Merge tag 'trace-v7.3-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-tracegrafted

Pull tracing fixes from Steven Rostedt:

 - Don't destroy user event fields when removal fails

   User event fields are destroyed before the event is removed from
   visibility. But that can fail leaving the still visible event with no
   fields. Move the destroying of the fields to after the event is
   successfully removed from visibility.

 - Initialize function graph state is fork before calling
   copy_exec_state()

   For non-CLONE_VM forks, copy_exec_state() allocates a new
   task_exec_state. If that allocation fails, ftrace_graph_exit_task()
   will free the tasks ret_stack pointer. Since that pointer is still
   using the parent's ret_stack, it mistakenly frees the parent's
   pointer too.

   Call ftrace_graph_init() on the task first which will NULL out the
   new tasks's ret_stack and if the copy fails, it will not free
   anything.

 - Remove FGRAPH_MAX_INDEX

   The macro FGRAPH_MAX_INDEX was added but never used. Remove it.

 - Save ent_size in function graph printing of nested functions

   The function graph tracer needs to look at the next event to see if
   the next event is the return of the current function entry. If it is,
   it prints a single line:

	ktime_get();

   Otherwise it prints it like a nested function:

	tick_nohz_irq_exit() {
	    ktime_get();
	    kcpustat_irq_exit();
	}

   In order to look at the next event, it must save the current event so
   that it has the information to print from it. It saves the event in
   the iterator descriptor called "ent". What it doesn't save is the
   ent_size of the event which is now used to know if the function graph
   arguments are to be printed. The peek doesn't save the size so the
   size used happens to be that of the size of the last event that was
   seen.

   Save the entry event size in the iterator descriptor so that the
   correct size is used.

 - Fix several errors with freeing data in the histogram code

   The histogram code had a lot of leaked or or incorrect accounting
   when failures happen. Correct them.

 - Fix histogram regression of .percent and .graph modifiers

   Up until 6.3 histogram values could have "percent" or "graph"
   modifiers that changed how they were printed. But a change that added
   restricting histograms values from being strings, stack traces and
   other modifiers inadvertently prevented them from using the percent
   and graph modifiers, which were legal use cases for values.

   Put back the percent and graph modifiers.

 - Fix various typos in the comments

 - Set the trace_clock before initializing a histogram with clock
   argument

   The histogram API allows the user to specific which trace clock to
   use via a "clock=" string. The histogram is set up first before the
   clock is checked. If the passed in clock is not valid, it exits
   without fully fixing up the histogram leaving it on the list and a
   use-after-free can trigger.

   Update the clock argument first and if it fails then exit gracefully
   before the histogram trigger is placed on any lists.

 - Restore :mod: trailer after parsing in ftrace_set_clr_event

   The function ftrace_set_clr_event() modifies the parse string and
   needs to put it back to what was passed in. It searches for ":mod:"
   via a strsep() but fails to put back the first ':' in the string.

   Add back the ':' in the passed in string.

 - Take trace_array reference when opening a tracer options file

   The options files are dynamically created and some tracers add their
   own options. When a tracer adds their own list of options, the
   trace_array holding them has an array to hold the list of options for
   each tracer. This array increases in size via a krealloc(), and the
   new entry gets a newly allocated array to hold the options of the new
   tracer being added.

   The element in each entry of the tracer's option array holds a
   pointer back to the trace_array, a pointer to the tracer it is
   associated to, a pointer to the flags of the option.

   The issue is that these arrays are freed when the trace_array is
   freed when its instance it represents is removed from the instances
   directory. There's a race that an open of one of these options files
   can happen when the instance is being removed.

   Add a new helper function to be called by the open function of the
   options file to iterate all existing trace_arrays under a lock and
   find the one that has the given option element in one of it's tracer
   arrays. If found, then update the associated trace_array's reference
   counter to keep it from being freed. If not found, have the open call
   return -ENODEV.

 - Disable interrupts when acquiring the lock in rb_wake_up_waiters()

   The function rb_wake_up_waiters() assumes it will be called in
   interrupt context and does not disable irqs when taking
   cpu_buffer->reader_lock, which can be called in hard interrupt
   context. The issue is in PREEMPT_RT, this function is called in
   thread context leaving this lock open to a deadlock.

   Take the lock with interrupts disabled.

 - Use rcu_assign_pointer() for tmp_ops filter hash

   The tmp_ops used in update_ftrace_direct_mod() assigns its
   filter_hash field directly, but that field is annotated as __rcu and
   sparse complains. Assign it with rcu_assign_pointer()

 - Fix use-after-free in enable_trigger_private_data_free()

   The trace_event_call is accessed through the event_trigger_data's
   trace_event_file pointer to put the trace_event_call on freeing. The
   issue is that the trace_event_file data may have been freed already
   causing a use-after-free. Add a field to the event_trigger_data that
   points directly to the trace_event_call so that it can decrement its
   reference directly without needing to go through the
   trace_event_file.

 - Fix accounting of buffer data remote headers

   trace_buffer_desc_size() and trace_remote_alloc_buffer() undercount
   the number of pages is needed for the asked for size as it doesn't
   take into account the meta data on each page. Add a helper function
   to do the calculation properly and use that in these functions.

 - Catch nr_page_va overflow in ring_buffer_desc sizing

   The number of pages per remote ring buffer is capped by
   ring_buffer_desc::nr_page_va (32 bits). A buffer_size large enough to
   overflow that field would silently allocate a descriptor smaller than
   what was asked for.

 - Do not resize the subbuf order if any per_cpu buffer is disabled

   The mmapping of ring buffers disables resizing the subbuffers, but it
   is done per-cpu whereas the subbuf size change is done for all the
   per_cpu buffers under the buffer->mutex. It could change the size of
   some while the mapping is happening on others. Have the resize of the
   subbuf order check all the per_cpu buffers under the lock to see if
   any of them is disabled before starting and causing an inconsistency
   between buffers that are being mapped.

