rust/compiler/rustc_codegen_llvm/src/coverageinfo/map_data.rs

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

312 lines
14 KiB
Rust
Raw Normal View History

pub use super::ffi::*;
use rustc_index::{IndexSlice, IndexVec};
use rustc_middle::bug;
use rustc_middle::mir::coverage::{
CodeRegion, CounterId, ExpressionId, MappedExpressionIndex, Op, Operand,
};
use rustc_middle::ty::Instance;
use rustc_middle::ty::TyCtxt;
#[derive(Clone, Debug, PartialEq)]
pub struct Expression {
lhs: Operand,
op: Op,
rhs: Operand,
region: Option<CodeRegion>,
}
/// Collects all of the coverage regions associated with (a) injected counters, (b) counter
/// expressions (additions or subtraction), and (c) unreachable regions (always counted as zero),
/// for a given Function. This struct also stores the `function_source_hash`,
/// computed during instrumentation, and forwarded with counters.
///
/// Note, it may be important to understand LLVM's definitions of `unreachable` regions versus "gap
/// regions" (or "gap areas"). A gap region is a code region within a counted region (either counter
/// or expression), but the line or lines in the gap region are not executable (such as lines with
/// only whitespace or comments). According to LLVM Code Coverage Mapping documentation, "A count
/// for a gap area is only used as the line execution count if there are no other regions on a
/// line."
#[derive(Debug)]
pub struct FunctionCoverage<'tcx> {
instance: Instance<'tcx>,
source_hash: u64,
coverage bug fixes and optimization support Adjusted LLVM codegen for code compiled with `-Zinstrument-coverage` to address multiple, somewhat related issues. Fixed a significant flaw in prior coverage solution: Every counter generated a new counter variable, but there should have only been one counter variable per function. This appears to have bloated .profraw files significantly. (For a small program, it increased the size by about 40%. I have not tested large programs, but there is anecdotal evidence that profraw files were way too large. This is a good fix, regardless, but hopefully it also addresses related issues. Fixes: #82144 Invalid LLVM coverage data produced when compiled with -C opt-level=1 Existing tests now work up to at least `opt-level=3`. This required a detailed analysis of the LLVM IR, comparisons with Clang C++ LLVM IR when compiled with coverage, and a lot of trial and error with codegen adjustments. The biggest hurdle was figuring out how to continue to support coverage results for unused functions and generics. Rust's coverage results have three advantages over Clang's coverage results: 1. Rust's coverage map does not include any overlapping code regions, making coverage counting unambiguous. 2. Rust generates coverage results (showing zero counts) for all unused functions, including generics. (Clang does not generate coverage for uninstantiated template functions.) 3. Rust's unused functions produce minimal stubbed functions in LLVM IR, sufficient for including in the coverage results; while Clang must generate the complete LLVM IR for each unused function, even though it will never be called. This PR removes the previous hack of attempting to inject coverage into some other existing function instance, and generates dedicated instances for each unused function. This change, and a few other adjustments (similar to what is required for `-C link-dead-code`, but with lower impact), makes it possible to support LLVM optimizations. Fixes: #79651 Coverage report: "Unexecuted instantiation:..." for a generic function from multiple crates Fixed by removing the aforementioned hack. Some "Unexecuted instantiation" notices are unavoidable, as explained in the `used_crate.rs` test, but `-Zinstrument-coverage` has new options to back off support for either unused generics, or all unused functions, which avoids the notice, at the cost of less coverage of unused functions. Fixes: #82875 Invalid LLVM coverage data produced with crate brotli_decompressor Fixed by disabling the LLVM function attribute that forces inlining, if `-Z instrument-coverage` is enabled. This attribute is applied to Rust functions with `#[inline(always)], and in some cases, the forced inlining breaks coverage instrumentation and reports.
