Auto merge of #29188 - nikomatsakis:remove-contraction, r=pnkfelix

This fixes #29048 (though I think adding better transactional support would be a better fix for that issue, but that is more difficult). It also simplifies region inference and changes the model to a pure data flow one, as discussed in [this internals thread](https://internals.rust-lang.org/t/rough-thoughts-on-the-impl-of-region-inference-mir-etc/2800). I am not 100% sure though if this PR is the right thing to do -- or at least maybe not at this moment, so thoughts on that would be appreciated.

r? @pnkfelix 
cc @arielb1
This commit is contained in:
bors 2015-10-29 11:14:27 +00:00
commit 427140f771
9 changed files with 178 additions and 452 deletions

View File

@ -61,12 +61,16 @@ impl fmt::Debug for DefId {
// Unfortunately, there seems to be no way to attempt to print
// a path for a def-id, so I'll just make a best effort for now
// and otherwise fallback to just printing the crate/node pair
try!(ty::tls::with_opt(|opt_tcx| {
if let Some(tcx) = opt_tcx {
try!(write!(f, " => {}", tcx.item_path_str(*self)));
}
Ok(())
}));
if self.is_local() { // (1)
// (1) side-step fact that not all external things have paths at
// the moment, such as type parameters
try!(ty::tls::with_opt(|opt_tcx| {
if let Some(tcx) = opt_tcx {
try!(write!(f, " => {}", tcx.item_path_str(*self)));
}
Ok(())
}));
}
write!(f, " }}")
}

View File

@ -65,7 +65,6 @@ use super::ValuePairs;
use super::region_inference::RegionResolutionError;
use super::region_inference::ConcreteFailure;
use super::region_inference::SubSupConflict;
use super::region_inference::SupSupConflict;
use super::region_inference::GenericBoundFailure;
use super::region_inference::GenericKind;
use super::region_inference::ProcessedErrors;
@ -258,13 +257,6 @@ pub trait ErrorReporting<'tcx> {
sup_origin: SubregionOrigin<'tcx>,
sup_region: Region);
fn report_sup_sup_conflict(&self,
var_origin: RegionVariableOrigin,
origin1: SubregionOrigin<'tcx>,
region1: Region,
origin2: SubregionOrigin<'tcx>,
region2: Region);
fn report_processed_errors(&self,
var_origin: &[RegionVariableOrigin],
trace_origin: &[(TypeTrace<'tcx>, TypeError<'tcx>)],
@ -313,14 +305,6 @@ impl<'a, 'tcx> ErrorReporting<'tcx> for InferCtxt<'a, 'tcx> {
sup_origin, sup_r);
}
SupSupConflict(var_origin,
origin1, r1,
origin2, r2) => {
self.report_sup_sup_conflict(var_origin,
origin1, r1,
origin2, r2);
}
ProcessedErrors(ref var_origins,
ref trace_origins,
ref same_regions) => {
@ -376,7 +360,6 @@ impl<'a, 'tcx> ErrorReporting<'tcx> for InferCtxt<'a, 'tcx> {
None => processed_errors.push((*error).clone()),
}
}
SupSupConflict(..) => processed_errors.push((*error).clone()),
_ => () // This shouldn't happen
}
}
@ -933,29 +916,6 @@ impl<'a, 'tcx> ErrorReporting<'tcx> for InferCtxt<'a, 'tcx> {
self.note_region_origin(&sub_origin);
}
fn report_sup_sup_conflict(&self,
var_origin: RegionVariableOrigin,
origin1: SubregionOrigin<'tcx>,
region1: Region,
origin2: SubregionOrigin<'tcx>,
region2: Region) {
self.report_inference_failure(var_origin);
self.tcx.note_and_explain_region(
"first, the lifetime must be contained by ",
region1,
"...");
self.note_region_origin(&origin1);
self.tcx.note_and_explain_region(
"but, the lifetime must also be contained by ",
region2,
"...");
self.note_region_origin(&origin2);
}
fn report_processed_errors(&self,
var_origins: &[RegionVariableOrigin],
trace_origins: &[(TypeTrace<'tcx>, TypeError<'tcx>)],

