lcnr
baf076825c
inline helper methods into with_new_goal
2023-08-03 15:15:48 +02:00
lcnr
8aca388af8
rewrite stack dependent overflow handling
2023-08-03 15:12:43 +02:00
lcnr
a745cbb042
handle overflow in the EvalCtxt separately
2023-08-03 14:41:44 +02:00
lcnr
ae3c353067
fix make_ambiguous_response_no_constraints
...
we previously had incorrect universes in the query response.
2023-08-03 14:30:13 +02:00
lcnr
f1753ff8f8
refactor builtin unsize handling, extend comments
2023-07-28 13:00:54 +02:00
lcnr
303af36be7
new solver: add a separate cache for coherence
2023-07-21 09:34:10 +02:00
Boxy
284b61417f
reviews
2023-07-04 14:56:09 +01:00
Boxy
040aa58d0a
add flag for disabling global cache and printing proof trees on error
2023-07-03 21:00:16 +01:00
Boxy
7a3665d016
dont use a trait
2023-06-19 09:01:37 +01:00
Boxy
3587d4ced8
say what kind of cache hit
2023-06-19 09:01:37 +01:00
Boxy
3009b2c647
initial info dump
2023-06-19 09:01:37 +01:00
Michael Goulet
f3c9c21658
Prepopulate opaques in canonical input
2023-05-25 03:21:22 +00:00
Maybe Waffle
e496fbec92
Split {Idx, IndexVec, IndexSlice} into their own modules
2023-04-24 13:53:35 +00:00
DaniPopes
677357d32b
Fix typos in compiler
2023-04-10 22:02:52 +02:00
Michael Goulet
786fc90855
Tweak debug outputs to make debugging new solver easier
2023-04-05 03:18:29 +00:00
Dylan DPC
14157561fb
Rollup merge of #109718 - scottmcm:indexvec-last, r=Nilstrieb
...
Rename `IndexVec::last` → `last_index`
As I've been trying to replace a `Vec` with an `IndexVec`, having `last` exist on both but returning very different types makes the transition a bit awkward -- the errors are later, where you get things like "there's no `ty` method on `mir::Field`" rather than a more localized error like "hey, there's no `last` on `IndexVec`".
So I propose renaming `last` to `last_index` to help distinguish `Vec::last`, which returns an element, and `IndexVec::last_index`, which returns an index.
(Similarly, `Iterator::last` also returns an element, not an index.)
2023-03-29 14:07:31 +05:30
Scott McMurray
843c5e361e
Rename IndexVec::last → last_index
...
As I've been trying to replace a `Vec` with an `IndexVec`, having `last` exist on both but returning very different types makes the transition a bit awkward -- the errors are later, where you get things like "there's no `ty` method on `mir::Field`" rather than a more localized error like "hey, there's no `last` on `IndexVec`".
So I propose renaming `last` to `last_index` to help distinguish `Vec::last`, which returns an element, and `IndexVec::last_index`, which returns an index.
(Similarly, `Iterator::last` also returns an element, not an index.)
2023-03-29 00:27:24 -07:00
lcnr
27a3b10ed2
check for intercrate mode when accessing the cache
2023-03-28 21:45:35 +02:00
Michael Goulet
53ec4bc631
Remove some stale FIXMEs in new solver
2023-03-25 01:51:19 +00:00
lcnr
a7ec045be8
disable global caching during coherence
2023-03-21 16:38:40 +01:00
lcnr
47f24a881b
new solver cleanup + coherence
2023-03-21 16:27:25 +01:00
Boxy
ed63201224
replace usage of evaluate_goal with a new add_goal
2023-03-16 14:58:35 +00:00
Michael Goulet
d21e4d8411
Actually cache goals
2023-03-10 23:49:42 +00:00
Michael Goulet
67698aa6ad
Move some solver stuff to middle
2023-03-10 23:46:38 +00:00
bors
0416b1a6f6
Auto merge of #108056 - matthiaskrgr:rollup-oa6bxvh, r=matthiaskrgr
...
Rollup of 9 pull requests
Successful merges:
- #107573 (Update the minimum external LLVM to 14)
- #107626 (Fix `x fix` on the standard library itself)
- #107673 (update ICU4X to 1.1.0)
- #107733 (Store metrics from `metrics.json` to CI PGO timer)
- #108007 (Use `is_str` instead of string kind comparison)
- #108033 (add an unstable `#[rustc_coinductive]` attribute)
- #108039 (Refactor refcounted structural_impls via functors)
- #108040 (Use derive attributes for uninteresting traversals)
- #108044 (interpret: rename Pointer::from_addr → from_addr_invalid)
Failed merges:
r? `@ghost`
`@rustbot` modify labels: rollup
2023-02-14 21:07:04 +00:00
Santiago Pastorino
26136c6224
Reduce visibility of some items
2023-02-14 10:17:07 -03:00
Santiago Pastorino
c8dae10f14
Check for overflow in evaluate_canonical_goal
2023-02-14 09:51:39 -03:00
lcnr
51671cd435
add test for coinduction in new solver
2023-02-14 12:18:33 +01:00
Santiago Pastorino
826bee7085
Implement repeat_while_none for both SearchGraph and EvalCtxt
2023-02-13 14:45:39 -03:00
Santiago Pastorino
873c83ba56
Extract try_move_finished_goal_to_global_cache from try_finalize_goal
2023-02-13 14:45:37 -03:00
lcnr
a5164605bc
correctly update goals in the cache
2023-02-08 19:11:14 +01:00
lcnr
85e6f38e79
assert that solver results are stable
2023-01-27 10:05:14 +01:00
lcnr
31ac29d989
update project to emulate a projection cache
2023-01-18 08:11:15 +01:00
lcnr
660c28391c
remove assembly context and impl a bit more
2023-01-18 08:11:15 +01:00
lcnr
bf7dbff921
instantiate canonical vars eagerly
2023-01-18 08:11:13 +01:00