2021-03-15 23:32:45 +00:00
is_used: bool,
counters: IndexVec<CounterId, Option<CodeRegion>>,
expressions: IndexVec<ExpressionId, Option<Expression>>,
unreachable_regions: Vec<CodeRegion>,
}
impl<'tcx> FunctionCoverage<'tcx> {
coverage bug fixes and optimization support Adjusted LLVM codegen for code compiled with `-Zinstrument-coverage` to address multiple, somewhat related issues. Fixed a significant flaw in prior coverage solution: Every counter generated a new counter variable, but there should have only been one counter variable per function. This appears to have bloated .profraw files significantly. (For a small program, it increased the size by about 40%. I have not tested large programs, but there is anecdotal evidence that profraw files were way too large. This is a good fix, regardless, but hopefully it also addresses related issues. Fixes: #82144 Invalid LLVM coverage data produced when compiled with -C opt-level=1 Existing tests now work up to at least `opt-level=3`. This required a detailed analysis of the LLVM IR, comparisons with Clang C++ LLVM IR when compiled with coverage, and a lot of trial and error with codegen adjustments. The biggest hurdle was figuring out how to continue to support coverage results for unused functions and generics. Rust's coverage results have three advantages over Clang's coverage results: 1. Rust's coverage map does not include any overlapping code regions, making coverage counting unambiguous. 2. Rust generates coverage results (showing zero counts) for all unused functions, including generics. (Clang does not generate coverage for uninstantiated template functions.) 3. Rust's unused functions produce minimal stubbed functions in LLVM IR, sufficient for including in the coverage results; while Clang must generate the complete LLVM IR for each unused function, even though it will never be called. This PR removes the previous hack of attempting to inject coverage into some other existing function instance, and generates dedicated instances for each unused function. This change, and a few other adjustments (similar to what is required for `-C link-dead-code`, but with lower impact), makes it possible to support LLVM optimizations. Fixes: #79651 Coverage report: "Unexecuted instantiation:..." for a generic function from multiple crates Fixed by removing the aforementioned hack. Some "Unexecuted instantiation" notices are unavoidable, as explained in the `used_crate.rs` test, but `-Zinstrument-coverage` has new options to back off support for either unused generics, or all unused functions, which avoids the notice, at the cost of less coverage of unused functions. Fixes: #82875 Invalid LLVM coverage data produced with crate brotli_decompressor Fixed by disabling the LLVM function attribute that forces inlining, if `-Z instrument-coverage` is enabled. This attribute is applied to Rust functions with `#[inline(always)], and in some cases, the forced inlining breaks coverage instrumentation and reports.
2021-03-15 23:32:45 +00:00
/// Creates a new set of coverage data for a used (called) function.
pub fn new(tcx: TyCtxt<'tcx>, instance: Instance<'tcx>) -> Self {
coverage bug fixes and optimization support Adjusted LLVM codegen for code compiled with `-Zinstrument-coverage` to address multiple, somewhat related issues. Fixed a significant flaw in prior coverage solution: Every counter generated a new counter variable, but there should have only been one counter variable per function. This appears to have bloated .profraw files significantly. (For a small program, it increased the size by about 40%. I have not tested large programs, but there is anecdotal evidence that profraw files were way too large. This is a good fix, regardless, but hopefully it also addresses related issues. Fixes: #82144 Invalid LLVM coverage data produced when compiled with -C opt-level=1 Existing tests now work up to at least `opt-level=3`. This required a detailed analysis of the LLVM IR, comparisons with Clang C++ LLVM IR when compiled with coverage, and a lot of trial and error with codegen adjustments. The biggest hurdle was figuring out how to continue to support coverage results for unused functions and generics. Rust's coverage results have three advantages over Clang's coverage results: 1. Rust's coverage map does not include any overlapping code regions, making coverage counting unambiguous. 2. Rust generates coverage results (showing zero counts) for all unused functions, including generics. (Clang does not generate coverage for uninstantiated template functions.) 3. Rust's unused functions produce minimal stubbed functions in LLVM IR, sufficient for including in the coverage results; while Clang must generate the complete LLVM IR for each unused function, even though it will never be called. This PR removes the previous hack of attempting to inject coverage into some other existing function instance, and generates dedicated instances for each unused function. This change, and a few other adjustments (similar to what is required for `-C link-dead-code`, but with lower impact), makes it possible to support LLVM optimizations. Fixes: #79651 Coverage report: "Unexecuted instantiation:..." for a generic function from multiple crates Fixed by removing the aforementioned hack. Some "Unexecuted instantiation" notices are unavoidable, as explained in the `used_crate.rs` test, but `-Zinstrument-coverage` has new options to back off support for either unused generics, or all unused functions, which avoids the notice, at the cost of less coverage of unused functions. Fixes: #82875 Invalid LLVM coverage data produced with crate brotli_decompressor Fixed by disabling the LLVM function attribute that forces inlining, if `-Z instrument-coverage` is enabled. This attribute is applied to Rust functions with `#[inline(always)], and in some cases, the forced inlining breaks coverage instrumentation and reports.