View File

@ -2,13 +2,12 @@ Region inference
# Terminology
Note that we use the terms region and lifetime interchangeably,
though the term `lifetime` is preferred.
Note that we use the terms region and lifetime interchangeably.
# Introduction
Region inference uses a somewhat more involved algorithm than type
inference. It is not the most efficient thing ever written though it
inference. It is not the most efficient thing ever written though it
seems to work well enough in practice (famous last words). The reason
that we use a different algorithm is because, unlike with types, it is
impractical to hand-annotate with regions (in some cases, there aren't
@ -25,22 +24,42 @@ once.
The constraints are always of one of three possible forms:
- ConstrainVarSubVar(R_i, R_j) states that region variable R_i
must be a subregion of R_j
- ConstrainRegSubVar(R, R_i) states that the concrete region R
(which must not be a variable) must be a subregion of the variable R_i
- ConstrainVarSubReg(R_i, R) is the inverse
- `ConstrainVarSubVar(Ri, Rj)` states that region variable Ri must be
a subregion of Rj
- `ConstrainRegSubVar(R, Ri)` states that the concrete region R (which
must not be a variable) must be a subregion of the variable Ri
- `ConstrainVarSubReg(Ri, R)` states the variable Ri shoudl be less
than the concrete region R. This is kind of deprecated and ought to
be replaced with a verify (they essentially play the same role).
In addition to constraints, we also gather up a set of "verifys"
(what, you don't think Verify is a noun? Get used to it my
friend!). These represent relations that must hold but which don't
influence inference proper. These take the form of:
- `VerifyRegSubReg(Ri, Rj)` indicates that Ri <= Rj must hold,
where Rj is not an inference variable (and Ri may or may not contain
one). This doesn't influence inference because we will already have
inferred Ri to be as small as possible, so then we just test whether
that result was less than Rj or not.
- `VerifyGenericBound(R, Vb)` is a more complex expression which tests
that the region R must satisfy the bound `Vb`. The bounds themselves
may have structure like "must outlive one of the following regions"
or "must outlive ALL of the following regions. These bounds arise
from constraints like `T: 'a` -- if we know that `T: 'b` and `T: 'c`
(say, from where clauses), then we can conclude that `T: 'a` if `'b:
'a` *or* `'c: 'a`.
# Building up the constraints
Variables and constraints are created using the following methods:
- `new_region_var()` creates a new, unconstrained region variable;
- `make_subregion(R_i, R_j)` states that R_i is a subregion of R_j
- `lub_regions(R_i, R_j) -> R_k` returns a region R_k which is
the smallest region that is greater than both R_i and R_j
- `glb_regions(R_i, R_j) -> R_k` returns a region R_k which is
the greatest region that is smaller than both R_i and R_j
- `make_subregion(Ri, Rj)` states that Ri is a subregion of Rj
- `lub_regions(Ri, Rj) -> Rk` returns a region Rk which is
the smallest region that is greater than both Ri and Rj
- `glb_regions(Ri, Rj) -> Rk` returns a region Rk which is
the greatest region that is smaller than both Ri and Rj
The actual region resolution algorithm is not entirely
obvious, though it is also not overly complex.
@ -54,14 +73,6 @@ Alternatively, you can call `commit()` which ends all snapshots.
Snapshots can be recursive---so you can start a snapshot when another
is in progress, but only the root snapshot can "commit".
# Resolving constraints
The constraint resolution algorithm is not super complex but also not
entirely obvious. Here I describe the problem somewhat abstractly,
then describe how the current code works. There may be other, smarter
ways of doing this with which I am unfamiliar and can't be bothered to
research at the moment. - NDM
## The problem
Basically our input is a directed graph where nodes can be divided
@ -83,31 +94,20 @@ Before resolution begins, we build up the constraints in a hashmap
that maps `Constraint` keys to spans. During resolution, we construct
the actual `Graph` structure that we describe here.
## Our current algorithm
## Computing the values for region variables
We divide region variables into two groups: Expanding and Contracting.
Expanding region variables are those that have a concrete region
predecessor (direct or indirect). Contracting region variables are
all others.
The algorithm is a simple dataflow algorithm. Each region variable
begins as empty. We iterate over the constraints, and for each constraint
we grow the relevant region variable to be as big as it must be to meet all the
constraints. This means the region variables can grow to be `'static` if
necessary.
We first resolve the values of Expanding region variables and then
process Contracting ones. We currently use an iterative, fixed-point
procedure (but read on, I believe this could be replaced with a linear
walk). Basically we iterate over the edges in the graph, ensuring
that, if the source of the edge has a value, then this value is a
subregion of the target value. If the target does not yet have a
value, it takes the value from the source. If the target already had
a value, then the resulting value is Least Upper Bound of the old and
new values. When we are done, each Expanding node will have the
smallest region that it could possibly have and still satisfy the
constraints.
## Verification
We next process the Contracting nodes. Here we again iterate over the
edges, only this time we move values from target to source (if the
source is a Contracting node). For each contracting node, we compute
its value as the GLB of all its successors. Basically contracting
nodes ensure that there is overlap between their successors; we will
ultimately infer the largest overlap possible.
After all constraints are fully propoagated, we do a "verification"
step where we walk over the verify bounds and check that they are
satisfied. These bounds represent the "maximal" values that a region
variable can take on, basically.
# The Region Hierarchy