* tag 'trace-v7.3-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace: (25 commits)
  ring-buffer: Check resize_disabled before publishing the new subbuf order
  tracing/remotes: Catch nr_page_va overflow in ring_buffer_desc sizing
  tracing/remotes: Account for ring buffer page header in size calculation
  tracing: Don't dereference trace_event_file in deferred trigger free
  ftrace: Use rcu_assign_pointer() for tmp_ops filter hash
  ring-buffer: Acquire the lock with irqsave in rb_wake_up_waiters()
  tracing: Take trace_array reference when opening a tracer options file
  tracing: Fix ring_buffer_read_page_size() kernel-doc
  tracing: Restore :mod: trailer after parsing in ftrace_set_clr_event()
  tracing: Fix memory corruption from a "STACKTRACE" histogram key
  tracing: Fix memory corruption from the histogram stacktrace modifier
  tracing: Undo the registration when enabling the histogram trigger fails
  tracing: Take the reference before publishing the named histogram trigger
  tracing: Set the trace clock before registering the histogram trigger
  tracing: Fix typo "preceeded" in comment
  tracing: Fix typo "availabe" in comment
  tracing: Let histogram values keep the percent and graph modifiers
  tracing: Keep the entry count when the histogram stats allocation fails
  tracing: Free histogram the field rejected for a bad modifier
  tracing: Free histogram the var ref when its initialization fails
  ...