2021-03-15 23:32:45 +00:00
Self::create(tcx, instance, true)
}
/// Creates a new set of coverage data for an unused (never called) function.
pub fn unused(tcx: TyCtxt<'tcx>, instance: Instance<'tcx>) -> Self {
Self::create(tcx, instance, false)
}
fn create(tcx: TyCtxt<'tcx>, instance: Instance<'tcx>, is_used: bool) -> Self {
let coverageinfo = tcx.coverageinfo(instance.def);
debug!(
"FunctionCoverage::create(instance={:?}) has coverageinfo={:?}. is_used={}",
coverage bug fixes and optimization support Adjusted LLVM codegen for code compiled with `-Zinstrument-coverage` to address multiple, somewhat related issues. Fixed a significant flaw in prior coverage solution: Every counter generated a new counter variable, but there should have only been one counter variable per function. This appears to have bloated .profraw files significantly. (For a small program, it increased the size by about 40%. I have not tested large programs, but there is anecdotal evidence that profraw files were way too large. This is a good fix, regardless, but hopefully it also addresses related issues. Fixes: #82144 Invalid LLVM coverage data produced when compiled with -C opt-level=1 Existing tests now work up to at least `opt-level=3`. This required a detailed analysis of the LLVM IR, comparisons with Clang C++ LLVM IR when compiled with coverage, and a lot of trial and error with codegen adjustments. The biggest hurdle was figuring out how to continue to support coverage results for unused functions and generics. Rust's coverage results have three advantages over Clang's coverage results: 1. Rust's coverage map does not include any overlapping code regions, making coverage counting unambiguous. 2. Rust generates coverage results (showing zero counts) for all unused functions, including generics. (Clang does not generate coverage for uninstantiated template functions.) 3. Rust's unused functions produce minimal stubbed functions in LLVM IR, sufficient for including in the coverage results; while Clang must generate the complete LLVM IR for each unused function, even though it will never be called. This PR removes the previous hack of attempting to inject coverage into some other existing function instance, and generates dedicated instances for each unused function. This change, and a few other adjustments (similar to what is required for `-C link-dead-code`, but with lower impact), makes it possible to support LLVM optimizations. Fixes: #79651 Coverage report: "Unexecuted instantiation:..." for a generic function from multiple crates Fixed by removing the aforementioned hack. Some "Unexecuted instantiation" notices are unavoidable, as explained in the `used_crate.rs` test, but `-Zinstrument-coverage` has new options to back off support for either unused generics, or all unused functions, which avoids the notice, at the cost of less coverage of unused functions. Fixes: #82875 Invalid LLVM coverage data produced with crate brotli_decompressor Fixed by disabling the LLVM function attribute that forces inlining, if `-Z instrument-coverage` is enabled. This attribute is applied to Rust functions with `#[inline(always)], and in some cases, the forced inlining breaks coverage instrumentation and reports.
2021-03-15 23:32:45 +00:00
instance, coverageinfo, is_used
);
Self {
instance,
source_hash: 0, // will be set with the first `add_counter()`
coverage bug fixes and optimization support Adjusted LLVM codegen for code compiled with `-Zinstrument-coverage` to address multiple, somewhat related issues. Fixed a significant flaw in prior coverage solution: Every counter generated a new counter variable, but there should have only been one counter variable per function. This appears to have bloated .profraw files significantly. (For a small program, it increased the size by about 40%. I have not tested large programs, but there is anecdotal evidence that profraw files were way too large. This is a good fix, regardless, but hopefully it also addresses related issues. Fixes: #82144 Invalid LLVM coverage data produced when compiled with -C opt-level=1 Existing tests now work up to at least `opt-level=3`. This required a detailed analysis of the LLVM IR, comparisons with Clang C++ LLVM IR when compiled with coverage, and a lot of trial and error with codegen adjustments. The biggest hurdle was figuring out how to continue to support coverage results for unused functions and generics. Rust's coverage results have three advantages over Clang's coverage results: 1. Rust's coverage map does not include any overlapping code regions, making coverage counting unambiguous. 2. Rust generates coverage results (showing zero counts) for all unused functions, including generics. (Clang does not generate coverage for uninstantiated template functions.) 3. Rust's unused functions produce minimal stubbed functions in LLVM IR, sufficient for including in the coverage results; while Clang must generate the complete LLVM IR for each unused function, even though it will never be called. This PR removes the previous hack of attempting to inject coverage into some other existing function instance, and generates dedicated instances for each unused function. This change, and a few other adjustments (similar to what is required for `-C link-dead-code`, but with lower impact), makes it possible to support LLVM optimizations. Fixes: #79651 Coverage report: "Unexecuted instantiation:..." for a generic function from multiple crates Fixed by removing the aforementioned hack. Some "Unexecuted instantiation" notices are unavoidable, as explained in the `used_crate.rs` test, but `-Zinstrument-coverage` has new options to back off support for either unused generics, or all unused functions, which avoids the notice, at the cost of less coverage of unused functions. Fixes: #82875 Invalid LLVM coverage data produced with crate brotli_decompressor Fixed by disabling the LLVM function attribute that forces inlining, if `-Z instrument-coverage` is enabled. This attribute is applied to Rust functions with `#[inline(always)], and in some cases, the forced inlining breaks coverage instrumentation and reports.