View File

@ -16,19 +16,16 @@ pub use self::UndoLogEntry::*;
pub use self::CombineMapType::*;
pub use self::RegionResolutionError::*;
pub use self::VarValue::*;
use self::Classification::*;
use super::{RegionVariableOrigin, SubregionOrigin, TypeTrace, MiscVariable};
use rustc_data_structures::graph::{self, Direction, NodeIndex};
use middle::free_region::FreeRegionMap;
use middle::region;
use middle::ty::{self, Ty};
use middle::ty::{BoundRegion, FreeRegion, Region, RegionVid};
use middle::ty::{ReEmpty, ReStatic, ReFree, ReEarlyBound};
use middle::ty::{ReLateBound, ReScope, ReVar, ReSkolemized, BrFresh};
use middle::ty::error::TypeError;
use middle::ty::relate::RelateResult;
use util::common::indenter;
use util::nodemap::{FnvHashMap, FnvHashSet};
@ -50,6 +47,8 @@ pub enum Constraint {
ConstrainRegSubVar(Region, RegionVid),
// Region variable is subregion of concrete region
//
// FIXME(#29436) -- should be remove in favor of a Verify
ConstrainVarSubReg(RegionVid, Region),
}
@ -144,15 +143,6 @@ pub enum RegionResolutionError<'tcx> {
SubregionOrigin<'tcx>, Region,
SubregionOrigin<'tcx>, Region),
/// `SupSupConflict(v, origin1, r1, origin2, r2)`:
///
/// Could not infer a value for `v` because `v <= r1` (due to
/// `origin1`) and `v <= r2` (due to `origin2`) and
/// `r1` and `r2` have no intersection.
SupSupConflict(RegionVariableOrigin,
SubregionOrigin<'tcx>, Region,
SubregionOrigin<'tcx>, Region),
/// For subsets of `ConcreteFailure` and `SubSupConflict`, we can derive
/// more specific errors message by suggesting to the user where they
/// should put a lifetime. In those cases we process and put those errors
@ -824,147 +814,14 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
}
}
}
fn glb_concrete_regions(&self,
free_regions: &FreeRegionMap,
a: Region,
b: Region)
-> RelateResult<'tcx, Region>
{
debug!("glb_concrete_regions({:?}, {:?})", a, b);
match (a, b) {
(ReLateBound(..), _) |
(_, ReLateBound(..)) |
(ReEarlyBound(..), _) |
(_, ReEarlyBound(..)) => {
self.tcx.sess.bug(
&format!("cannot relate bound region: GLB({:?}, {:?})",
a,
b));
}
(ReStatic, r) | (r, ReStatic) => {
// static lives longer than everything else
Ok(r)
}
(ReEmpty, _) | (_, ReEmpty) => {
// nothing lives shorter than everything else
Ok(ReEmpty)
}
(ReVar(v_id), _) |
(_, ReVar(v_id)) => {
self.tcx.sess.span_bug(
(*self.var_origins.borrow())[v_id.index as usize].span(),
&format!("glb_concrete_regions invoked with \
non-concrete regions: {:?}, {:?}",
a,
b));
}
(ReFree(fr), ReScope(s_id)) |
(ReScope(s_id), ReFree(fr)) => {
let s = ReScope(s_id);
// Free region is something "at least as big as
// `fr.scope_id`." If we find that the scope `fr.scope_id` is bigger
// than the scope `s_id`, then we can say that the GLB
// is the scope `s_id`. Otherwise, as we do not know
// big the free region is precisely, the GLB is undefined.
if self.tcx.region_maps.nearest_common_ancestor(fr.scope, s_id) == fr.scope ||
free_regions.is_static(fr) {
Ok(s)
} else {
Err(TypeError::RegionsNoOverlap(b, a))
}
}
(ReScope(a_id), ReScope(b_id)) => {
self.intersect_scopes(a, b, a_id, b_id)
}
(ReFree(ref a_fr), ReFree(ref b_fr)) => {
self.glb_free_regions(free_regions, a_fr, b_fr)
}
// For these types, we cannot define any additional
// relationship:
(ReSkolemized(..), _) |
(_, ReSkolemized(..)) => {
if a == b {
Ok(a)
} else {
Err(TypeError::RegionsNoOverlap(b, a))
}
}
}
}
/// Computes a region that is enclosed by both free region arguments, if any. Guarantees that
/// if the same two regions are given as argument, in any order, a consistent result is