1 files changed, 1258 insertions, 0 deletions

diff --git a/rust/proc-macro2/fallback.rs b/rust/proc-macro2/fallback.rs
new file mode 100644
index 000000000..9b43c97df
--- /dev/null
+++ b/rust/proc-macro2/fallback.rs
@@ -0,0 +1,1258 @@
+// SPDX-License-Identifier: Apache-2.0 OR MIT
+
+#[cfg(wrap_proc_macro)]
+use crate::imp;
+#[cfg(span_locations)]
+use crate::location::LineColumn;
+use crate::parse::{self, Cursor};
+use crate::rcvec::{RcVec, RcVecBuilder, RcVecIntoIter, RcVecMut};
+use crate::{Delimiter, Spacing, TokenTree};
+#[cfg(all(span_locations, not(fuzzing)))]
+use alloc::collections::BTreeMap;
+#[cfg(all(span_locations, not(fuzzing)))]
+use core::cell::RefCell;
+#[cfg(span_locations)]
+use core::cmp;
+#[cfg(all(span_locations, not(fuzzing)))]
+use core::cmp::Ordering;
+use core::fmt::{self, Debug, Display, Write};
+use core::mem::ManuallyDrop;
+#[cfg(span_locations)]
+use core::ops::Range;
+use core::ops::RangeBounds;
+use core::ptr;
+use core::str;
+#[cfg(feature = "proc-macro")]
+use core::str::FromStr;
+use std::ffi::CStr;
+#[cfg(wrap_proc_macro)]
+use std::panic;
+#[cfg(span_locations)]
+use std::path::PathBuf;
+
+/// Force use of proc-macro2's fallback implementation of the API for now, even
+/// if the compiler's implementation is available.
+pub fn force() {
+    #[cfg(wrap_proc_macro)]
+    crate::detection::force_fallback();
+}
+
+/// Resume using the compiler's implementation of the proc macro API if it is
+/// available.
+pub fn unforce() {
+    #[cfg(wrap_proc_macro)]
+    crate::detection::unforce_fallback();
+}
+
+#[derive(Clone)]
+pub(crate) struct TokenStream {
+    inner: RcVec<TokenTree>,
+}
+
+#[derive(Debug)]
+pub(crate) struct LexError {
+    pub(crate) span: Span,
+}
+
+impl LexError {
+    pub(crate) fn span(&self) -> Span {
+        self.span
+    }
+
+    pub(crate) fn call_site() -> Self {
+        LexError {
+            span: Span::call_site(),
+        }
+    }
+}
+
+impl TokenStream {
+    pub(crate) fn new() -> Self {
+        TokenStream {
+            inner: RcVecBuilder::new().build(),
+        }
+    }
+
+    pub(crate) fn from_str_checked(src: &str) -> Result<Self, LexError> {
+        // Create a dummy file & add it to the source map
+        let mut cursor = get_cursor(src);
+
+        // Strip a byte order mark if present
+        const BYTE_ORDER_MARK: &str = "\u{feff}";
+        if cursor.starts_with(BYTE_ORDER_MARK) {
+            cursor = cursor.advance(BYTE_ORDER_MARK.len());
+        }
+
+        parse::token_stream(cursor)
+    }
+
+    #[cfg(feature = "proc-macro")]
+    pub(crate) fn from_str_unchecked(src: &str) -> Self {
+        Self::from_str_checked(src).unwrap()
+    }
+
+    pub(crate) fn is_empty(&self) -> bool {
+        self.inner.len() == 0
+    }
+
+    fn take_inner(self) -> RcVecBuilder<TokenTree> {
+        let nodrop = ManuallyDrop::new(self);
+        unsafe { ptr::read(&nodrop.inner) }.make_owned()
+    }
+}
+
+fn push_token_from_proc_macro(mut vec: RcVecMut<TokenTree>, token: TokenTree) {
+    // https://github.com/dtolnay/proc-macro2/issues/235
+    match token {
+        TokenTree::Literal(crate::Literal {
+            #[cfg(wrap_proc_macro)]
+                inner: crate::imp::Literal::Fallback(literal),
+            #[cfg(not(wrap_proc_macro))]
+                inner: literal,
+            ..
+        }) if literal.repr.starts_with('-') => {
+            push_negative_literal(vec, literal);
+        }
+        _ => vec.push(token),
+    }
+
+    #[cold]
+    fn push_negative_literal(mut vec: RcVecMut<TokenTree>, mut literal: Literal) {
+        literal.repr.remove(0);
+        let mut punct = crate::Punct::new('-', Spacing::Alone);
+        punct.set_span(crate::Span::_new_fallback(literal.span));
+        vec.push(TokenTree::Punct(punct));
+        vec.push(TokenTree::Literal(crate::Literal::_new_fallback(literal)));
+    }
+}
+
+// Nonrecursive to prevent stack overflow.
+impl Drop for TokenStream {
+    fn drop(&mut self) {
+        let mut stack = Vec::new();
+        let mut current = match self.inner.get_mut() {
+            Some(inner) => inner.take().into_iter(),
+            None => return,
+        };
+        loop {
+            while let Some(token) = current.next() {
+                let group = match token {
+                    TokenTree::Group(group) => group.inner,
+                    _ => continue,
+                };
+                #[cfg(wrap_proc_macro)]
+                let group = match group {
+                    crate::imp::Group::Fallback(group) => group,
+                    crate::imp::Group::Compiler(_) => continue,
+                };
+                let mut group = group;
+                if let Some(inner) = group.stream.inner.get_mut() {
+                    stack.push(current);
+                    current = inner.take().into_iter();
+                }
+            }
+            match stack.pop() {
+                Some(next) => current = next,
+                None => return,
+            }
+        }
+    }
+}
+
+pub(crate) struct TokenStreamBuilder {
+    inner: RcVecBuilder<TokenTree>,
+}
+
+impl TokenStreamBuilder {
+    pub(crate) fn new() -> Self {
+        TokenStreamBuilder {
+            inner: RcVecBuilder::new(),
+        }
+    }
+
+    pub(crate) fn with_capacity(cap: usize) -> Self {
+        TokenStreamBuilder {
+            inner: RcVecBuilder::with_capacity(cap),
+        }
+    }
+
+    pub(crate) fn push_token_from_parser(&mut self, tt: TokenTree) {
+        self.inner.push(tt);
+    }
+
+    pub(crate) fn build(self) -> TokenStream {
+        TokenStream {
+            inner: self.inner.build(),
+        }
+    }
+}
+
+#[cfg(span_locations)]
+fn get_cursor(src: &str) -> Cursor {
+    #[cfg(fuzzing)]
+    return Cursor { rest: src, off: 1 };
+
+    // Create a dummy file & add it to the source map
+    #[cfg(not(fuzzing))]
+    SOURCE_MAP.with(|sm| {
+        let mut sm = sm.borrow_mut();