2021-03-15 23:32:45 +00:00
is_used,
counters: IndexVec::from_elem_n(None, coverageinfo.num_counters as usize),
expressions: IndexVec::from_elem_n(None, coverageinfo.num_expressions as usize),
unreachable_regions: Vec::new(),
}
}
coverage bug fixes and optimization support Adjusted LLVM codegen for code compiled with `-Zinstrument-coverage` to address multiple, somewhat related issues. Fixed a significant flaw in prior coverage solution: Every counter generated a new counter variable, but there should have only been one counter variable per function. This appears to have bloated .profraw files significantly. (For a small program, it increased the size by about 40%. I have not tested large programs, but there is anecdotal evidence that profraw files were way too large. This is a good fix, regardless, but hopefully it also addresses related issues. Fixes: #82144 Invalid LLVM coverage data produced when compiled with -C opt-level=1 Existing tests now work up to at least `opt-level=3`. This required a detailed analysis of the LLVM IR, comparisons with Clang C++ LLVM IR when compiled with coverage, and a lot of trial and error with codegen adjustments. The biggest hurdle was figuring out how to continue to support coverage results for unused functions and generics. Rust's coverage results have three advantages over Clang's coverage results: 1. Rust's coverage map does not include any overlapping code regions, making coverage counting unambiguous. 2. Rust generates coverage results (showing zero counts) for all unused functions, including generics. (Clang does not generate coverage for uninstantiated template functions.) 3. Rust's unused functions produce minimal stubbed functions in LLVM IR, sufficient for including in the coverage results; while Clang must generate the complete LLVM IR for each unused function, even though it will never be called. This PR removes the previous hack of attempting to inject coverage into some other existing function instance, and generates dedicated instances for each unused function. This change, and a few other adjustments (similar to what is required for `-C link-dead-code`, but with lower impact), makes it possible to support LLVM optimizations. Fixes: #79651 Coverage report: "Unexecuted instantiation:..." for a generic function from multiple crates Fixed by removing the aforementioned hack. Some "Unexecuted instantiation" notices are unavoidable, as explained in the `used_crate.rs` test, but `-Zinstrument-coverage` has new options to back off support for either unused generics, or all unused functions, which avoids the notice, at the cost of less coverage of unused functions. Fixes: #82875 Invalid LLVM coverage data produced with crate brotli_decompressor Fixed by disabling the LLVM function attribute that forces inlining, if `-Z instrument-coverage` is enabled. This attribute is applied to Rust functions with `#[inline(always)], and in some cases, the forced inlining breaks coverage instrumentation and reports.
2021-03-15 23:32:45 +00:00
/// Returns true for a used (called) function, and false for an unused function.
pub fn is_used(&self) -> bool {
self.is_used
}
2020-10-25 18:13:16 +00:00
/// Sets the function source hash value. If called multiple times for the same function, all
/// calls should have the same hash value.
pub fn set_function_source_hash(&mut self, source_hash: u64) {
if self.source_hash == 0 {
self.source_hash = source_hash;
} else {
debug_assert_eq!(source_hash, self.source_hash);
}
}
/// Adds a code region to be counted by an injected counter intrinsic.
pub fn add_counter(&mut self, id: CounterId, region: CodeRegion) {
if let Some(previous_region) = self.counters[id].replace(region.clone()) {
assert_eq!(previous_region, region, "add_counter: code region for id changed");
}
}
/// Both counters and "counter expressions" (or simply, "expressions") can be operands in other
/// expressions. These are tracked as separate variants of `Operand`, so there is no ambiguity
/// between operands that are counter IDs and operands that are expression IDs.
pub fn add_counter_expression(
&mut self,
expression_id: ExpressionId,
lhs: Operand,
op: Op,
rhs: Operand,
region: Option<CodeRegion>,
) {
debug!(
"add_counter_expression({:?}, lhs={:?}, op={:?}, rhs={:?} at {:?}",
expression_id, lhs, op, rhs, region
);
debug_assert!(
expression_id.as_usize() < self.expressions.len(),
"expression_id {} is out of range for expressions.len() = {}
for {:?}",
expression_id.as_usize(),
self.expressions.len(),
self,
);
if let Some(previous_expression) = self.expressions[expression_id].replace(Expression {
lhs,
op,
rhs,
region: region.clone(),
}) {
assert_eq!(
previous_expression,
Expression { lhs, op, rhs, region },
"add_counter_expression: expression for id changed"
);
}
}
/// Add a region that will be marked as "unreachable", with a constant "zero counter".
