04:00
Niko Matsakis
f23a7bc98a move to only def-id passes
this temporary disables `inline`
2017-05-02 14:01:01 -04:00
Niko Matsakis
668886a6cc rewrite Passes to have sets of passes
Also, store the completed set of passes in the tcx.
2017-05-02 14:01:01 -04:00
Niko Matsakis
e9e6ccc042 introduce DefIdPass and remove all impls of Pass but Inline 2017-05-02 14:01:01 -04:00
Niko Matsakis
46b342fbc0 simplify the MirPass traits and passes dramatically
Overall goal: reduce the amount of context a mir pass needs so that it
resembles a query.

- The hooks are no longer "threaded down" to the pass, but rather run
  automatically from the top-level (we also thread down the current pass
  number, so that the files are sorted better).
  - The hook now receives a *single* callback, rather than a callback per-MIR.
- The traits are no longer lifetime parameters, which moved to the
  methods -- given that we required
  `for<'tcx>` objecs, there wasn't much point to that.
- Several passes now store a `String` instead of a `&'l str` (again, no
  point).
2017-05-02 14:01:01 -04:00
Niko Matsakis
11b6b0663a rework MirPass API to be stateless and extract helper fns 2017-05-02 14:01:00 -04:00
Niko Matsakis
0e5e2f3634 introduce mir_keys()
Each MIR key is a DefId that has MIR associated with it
2017-05-02 14:01:00 -04:00
bors
ed1f26ddda Auto merge of #41702 - frewsxcv:rollup, r=frewsxcv
Rollup of 6 pull requests

- Successful merges: #41661, #41662, #41673, #41688, #41692, #41693
- Failed merges:
2017-05-02 14:56:11 +00:00
Corey Farwell
e0bfd19add Rollup merge of #41693 - est31:anon_params_removal, r=eddyb
Removal pass for anonymous parameters

Removes occurences of anonymous parameters from the
rustc codebase, as they are to be deprecated.

See issue #41686 and RFC 1685.

r? @frewsxcv
2017-05-02 09:09:59 -04:00
Corey Farwell
4733ee7eb4 Rollup merge of #41692 - est31:anon_params, r=eddyb
Add a lint to disallow anonymous parameters

Adds a (allow by default) lint to disallow anonymous parameters, like it was decided in RFC 1685 (rust-lang/rfcs#1685).

cc tracking issue #41686
2017-05-02 09:09:58 -04:00
Corey Farwell
7dbb93a391 Rollup merge of #41688 - rust-lang:hex-value-process-exit, r=sfackler
Fix incorrect hex value in doc comment example.

Fixes https://github.com/rust-lang/rust/issues/41682.
2017-05-02 09:09:57 -04:00
Corey Farwell
b8fc516486 Rollup merge of #41673 - bitshifter:repr-align-codegen-test, r=arielb1
Add simple `[repr(align)]` codegen test.

Checks alloca and memcpy are aligned correctly.

Test added to as additional check for #33626.
2017-05-02 09:09:56 -04:00