diff --git a/src/abbrs.rs b/src/abbrs.rs
index 002e5eefd..f742b7bb0 100644
--- a/src/abbrs.rs
+++ b/src/abbrs.rs
@@ -149,7 +149,7 @@ pub struct Replacement {
 
 impl Replacement {
     /// Construct a replacement from a replacer.
-    /// The \p range is the range of the text matched by the replacer in the command line.
+    /// The `range` is the range of the text matched by the replacer in the command line.
     /// The text is passed in separately as it may be the output of the replacer's function.
     pub fn new(range: SourceRange, mut text: WString, set_cursor_marker: Option<WString>) -> Self {
         let mut cursor = None;
@@ -184,7 +184,7 @@ pub struct AbbreviationSet {
 
 impl AbbreviationSet {
     /// \return the list of replacers for an input token, in priority order.
-    /// The \p position is given to describe where the token was found.
+    /// The `position` is given to describe where the token was found.
     pub fn r#match(&self, token: &wstr, position: Position, cmd: &wstr) -> Vec<Replacer> {
         let mut result = vec![];
 
@@ -270,7 +270,7 @@ impl AbbreviationSet {
 }
 
 /// \return the list of replacers for an input token, in priority order, using the global set.
-/// The \p position is given to describe where the token was found.
+/// The `position` is given to describe where the token was found.
 pub fn abbrs_match(token: &wstr, position: Position, cmd: &wstr) -> Vec<Replacer> {
     with_abbrs(|set| set.r#match(token, position, cmd))
         .into_iter()
diff --git a/src/ast.rs b/src/ast.rs
index 2af3dd07c..cd36db16f 100644
--- a/src/ast.rs
+++ b/src/ast.rs
@@ -518,7 +518,7 @@ macro_rules! implement_leaf {
             }
         }
         impl $name {
-            /// Set the parent fields of all nodes in the tree rooted at \p self.
+            /// Set the parent fields of all nodes in the tree rooted at `self`.
             fn set_parents(&mut self) {}
         }
     };
@@ -672,7 +672,7 @@ macro_rules! define_list_node {
             }
         }
         impl $name {
-            /// Set the parent fields of all nodes in the tree rooted at \p self.
+            /// Set the parent fields of all nodes in the tree rooted at `self`.
             fn set_parents(&mut self) {
                 for i in 0..self.count() {
                     self[i].parent = Some(self);
@@ -771,7 +771,7 @@ macro_rules! implement_acceptor_for_branch {
             }
         }
         impl $name {
-            /// Set the parent fields of all nodes in the tree rooted at \p self.
+            /// Set the parent fields of all nodes in the tree rooted at `self`.
             fn set_parents(&mut self) {
                 $(
                     set_parent_of_field!(self, $field_name, $field_type);
@@ -2476,9 +2476,9 @@ impl Default for Ast {
 }
 
 impl Ast {
-    /// Construct an ast by parsing \p src as a job list.
-    /// The ast attempts to produce \p type as the result.
-    /// \p type may only be JobList or FreestandingArgumentList.
+    /// Construct an ast by parsing `src` as a job list.
+    /// The ast attempts to produce `type` as the result.
+    /// `type` may only be JobList or FreestandingArgumentList.
     pub fn parse(
         src: &wstr,
         flags: ParseTreeFlags,
@@ -2511,7 +2511,7 @@ impl Ast {
     }
 
     /// \return a textual representation of the tree.
-    /// Pass the original source as \p orig.
+    /// Pass the original source as `orig`.
     pub fn dump(&self, orig: &wstr) -> WString {
         let mut result = WString::new();
 
@@ -2749,7 +2749,7 @@ macro_rules! internal_error {
     };
 }
 
-/// Report an error based on \p fmt for the tokens' range
+/// Report an error based on `fmt` for the tokens' range
 macro_rules! parse_error {
     (
         $self:ident,
@@ -2764,7 +2764,7 @@ macro_rules! parse_error {
     }
 }
 
-/// Report an error based on \p fmt for the source range \p range.
+/// Report an error based on `fmt` for the source range `range`.
 macro_rules! parse_error_range {
     (
         $self:ident,
@@ -3134,7 +3134,7 @@ impl<'s> Populator<'s> {
         self.flags.contains(ParseTreeFlags::LEAVE_UNTERMINATED)
     }
 
-    /// \return whether a list type \p type allows arbitrary newlines in it.
+    /// \return whether a list type `type` allows arbitrary newlines in it.
     fn list_type_chomps_newlines(&self, typ: Type) -> bool {
         match typ {
             Type::argument_list => {
@@ -3187,7 +3187,7 @@ impl<'s> Populator<'s> {
         }
     }
 