pub fn add_unreachable_region(&mut self, region: CodeRegion) {
self.unreachable_regions.push(region)
}
/// Return the source hash, generated from the HIR node structure, and used to indicate whether
/// or not the source code structure changed between different compilations.
pub fn source_hash(&self) -> u64 {
self.source_hash
}
/// Generate an array of CounterExpressions, and an iterator over all `Counter`s and their
/// associated `Regions` (from which the LLVM-specific `CoverageMapGenerator` will create
/// `CounterMappingRegion`s.
pub fn get_expressions_and_counter_regions(
&self,
) -> (Vec<CounterExpression>, impl Iterator<Item = (Counter, &CodeRegion)>) {
assert!(
coverage bug fixes and optimization support Adjusted LLVM codegen for code compiled with `-Zinstrument-coverage` to address multiple, somewhat related issues. Fixed a significant flaw in prior coverage solution: Every counter generated a new counter variable, but there should have only been one counter variable per function. This appears to have bloated .profraw files significantly. (For a small program, it increased the size by about 40%. I have not tested large programs, but there is anecdotal evidence that profraw files were way too large. This is a good fix, regardless, but hopefully it also addresses related issues. Fixes: #82144 Invalid LLVM coverage data produced when compiled with -C opt-level=1 Existing tests now work up to at least `opt-level=3`. This required a detailed analysis of the LLVM IR, comparisons with Clang C++ LLVM IR when compiled with coverage, and a lot of trial and error with codegen adjustments. The biggest hurdle was figuring out how to continue to support coverage results for unused functions and generics. Rust's coverage results have three advantages over Clang's coverage results: 1. Rust's coverage map does not include any overlapping code regions, making coverage counting unambiguous. 2. Rust generates coverage results (showing zero counts) for all unused functions, including generics. (Clang does not generate coverage for uninstantiated template functions.) 3. Rust's unused functions produce minimal stubbed functions in LLVM IR, sufficient for including in the coverage results; while Clang must generate the complete LLVM IR for each unused function, even though it will never be called. This PR removes the previous hack of attempting to inject coverage into some other existing function instance, and generates dedicated instances for each unused function. This change, and a few other adjustments (similar to what is required for `-C link-dead-code`, but with lower impact), makes it possible to support LLVM optimizations. Fixes: #79651 Coverage report: "Unexecuted instantiation:..." for a generic function from multiple crates Fixed by removing the aforementioned hack. Some "Unexecuted instantiation" notices are unavoidable, as explained in the `used_crate.rs` test, but `-Zinstrument-coverage` has new options to back off support for either unused generics, or all unused functions, which avoids the notice, at the cost of less coverage of unused functions. Fixes: #82875 Invalid LLVM coverage data produced with crate brotli_decompressor Fixed by disabling the LLVM function attribute that forces inlining, if `-Z instrument-coverage` is enabled. This attribute is applied to Rust functions with `#[inline(always)], and in some cases, the forced inlining breaks coverage instrumentation and reports.
2021-03-15 23:32:45 +00:00
self.source_hash != 0 || !self.is_used,
"No counters provided the source_hash for used function: {:?}",
self.instance
);
let counter_regions = self.counter_regions();
let (counter_expressions, expression_regions) = self.expressions_with_regions();
let unreachable_regions = self.unreachable_regions();
let counter_regions =
counter_regions.chain(expression_regions.into_iter().chain(unreachable_regions));
(counter_expressions, counter_regions)
}
fn counter_regions(&self) -> impl Iterator<Item = (Counter, &CodeRegion)> {
self.counters.iter_enumerated().filter_map(|(index, entry)| {
// Option::map() will return None to filter out missing counters. This may happen
// if, for example, a MIR-instrumented counter is removed during an optimization.
entry.as_ref().map(|region| (Counter::counter_value_reference(index), region))
})
}
fn expressions_with_regions(
&self,
) -> (Vec<CounterExpression>, impl Iterator<Item = (Counter, &CodeRegion)>) {
let mut counter_expressions = Vec::with_capacity(self.expressions.len());
let mut expression_regions = Vec::with_capacity(self.expressions.len());
let mut new_indexes = IndexVec::from_elem_n(None, self.expressions.len());
// This closure converts any `Expression` operand (`lhs` or `rhs` of the `Op::Add` or
// `Op::Subtract` operation) into its native `llvm::coverage::Counter::CounterKind` type
// and value.