/// returned.
fn glb_free_regions(&self,
free_regions: &FreeRegionMap,
a: &FreeRegion,
b: &FreeRegion)
-> RelateResult<'tcx, ty::Region>
{
return match a.cmp(b) {
Less => helper(self, free_regions, a, b),
Greater => helper(self, free_regions, b, a),
Equal => Ok(ty::ReFree(*a))
};
fn helper<'a, 'tcx>(this: &RegionVarBindings<'a, 'tcx>,
free_regions: &FreeRegionMap,
a: &FreeRegion,
b: &FreeRegion) -> RelateResult<'tcx, ty::Region>
{
if free_regions.sub_free_region(*a, *b) {
Ok(ty::ReFree(*a))
} else if free_regions.sub_free_region(*b, *a) {
Ok(ty::ReFree(*b))
} else {
this.intersect_scopes(ty::ReFree(*a), ty::ReFree(*b),
a.scope, b.scope)
}
}
}
fn intersect_scopes(&self,
region_a: ty::Region,
region_b: ty::Region,
scope_a: region::CodeExtent,
scope_b: region::CodeExtent)
-> RelateResult<'tcx, Region>
{
// We want to generate the intersection of two
// scopes or two free regions. So, if one of
// these scopes is a subscope of the other, return
// it. Otherwise fail.
debug!("intersect_scopes(scope_a={:?}, scope_b={:?}, region_a={:?}, region_b={:?})",
scope_a, scope_b, region_a, region_b);
let r_id = self.tcx.region_maps.nearest_common_ancestor(scope_a, scope_b);
if r_id == scope_a {
Ok(ReScope(scope_b))
} else if r_id == scope_b {
Ok(ReScope(scope_a))
} else {
Err(TypeError::RegionsNoOverlap(region_a, region_b))
}
}
}
// ______________________________________________________________________
#[derive(Copy, Clone, PartialEq, Debug)]
enum Classification { Expanding, Contracting }
#[derive(Copy, Clone, Debug)]
pub enum VarValue { NoValue, Value(Region), ErrorValue }
pub enum VarValue { Value(Region), ErrorValue }
struct VarData {
classification: Classification,
value: VarValue,
}
@ -1005,12 +862,7 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
fn construct_var_data(&self) -> Vec<VarData> {
(0..self.num_vars() as usize).map(|_| {
VarData {
// All nodes are initially classified as contracting; during
// the expansion phase, we will shift the classification for
// those nodes that have a concrete region predecessor to
// Expanding.
classification: Contracting,
value: NoValue,
value: Value(ty::ReEmpty),
}
}).collect()
}
@ -1062,11 +914,11 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
}
ConstrainVarSubVar(a_vid, b_vid) => {
match var_data[a_vid.index as usize].value {
NoValue | ErrorValue => false,
Value(a_region) => {
let b_node = &mut var_data[b_vid.index as usize];
self.expand_node(free_regions, a_region, b_vid, b_node)
}
ErrorValue => false,
Value(a_region) => {
let b_node = &mut var_data[b_vid.index as usize];
self.expand_node(free_regions, a_region, b_vid, b_node)
}
}
}
ConstrainVarSubReg(..) => {
@ -1100,16 +952,7 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
_ => { }
}
b_data.classification = Expanding;
match b_data.value {
NoValue => {
debug!("Setting initial value of {:?} to {:?}",
b_vid, a_region);
b_data.value = Value(a_region);
return true;
}
Value(cur_region) => {
let lub = self.lub_concrete_regions(free_regions, a_region, cur_region);
if lub == cur_region {
@ -1131,6 +974,7 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
}
}
// FIXME(#29436) -- this fn would just go away if we removed ConstrainVarSubReg
fn contraction(&self,
free_regions: &FreeRegionMap,
var_data: &mut [VarData]) {
@ -1142,104 +986,31 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
.unwrap()
);
match *constraint {
ConstrainRegSubVar(..) => {
// This is an expansion constraint. Ignore.
false
}
ConstrainVarSubVar(a_vid, b_vid) => {