-    /// \return whether a list type \p type allows arbitrary semicolons in it.
+    /// \return whether a list type `type` allows arbitrary semicolons in it.
     fn list_type_chomps_semis(&self, typ: Type) -> bool {
         match typ {
             Type::argument_list => {
@@ -3264,7 +3264,7 @@ impl<'s> Populator<'s> {
         typ == Type::job_list && self.flags.contains(ParseTreeFlags::CONTINUE_AFTER_ERROR)
     }
 
-    /// \return a reference to a non-comment token at index \p idx.
+    /// \return a reference to a non-comment token at index `idx`.
     fn peek_token(&mut self, idx: usize) -> &ParseToken {
         self.tokens.peek(idx)
     }
diff --git a/src/autoload.rs b/src/autoload.rs
index 9b84cf1dd..88d6c3088 100644
--- a/src/autoload.rs
+++ b/src/autoload.rs
@@ -38,7 +38,7 @@ pub struct Autoload {
 }
 
 impl Autoload {
-    /// Construct an autoloader that loads from the paths given by \p env_var_name.
+    /// Construct an autoloader that loads from the paths given by `env_var_name`.
     pub fn new(env_var_name: &'static wstr) -> Self {
         Self {
             env_var_name,
@@ -219,8 +219,8 @@ impl AutoloadFileCache {
         &self.dirs
     }
 
-    /// Check if a command \p cmd can be loaded.
-    /// If \p allow_stale is true, allow stale entries; otherwise discard them.
+    /// Check if a command `cmd` can be loaded.
+    /// If `allow_stale` is true, allow stale entries; otherwise discard them.
     /// This returns an autoloadable file, or none() if there is no such file.
     fn check(&mut self, cmd: &wstr, allow_stale: bool) -> Option<AutoloadableFile> {
         // Check hits.
diff --git a/src/bin/fish_indent.rs b/src/bin/fish_indent.rs
index a013595fa..db8978b1d 100644
--- a/src/bin/fish_indent.rs
+++ b/src/bin/fish_indent.rs
@@ -478,7 +478,7 @@ impl<'source, 'ast> PrettyPrinterState<'source, 'ast> {
         }
     }
 
-    /// \return whether a range \p r overlaps an error range from our ast.
+    /// \return whether a range `r` overlaps an error range from our ast.
     fn range_contained_error(&self, r: SourceRange) -> bool {
         let errs = self.errors.as_ref().unwrap();
         let range_is_before = |x: SourceRange, y: SourceRange| x.end().cmp(&y.start());
@@ -526,7 +526,7 @@ impl<'source, 'ast> PrettyPrinterState<'source, 'ast> {
         added_newline
     }
 
-    /// Given a string \p input, remove unnecessary quotes, etc.
+    /// Given a string `input`, remove unnecessary quotes, etc.
     fn clean_text(&self, input: &wstr) -> WString {
         // Unescape the string - this leaves special markers around if there are any
         // expansions or anything. We specifically tell it to not compute backslash-escapes
@@ -557,7 +557,7 @@ impl<'source, 'ast> PrettyPrinterState<'source, 'ast> {
     }
 
     // Emit a range of original text. This indents as needed, and also inserts preceding gap text.
-    // If \p tolerate_line_splitting is set, then permit escaped newlines; otherwise collapse such
+    // If `tolerate_line_splitting` is set, then permit escaped newlines; otherwise collapse such
     // lines.
     fn emit_text(&mut self, r: SourceRange, flags: GapFlags) {
         self.emit_gap_text_before(r, flags);
diff --git a/src/builtins/printf.rs b/src/builtins/printf.rs
index 037889bca..59a1afb69 100644
--- a/src/builtins/printf.rs
+++ b/src/builtins/printf.rs
@@ -60,17 +60,17 @@ use crate::wutil::{
 use printf_compat::args::ToArg;
 use printf_compat::printf::sprintf_locale;
 
-/// \return true if \p c is an octal digit.
+/// \return true if `c` is an octal digit.
 fn is_octal_digit(c: char) -> bool {
     ('0'..='7').contains(&c)
 }
 
-/// \return true if \p c is a decimal digit.
+/// \return true if `c` is a decimal digit.
 fn iswdigit(c: char) -> bool {
     c.is_ascii_digit()
 }
 
-/// \return true if \p c is a hexadecimal digit.
+/// \return true if `c` is a hexadecimal digit.
 fn iswxdigit(c: char) -> bool {
     c.is_ascii_hexdigit()
 }
@@ -96,7 +96,7 @@ struct builtin_printf_state_t<'a, 'b> {
 }
 