//
// Expressions will be returned from this function in a sequential vector (array) of
// `CounterExpression`, so the expression IDs must be mapped from their original,
// potentially sparse set of indexes.
//
// An `Expression` as an operand will have already been encountered as an `Expression` with
// operands, so its new_index will already have been generated (as a 1-up index value).
// (If an `Expression` as an operand does not have a corresponding new_index, it was
// probably optimized out, after the expression was injected into the MIR, so it will
// get a `CounterKind::Zero` instead.)
//
// In other words, an `Expression`s at any given index can include other expressions as
// operands, but expression operands can only come from the subset of expressions having
// `expression_index`s lower than the referencing `Expression`. Therefore, it is
// reasonable to look up the new index of an expression operand while the `new_indexes`
// vector is only complete up to the current `ExpressionIndex`.
type NewIndexes = IndexSlice<ExpressionId, Option<MappedExpressionIndex>>;
let id_to_counter = |new_indexes: &NewIndexes, operand: Operand| match operand {
Operand::Zero => Some(Counter::zero()),
Operand::Counter(id) => Some(Counter::counter_value_reference(id)),
Operand::Expression(id) => {
self.expressions
.get(id)
.expect("expression id is out of range")
.as_ref()
// If an expression was optimized out, assume it would have produced a count
// of zero. This ensures that expressions dependent on optimized-out
// expressions are still valid.
.map_or(Some(Counter::zero()), |_| new_indexes[id].map(Counter::expression))
}
};
for (original_index, expression) in
self.expressions.iter_enumerated().filter_map(|(original_index, entry)| {
// Option::map() will return None to filter out missing expressions. This may happen
// if, for example, a MIR-instrumented expression is removed during an optimization.
entry.as_ref().map(|expression| (original_index, expression))
})
{
let optional_region = &expression.region;
let Expression { lhs, op, rhs, .. } = *expression;
if let Some(Some((lhs_counter, mut rhs_counter))) = id_to_counter(&new_indexes, lhs)
.map(|lhs_counter| {
id_to_counter(&new_indexes, rhs).map(|rhs_counter| (lhs_counter, rhs_counter))
})
{
if lhs_counter.is_zero() && op.is_subtract() {
// The left side of a subtraction was probably optimized out. As an example,
// a branch condition might be evaluated as a constant expression, and the
// branch could be removed, dropping unused counters in the process.
//
// Since counters are unsigned, we must assume the result of the expression
// can be no more and no less than zero. An expression known to evaluate to zero
// does not need to be added to the coverage map.
//
// Coverage test `loops_branches.rs` includes multiple variations of branches
// based on constant conditional (literal `true` or `false`), and demonstrates
// that the expected counts are still correct.
debug!(
"Expression subtracts from zero (assume unreachable): \
original_index={:?}, lhs={:?}, op={:?}, rhs={:?}, region={:?}",
original_index, lhs, op, rhs, optional_region,
);
rhs_counter = Counter::zero();
}
debug_assert!(
lhs_counter.is_zero()
// Note: with `as usize` the ID _could_ overflow/wrap if `usize = u16`
|| ((lhs_counter.zero_based_id() as usize)
<= usize::max(self.counters.len(), self.expressions.len())),
"lhs id={} > both counters.len()={} and expressions.len()={}
({:?} {:?} {:?})",
lhs_counter.zero_based_id(),
self.counters.len(),
self.expressions.len(),
lhs_counter,
op,
rhs_counter,
);
debug_assert!(
rhs_counter.is_zero()
// Note: with `as usize` the ID _could_ overflow/wrap if `usize = u16`
|| ((rhs_counter.zero_based_id() as usize)
<= usize::max(self.counters.len(), self.expressions.len())),
"rhs id={} > both counters.len()={} and expressions.len()={}
({:?} {:?} {:?})",
rhs_counter.zero_based_id(),
self.counters.len(),
self.expressions.len(),
lhs_counter,
op,
rhs_counter,
);
// Both operands exist. `Expression` operands exist in `self.expressions` and have
// been assigned a `new_index`.