+        let span = sm.add_file(src);
+        Cursor {
+            rest: src,
+            off: span.lo,
+        }
+    })
+}
+
+#[cfg(not(span_locations))]
+fn get_cursor(src: &str) -> Cursor {
+    Cursor { rest: src }
+}
+
+impl Display for LexError {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        f.write_str("cannot parse string into token stream")
+    }
+}
+
+impl Display for TokenStream {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        let mut joint = false;
+        for (i, tt) in self.inner.iter().enumerate() {
+            if i != 0 && !joint {
+                write!(f, " ")?;
+            }
+            joint = false;
+            match tt {
+                TokenTree::Group(tt) => Display::fmt(tt, f),
+                TokenTree::Ident(tt) => Display::fmt(tt, f),
+                TokenTree::Punct(tt) => {
+                    joint = tt.spacing() == Spacing::Joint;
+                    Display::fmt(tt, f)
+                }
+                TokenTree::Literal(tt) => Display::fmt(tt, f),
+            }?;
+        }
+
+        Ok(())
+    }
+}
+
+impl Debug for TokenStream {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        f.write_str("TokenStream ")?;
+        f.debug_list().entries(self.clone()).finish()
+    }
+}
+
+#[cfg(feature = "proc-macro")]
+impl From<proc_macro::TokenStream> for TokenStream {
+    fn from(inner: proc_macro::TokenStream) -> Self {
+        TokenStream::from_str_unchecked(&inner.to_string())
+    }
+}
+
+#[cfg(feature = "proc-macro")]
+impl From<TokenStream> for proc_macro::TokenStream {
+    fn from(inner: TokenStream) -> Self {
+        proc_macro::TokenStream::from_str_unchecked(&inner.to_string())
+    }
+}
+
+impl From<TokenTree> for TokenStream {
+    fn from(tree: TokenTree) -> Self {
+        let mut stream = RcVecBuilder::new();
+        push_token_from_proc_macro(stream.as_mut(), tree);
+        TokenStream {
+            inner: stream.build(),
+        }
+    }
+}
+
+impl FromIterator<TokenTree> for TokenStream {
+    fn from_iter<I: IntoIterator<Item = TokenTree>>(tokens: I) -> Self {
+        let mut stream = TokenStream::new();
+        stream.extend(tokens);
+        stream
+    }
+}
+
+impl FromIterator<TokenStream> for TokenStream {
+    fn from_iter<I: IntoIterator<Item = TokenStream>>(streams: I) -> Self {
+        let mut v = RcVecBuilder::new();
+
+        for stream in streams {
+            v.extend(stream.take_inner());
+        }
+
+        TokenStream { inner: v.build() }
+    }
+}
+
+impl Extend<TokenTree> for TokenStream {
+    fn extend<I: IntoIterator<Item = TokenTree>>(&mut self, tokens: I) {
+        let mut vec = self.inner.make_mut();
+        tokens
+            .into_iter()
+            .for_each(|token| push_token_from_proc_macro(vec.as_mut(), token));
+    }
+}
+
+impl Extend<TokenStream> for TokenStream {
+    fn extend<I: IntoIterator<Item = TokenStream>>(&mut self, streams: I) {
+        self.inner.make_mut().extend(streams.into_iter().flatten());
+    }
+}
+
+pub(crate) type TokenTreeIter = RcVecIntoIter<TokenTree>;
+
+impl IntoIterator for TokenStream {
+    type Item = TokenTree;
+    type IntoIter = TokenTreeIter;
+
+    fn into_iter(self) -> TokenTreeIter {
+        self.take_inner().into_iter()
+    }
+}
+
+#[cfg(all(span_locations, not(fuzzing)))]
+thread_local! {
+    static SOURCE_MAP: RefCell<SourceMap> = RefCell::new(SourceMap {
+        // Start with a single dummy file which all call_site() and def_site()
+        // spans reference.
+        files: vec![FileInfo {
+            source_text: String::new(),
+            span: Span { lo: 0, hi: 0 },
+            lines: vec![0],
+            char_index_to_byte_offset: BTreeMap::new(),
+        }],
+    });
+}
+
+#[cfg(span_locations)]
+pub(crate) fn invalidate_current_thread_spans() {
+    #[cfg(not(fuzzing))]
+    SOURCE_MAP.with(|sm| sm.borrow_mut().files.truncate(1));
+}
+
+#[cfg(all(span_locations, not(fuzzing)))]
+struct FileInfo {
+    source_text: String,
+    span: Span,
+    lines: Vec<usize>,
+    char_index_to_byte_offset: BTreeMap<usize, usize>,
+}
+
+#[cfg(all(span_locations, not(fuzzing)))]
+impl FileInfo {
+    fn offset_line_column(&self, offset: usize) -> LineColumn {
+        assert!(self.span_within(Span {
+            lo: offset as u32,
+            hi: offset as u32,
+        }));
+        let offset = offset - self.span.lo as usize;
+        match self.lines.binary_search(&offset) {
+            Ok(found) => LineColumn {
+                line: found + 1,
+                column: 0,
+            },
+            Err(idx) => LineColumn {
+                line: idx,
+                column: offset - self.lines[idx - 1],
+            },
+        }
+    }
+
+    fn span_within(&self, span: Span) -> bool {
+        span.lo >= self.span.lo && span.hi <= self.span.hi
+    }
+
+    fn byte_range(&mut self, span: Span) -> Range<usize> {
+        let lo_char = (span.lo - self.span.lo) as usize;
+
+        // Look up offset of the largest already-computed char index that is
+        // less than or equal to the current requested one. We resume counting
+        // chars from that point.
+        let (&last_char_index, &last_byte_offset) = self
+            .char_index_to_byte_offset
+            .range(..=lo_char)
+            .next_back()
+            .unwrap_or((&0, &0));
+
+        let lo_byte = if last_char_index == lo_char {
+            last_byte_offset
+        } else {
+            let total_byte_offset = match self.source_text[last_byte_offset..]
+                .char_indices()
+                .nth(lo_char - last_char_index)
+            {
+                Some((additional_offset, _ch)) => last_byte_offset + additional_offset,
+                None => self.source_text.len(),
+            };
+            self.char_index_to_byte_offset
+                .insert(lo_char, total_byte_offset);
+            total_byte_offset
+        };
+
+        let trunc_lo = &self.source_text[lo_byte..];
