mikros/rust
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
11,555 Commits 2 Branches 0 Tags 1.4 GiB
8a880230a5
Commit Graph

30 Commits

Author SHA1 Message Date
Florian Diebold
2077004463 Remove TypeCtor interning 
Our TypeCtor and Chalk's TypeName match now!
 2020-07-15 22:07:13 +02:00
Florian Diebold
a48843a16a Use Chalk closure support 2020-07-15 22:04:49 +02:00
Aleksey Kladov
4bbc385277 Switch to fully dynamically dispatched salsa 
This improves compile times quite a bit
 2020-07-07 10:14:48 +02:00
Jonas Schievink
6bde542a39 Split CrateImplDefs in inherent and trait impls 
This makes the intention of inherent vs. trait impls somewhat more
clear and also fixes (?) an issue where trait impls with an unresolved
trait were added as inherent impls instead (hence the test changes).
 2020-07-01 17:15:20 +02:00
Florian Diebold
8e8d2ffecb (Partially) fix handling of type params depending on type params 
If the first type parameter gets inferred, that's still not handled correctly;
it'll require some more refactoring: E.g. if we have `Thing<T, F=fn() -> T>` and
then instantiate `Thing<_>`, that gets turned into `Thing<_, fn() -> _>` before
the `_` is instantiated into a type variable -- so afterwards, we have two type
variables without any connection to each other.
 2020-06-29 16:10:20 +02:00
Jonas Schievink
ebd8233b3e Replace impls_in_trait with CrateImplDefs 2020-06-19 01:29:34 +02:00
Florian Diebold
02962b374e Implement return position impl trait / opaque type support 
This is working, but I'm not that happy with how the lowering works. We might
need an additional representation between `TypeRef` and `Ty` where names are
resolved and `impl Trait` bounds are separated out, but things like inference
variables don't exist and `impl Trait` is always represented the same
way.

Also note that this doesn't implement correct handling of RPIT *inside* the
function (which involves turning the `impl Trait`s into variables and creating
obligations for them). That intermediate representation might help there as
well.
 2020-06-05 17:08:27 +02:00
Florian Diebold
bfbc210bc1 Use Chalk's built-in representation of function item types 2020-05-22 21:05:13 +02:00
Florian Diebold
27fe68ad5c Use TypeCtorId as AdtId directly, and rename the type alias StructId -> AdtId 2020-05-22 19:52:06 +02:00
Aleksey Kladov
970dbf8717 Rename StructField -> Field 2020-04-25 14:23:34 +02:00
Florian Diebold
0be68a4825 Update Chalk, and cache Chalk env elaboration through a query 
This should fix some of the worst performance problems.
 2020-04-20 19:16:01 +02:00
Florian Diebold
a2783df3f0 Look up impls by self type 
This speeds up inference in analysis-stats by ~30% (even more with the recursive
solver).
 2020-04-11 18:04:09 +02:00
Aleksey Kladov
9f53cec1da Cleanup memory usage stats 2020-03-25 19:35:46 +01:00
Aleksey Kladov
9faea2364d Use dyn Trait for working with databse 
It improves compile time in `--release` mode quite a bit, it doesn't
really slow things down and, conceptually, it seems closer to what we
want the physical architecture to look like (we don't want to
monomorphise EVERYTHING in a single leaf crate).
 2020-03-16 17:42:30 +01:00
Aleksey Kladov
9abf0d9659 Normalize waiting queries names 2020-03-07 00:18:04 +01:00
Florian Diebold
9ce30281f6 Don't reuse the Chalk solver 
This slows down analysis-stats a bit (~5% in my measurement), but improves
incremental checking a lot because we can reuse trait solve results.
 2020-03-06 23:04:14 +01:00
Aleksey Kladov
b33b843f40 Allow specifying additional info on call to profile 2020-03-06 17:36:51 +01:00
Aleksey Kladov
59f91f2f9b Less confusing profile names 2020-03-06 16:40:38 +01:00
Aleksey Kladov
a1e1869554 Rename ast::ImplBlock -> ast::ImplDef 2020-02-29 21:33:15 +01:00
Florian Diebold
dded90a748 Formatting 2020-02-07 18:28:10 +01:00
Florian Diebold
a3d8cffde3 Use variables in predicates as well 2020-02-07 18:28:10 +01:00
Florian Diebold
ed25cf70d5 Change Ty::Param to contain param ID 2020-02-07 18:28:10 +01:00
Florian Diebold
16c6937447 Lower impl trait to variables, move away from using placeholders where they don't belong 2020-02-07 18:28:10 +01:00
Michal Terepeta
d6c2a59538 Split infer query into two for better profiling 
This is the same change as we did with `crate_def_map` and it does seem
that we mostly spend time in salsa, without recomputing much on
rust-analyzer side.

Example output:

```
 233ms - handle_inlay_hints
      163ms - get_inlay_hints
          163ms - SourceAnalyzer::new
               67ms - def_with_body_from_child_node
                   67ms - analyze_container
                       67ms - analyze_container
                           67ms - Module::from_definition
                               67ms - Module::from_file
                                   67ms - crate_def_map
                                        0ms - parse_macro_query (6 calls)
                                        0ms - raw_items_query (1 calls)
                                       66ms - ???
                            0ms - crate_def_map (1 calls)
                        0ms - crate_def_map (1 calls)
               96ms - infer
                    2ms - trait_solve_query (2 calls)
                   94ms - ???
                0ms - body_with_source_map_query (1 calls)
                0ms - crate_def_map (1 calls)
      [...]
```

Signed-off-by: Michal Terepeta <michal.terepeta@gmail.com>
 2020-01-03 19:58:11 +01:00
Florian Diebold
67a2555f6d Update Chalk, clean up Chalk integration a bit 2019-12-23 00:08:03 +01:00
Florian Diebold
1c622e9fed Add cycle recovery for type aliases 2019-11-30 12:57:32 +01:00
Florian Diebold
3ca40f7c08 Add cycle recovery for generic predicates 2019-11-30 12:57:32 +01:00
Florian Diebold
cf6809645e Handle cycles in impl types better 
- impl Trait<Self> for S is allowed
 - impl Trait for S<Self> is an invalid cycle, but we can add cycle recovery for
   it in Salsa now
 2019-11-30 12:57:32 +01:00
Aleksey Kladov
8d34696826 Memoize impl resolutions 2019-11-27 22:12:09 +03:00
Aleksey Kladov
a87579500a Move Ty 2019-11-27 21:16:00 +03:00