let mapped_expression_index =
MappedExpressionIndex::from(counter_expressions.len());
Updates to experimental coverage counter injection This is a combination of 18 commits. Commit #2: Additional examples and some small improvements. Commit #3: fixed mir-opt non-mir extensions and spanview title elements Corrected a fairly recent assumption in runtest.rs that all MIR dump files end in .mir. (It was appending .mir to the graphviz .dot and spanview .html file names when generating blessed output files. That also left outdated files in the baseline alongside the files with the incorrect names, which I've now removed.) Updated spanview HTML title elements to match their content, replacing a hardcoded and incorrect name that was left in accidentally when originally submitted. Commit #4: added more test examples also improved Makefiles with support for non-zero exit status and to force validation of tests unless a specific test overrides it with a specific comment. Commit #5: Fixed rare issues after testing on real-world crate Commit #6: Addressed PR feedback, and removed temporary -Zexperimental-coverage -Zinstrument-coverage once again supports the latest capabilities of LLVM instrprof coverage instrumentation. Also fixed a bug in spanview. Commit #7: Fix closure handling, add tests for closures and inner items And cleaned up other tests for consistency, and to make it more clear where spans start/end by breaking up lines. Commit #8: renamed "typical" test results "expected" Now that the `llvm-cov show` tests are improved to normally expect matching actuals, and to allow individual tests to override that expectation. Commit #9: test coverage of inline generic struct function Commit #10: Addressed review feedback * Removed unnecessary Unreachable filter. * Replaced a match wildcard with remining variants. * Added more comments to help clarify the role of successors() in the CFG traversal Commit #11: refactoring based on feedback * refactored `fn coverage_spans()`. * changed the way I expand an empty coverage span to improve performance * fixed a typo that I had accidently left in, in visit.rs Commit #12: Optimized use of SourceMap and SourceFile Commit #13: Fixed a regression, and synched with upstream Some generated test file names changed due to some new change upstream. Commit #14: Stripping out crate disambiguators from demangled names These can vary depending on the test platform. Commit #15: Ignore llvm-cov show diff on test with generics, expand IO error message Tests with generics produce llvm-cov show results with demangled names that can include an unstable "crate disambiguator" (hex value). The value changes when run in the Rust CI Windows environment. I added a sed filter to strip them out (in a prior commit), but sed also appears to fail in the same environment. Until I can figure out a workaround, I'm just going to ignore this specific test result. I added a FIXME to follow up later, but it's not that critical. I also saw an error with Windows GNU, but the IO error did not specify a path for the directory or file that triggered the error. I updated the error messages to provide more info for next, time but also noticed some other tests with similar steps did not fail. Looks spurious. Commit #16: Modify rust-demangler to strip disambiguators by default Commit #17: Remove std::process::exit from coverage tests Due to Issue #77553, programs that call std::process::exit() do not generate coverage results on Windows MSVC. Commit #18: fix: test file paths exceeding Windows max path len
2020-09-01 23:15:17 +00:00
let expression = CounterExpression::new(
lhs_counter,
match op {
Op::Add => ExprKind::Add,
Op::Subtract => ExprKind::Subtract,
},
rhs_counter,
Updates to experimental coverage counter injection This is a combination of 18 commits. Commit #2: Additional examples and some small improvements. Commit #3: fixed mir-opt non-mir extensions and spanview title elements Corrected a fairly recent assumption in runtest.rs that all MIR dump files end in .mir. (It was appending .mir to the graphviz .dot and spanview .html file names when generating blessed output files. That also left outdated files in the baseline alongside the files with the incorrect names, which I've now removed.) Updated spanview HTML title elements to match their content, replacing a hardcoded and incorrect name that was left in accidentally when originally submitted. Commit #4: added more test examples also improved Makefiles with support for non-zero exit status and to force validation of tests unless a specific test overrides it with a specific comment. Commit #5: Fixed rare issues after testing on real-world crate Commit #6: Addressed PR feedback, and removed temporary -Zexperimental-coverage -Zinstrument-coverage once again supports the latest capabilities of LLVM instrprof coverage instrumentation. Also fixed a bug in spanview. Commit #7: Fix closure handling, add tests for closures and inner items And cleaned up other tests for consistency, and to make it more clear where spans start/end by breaking up lines. Commit #8: renamed "typical" test results "expected" Now that the `llvm-cov show` tests are improved to normally expect matching actuals, and to allow individual tests to override that expectation. Commit #9: test coverage of inline generic struct function Commit #10: Addressed review feedback * Removed unnecessary Unreachable filter. * Replaced a match wildcard with remining variants. * Added more comments to help clarify the role of successors() in the CFG traversal Commit #11: refactoring based on feedback * refactored `fn coverage_spans()`. * changed the way I expand an empty coverage span to improve performance * fixed a typo