 /// Convert to a scalar type. \return the result of conversion, and the end of the converted string.
-/// On conversion failure, \p end is not modified.
+/// On conversion failure, `end` is not modified.
 trait RawStringToScalarType: Copy + std::convert::From<u32> {
     /// Convert from a string to our self type.
     /// \return the result of conversion, and the remainder of the string.
diff --git a/src/builtins/set.rs b/src/builtins/set.rs
index ba9e50d03..5ea9c389c 100644
--- a/src/builtins/set.rs
+++ b/src/builtins/set.rs
@@ -387,7 +387,7 @@ impl<'a> SplitVar<'a> {
 }
 
 /// Extract indexes from an argument of the form `var_name[index1 index2...]`.
-/// The argument \p arg is split into a variable name and list of indexes, which is returned by
+/// The argument `arg` is split into a variable name and list of indexes, which is returned by
 /// reference. Indexes are "expanded" in the sense that range expressions .. and negative values are
 /// handled.
 ///
@@ -809,7 +809,7 @@ fn env_result_to_status(retval: EnvStackSetResult) -> Option<c_int> {
     })
 }
 
-/// Return a list of new values for the variable \p varname, respecting the \p opts.
+/// Return a list of new values for the variable `varname`, respecting the `opts`.
 /// This handles the simple case where there are no indexes.
 fn new_var_values(
     varname: &wstr,
@@ -859,7 +859,7 @@ fn new_var_values_by_index(split: &SplitVar, argv: &[&wstr]) -> Vec<WString> {
         result = var.as_list().to_owned();
     }
 
-    // For each (index, argument) pair, set the element in our \p result to the replacement string.
+    // For each (index, argument) pair, set the element in our `result` to the replacement string.
     // Extend the list with empty strings as needed. The indexes are 1-based.
     for (i, arg) in argv.iter().copied().enumerate() {
         let lidx = usize::try_from(split.indexes[i]).unwrap();
diff --git a/src/builtins/wait.rs b/src/builtins/wait.rs
index 1859e14ed..81e41e449 100644
--- a/src/builtins/wait.rs
+++ b/src/builtins/wait.rs
@@ -30,7 +30,7 @@ enum WaitHandleQuery<'a> {
 }
 
 /// Walk the list of jobs, looking for a process with the given pid or proc name.
-/// Append all matching wait handles to \p handles.
+/// Append all matching wait handles to `handles`.
 /// \return true if we found a matching job (even if not waitable), false if not.
 fn find_wait_handles(
     query: WaitHandleQuery<'_>,
@@ -92,7 +92,7 @@ fn is_completed(wh: &WaitHandleRef) -> bool {
 }
 
 /// Wait for the given wait handles to be marked as completed.
-/// If \p any_flag is set, wait for the first one; otherwise wait for all.
+/// If `any_flag` is set, wait for the first one; otherwise wait for all.
 /// \return a status code.
 fn wait_for_completion(parser: &Parser, whs: &[WaitHandleRef], any_flag: bool) -> Option<c_int> {
     if whs.is_empty() {
diff --git a/src/common.rs b/src/common.rs
index 988b1d54f..78d9b040b 100644
--- a/src/common.rs
+++ b/src/common.rs
@@ -1249,7 +1249,7 @@ pub fn wcs2zstring(input: &wstr) -> CString {
     CString::new(until_nul).unwrap()
 }
 
-/// Like wcs2string, but appends to \p receiver instead of returning a new string.
+/// Like wcs2string, but appends to `receiver` instead of returning a new string.
 pub fn wcs2string_appending(output: &mut Vec<u8>, input: &wstr) {
     output.reserve(input.len());
     wcs2string_callback(input, |buff| {
@@ -1258,7 +1258,7 @@ pub fn wcs2string_appending(output: &mut Vec<u8>, input: &wstr) {
     });
 }
 
-/// \return the count of initial characters in \p in which are ASCII.
+/// \return the count of initial characters in `in` which are ASCII.
 fn count_ascii_prefix(inp: &[u8]) -> usize {
     // The C++ version had manual vectorization.
     inp.iter().take_while(|c| c.is_ascii()).count()
@@ -1553,7 +1553,7 @@ pub fn read_loop<Fd: AsRawFd>(fd: &Fd, buf: &mut [u8]) -> std::io::Result<usize>
     }
 }
 
-/// Write the given paragraph of output, redoing linebreaks to fit \p termsize.
+/// Write the given paragraph of output, redoing linebreaks to fit `termsize`.
 pub fn reformat_for_screen(msg: &wstr, termsize: &Termsize) -> WString {
     let mut buff = WString::new();
 
diff --git a/src/complete.rs b/src/complete.rs
index d8f604cff..14af2da4d 100644
--- a/src/complete.rs
+++ b/src/complete.rs
@@ -1968,11 +1968,11 @@ impl<'ctx> Completer<'ctx> {
         );
     }
 