+        let char_len = (span.hi - span.lo) as usize;
+        lo_byte..match trunc_lo.char_indices().nth(char_len) {
+            Some((offset, _ch)) => lo_byte + offset,
+            None => self.source_text.len(),
+        }
+    }
+
+    fn source_text(&mut self, span: Span) -> String {
+        let byte_range = self.byte_range(span);
+        self.source_text[byte_range].to_owned()
+    }
+}
+
+/// Computes the offsets of each line in the given source string
+/// and the total number of characters
+#[cfg(all(span_locations, not(fuzzing)))]
+fn lines_offsets(s: &str) -> (usize, Vec<usize>) {
+    let mut lines = vec![0];
+    let mut total = 0;
+
+    for ch in s.chars() {
+        total += 1;
+        if ch == '\n' {
+            lines.push(total);
+        }
+    }
+
+    (total, lines)
+}
+
+#[cfg(all(span_locations, not(fuzzing)))]
+struct SourceMap {
+    files: Vec<FileInfo>,
+}
+
+#[cfg(all(span_locations, not(fuzzing)))]
+impl SourceMap {
+    fn next_start_pos(&self) -> u32 {
+        // Add 1 so there's always space between files.
+        //
+        // We'll always have at least 1 file, as we initialize our files list
+        // with a dummy file.
+        self.files.last().unwrap().span.hi + 1
+    }
+
+    fn add_file(&mut self, src: &str) -> Span {
+        let (len, lines) = lines_offsets(src);
+        let lo = self.next_start_pos();
+        let span = Span {
+            lo,
+            hi: lo + (len as u32),
+        };
+
+        self.files.push(FileInfo {
+            source_text: src.to_owned(),
+            span,
+            lines,
+            // Populated lazily by source_text().
+            char_index_to_byte_offset: BTreeMap::new(),
+        });
+
+        span
+    }
+
+    fn find(&self, span: Span) -> usize {
+        match self.files.binary_search_by(|file| {
+            if file.span.hi < span.lo {
+                Ordering::Less
+            } else if file.span.lo > span.hi {
+                Ordering::Greater
+            } else {
+                assert!(file.span_within(span));
+                Ordering::Equal
+            }
+        }) {
+            Ok(i) => i,
+            Err(_) => unreachable!("Invalid span with no related FileInfo!"),
+        }
+    }
+
+    fn filepath(&self, span: Span) -> String {
+        let i = self.find(span);
+        if i == 0 {
+            "<unspecified>".to_owned()
+        } else {
+            format!("<parsed string {}>", i)
+        }
+    }
+
+    fn fileinfo(&self, span: Span) -> &FileInfo {
+        let i = self.find(span);
+        &self.files[i]
+    }
+
+    fn fileinfo_mut(&mut self, span: Span) -> &mut FileInfo {
+        let i = self.find(span);
+        &mut self.files[i]
+    }
+}
+
+#[derive(Clone, Copy, PartialEq, Eq)]
+pub(crate) struct Span {
+    #[cfg(span_locations)]
+    pub(crate) lo: u32,
+    #[cfg(span_locations)]
+    pub(crate) hi: u32,
+}
+
+impl Span {
+    #[cfg(not(span_locations))]
+    pub(crate) fn call_site() -> Self {
+        Span {}
+    }
+
+    #[cfg(span_locations)]
+    pub(crate) fn call_site() -> Self {
+        Span { lo: 0, hi: 0 }
+    }
+
+    pub(crate) fn mixed_site() -> Self {
+        Span::call_site()
+    }
+
+    #[cfg(procmacro2_semver_exempt)]
+    pub(crate) fn def_site() -> Self {
+        Span::call_site()
+    }
+
+    pub(crate) fn resolved_at(&self, _other: Span) -> Span {
+        // Stable spans consist only of line/column information, so
+        // `resolved_at` and `located_at` only select which span the
+        // caller wants line/column information from.
+        *self
+    }
+
+    pub(crate) fn located_at(&self, other: Span) -> Span {
+        other
+    }
+
+    #[cfg(span_locations)]
+    pub(crate) fn byte_range(&self) -> Range<usize> {
+        #[cfg(fuzzing)]
+        return 0..0;
+
+        #[cfg(not(fuzzing))]
+        {
+            if self.is_call_site() {
+                0..0
+            } else {
+                SOURCE_MAP.with(|sm| sm.borrow_mut().fileinfo_mut(*self).byte_range(*self))
+            }
+        }
+    }
+
+    #[cfg(span_locations)]
+    pub(crate) fn start(&self) -> LineColumn {
+        #[cfg(fuzzing)]
+        return LineColumn { line: 0, column: 0 };
+
+        #[cfg(not(fuzzing))]
+        SOURCE_MAP.with(|sm| {
+            let sm = sm.borrow();
+            let fi = sm.fileinfo(*self);
+            fi.offset_line_column(self.lo as usize)
+        })
+    }
+
+    #[cfg(span_locations)]
+    pub(crate) fn end(&self) -> LineColumn {
+        #[cfg(fuzzing)]
+        return LineColumn { line: 0, column: 0 };
+
+        #[cfg(not(fuzzing))]
+        SOURCE_MAP.with(|sm| {
+            let sm = sm.borrow();
+            let fi = sm.fileinfo(*self);
+            fi.offset_line_column(self.hi as usize)
+        })
+    }
+
+    #[cfg(span_locations)]
+    pub(crate) fn file(&self) -> String {
+        #[cfg(fuzzing)]
+        return "<unspecified>".to_owned();
+
+        #[cfg(not(fuzzing))]
+        SOURCE_MAP.with(|sm| {
+            let sm = sm.borrow();
+            sm.filepath(*self)
+        })
+    }
+
+    #[cfg(span_locations)]
+    pub(crate) fn local_file(&self) -> Option<PathBuf> {
+        None
+    }
+
+    #[cfg(not(span_locations))]
+    pub(crate) fn join(&self, _other: Span) -> Option<Span> {
+        Some(Span {})
+    }
+
+    #[cfg(span_locations)]
+    pub(crate) fn join(&self, other: Span) -> Option<Span> {
+        #[cfg(fuzzing)]
+        return {
+            let _ = other;
+            None
+        };
+
+        #[cfg(not(fuzzing))]
+        SOURCE_MAP.with(|sm| {
+            let sm = sm.borrow();
+            // If `other` is not within the same FileInfo as us, return None.
+            if !sm.fileinfo(*self).span_within(other) {
+                return None;
+            }
+            Some(Span {
+                lo: cmp::min(self.lo, other.lo),
+                hi: cmp::max(self.hi, other.hi),
+            })
