rust/src/librustc_mir/transform/simplify.rs

378 lines
12 KiB
Rust
Raw Normal View History

//! A number of passes which remove various redundancies in the CFG.
//!
//! The `SimplifyCfg` pass gets rid of unnecessary blocks in the CFG, whereas the `SimplifyLocals`
//! gets rid of all the unnecessary local variable declarations.
//!
//! The `SimplifyLocals` pass is kinda expensive and therefore not very suitable to be run often.
//! Most of the passes should not care or be impacted in meaningful ways due to extra locals
2018-05-08 08:10:16 -05:00
//! either, so running the pass once, right before codegen, should suffice.
//!
//! On the other side of the spectrum, the `SimplifyCfg` pass is considerably cheap to run, thus
//! one should run it after every pass which may modify CFG in significant ways. This pass must
//! also be run before any analysis passes because it removes dead blocks, and some of these can be
//! ill-typed.
//!
//! The cause of this typing issue is typeck allowing most blocks whose end is not reachable have
//! an arbitrary return type, rather than having the usual () return type (as a note, typeck's
//! notion of reachability is in fact slightly weaker than MIR CFG reachability - see #31617). A
//! standard example of the situation is:
//!
//! ```rust
//! fn example() {
//! let _a: char = { return; };
//! }
//! ```
//!
//! Here the block (`{ return; }`) has the return type `char`, rather than `()`, but the MIR we
//! naively generate still contains the `_a = ()` write in the unreachable block "after" the
//! return.
Merge indexed_set.rs into bitvec.rs, and rename it bit_set.rs. Currently we have two files implementing bitsets (and 2D bit matrices). This commit combines them into one, taking the best features from each. This involves renaming a lot of things. The high level changes are as follows. - bitvec.rs --> bit_set.rs - indexed_set.rs --> (removed) - BitArray + IdxSet --> BitSet (merged, see below) - BitVector --> GrowableBitSet - {,Sparse,Hybrid}IdxSet --> {,Sparse,Hybrid}BitSet - BitMatrix --> BitMatrix - SparseBitMatrix --> SparseBitMatrix The changes within the bitset types themselves are as follows. ``` OLD OLD NEW BitArray<C> IdxSet<T> BitSet<T> -------- ------ ------ grow - grow new - (remove) new_empty new_empty new_empty new_filled new_filled new_filled - to_hybrid to_hybrid clear clear clear set_up_to set_up_to set_up_to clear_above - clear_above count - count contains(T) contains(&T) contains(T) contains_all - superset is_empty - is_empty insert(T) add(&T) insert(T) insert_all - insert_all() remove(T) remove(&T) remove(T) words words words words_mut words_mut words_mut - overwrite overwrite merge union union - subtract subtract - intersect intersect iter iter iter ``` In general, when choosing names I went with: - names that are more obvious (e.g. `BitSet` over `IdxSet`). - names that are more like the Rust libraries (e.g. `T` over `C`, `insert` over `add`); - names that are more set-like (e.g. `union` over `merge`, `superset` over `contains_all`, `domain_size` over `num_bits`). Also, using `T` for index arguments seems more sensible than `&T` -- even though the latter is standard in Rust collection types -- because indices are always copyable. It also results in fewer `&` and `*` sigils in practice.
2018-09-14 00:07:25 -05:00
use rustc_data_structures::bit_set::BitSet;
use rustc_data_structures::indexed_vec::{Idx, IndexVec};
use rustc::ty::TyCtxt;
2016-09-19 15:50:00 -05:00
use rustc::mir::*;
2017-12-01 06:31:47 -06:00
use rustc::mir::visit::{MutVisitor, Visitor, PlaceContext};
use rustc::session::config::DebugInfo;
use std::borrow::Cow;
2019-02-07 15:28:15 -06:00
use crate::transform::{MirPass, MirSource};
2016-02-26 10:05:50 -06:00
pub struct SimplifyCfg { label: String }
impl SimplifyCfg {
pub fn new(label: &str) -> Self {
SimplifyCfg { label: format!("SimplifyCfg-{}", label) }
}
}
pub fn simplify_cfg(body: &mut Body<'_>) {
CfgSimplifier::new(body).simplify();
remove_dead_blocks(body);
2017-03-09 12:36:01 -06:00
// FIXME: Should probably be moved into some kind of pass manager
body.basic_blocks_mut().raw.shrink_to_fit();
2017-03-09 12:36:01 -06:00
}
impl MirPass for SimplifyCfg {
fn name<'a>(&'a self) -> Cow<'a, str> {
Cow::Borrowed(&self.label)
}
fn run_pass<'tcx>(
&self,
_tcx: TyCtxt<'tcx, 'tcx>,
_src: MirSource<'tcx>,
body: &mut Body<'tcx>,
) {
debug!("SimplifyCfg({:?}) - simplifying {:?}", self.label, body);
simplify_cfg(body);
}
}
pub struct CfgSimplifier<'a, 'tcx: 'a> {
basic_blocks: &'a mut IndexVec<BasicBlock, BasicBlockData<'tcx>>,
pred_count: IndexVec<BasicBlock, u32>
}
impl<'a, 'tcx: 'a> CfgSimplifier<'a, 'tcx> {
pub fn new(body: &'a mut Body<'tcx>) -> Self {
let mut pred_count = IndexVec::from_elem(0u32, body.basic_blocks());
// we can't use mir.predecessors() here because that counts
// dead blocks, which we don't want to.