-    // Invoke command-specific completions given by \p arg_data.
+    // Invoke command-specific completions given by `arg_data`.
     // Then, for each target wrapped by the given command, update the command
     // line with that target and invoke this recursively.
-    // The command whose completions to use is given by \p cmd. The full command line is given by \p
-    // cmdline and the command's range in it is given by \p cmdrange. Note: the command range
+    // The command whose completions to use is given by `cmd`. The full command line is given by \p
+    // cmdline and the command's range in it is given by `cmdrange`. Note: the command range
     // may have a different length than the command itself, because the command is unescaped (i.e.
     // quotes removed).
     fn walk_wrap_chain(
diff --git a/src/env/environment.rs b/src/env/environment.rs
index 98c8fd3c0..97c84a8e1 100644
--- a/src/env/environment.rs
+++ b/src/env/environment.rs
@@ -327,7 +327,7 @@ impl EnvStack {
     }
 
     /// Synchronizes universal variable changes.
-    /// If \p always is set, perform synchronization even if there's no pending changes from this
+    /// If `always` is set, perform synchronization even if there's no pending changes from this
     /// instance (that is, look for changes from other fish instances).
     /// \return a list of events for changed variables.
     #[allow(clippy::vec_box)]
diff --git a/src/env/environment_impl.rs b/src/env/environment_impl.rs
index f5acfb38b..dd6315faa 100644
--- a/src/env/environment_impl.rs
+++ b/src/env/environment_impl.rs
@@ -480,7 +480,7 @@ impl EnvScopedImpl {
         let query = Query::new(flags);
         let mut names: HashSet<WString> = HashSet::new();
 
-        // Helper to add the names of variables from \p envs to names, respecting show_exported and
+        // Helper to add the names of variables from `envs` to names, respecting show_exported and
         // show_unexported.
         let add_keys = |envs: &VarTable, names: &mut HashSet<WString>| {
             for (key, val) in envs.iter() {
@@ -707,7 +707,7 @@ impl EnvStackImpl {
         })
     }
 
-    /// Set a variable under the name \p key, using the given \p mode, setting its value to \p val.
+    /// Set a variable under the name `key`, using the given `mode`, setting its value to `val`.
     pub fn set(&mut self, key: &wstr, mode: EnvMode, mut val: Vec<WString>) -> ModResult {
         let query = Query::new(mode);
         // Handle electric and read-only variables.
@@ -788,7 +788,7 @@ impl EnvStackImpl {
         result
     }
 
-    /// Remove a variable under the name \p key.
+    /// Remove a variable under the name `key`.
     pub fn remove(&mut self, key: &wstr, mode: EnvMode) -> ModResult {
         let query = Query::new(mode);
         // Users can't remove read-only keys.
@@ -891,7 +891,7 @@ impl EnvStackImpl {
         var_names
     }
 
-    /// Find the first node in the chain starting at \p node which contains the given key \p key.
+    /// Find the first node in the chain starting at `node` which contains the given key `key`.
     fn find_in_chain(node: &EnvNodeRef, key: &wstr) -> Option<EnvNodeRef> {
         #[allow(clippy::manual_find)]
         for cur in node.iter() {
@@ -902,7 +902,7 @@ impl EnvStackImpl {
         None
     }
 
-    /// Remove a variable from the chain \p node.
+    /// Remove a variable from the chain `node`.
     /// Return true if the variable was found and removed.
     fn remove_from_chain(node: &mut EnvNodeRef, key: &wstr) -> bool {
         for cur in node.iter() {
@@ -917,9 +917,9 @@ impl EnvStackImpl {
         false
     }
 
-    /// Try setting \p key as an electric or readonly variable, whose value is provided by reference in \p val.
+    /// Try setting `key` as an electric or readonly variable, whose value is provided by reference in `val`.
     /// Return an error code, or NOne if not an electric or readonly variable.
-    /// \p val will not be modified upon a None return.
+    /// `val` will not be modified upon a None return.
     fn try_set_electric(
         &mut self,
         key: &wstr,
@@ -1015,7 +1015,7 @@ impl EnvStackImpl {
         locked_uvars.set(key, new_var);
     }
 