+        })
+    }
+
+    #[cfg(not(span_locations))]
+    pub(crate) fn source_text(&self) -> Option<String> {
+        None
+    }
+
+    #[cfg(span_locations)]
+    pub(crate) fn source_text(&self) -> Option<String> {
+        #[cfg(fuzzing)]
+        return None;
+
+        #[cfg(not(fuzzing))]
+        {
+            if self.is_call_site() {
+                None
+            } else {
+                Some(SOURCE_MAP.with(|sm| sm.borrow_mut().fileinfo_mut(*self).source_text(*self)))
+            }
+        }
+    }
+
+    #[cfg(not(span_locations))]
+    pub(crate) fn first_byte(self) -> Self {
+        self
+    }
+
+    #[cfg(span_locations)]
+    pub(crate) fn first_byte(self) -> Self {
+        Span {
+            lo: self.lo,
+            hi: cmp::min(self.lo.saturating_add(1), self.hi),
+        }
+    }
+
+    #[cfg(not(span_locations))]
+    pub(crate) fn last_byte(self) -> Self {
+        self
+    }
+
+    #[cfg(span_locations)]
+    pub(crate) fn last_byte(self) -> Self {
+        Span {
+            lo: cmp::max(self.hi.saturating_sub(1), self.lo),
+            hi: self.hi,
+        }
+    }
+
+    #[cfg(span_locations)]
+    fn is_call_site(&self) -> bool {
+        self.lo == 0 && self.hi == 0
+    }
+}
+
+impl Debug for Span {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        #[cfg(span_locations)]
+        return write!(f, "bytes({}..{})", self.lo, self.hi);
+
+        #[cfg(not(span_locations))]
+        write!(f, "Span")
+    }
+}
+
+pub(crate) fn debug_span_field_if_nontrivial(debug: &mut fmt::DebugStruct, span: Span) {
+    #[cfg(span_locations)]
+    {
+        if span.is_call_site() {
+            return;
+        }
+    }
+
+    if cfg!(span_locations) {
+        debug.field("span", &span);
+    }
+}
+
+#[derive(Clone)]
+pub(crate) struct Group {
+    delimiter: Delimiter,
+    stream: TokenStream,
+    span: Span,
+}
+
+impl Group {
+    pub(crate) fn new(delimiter: Delimiter, stream: TokenStream) -> Self {
+        Group {
+            delimiter,
+            stream,
+            span: Span::call_site(),
+        }
+    }
+
+    pub(crate) fn delimiter(&self) -> Delimiter {
+        self.delimiter
+    }
+
+    pub(crate) fn stream(&self) -> TokenStream {
+        self.stream.clone()
+    }
+
+    pub(crate) fn span(&self) -> Span {
+        self.span
+    }
+
+    pub(crate) fn span_open(&self) -> Span {
+        self.span.first_byte()
+    }
+
+    pub(crate) fn span_close(&self) -> Span {
+        self.span.last_byte()
+    }
+
+    pub(crate) fn set_span(&mut self, span: Span) {
+        self.span = span;
+    }
+}
+
+impl Display for Group {
+    // We attempt to match libproc_macro's formatting.
+    // Empty parens: ()
+    // Nonempty parens: (...)
+    // Empty brackets: []
+    // Nonempty brackets: [...]
+    // Empty braces: { }
+    // Nonempty braces: { ... }
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        let (open, close) = match self.delimiter {
+            Delimiter::Parenthesis => ("(", ")"),
+            Delimiter::Brace => ("{ ", "}"),
+            Delimiter::Bracket => ("[", "]"),
+            Delimiter::None => ("", ""),
+        };
+
+        f.write_str(open)?;
+        Display::fmt(&self.stream, f)?;
+        if self.delimiter == Delimiter::Brace && !self.stream.inner.is_empty() {
+            f.write_str(" ")?;
+        }
+        f.write_str(close)?;
+
+        Ok(())
+    }
+}
+
+impl Debug for Group {
+    fn fmt(&self, fmt: &mut fmt::Formatter) -> fmt::Result {
+        let mut debug = fmt.debug_struct("Group");
+        debug.field("delimiter", &self.delimiter);
+        debug.field("stream", &self.stream);
+        debug_span_field_if_nontrivial(&mut debug, self.span);
+        debug.finish()
+    }
+}
+
+#[derive(Clone)]
+pub(crate) struct Ident {
+    sym: Box<str>,
+    span: Span,
+    raw: bool,
+}
+
+impl Ident {
+    #[track_caller]
+    pub(crate) fn new_checked(string: &str, span: Span) -> Self {
+        validate_ident(string);
+        Ident::new_unchecked(string, span)
+    }
+
+    pub(crate) fn new_unchecked(string: &str, span: Span) -> Self {
+        Ident {
+            sym: Box::from(string),
+            span,
+            raw: false,
+        }
+    }
+
+    #[track_caller]
+    pub(crate) fn new_raw_checked(string: &str, span: Span) -> Self {
+        validate_ident_raw(string);
+        Ident::new_raw_unchecked(string, span)
+    }
+
+    pub(crate) fn new_raw_unchecked(string: &str, span: Span) -> Self {
+        Ident {
+            sym: Box::from(string),
+            span,
+            raw: true,
+        }
+    }
+
+    pub(crate) fn span(&self) -> Span {
+        self.span
+    }
+
+    pub(crate) fn set_span(&mut self, span: Span) {
+        self.span = span;
+    }
+}
+
+pub(crate) fn is_ident_start(c: char) -> bool {
+    c == '_' || c.is_ascii_alphabetic()
+}
+
+pub(crate) fn is_ident_continue(c: char) -> bool {
+    c == '_' || c.is_ascii_alphanumeric()
+}
+
+#[track_caller]
+fn validate_ident(string: &str) {
+    if string.is_empty() {
+        panic!("Ident is not allowed to be empty; use Option<Ident>");
+    }
+
+    if string.bytes().all(|digit| b'0' <= digit && digit <= b'9') {
+        panic!("Ident cannot be a number; use Literal instead");
+    }
+
+    fn ident_ok(string: &str) -> bool {
+        let mut chars = string.chars();
+        let first = chars.next().unwrap();
+        if !is_ident_start(first) {
+            return false;
+        }
+        for ch in chars {
+            if !is_ident_continue(ch) {
+                return false;
+            }
+        }
+        true
+    }
+
+    if !ident_ok(string) {
+        panic!("{:?} is not a valid Ident", string);
+    }
+}
+
+#[track_caller]
+fn validate_ident_raw(string: &str) {
+    validate_ident(string);
+
+    match string {
+        "_" | "super" | "self" | "Self" | "crate" => {
+            panic!("`r#{}` cannot be a raw identifier", string);
+        }