2016-10-04 08:23:01 -05:00
pred_count[START_BLOCK] = 1;
for (_, data) in traversal::preorder(body) {
if let Some(ref term) = data.terminator {
for &tgt in term.successors() {
pred_count[tgt] += 1;
}
}
}
let basic_blocks = body.basic_blocks_mut();
CfgSimplifier {
basic_blocks,
pred_count,
}
}
pub fn simplify(mut self) {
self.strip_nops();
let mut start = START_BLOCK;
loop {
let mut changed = false;
self.collapse_goto_chain(&mut start, &mut changed);
for bb in self.basic_blocks.indices() {
if self.pred_count[bb] == 0 {
continue
}
debug!("simplifying {:?}", bb);
let mut terminator = self.basic_blocks[bb].terminator.take()
.expect("invalid terminator state");
for successor in terminator.successors_mut() {
self.collapse_goto_chain(successor, &mut changed);
}
let mut new_stmts = vec![];
let mut inner_changed = true;
while inner_changed {
inner_changed = false;
inner_changed |= self.simplify_branch(&mut terminator);
inner_changed |= self.merge_successor(&mut new_stmts, &mut terminator);
changed |= inner_changed;
}
self.basic_blocks[bb].statements.extend(new_stmts);
self.basic_blocks[bb].terminator = Some(terminator);
changed |= inner_changed;
}
if !changed { break }
}
if start != START_BLOCK {
debug_assert!(self.pred_count[START_BLOCK] == 0);
self.basic_blocks.swap(START_BLOCK, start);
self.pred_count.swap(START_BLOCK, start);
// pred_count == 1 if the start block has no predecessor _blocks_.
if self.pred_count[START_BLOCK] > 1 {
for (bb, data) in self.basic_blocks.iter_enumerated_mut() {
if self.pred_count[bb] == 0 {
continue;
}
for target in data.terminator_mut().successors_mut() {
if *target == start {
*target = START_BLOCK;
}
}
}
}
}
}
// Collapse a goto chain starting from `start`
fn collapse_goto_chain(&mut self, start: &mut BasicBlock, changed: &mut bool) {
let mut terminator = match self.basic_blocks[*start] {
BasicBlockData {
ref statements,
terminator: ref mut terminator @ Some(Terminator {
kind: TerminatorKind::Goto { .. }, ..
}), ..
} if statements.is_empty() => terminator.take(),
// if `terminator` is None, this means we are in a loop. In that
// case, let all the loop collapse to its entry.
_ => return
};
let target = match terminator {
Some(Terminator { kind: TerminatorKind::Goto { ref mut target }, .. }) => {
self.collapse_goto_chain(target, changed);
*target
}
_ => unreachable!()
};
self.basic_blocks[*start].terminator = terminator;
debug!("collapsing goto chain from {:?} to {:?}", *start, target);
*changed |= *start != target;
if self.pred_count[*start] == 1 {
// This is the last reference to *start, so the pred-count to
// to target is moved into the current block.
self.pred_count[*start] = 0;
} else {
self.pred_count[target] += 1;
self.pred_count[*start] -= 1;
}
*start = target;
}
// merge a block with 1 `goto` predecessor to its parent
fn merge_successor(&mut self,
new_stmts: &mut Vec<Statement<'tcx>>,
terminator: &mut Terminator<'tcx>)
-> bool
{
let target = match terminator.kind {
TerminatorKind::Goto { target }
if self.pred_count[target] == 1
=> target,
_ => return false
};
debug!("merging block {:?} into {:?}", target, terminator);
*terminator = match self.basic_blocks[target].terminator.take() {
Some(terminator) => terminator,
None => {
// unreachable loop - this should not be possible, as we
// don't strand blocks, but handle it correctly.