-    /// Set a variable in a given node \p node.
+    /// Set a variable in a given node `node`.
     fn set_in_node(node: &mut EnvNodeRef, key: &wstr, mut val: Vec<WString>, flags: VarFlags) {
         // Read the var from the node. In C++ this was node->env[key] which establishes a default.
         let mut node_ref = node.borrow_mut();
diff --git a/src/env_universal_common.rs b/src/env_universal_common.rs
index 497956de3..5ed86a660 100644
--- a/src/env_universal_common.rs
+++ b/src/env_universal_common.rs
@@ -250,7 +250,7 @@ impl EnvUniversal {
         }
     }
 
-    /// Populate a variable table \p out_vars from a \p s string.
+    /// Populate a variable table `out_vars` from a `s` string.
     /// This is exposed for testing only.
     /// \return the format of the file that we read.
     pub fn populate_variables(s: &[u8], out_vars: &mut VarTable) -> UvarFormat {
@@ -290,7 +290,7 @@ impl EnvUniversal {
     }
 
     /// Guess a file format. Exposed for testing only.
-    /// \return the format corresponding to file contents \p s.
+    /// \return the format corresponding to file contents `s`.
     pub fn format_for_contents(s: &[u8]) -> UvarFormat {
         // Walk over leading comments, looking for one like '# version'
         let iter = LineIterator::new(s);
diff --git a/src/exec.rs b/src/exec.rs
index 0e70fe224..87e4aff17 100644
--- a/src/exec.rs
+++ b/src/exec.rs
@@ -61,7 +61,7 @@ use std::slice;
 use std::sync::atomic::Ordering;
 use std::sync::{atomic::AtomicUsize, Arc};
 
-/// Execute the processes specified by \p j in the parser \p.
+/// Execute the processes specified by `j` in the parser \p.
 /// On a true return, the job was successfully launched and the parser will take responsibility for
 /// cleaning it up. On a false return, the job could not be launched and the caller must clean it
 /// up.
@@ -268,7 +268,7 @@ pub fn exec_subshell(
 
 /// Like exec_subshell, but only returns expansion-breaking errors. That is, a zero return means
 /// "success" (even though the command may have failed), a non-zero return means that we should
-/// halt expansion. If the \p pgid is supplied, then any spawned external commands should join that
+/// halt expansion. If the `pgid` is supplied, then any spawned external commands should join that
 /// pgroup.
 pub fn exec_subshell_for_expand(
     cmd: &wstr,
@@ -303,7 +303,7 @@ static FORK_COUNT: AtomicUsize = AtomicUsize::new(0);
 /// etc.
 type LaunchResult = Result<(), ()>;
 
-/// Given an error \p err returned from either posix_spawn or exec, \return a process exit code.
+/// Given an error `err` returned from either posix_spawn or exec, \return a process exit code.
 fn exit_code_from_exec_error(err: libc::c_int) -> libc::c_int {
     assert!(err != 0, "Zero is success, not an error");
     match err {
@@ -512,8 +512,8 @@ fn internal_exec(vars: &EnvStack, j: &Job, block_io: IoChain) {
     }
 }
 
-/// Construct an internal process for the process p. In the background, write the data \p outdata to
-/// stdout and \p errdata to stderr, respecting the io chain \p ios. For example if target_fd is 1
+/// Construct an internal process for the process p. In the background, write the data `outdata` to
+/// stdout and `errdata` to stderr, respecting the io chain `ios`. For example if target_fd is 1
 /// (stdout), and there is a dup2 3->1, then we need to write to fd 3. Then exit the internal
 /// process.
 fn run_internal_process(p: &Process, outdata: Vec<u8>, errdata: Vec<u8>, ios: &IoChain) {
@@ -625,7 +625,7 @@ fn run_internal_process(p: &Process, outdata: Vec<u8>, errdata: Vec<u8>, ios: &I
     });
 }
 
-/// If \p outdata or \p errdata are both empty, then mark the process as completed immediately.
+/// If `outdata` or `errdata` are both empty, then mark the process as completed immediately.
 /// Otherwise, run an internal process.
 fn run_internal_process_or_short_circuit(
     parser: &Parser,
@@ -662,7 +662,7 @@ fn run_internal_process_or_short_circuit(
     }
 }
 
-/// Call fork() as part of executing a process \p p in a job \j. Execute \p child_action in the
+/// Call fork() as part of executing a process `p` in a job \j. Execute `child_action` in the
 /// context of the child.
 fn fork_child_for_process(
     job: &Job,
@@ -746,7 +746,7 @@ fn fork_child_for_process(
 
 /// \return an newly allocated output stream for the given fd, which is typically stdout or stderr.
 /// This inspects the io_chain and decides what sort of output stream to return.
-/// If \p piped_output_needs_buffering is set, and if the output is going to a pipe, then the other
+/// If `piped_output_needs_buffering` is set, and if the output is going to a pipe, then the other
 /// end then synchronously writing to the pipe risks deadlock, so we must buffer it.
 fn create_output_stream_for_builtin(
     fd: RawFd,
@@ -949,7 +949,7 @@ fn function_restore_environment(parser: &Parser, block: BlockId) {
 }
 