+        _ => {}
+    }
+}
+
+impl PartialEq for Ident {
+    fn eq(&self, other: &Ident) -> bool {
+        self.sym == other.sym && self.raw == other.raw
+    }
+}
+
+impl<T> PartialEq<T> for Ident
+where
+    T: ?Sized + AsRef<str>,
+{
+    fn eq(&self, other: &T) -> bool {
+        let other = other.as_ref();
+        if self.raw {
+            other.starts_with("r#") && *self.sym == other[2..]
+        } else {
+            *self.sym == *other
+        }
+    }
+}
+
+impl Display for Ident {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        if self.raw {
+            f.write_str("r#")?;
+        }
+        Display::fmt(&self.sym, f)
+    }
+}
+
+#[allow(clippy::missing_fields_in_debug)]
+impl Debug for Ident {
+    // Ident(proc_macro), Ident(r#union)
+    #[cfg(not(span_locations))]
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        let mut debug = f.debug_tuple("Ident");
+        debug.field(&format_args!("{}", self));
+        debug.finish()
+    }
+
+    // Ident {
+    //     sym: proc_macro,
+    //     span: bytes(128..138)
+    // }
+    #[cfg(span_locations)]
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        let mut debug = f.debug_struct("Ident");
+        debug.field("sym", &format_args!("{}", self));
+        debug_span_field_if_nontrivial(&mut debug, self.span);
+        debug.finish()
+    }
+}
+
+#[derive(Clone)]
+pub(crate) struct Literal {
+    pub(crate) repr: String,
+    span: Span,
+}
+
+macro_rules! suffixed_numbers {
+    ($($name:ident => $kind:ident,)*) => ($(
+        pub(crate) fn $name(n: $kind) -> Literal {
+            Literal::_new(format!(concat!("{}", stringify!($kind)), n))
+        }
+    )*)
+}
+
+macro_rules! unsuffixed_numbers {
+    ($($name:ident => $kind:ident,)*) => ($(
+        pub(crate) fn $name(n: $kind) -> Literal {
+            Literal::_new(n.to_string())
+        }
+    )*)
+}
+
+impl Literal {
+    pub(crate) fn _new(repr: String) -> Self {
+        Literal {
+            repr,
+            span: Span::call_site(),
+        }
+    }
+
+    pub(crate) fn from_str_checked(repr: &str) -> Result<Self, LexError> {
+        let mut cursor = get_cursor(repr);
+        #[cfg(span_locations)]
+        let lo = cursor.off;
+
+        let negative = cursor.starts_with_char('-');
+        if negative {
+            cursor = cursor.advance(1);
+            if !cursor.starts_with_fn(|ch| ch.is_ascii_digit()) {
+                return Err(LexError::call_site());
+            }
+        }
+
+        if let Ok((rest, mut literal)) = parse::literal(cursor) {
+            if rest.is_empty() {
+                if negative {
+                    literal.repr.insert(0, '-');
+                }
+                literal.span = Span {
+                    #[cfg(span_locations)]
+                    lo,
+                    #[cfg(span_locations)]
+                    hi: rest.off,
+                };
+                return Ok(literal);
+            }
+        }
+        Err(LexError::call_site())
+    }
+
+    pub(crate) unsafe fn from_str_unchecked(repr: &str) -> Self {
+        Literal::_new(repr.to_owned())
+    }
+
+    suffixed_numbers! {
+        u8_suffixed => u8,
+        u16_suffixed => u16,
+        u32_suffixed => u32,
+        u64_suffixed => u64,
+        u128_suffixed => u128,
+        usize_suffixed => usize,
+        i8_suffixed => i8,
+        i16_suffixed => i16,
+        i32_suffixed => i32,
+        i64_suffixed => i64,
+        i128_suffixed => i128,
+        isize_suffixed => isize,
+
+        f32_suffixed => f32,
+        f64_suffixed => f64,
+    }
+
+    unsuffixed_numbers! {
+        u8_unsuffixed => u8,
+        u16_unsuffixed => u16,
+        u32_unsuffixed => u32,
+        u64_unsuffixed => u64,
+        u128_unsuffixed => u128,
+        usize_unsuffixed => usize,
+        i8_unsuffixed => i8,
+        i16_unsuffixed => i16,
+        i32_unsuffixed => i32,
+        i64_unsuffixed => i64,
+        i128_unsuffixed => i128,
+        isize_unsuffixed => isize,
+    }
+
+    pub(crate) fn f32_unsuffixed(f: f32) -> Literal {
+        let mut s = f.to_string();
+        if !s.contains('.') {
+            s.push_str(".0");
+        }
+        Literal::_new(s)
+    }
+
+    pub(crate) fn f64_unsuffixed(f: f64) -> Literal {
+        let mut s = f.to_string();
+        if !s.contains('.') {
+            s.push_str(".0");
+        }
+        Literal::_new(s)
+    }
+
+    pub(crate) fn string(string: &str) -> Literal {
+        let mut repr = String::with_capacity(string.len() + 2);
+        repr.push('"');
+        escape_utf8(string, &mut repr);
+        repr.push('"');
+        Literal::_new(repr)
+    }
+
+    pub(crate) fn character(ch: char) -> Literal {
+        let mut repr = String::new();
+        repr.push('\'');
+        if ch == '"' {
+            // escape_debug turns this into '\"' which is unnecessary.
+            repr.push(ch);
+        } else {
+            repr.extend(ch.escape_debug());
+        }
+        repr.push('\'');
+        Literal::_new(repr)
+    }
+
+    pub(crate) fn byte_character(byte: u8) -> Literal {
+        let mut repr = "b'".to_string();
+        #[allow(clippy::match_overlapping_arm)]
+        match byte {
+            b'\0' => repr.push_str(r"\0"),
+            b'\t' => repr.push_str(r"\t"),
+            b'\n' => repr.push_str(r"\n"),
+            b'\r' => repr.push_str(r"\r"),
+            b'\'' => repr.push_str(r"\'"),
+            b'\\' => repr.push_str(r"\\"),
+            b'\x20'..=b'\x7E' => repr.push(byte as char),
+            _ => {
+                let _ = write!(repr, r"\x{:02X}", byte);
+            }
+        }
+        repr.push('\'');
+        Literal::_new(repr)
+    }
+
+    pub(crate) fn byte_string(bytes: &[u8]) -> Literal {
+        let mut repr = "b\"".to_string();
+        let mut bytes = bytes.iter();
+        while let Some(&b) = bytes.next() {
+            #[allow(clippy::match_overlapping_arm)]