match var_data[b_vid.index as usize].value {
NoValue | ErrorValue => false,
Value(b_region) => {
ConstrainRegSubVar(..) |
ConstrainVarSubVar(..) => {
// Expansion will ensure that these constraints hold. Ignore.
}
ConstrainVarSubReg(a_vid, b_region) => {
let a_data = &mut var_data[a_vid.index as usize];
self.contract_node(free_regions, a_vid, a_data, b_region)
}
}
}
ConstrainVarSubReg(a_vid, b_region) => {
let a_data = &mut var_data[a_vid.index as usize];
self.contract_node(free_regions, a_vid, a_data, b_region)
}
}
})
}
debug!("contraction: {:?} == {:?}, {:?}", a_vid, a_data.value, b_region);
fn contract_node(&self,
free_regions: &FreeRegionMap,
a_vid: RegionVid,
a_data: &mut VarData,
b_region: Region)
-> bool {
debug!("contract_node({:?} == {:?}/{:?}, {:?})",
a_vid, a_data.value,
a_data.classification, b_region);
let a_region = match a_data.value {
ErrorValue => return false,
Value(a_region) => a_region,
};
return match a_data.value {
NoValue => {
assert_eq!(a_data.classification, Contracting);
a_data.value = Value(b_region);
true // changed
}
ErrorValue => false, // no change
Value(a_region) => {
match a_data.classification {
Expanding =>
check_node(self, free_regions, a_vid, a_data, a_region, b_region),
Contracting =>
adjust_node(self, free_regions, a_vid, a_data, a_region, b_region),
}
}
};
fn check_node(this: &RegionVarBindings,
free_regions: &FreeRegionMap,
a_vid: RegionVid,
a_data: &mut VarData,
a_region: Region,
b_region: Region)
-> bool
{
if !free_regions.is_subregion_of(this.tcx, a_region, b_region) {
debug!("Setting {:?} to ErrorValue: {:?} not subregion of {:?}",
a_vid,
a_region,
b_region);
a_data.value = ErrorValue;
}
false
}
fn adjust_node(this: &RegionVarBindings,
free_regions: &FreeRegionMap,
a_vid: RegionVid,
a_data: &mut VarData,
a_region: Region,
b_region: Region)
-> bool {
match this.glb_concrete_regions(free_regions, a_region, b_region) {
Ok(glb) => {
if glb == a_region {
false
} else {
debug!("Contracting value of {:?} from {:?} to {:?}",
a_vid,
a_region,
glb);
a_data.value = Value(glb);
true
if !free_regions.is_subregion_of(self.tcx, a_region, b_region) {
debug!("Setting {:?} to ErrorValue: {:?} not subregion of {:?}",
a_vid,
a_region,
b_region);
a_data.value = ErrorValue;
}
}
Err(_) => {
debug!("Setting {:?} to ErrorValue: no glb of {:?}, {:?}",
a_vid,
a_region,
b_region);
a_data.value = ErrorValue;
false
}
}
}
false
})
}
fn collect_concrete_region_errors(&self,
@ -1308,12 +1079,6 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
Value(_) => {
/* Inference successful */
}
NoValue => {
/* Unconstrained inference: do not report an error
until the value of this variable is requested.
After all, sometimes we make region variables but never
really use their values. */
}
ErrorValue => {
/* Inference impossible, this value contains
inconsistent constraints.
@ -1339,18 +1104,8 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
this portion of the code and think hard about it. =) */
let node_vid = RegionVid { index: idx as u32 };
match var_data[idx].classification {
Expanding => {
self.collect_error_for_expanding_node(
free_regions, graph, var_data, &mut dup_vec,
node_vid, errors);
}
Contracting => {
self.collect_error_for_contracting_node(
free_regions, graph, var_data, &mut dup_vec,
node_vid, errors);
}
}
self.collect_error_for_expanding_node(
free_regions, graph, &mut dup_vec, node_vid, errors);
}
}
}
@ -1396,7 +1151,6 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
fn collect_error_for_expanding_node(&self,
