1
Fork 0

std: Use rustc_demangle from crates.io

No more need to duplicate the demangling routine between crates.io and
the standard library, we can use the exact same one!
This commit is contained in:
Alex Crichton 2018-12-14 14:37:51 -08:00
parent 1897657ef0
commit fcc8bb41e9
5 changed files with 27 additions and 236 deletions

View file

@ -14,11 +14,12 @@
use env;
use io::prelude::*;
use io;
use path::{self, Path};
use ptr;
use rustc_demangle::demangle;
use str;
use sync::atomic::{self, Ordering};
use path::{self, Path};
use sys::mutex::Mutex;
use ptr;
pub use sys::backtrace::{
unwind_backtrace,
@ -191,7 +192,14 @@ fn output(w: &mut dyn Write, idx: usize, frame: Frame,
PrintFormat::Short => write!(w, " {:2}: ", idx)?,
}
match s {
Some(string) => demangle(w, string, format)?,
Some(string) => {
let symbol = demangle(string);
match format {
PrintFormat::Full => write!(w, "{}", symbol)?,
// strip the trailing hash if short mode
PrintFormat::Short => write!(w, "{:#}", symbol)?,
}
}
None => w.write_all(b"<unknown>")?,
}
w.write_all(b"\n")
@ -235,228 +243,3 @@ fn output_fileline(w: &mut dyn Write,
w.write_all(b"\n")
}
// All rust symbols are in theory lists of "::"-separated identifiers. Some
// assemblers, however, can't handle these characters in symbol names. To get
// around this, we use C++-style mangling. The mangling method is:
//
// 1. Prefix the symbol with "_ZN"
// 2. For each element of the path, emit the length plus the element
// 3. End the path with "E"
//
// For example, "_ZN4testE" => "test" and "_ZN3foo3barE" => "foo::bar".
//
// We're the ones printing our backtraces, so we can't rely on anything else to
// demangle our symbols. It's *much* nicer to look at demangled symbols, so
// this function is implemented to give us nice pretty output.
//
// Note that this demangler isn't quite as fancy as it could be. We have lots
// of other information in our symbols like hashes, version, type information,
// etc. Additionally, this doesn't handle glue symbols at all.
pub fn demangle(writer: &mut dyn Write, mut s: &str, format: PrintFormat) -> io::Result<()> {
// During ThinLTO LLVM may import and rename internal symbols, so strip out
// those endings first as they're one of the last manglings applied to
// symbol names.
let llvm = ".llvm.";
if let Some(i) = s.find(llvm) {
let candidate = &s[i + llvm.len()..];
let all_hex = candidate.chars().all(|c| {
match c {
'A' ..= 'F' | '0' ..= '9' => true,
_ => false,
}
});
if all_hex {
s = &s[..i];
}
}
// Validate the symbol. If it doesn't look like anything we're
// expecting, we just print it literally. Note that we must handle non-rust
// symbols because we could have any function in the backtrace.
let mut valid = true;
let mut inner = s;
if s.len() > 4 && s.starts_with("_ZN") && s.ends_with("E") {
inner = &s[3 .. s.len() - 1];
// On Windows, dbghelp strips leading underscores, so we accept "ZN...E" form too.
} else if s.len() > 3 && s.starts_with("ZN") && s.ends_with("E") {
inner = &s[2 .. s.len() - 1];
} else {
valid = false;
}
if valid {
let mut chars = inner.chars();
while valid {
let mut i = 0;
for c in chars.by_ref() {
if c.is_numeric() {
i = i * 10 + c as usize - '0' as usize;
} else {
break
}
}
if i == 0 {
valid = chars.next().is_none();
break