return false
}
};
new_stmts.extend(self.basic_blocks[target].statements.drain(..));
self.pred_count[target] = 0;
true
}
// turn a branch with all successors identical to a goto
fn simplify_branch(&mut self, terminator: &mut Terminator<'tcx>) -> bool {
match terminator.kind {
TerminatorKind::SwitchInt { .. } => {},
_ => return false
};
let first_succ = {
if let Some(&first_succ) = terminator.successors().nth(0) {
if terminator.successors().all(|s| *s == first_succ) {
let count = terminator.successors().count();
self.pred_count[first_succ] -= (count - 1) as u32;
first_succ
} else {
return false
}
} else {
return false
}
};
debug!("simplifying branch {:?}", terminator);
terminator.kind = TerminatorKind::Goto { target: first_succ };
true
}
fn strip_nops(&mut self) {
for blk in self.basic_blocks.iter_mut() {
blk.statements.retain(|stmt| if let StatementKind::Nop = stmt.kind {
false
} else {
true
})
}
}
}
pub fn remove_dead_blocks(body: &mut Body<'_>) {
let mut seen = BitSet::new_empty(body.basic_blocks().len());
for (bb, _) in traversal::preorder(body) {
seen.insert(bb.index());
}
let basic_blocks = body.basic_blocks_mut();
2016-06-07 13:20:50 -05:00
let num_blocks = basic_blocks.len();
let mut replacements : Vec<_> = (0..num_blocks).map(BasicBlock::new).collect();
let mut used_blocks = 0;
for alive_index in seen.iter() {
replacements[alive_index] = BasicBlock::new(used_blocks);
if alive_index != used_blocks {
// Swap the next alive block data with the current available slot. Since alive_index is
// non-decreasing this is a valid operation.
2016-06-07 13:20:50 -05:00
basic_blocks.raw.swap(alive_index, used_blocks);
}
used_blocks += 1;
}
2016-06-07 13:20:50 -05:00
basic_blocks.raw.truncate(used_blocks);
2016-06-07 13:20:50 -05:00
for block in basic_blocks {
for target in block.terminator_mut().successors_mut() {
*target = replacements[target.index()];
}
}
}
pub struct SimplifyLocals;
impl MirPass for SimplifyLocals {
fn run_pass<'tcx>(&self, tcx: TyCtxt<'tcx, 'tcx>, _: MirSource<'tcx>, body: &mut Body<'tcx>) {
let mut marker = DeclMarker { locals: BitSet::new_empty(body.local_decls.len()) };
marker.visit_body(body);
// Return pointer and arguments are always live
marker.locals.insert(RETURN_PLACE);
for arg in body.args_iter() {
marker.locals.insert(arg);
}
// We may need to keep dead user variables live for debuginfo.
if tcx.sess.opts.debuginfo == DebugInfo::Full {
for local in body.vars_iter() {
marker.locals.insert(local);
}
}
let map = make_local_map(&mut body.local_decls, marker.locals);
// Update references to all vars and tmps now
LocalUpdater { map }.visit_body(body);
body.local_decls.shrink_to_fit();
}
}
/// Construct the mapping while swapping out unused stuff out from the `vec`.