 // The "performer" function of a block or function process.
-// This accepts a place to execute as \p parser and then executes the result, returning a status.
+// This accepts a place to execute as `parser` and then executes the result, returning a status.
 // This is factored out in this funny way in preparation for concurrent execution.
 type ProcPerformer = dyn FnOnce(
     &Parser,
@@ -1030,7 +1030,7 @@ fn get_performer_for_process(
 }
 
 /// Execute a block node or function "process".
-/// \p piped_output_needs_buffering if true, buffer the output.
+/// `piped_output_needs_buffering` if true, buffer the output.
 fn exec_block_or_func_process(
     parser: &Parser,
     j: &Job,
@@ -1184,10 +1184,10 @@ struct PartialPipes {
     write: Option<OwnedFd>,
 }
 
-/// Executes a process \p \p in \p job, using the pipes \p pipes (which may have invalid fds if this
+/// Executes a process \p `in` `job`, using the pipes `pipes` (which may have invalid fds if this
 /// is the first or last process).
-/// \p deferred_pipes represents the pipes from our deferred process; if set ensure they get closed
-/// in any child. If \p is_deferred_run is true, then this is a deferred run; this affects how
+/// `deferred_pipes` represents the pipes from our deferred process; if set ensure they get closed
+/// in any child. If `is_deferred_run` is true, then this is a deferred run; this affects how
 /// certain buffering works.
 /// An error return here indicates that the process failed to launch, and the rest of
 /// the pipeline should be cancelled.
@@ -1356,7 +1356,7 @@ fn get_deferred_process(j: &Job) -> Option<usize> {
     None
 }
 
-/// Given that we failed to execute process \p failed_proc in job \p job, mark that process and
+/// Given that we failed to execute process `failed_proc` in job `job`, mark that process and
 /// every subsequent process in the pipeline as aborted before launch.
 fn abort_pipeline_from(job: &Job, offset: usize) {
     for p in job.processes().iter().skip(offset) {
@@ -1392,7 +1392,7 @@ fn allow_exec_with_background_jobs(parser: &Parser) -> bool {
     }
 }
 
-/// Populate \p lst with the output of \p buffer, perhaps splitting lines according to \p split.
+/// Populate `lst` with the output of `buffer`, perhaps splitting lines according to `split`.
 fn populate_subshell_output(lst: &mut Vec<WString>, buffer: &SeparatedBuffer, split: bool) {
     // Walk over all the elements.
     for elem in buffer.elements() {
@@ -1430,12 +1430,12 @@ fn populate_subshell_output(lst: &mut Vec<WString>, buffer: &SeparatedBuffer, sp
     }
 }
 
-/// Execute \p cmd in a subshell in \p parser. If \p lst is not null, populate it with the output.
-/// Return $status in \p out_status.
-/// If \p job_group is set, any spawned commands should join that job group.
-/// If \p apply_exit_status is false, then reset $status back to its original value.
-/// \p is_subcmd controls whether we apply a read limit.
-/// \p break_expand is used to propagate whether the result should be "expansion breaking" in the
+/// Execute `cmd` in a subshell in `parser`. If `lst` is not null, populate it with the output.
+/// Return $status in `out_status`.
+/// If `job_group` is set, any spawned commands should join that job group.
+/// If `apply_exit_status` is false, then reset $status back to its original value.
+/// `is_subcmd` controls whether we apply a read limit.
+/// `break_expand` is used to propagate whether the result should be "expansion breaking" in the
 /// sense that subshells used during string expansion should halt that expansion. \return the value
 /// of $status.
 fn exec_subshell_internal(
diff --git a/src/expand.rs b/src/expand.rs
index 8ebbd03ee..35c2eb832 100644
--- a/src/expand.rs
+++ b/src/expand.rs
@@ -207,7 +207,7 @@ pub fn expand_one(
 /// If the expansion resulted in no or an empty command, the command will be an empty string. Note
 /// that API does not distinguish between expansion resulting in an empty command (''), and
 /// expansion resulting in no command (e.g. unset variable).
-/// If \p skip_wildcards is true, then do not do wildcard expansion
+/// If `skip_wildcards` is true, then do not do wildcard expansion
 /// \return an expand error.
 pub fn expand_to_command_and_args(
     instr: &wstr,
@@ -792,7 +792,7 @@ fn expand_variables(
     ExpandResult::ok()
 }
 