+            match b {
+                b'\0' => repr.push_str(match bytes.as_slice().first() {
+                    // circumvent clippy::octal_escapes lint
+                    Some(b'0'..=b'7') => r"\x00",
+                    _ => r"\0",
+                }),
+                b'\t' => repr.push_str(r"\t"),
+                b'\n' => repr.push_str(r"\n"),
+                b'\r' => repr.push_str(r"\r"),
+                b'"' => repr.push_str("\\\""),
+                b'\\' => repr.push_str(r"\\"),
+                b'\x20'..=b'\x7E' => repr.push(b as char),
+                _ => {
+                    let _ = write!(repr, r"\x{:02X}", b);
+                }
+            }
+        }
+        repr.push('"');
+        Literal::_new(repr)
+    }
+
+    pub(crate) fn c_string(string: &CStr) -> Literal {
+        let mut repr = "c\"".to_string();
+        let mut bytes = string.to_bytes();
+        while !bytes.is_empty() {
+            let (valid, invalid) = match str::from_utf8(bytes) {
+                Ok(all_valid) => {
+                    bytes = b"";
+                    (all_valid, bytes)
+                }
+                Err(utf8_error) => {
+                    let (valid, rest) = bytes.split_at(utf8_error.valid_up_to());
+                    let valid = str::from_utf8(valid).unwrap();
+                    let invalid = utf8_error
+                        .error_len()
+                        .map_or(rest, |error_len| &rest[..error_len]);
+                    bytes = &bytes[valid.len() + invalid.len()..];
+                    (valid, invalid)
+                }
+            };
+            escape_utf8(valid, &mut repr);
+            for &byte in invalid {
+                let _ = write!(repr, r"\x{:02X}", byte);
+            }
+        }
+        repr.push('"');
+        Literal::_new(repr)
+    }
+
+    pub(crate) fn span(&self) -> Span {
+        self.span
+    }
+
+    pub(crate) fn set_span(&mut self, span: Span) {
+        self.span = span;
+    }
+
+    pub(crate) fn subspan<R: RangeBounds<usize>>(&self, range: R) -> Option<Span> {
+        #[cfg(not(span_locations))]
+        {
+            let _ = range;
+            None
+        }
+
+        #[cfg(span_locations)]
+        {
+            use core::ops::Bound;
+
+            let lo = match range.start_bound() {
+                Bound::Included(start) => {
+                    let start = u32::try_from(*start).ok()?;
+                    self.span.lo.checked_add(start)?
+                }
+                Bound::Excluded(start) => {
+                    let start = u32::try_from(*start).ok()?;
+                    self.span.lo.checked_add(start)?.checked_add(1)?
+                }
+                Bound::Unbounded => self.span.lo,
+            };
+            let hi = match range.end_bound() {
+                Bound::Included(end) => {
+                    let end = u32::try_from(*end).ok()?;
+                    self.span.lo.checked_add(end)?.checked_add(1)?
+                }
+                Bound::Excluded(end) => {
+                    let end = u32::try_from(*end).ok()?;
+                    self.span.lo.checked_add(end)?
+                }
+                Bound::Unbounded => self.span.hi,
+            };
+            if lo <= hi && hi <= self.span.hi {
+                Some(Span { lo, hi })
+            } else {
+                None
+            }
+        }
+    }
+}
+
+impl Display for Literal {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        Display::fmt(&self.repr, f)
+    }
+}
+
+impl Debug for Literal {
+    fn fmt(&self, fmt: &mut fmt::Formatter) -> fmt::Result {
+        let mut debug = fmt.debug_struct("Literal");
+        debug.field("lit", &format_args!("{}", self.repr));
+        debug_span_field_if_nontrivial(&mut debug, self.span);
+        debug.finish()
+    }
+}
+
+fn escape_utf8(string: &str, repr: &mut String) {
+    let mut chars = string.chars();
+    while let Some(ch) = chars.next() {
+        if ch == '\0' {
+            repr.push_str(
+                if chars
+                    .as_str()
+                    .starts_with(|next| '0' <= next && next <= '7')
+                {
+                    // circumvent clippy::octal_escapes lint
+                    r"\x00"
+                } else {
+                    r"\0"
+                },
+            );
+        } else if ch == '\'' {
+            // escape_debug turns this into "\'" which is unnecessary.
+            repr.push(ch);
+        } else {
+            repr.extend(ch.escape_debug());
+        }
+    }
+}
+
+#[cfg(feature = "proc-macro")]
+pub(crate) trait FromStr2: FromStr<Err = proc_macro::LexError> {
+    #[cfg(wrap_proc_macro)]
+    fn valid(src: &str) -> bool;
+
+    #[cfg(wrap_proc_macro)]
+    fn from_str_checked(src: &str) -> Result<Self, imp::LexError> {
+        // Validate using fallback parser, because rustc is incapable of
+        // returning a recoverable Err for certain invalid token streams, and
+        // will instead permanently poison the compilation.
+        if !Self::valid(src) {
+            return Err(imp::LexError::CompilerPanic);
+        }
+
+        // Catch panic to work around https://github.com/rust-lang/rust/issues/58736.
+        match panic::catch_unwind(|| Self::from_str(src)) {
+            Ok(Ok(ok)) => Ok(ok),
+            Ok(Err(lex)) => Err(imp::LexError::Compiler(lex)),
+            Err(_panic) => Err(imp::LexError::CompilerPanic),
+        }
+    }
+
+    fn from_str_unchecked(src: &str) -> Self {
+        Self::from_str(src).unwrap()
+    }
+}
+
+#[cfg(feature = "proc-macro")]
+impl FromStr2 for proc_macro::TokenStream {
+    #[cfg(wrap_proc_macro)]
+    fn valid(src: &str) -> bool {
+        TokenStream::from_str_checked(src).is_ok()
+    }
+}
+
+#[cfg(feature = "proc-macro")]
+impl FromStr2 for proc_macro::Literal {
+    #[cfg(wrap_proc_macro)]
+    fn valid(src: &str) -> bool {
+        Literal::from_str_checked(src).is_ok()
+    }
+}