2019-06-11 04:47:30 -05:00
fn make_local_map<V>(
vec: &mut IndexVec<Local, V>,
Merge indexed_set.rs into bitvec.rs, and rename it bit_set.rs. Currently we have two files implementing bitsets (and 2D bit matrices). This commit combines them into one, taking the best features from each. This involves renaming a lot of things. The high level changes are as follows. - bitvec.rs --> bit_set.rs - indexed_set.rs --> (removed) - BitArray + IdxSet --> BitSet (merged, see below) - BitVector --> GrowableBitSet - {,Sparse,Hybrid}IdxSet --> {,Sparse,Hybrid}BitSet - BitMatrix --> BitMatrix - SparseBitMatrix --> SparseBitMatrix The changes within the bitset types themselves are as follows. ``` OLD OLD NEW BitArray<C> IdxSet<T> BitSet<T> -------- ------ ------ grow - grow new - (remove) new_empty new_empty new_empty new_filled new_filled new_filled - to_hybrid to_hybrid clear clear clear set_up_to set_up_to set_up_to clear_above - clear_above count - count contains(T) contains(&T) contains(T) contains_all - superset is_empty - is_empty insert(T) add(&T) insert(T) insert_all - insert_all() remove(T) remove(&T) remove(T) words words words words_mut words_mut words_mut - overwrite overwrite merge union union - subtract subtract - intersect intersect iter iter iter ``` In general, when choosing names I went with: - names that are more obvious (e.g. `BitSet` over `IdxSet`). - names that are more like the Rust libraries (e.g. `T` over `C`, `insert` over `add`); - names that are more set-like (e.g. `union` over `merge`, `superset` over `contains_all`, `domain_size` over `num_bits`). Also, using `T` for index arguments seems more sensible than `&T` -- even though the latter is standard in Rust collection types -- because indices are always copyable. It also results in fewer `&` and `*` sigils in practice.
2018-09-14 00:07:25 -05:00
mask: BitSet<Local>,
) -> IndexVec<Local, Option<Local>> {
let mut map: IndexVec<Local, Option<Local>> = IndexVec::from_elem(None, &*vec);
let mut used = Local::new(0);
for alive_index in mask.iter() {
map[alive_index] = Some(used);
if alive_index != used {
vec.swap(alive_index, used);
}
used.increment_by(1);
}
vec.truncate(used.index());
map
}
struct DeclMarker {
Merge indexed_set.rs into bitvec.rs, and rename it bit_set.rs. Currently we have two files implementing bitsets (and 2D bit matrices). This commit combines them into one, taking the best features from each. This involves renaming a lot of things. The high level changes are as follows. - bitvec.rs --> bit_set.rs - indexed_set.rs --> (removed) - BitArray + IdxSet --> BitSet (merged, see below) - BitVector --> GrowableBitSet - {,Sparse,Hybrid}IdxSet --> {,Sparse,Hybrid}BitSet - BitMatrix --> BitMatrix - SparseBitMatrix --> SparseBitMatrix The changes within the bitset types themselves are as follows. ``` OLD OLD NEW BitArray<C> IdxSet<T> BitSet<T> -------- ------ ------ grow - grow new - (remove) new_empty new_empty new_empty new_filled new_filled new_filled - to_hybrid to_hybrid clear clear clear set_up_to set_up_to set_up_to clear_above - clear_above count - count contains(T) contains(&T) contains(T) contains_all - superset is_empty - is_empty insert(T) add(&T) insert(T) insert_all - insert_all() remove(T) remove(&T) remove(T) words words words words_mut words_mut words_mut - overwrite overwrite merge union union - subtract subtract - intersect intersect iter iter iter ``` In general, when choosing names I went with: - names that are more obvious (e.g. `BitSet` over `IdxSet`). - names that are more like the Rust libraries (e.g. `T` over `C`, `insert` over `add`); - names that are more set-like (e.g. `union` over `merge`, `superset` over `contains_all`, `domain_size` over `num_bits`). Also, using `T` for index arguments seems more sensible than `&T` -- even though the latter is standard in Rust collection types -- because indices are always copyable. It also results in fewer `&` and `*` sigils in practice.
2018-09-14 00:07:25 -05:00
pub locals: BitSet<Local>,
}
impl<'tcx> Visitor<'tcx> for DeclMarker {
fn visit_local(&mut self, local: &Local, ctx: PlaceContext, _: Location) {
// Ignore storage markers altogether, they get removed along with their otherwise unused
// decls.
// FIXME: Extend this to all non-uses.
if !ctx.is_storage_marker() {
self.locals.insert(*local);
}
}
}
struct LocalUpdater {
map: IndexVec<Local, Option<Local>>,
}
impl<'tcx> MutVisitor<'tcx> for LocalUpdater {
fn visit_basic_block_data(&mut self, block: BasicBlock, data: &mut BasicBlockData<'tcx>) {
// Remove unnecessary StorageLive and StorageDead annotations.
data.statements.retain(|stmt| {
match stmt.kind {
StatementKind::StorageLive(l) | StatementKind::StorageDead(l) => {
self.map[l].is_some()
}
_ => true
}
});
self.super_basic_block_data(block, data);
}
fn visit_local(&mut self, l: &mut Local, _: PlaceContext, _: Location) {
*l = self.map[*l].unwrap();
}
}