-/// Perform brace expansion, placing the expanded strings into \p out.
+/// Perform brace expansion, placing the expanded strings into `out`.
 fn expand_braces(
     input: WString,
     flags: ExpandFlags,
@@ -916,8 +916,8 @@ fn expand_braces(
     ExpandResult::ok()
 }
 
-/// Expand a command substitution \p input, executing on \p ctx, and inserting the results into
-/// \p out_list, or any errors into \p errors. \return an expand result.
+/// Expand a command substitution `input`, executing on `ctx`, and inserting the results into
+/// `out_list`, or any errors into `errors`. \return an expand result.
 pub fn expand_cmdsubst(
     input: WString,
     ctx: &OperationContext,
diff --git a/src/fd_monitor.rs b/src/fd_monitor.rs
index 2bfac02d2..b4f310cad 100644
--- a/src/fd_monitor.rs
+++ b/src/fd_monitor.rs
@@ -138,7 +138,7 @@ impl FdEventSignaller {
     }
 
     /// Perform a poll to see if an event is received.
-    /// If \p wait is set, wait until it is readable; this does not consume the event
+    /// If `wait` is set, wait until it is readable; this does not consume the event
     /// but guarantees that the next call to wait() will not block.
     /// \return true if readable, false if not readable, or not interrupted by a signal.
     pub fn poll(&self, wait: bool /* = false */) -> bool {
diff --git a/src/fds.rs b/src/fds.rs
index a9737998b..6833f551e 100644
--- a/src/fds.rs
+++ b/src/fds.rs
@@ -176,7 +176,7 @@ pub fn make_autoclose_pipes() -> nix::Result<AutoClosePipes> {
 
 /// If the given fd is in the "user range", move it to a new fd in the "high range".
 /// zsh calls this movefd().
-/// \p input_has_cloexec describes whether the input has CLOEXEC already set, so we can avoid
+/// `input_has_cloexec` describes whether the input has CLOEXEC already set, so we can avoid
 /// setting it again.
 /// \return the fd, which always has CLOEXEC set; or an invalid fd on failure, in
 /// which case an error will have been printed, and the input fd closed.
@@ -202,7 +202,7 @@ fn heightenize_fd(fd: OwnedFd, input_has_cloexec: bool) -> nix::Result<OwnedFd>
     Ok(unsafe { OwnedFd::from_raw_fd(newfd) })
 }
 
-/// Sets CLO_EXEC on a given fd according to the value of \p should_set.
+/// Sets CLO_EXEC on a given fd according to the value of `should_set`.
 pub fn set_cloexec(fd: RawFd, should_set: bool /* = true */) -> c_int {
     // Note we don't want to overwrite existing flags like O_NONBLOCK which may be set. So fetch the
     // existing flags and modify them.
@@ -270,7 +270,7 @@ pub fn open_dir(path: &CStr, mode: nix::sys::stat::Mode) -> nix::Result<OwnedFd>
     open_cloexec(path, OFlag::O_RDONLY | OFlag::O_DIRECTORY, mode).map(OwnedFd::from)
 }
 
-/// Close a file descriptor \p fd, retrying on EINTR.
+/// Close a file descriptor `fd`, retrying on EINTR.
 pub fn exec_close(fd: RawFd) {
     assert!(fd >= 0, "Invalid fd");
     while unsafe { libc::close(fd) } == -1 {
diff --git a/src/flog.rs b/src/flog.rs
index fed77dddd..d223ee964 100644
--- a/src/flog.rs
+++ b/src/flog.rs
@@ -228,7 +228,7 @@ fn apply_one_wildcard(wc_esc: &wstr, sense: bool) {
     }
 }
 
-/// Set the active flog categories according to the given wildcard \p wc.
+/// Set the active flog categories according to the given wildcard `wc`.
 pub fn activate_flog_categories_by_pattern(wc_ptr: &wstr) {
     let mut wc: WString = wc_ptr.into();
     // Normalize underscores to dashes, allowing the user to be sloppy.
diff --git a/src/function.rs b/src/function.rs
index e89cba16e..09e164a2a 100644
--- a/src/function.rs
+++ b/src/function.rs
@@ -177,7 +177,7 @@ fn autoload_names(names: &mut HashSet<WString>, get_hidden: bool) {
     }
 }
 
-/// Add a function. This may mutate \p props to set is_autoload.
+/// Add a function. This may mutate `props` to set is_autoload.
 pub fn add(name: WString, props: Arc<FunctionProperties>) {
     let mut funcset = FUNCTION_SET.lock().unwrap();
 
@@ -221,7 +221,7 @@ pub fn get_props_autoload(name: &wstr, parser: &Parser) -> Option<Arc<FunctionPr
     get_props(name)
 }
 