free_regions: &FreeRegionMap,
graph: &RegionGraph,
var_data: &[VarData],
dup_vec: &mut [u32],
node_idx: RegionVid,
errors: &mut Vec<RegionResolutionError<'tcx>>)
@ -1404,11 +1158,9 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
// Errors in expanding nodes result from a lower-bound that is
// not contained by an upper-bound.
let (mut lower_bounds, lower_dup) =
self.collect_concrete_regions(graph, var_data, node_idx,
graph::INCOMING, dup_vec);
self.collect_concrete_regions(graph, node_idx, graph::INCOMING, dup_vec);
let (mut upper_bounds, upper_dup) =
self.collect_concrete_regions(graph, var_data, node_idx,
graph::OUTGOING, dup_vec);
self.collect_concrete_regions(graph, node_idx, graph::OUTGOING, dup_vec);
if lower_dup || upper_dup {
return;
@ -1459,59 +1211,8 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
upper_bounds));
}
fn collect_error_for_contracting_node(
&self,
free_regions: &FreeRegionMap,
graph: &RegionGraph,
var_data: &[VarData],
dup_vec: &mut [u32],
node_idx: RegionVid,
errors: &mut Vec<RegionResolutionError<'tcx>>)
{
// Errors in contracting nodes result from two upper-bounds
// that have no intersection.
let (upper_bounds, dup_found) =
self.collect_concrete_regions(graph, var_data, node_idx,
graph::OUTGOING, dup_vec);
if dup_found {
return;
}
for upper_bound_1 in &upper_bounds {
for upper_bound_2 in &upper_bounds {
match self.glb_concrete_regions(free_regions,
upper_bound_1.region,
upper_bound_2.region) {
Ok(_) => {}
Err(_) => {
let origin = (*self.var_origins.borrow())[node_idx.index as usize].clone();
debug!("region inference error at {:?} for {:?}: \
SupSupConflict sub: {:?} sup: {:?}",
origin, node_idx, upper_bound_1.region, upper_bound_2.region);
errors.push(SupSupConflict(
origin,
upper_bound_1.origin.clone(),
upper_bound_1.region,
upper_bound_2.origin.clone(),
upper_bound_2.region));
return;
}
}
}
}
self.tcx.sess.span_bug(
(*self.var_origins.borrow())[node_idx.index as usize].span(),
&format!("collect_error_for_contracting_node() could not find error \
for var {:?}, upper_bounds={:?}",
node_idx,
upper_bounds));
}
fn collect_concrete_regions(&self,
graph: &RegionGraph,
var_data: &[VarData],
orig_node_idx: RegionVid,
dir: Direction,
dup_vec: &mut [u32])
@ -1536,7 +1237,6 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
while !state.stack.is_empty() {
let node_idx = state.stack.pop().unwrap();
let classification = var_data[node_idx.index as usize].classification;
// check whether we've visited this node on some previous walk
if dup_vec[node_idx.index as usize] == u32::MAX {
@ -1545,17 +1245,12 @@ impl<'a, 'tcx> RegionVarBindings<'a, 'tcx> {
state.dup_found = true;
}
debug!("collect_concrete_regions(orig_node_idx={:?}, node_idx={:?}, \
classification={:?})",
orig_node_idx, node_idx, classification);
debug!("collect_concrete_regions(orig_node_idx={:?}, node_idx={:?})",
orig_node_idx, node_idx);
// figure out the direction from which this node takes its
// values, and search for concrete regions etc in that direction
let dir = match classification {
Expanding => graph::INCOMING,
Contracting => graph::OUTGOING,
};
let dir = graph::INCOMING;
process_edges(self, &mut state, graph, node_idx, dir);
}
@ -1638,7 +1333,6 @@ fn normalize(values: &Vec<VarValue>, r: ty::Region) -> ty::Region {
fn lookup(values: &Vec<VarValue>, rid: ty::RegionVid) -> ty::Region {
match values[rid.index as usize] {
Value(r) => r,
NoValue => ReEmpty, // No constraints, return ty::ReEmpty
ErrorValue => ReStatic, // Previously reported error.
}
}