that I had accidently left in, in visit.rs Commit #12: Optimized use of SourceMap and SourceFile Commit #13: Fixed a regression, and synched with upstream Some generated test file names changed due to some new change upstream. Commit #14: Stripping out crate disambiguators from demangled names These can vary depending on the test platform. Commit #15: Ignore llvm-cov show diff on test with generics, expand IO error message Tests with generics produce llvm-cov show results with demangled names that can include an unstable "crate disambiguator" (hex value). The value changes when run in the Rust CI Windows environment. I added a sed filter to strip them out (in a prior commit), but sed also appears to fail in the same environment. Until I can figure out a workaround, I'm just going to ignore this specific test result. I added a FIXME to follow up later, but it's not that critical. I also saw an error with Windows GNU, but the IO error did not specify a path for the directory or file that triggered the error. I updated the error messages to provide more info for next, time but also noticed some other tests with similar steps did not fail. Looks spurious. Commit #16: Modify rust-demangler to strip disambiguators by default Commit #17: Remove std::process::exit from coverage tests Due to Issue #77553, programs that call std::process::exit() do not generate coverage results on Windows MSVC. Commit #18: fix: test file paths exceeding Windows max path len
2020-09-01 23:15:17 +00:00
);
debug!(
"Adding expression {:?} = {:?}, region: {:?}",
mapped_expression_index, expression, optional_region
Updates to experimental coverage counter injection This is a combination of 18 commits. Commit #2: Additional examples and some small improvements. Commit #3: fixed mir-opt non-mir extensions and spanview title elements Corrected a fairly recent assumption in runtest.rs that all MIR dump files end in .mir. (It was appending .mir to the graphviz .dot and spanview .html file names when generating blessed output files. That also left outdated files in the baseline alongside the files with the incorrect names, which I've now removed.) Updated spanview HTML title elements to match their content, replacing a hardcoded and incorrect name that was left in accidentally when originally submitted. Commit #4: added more test examples also improved Makefiles with support for non-zero exit status and to force validation of tests unless a specific test overrides it with a specific comment. Commit #5: Fixed rare issues after testing on real-world crate Commit #6: Addressed PR feedback, and removed temporary -Zexperimental-coverage -Zinstrument-coverage once again supports the latest capabilities of LLVM instrprof coverage instrumentation. Also fixed a bug in spanview. Commit #7: Fix closure handling, add tests for closures and inner items And cleaned up other tests for consistency, and to make it more clear where spans start/end by breaking up lines. Commit #8: renamed "typical" test results "expected" Now that the `llvm-cov show` tests are improved to normally expect matching actuals, and to allow individual tests to override that expectation. Commit #9: test coverage of inline generic struct function Commit #10: Addressed review feedback * Removed unnecessary Unreachable filter. * Replaced a match wildcard with remining variants. * Added more comments to help clarify the role of successors() in the CFG traversal Commit #11: refactoring based on feedback * refactored `fn coverage_spans()`. * changed the way I expand an empty coverage span to improve performance * fixed a typo that I had accidently left in, in visit.rs Commit #12: Optimized use of SourceMap and SourceFile Commit #13: Fixed a regression, and synched with upstream Some generated test file names changed due to some new change upstream. Commit #14: Stripping out crate disambiguators from demangled names These can vary depending on the test platform. Commit #15: Ignore llvm-cov show diff on test with generics, expand IO error message Tests with generics produce llvm-cov show results with demangled names that can include an unstable "crate disambiguator" (hex value). The value changes when run in the Rust CI Windows environment. I added a sed filter to strip them out (in a prior commit), but sed also appears to fail in the same environment. Until I can figure out a workaround, I'm just going to ignore this specific test result. I added a FIXME to follow up later, but it's not that critical. I also saw an error with Windows GNU, but the IO error did not specify a path for the directory or file that triggered the error. I updated the error messages to provide more info for next, time but also noticed some other tests with similar steps did not fail. Looks spurious. Commit #16: Modify rust-demangler to strip disambiguators by default Commit #17: Remove std::process::exit from coverage tests Due to Issue #77553, programs that call std::process::exit() do not generate coverage results on Windows MSVC. Commit #18: fix: test file paths exceeding Windows max path len
2020-09-01 23:15:17 +00:00
);
counter_expressions.push(expression);
new_indexes[original_index] = Some(mapped_expression_index);
if let Some(region) = optional_region {
expression_regions.push((Counter::expression(mapped_expression_index), region));
}
} else {
bug!(
"expression has one or more missing operands \
original_index={:?}, lhs={:?}, op={:?}, rhs={:?}, region={:?}",
original_index,
lhs,
op,
rhs,
optional_region,
);
}
}
(counter_expressions, expression_regions.into_iter())
}
fn unreachable_regions(&self) -> impl Iterator<Item = (Counter, &CodeRegion)> {
self.unreachable_regions.iter().map(|region| (Counter::zero(), region))
}
}