} else if chars.by_ref().take(i - 1).count() != i - 1 {
valid = false;
}
}
}
// Alright, let's do this.
if !valid {
writer.write_all(s.as_bytes())?;
} else {
// remove the `::hfc2edb670e5eda97` part at the end of the symbol.
if format == PrintFormat::Short {
// The symbol in still mangled.
let mut split = inner.rsplitn(2, "17h");
match (split.next(), split.next()) {
(Some(addr), rest) => {
if addr.len() == 16 &&
addr.chars().all(|c| c.is_digit(16))
{
inner = rest.unwrap_or("");
}
}
_ => (),
}
}
let mut first = true;
while !inner.is_empty() {
if !first {
writer.write_all(b"::")?;
} else {
first = false;
}
let mut rest = inner;
while rest.chars().next().unwrap().is_numeric() {
rest = &rest[1..];
}
let i: usize = inner[.. (inner.len() - rest.len())].parse().unwrap();
inner = &rest[i..];
rest = &rest[..i];
if rest.starts_with("_$") {
rest = &rest[1..];
}
while !rest.is_empty() {
if rest.starts_with(".") {
if let Some('.') = rest[1..].chars().next() {
writer.write_all(b"::")?;
rest = &rest[2..];
} else {
writer.write_all(b".")?;
rest = &rest[1..];
}
} else if rest.starts_with("$") {
macro_rules! demangle {
($($pat:expr => $demangled:expr),*) => ({
$(if rest.starts_with($pat) {
writer.write_all($demangled)?;
rest = &rest[$pat.len()..];
} else)*
{
writer.write_all(rest.as_bytes())?;
break;
}
})
}
// see src/librustc/back/link.rs for these mappings
demangle! (
"$SP$" => b"@",
"$BP$" => b"*",
"$RF$" => b"&",
"$LT$" => b"<",
"$GT$" => b">",
"$LP$" => b"(",
"$RP$" => b")",
"$C$" => b",",
// in theory we can demangle any Unicode code point, but
// for simplicity we just catch the common ones.
"$u7e$" => b"~",
"$u20$" => b" ",
"$u27$" => b"'",
"$u5b$" => b"[",
"$u5d$" => b"]",
"$u7b$" => b"{",
"$u7d$" => b"}",
"$u3b$" => b";",
"$u2b$" => b"+",
"$u22$" => b"\""
)
} else {
let idx = match rest.char_indices().find(|&(_, c)| c == '$' || c == '.') {
None => rest.len(),
Some((i, _)) => i,
};
writer.write_all(rest[..idx].as_bytes())?;
rest = &rest[idx..];
}
}
}
}
Ok(())
}
#[cfg(test)]
mod tests {
use sys_common;
macro_rules! t { ($a:expr, $b:expr) => ({
let mut m = Vec::new();
sys_common::backtrace::demangle(&mut m,
$a,
super::PrintFormat::Full).unwrap();
assert_eq!(String::from_utf8(m).unwrap(), $b);
}) }
#[test]
fn demangle() {
t!("test", "test");
t!("_ZN4testE", "test");
t!("_ZN4test", "_ZN4test");
t!("_ZN4test1a2bcE", "test::a::bc");
}
#[test]
fn demangle_dollars() {
t!("_ZN4$RP$E", ")");
t!("_ZN8$RF$testE", "&test");
t!("_ZN8$BP$test4foobE", "*test::foob");
t!("_ZN9$u20$test4foobE", " test::foob");
t!("_ZN35Bar$LT$$u5b$u32$u3b$$u20$4$u5d$$GT$E", "Bar<[u32; 4]>");
}
#[test]
fn demangle_many_dollars() {
t!("_ZN13test$u20$test4foobE", "test test::foob");
t!("_ZN12test$BP$test4foobE", "test*test::foob");
}
#[test]
fn demangle_windows() {
t!("ZN4testE", "test");
t!("ZN13test$u20$test4foobE", "test test::foob");
t!("ZN12test$RF$test4foobE", "test&test::foob");
}
#[test]
fn demangle_elements_beginning_with_underscore() {
t!("_ZN13_$LT$test$GT$E", "<test>");
t!("_ZN28_$u7b$$u7b$closure$u7d$$u7d$E", "{{closure}}");
t!("_ZN15__STATIC_FMTSTRE", "__STATIC_FMTSTR");
}
#[test]
fn demangle_trait_impls() {
t!("_ZN71_$LT$Test$u20$$u2b$$u20$$u27$static$u20$as$u20$foo..Bar$LT$Test$GT$$GT$3barE",
"<Test + 'static as foo::Bar<Test>>::bar");
}
}