View File

@ -351,6 +351,11 @@ impl<'a, 'tcx> Env<'a, 'tcx> {
self.tcx().types.isize)
}
pub fn t_rptr_empty(&self) -> Ty<'tcx> {
self.infcx.tcx.mk_imm_ref(self.infcx.tcx.mk_region(ty::ReEmpty),
self.tcx().types.isize)
}
pub fn dummy_type_trace(&self) -> infer::TypeTrace<'tcx> {
infer::TypeTrace::dummy(self.tcx())
}
@ -593,16 +598,15 @@ fn lub_free_free() {
#[test]
fn lub_returning_scope() {
test_env(EMPTY_SOURCE_STR,
errors(&["cannot infer an appropriate lifetime"]), |env| {
env.create_simple_region_hierarchy();
let t_rptr_scope10 = env.t_rptr_scope(10);
let t_rptr_scope11 = env.t_rptr_scope(11);
// this should generate an error when regions are resolved
env.make_lub_ty(env.t_fn(&[], t_rptr_scope10),
env.t_fn(&[], t_rptr_scope11));
})
test_env(EMPTY_SOURCE_STR, errors(&[]), |env| {
env.create_simple_region_hierarchy();
let t_rptr_scope10 = env.t_rptr_scope(10);
let t_rptr_scope11 = env.t_rptr_scope(11);
let t_rptr_empty = env.t_rptr_empty();
env.check_lub(env.t_fn(&[t_rptr_scope10], env.tcx().types.isize),
env.t_fn(&[t_rptr_scope11], env.tcx().types.isize),
env.t_fn(&[t_rptr_empty], env.tcx().types.isize));
});
}
#[test]

View File

@ -259,17 +259,30 @@ pub fn check_pat<'a, 'tcx>(pcx: &pat_ctxt<'a, 'tcx>,
}
}
hir::PatRegion(ref inner, mutbl) => {
let inner_ty = fcx.infcx().next_ty_var();
let mt = ty::TypeAndMut { ty: inner_ty, mutbl: mutbl };
let region = fcx.infcx().next_region_var(infer::PatternRegion(pat.span));
let rptr_ty = tcx.mk_ref(tcx.mk_region(region), mt);
let expected = fcx.infcx().shallow_resolve(expected);
if check_dereferencable(pcx, pat.span, expected, &**inner) {
// `demand::subtype` would be good enough, but using
// `eqtype` turns out to be equally general. See (*)
// below for details.
demand::eqtype(fcx, pat.span, expected, rptr_ty);
// Take region, inner-type from expected type if we
// can, to avoid creating needless variables. This
// also helps with the bad interactions of the given
// hack detailed in (*) below.
let (rptr_ty, inner_ty) = match expected.sty {
ty::TyRef(_, mt) if mt.mutbl == mutbl => {
(expected, mt.ty)
}
_ => {
let inner_ty = fcx.infcx().next_ty_var();
let mt = ty::TypeAndMut { ty: inner_ty, mutbl: mutbl };
let region = fcx.infcx().next_region_var(infer::PatternRegion(pat.span));
let rptr_ty = tcx.mk_ref(tcx.mk_region(region), mt);
demand::eqtype(fcx, pat.span, expected, rptr_ty);
(rptr_ty, inner_ty)
}
};
fcx.write_ty(pat.id, rptr_ty);
check_pat(pcx, &**inner, inner_ty);
} else {

View File

@ -1182,9 +1182,10 @@ fn link_fn_args(rcx: &Rcx, body_scope: CodeExtent, args: &[hir::Arg]) {
let arg_ty = rcx.fcx.node_ty(arg.id);
let re_scope = ty::ReScope(body_scope);
let arg_cmt = mc.cat_rvalue(arg.id, arg.ty.span, re_scope, arg_ty);
debug!("arg_ty={:?} arg_cmt={:?}",
debug!("arg_ty={:?} arg_cmt={:?} arg={:?}",
arg_ty,
arg_cmt);
arg_cmt,
arg);
link_pattern(rcx, mc, arg_cmt, &*arg.pat);
}
}
@ -1527,9 +1528,10 @@ pub fn type_must_outlive<'a, 'tcx>(rcx: &Rcx<'a, 'tcx>,
{
let ty = rcx.resolve_type(ty);
debug!("type_must_outlive(ty={:?}, region={:?})",
debug!("type_must_outlive(ty={:?}, region={:?}, origin={:?})",
ty,
region);
region,
origin);
assert!(!ty.has_escaping_regions());

View File

@ -0,0 +1,28 @@
// Copyright 2015 The Rust Project Developers. See the COPYRIGHT
// file at the top-level directory of this distribution and at
// http://rust-lang.org/COPYRIGHT.
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
// except according to those terms.
// Regression test: issue had to do with "givens" in region inference,
// which were not being considered during the contraction phase.
// error-pattern:explicit panic
struct Parser<'i: 't, 't>(&'i u8, &'t u8);
impl<'i, 't> Parser<'i, 't> {
fn parse_nested_block<F, T>(&mut self, parse: F) -> Result<T, ()>
where for<'tt> F: FnOnce(&mut Parser<'i, 'tt>) -> T { panic!() }
fn expect_exhausted(&mut self) -> Result<(), ()> { Ok(()) }
}
fn main() {
let x = 0u8;
Parser(&x, &x).parse_nested_block(|input| input.expect_exhausted()).unwrap();
}

View File

@ -0,0 +1,21 @@
// Copyright 2014 The Rust Project Developers. See the COPYRIGHT
// file at the top-level directory of this distribution and at
// http://rust-lang.org/COPYRIGHT.
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
// except according to those terms.
pub struct Chan;
pub struct ChanSelect<'c, T> {
chans: Vec<(&'c Chan, T)>,
}
impl<'c, T> ChanSelect<'c, T> {
pub fn add_recv_ret(&mut self, chan: &'c Chan, ret: T)
{
self.chans.push((chan, ret));
}
}
fn main() {}