-/// Returns true if the function named \p cmd exists.
+/// Returns true if the function named `cmd` exists.
 /// This may autoload.
 pub fn exists(cmd: &wstr, parser: &Parser) -> bool {
     parser.assert_can_execute();
@@ -231,7 +231,7 @@ pub fn exists(cmd: &wstr, parser: &Parser) -> bool {
     get_props_autoload(cmd, parser).is_some()
 }
 
-/// Returns true if the function \p cmd either is loaded, or exists on disk in an autoload
+/// Returns true if the function `cmd` either is loaded, or exists on disk in an autoload
 /// directory.
 pub fn exists_no_autoload(cmd: &wstr) -> bool {
     if !valid_func_name(cmd) {
diff --git a/src/highlight.rs b/src/highlight.rs
index 9f7609894..f2e331b11 100644
--- a/src/highlight.rs
+++ b/src/highlight.rs
@@ -316,7 +316,7 @@ fn autosuggest_parse_command(
     None
 }
 
-/// Given an item \p item from the history which is a proposed autosuggestion, return whether the
+/// Given an item `item` from the history which is a proposed autosuggestion, return whether the
 /// autosuggestion is valid. It may not be valid if e.g. it is attempting to cd into a directory
 /// which does not exist.
 pub fn autosuggest_validate_from_history(
diff --git a/src/input.rs b/src/input.rs
index bd805e513..86b9401c5 100644
--- a/src/input.rs
+++ b/src/input.rs
@@ -638,7 +638,7 @@ impl EventQueuePeeker<'_> {
     }
 
     /// Check if the next event is the given character. This advances the index on success only.
-    /// If \p escaped is set, then return false if this (or any other) character had a timeout.
+    /// If `escaped` is set, then return false if this (or any other) character had a timeout.
     fn next_is_char(&mut self, style: &KeyNameStyle, key: Key, escaped: bool) -> bool {
         assert!(
             self.idx <= self.peeked.len(),
@@ -773,7 +773,7 @@ impl Drop for EventQueuePeeker<'_> {
     }
 }
 
-/// \return true if a given \p peeker matches a given sequence of char events given by \p str.
+/// \return true if a given `peeker` matches a given sequence of char events given by `str`.
 fn try_peek_sequence(peeker: &mut EventQueuePeeker, style: &KeyNameStyle, seq: &[Key]) -> bool {
     assert!(
         !seq.is_empty(),
diff --git a/src/io.rs b/src/io.rs
index 316225189..7214e0498 100644
--- a/src/io.rs
+++ b/src/io.rs
@@ -115,7 +115,7 @@ impl SeparatedBuffer {
         &self.elements
     }
 
-    /// Append the given data with separation type \p sep.
+    /// Append the given data with separation type `sep`.
     pub fn append(&mut self, data: &[u8], sep: SeparationType) -> bool {
         if !self.try_add_size(data.len()) {
             return false;
@@ -145,7 +145,7 @@ impl SeparatedBuffer {
         !self.elements.is_empty() && !self.elements.last().unwrap().is_explicitly_separated()
     }
 
-    /// Mark that we are about to add the given size \p delta to the buffer. \return true if we
+    /// Mark that we are about to add the given size `delta` to the buffer. \return true if we
     /// succeed, false if we exceed buffer_limit.
     fn try_add_size(&mut self, delta: usize) -> bool {
         if self.discard {
diff --git a/src/output.rs b/src/output.rs
index e3a01a6bc..a59281937 100644
--- a/src/output.rs
+++ b/src/output.rs
@@ -385,7 +385,7 @@ impl Outputter {
         &self.contents
     }
 
-    /// Output any buffered data to the given \p fd.
+    /// Output any buffered data to the given `fd`.
     fn flush_to(&mut self, fd: RawFd) {
         if fd >= 0 && !self.contents.is_empty() {
             let _ = common::write_loop(&fd, &self.contents);
diff --git a/src/parse_execution.rs b/src/parse_execution.rs
index 0964e6d97..969508cf9 100644
--- a/src/parse_execution.rs
+++ b/src/parse_execution.rs
@@ -113,7 +113,7 @@ impl ParseExecutionContext {
     }
 }
 
-// Report an error, setting $status to \p status. Always returns
+// Report an error, setting $status to `status`. Always returns
 // 'end_execution_reason_t::error'.
 macro_rules! report_error {
     ( $self:ident, $ctx:expr, $status:expr, $node:expr, $fmt:expr $(, $arg:expr )* $(,)? ) => {
diff --git a/src/parse_util.rs b/src/parse_util.rs
index ae182203b..d6d62f031 100644
--- a/src/parse_util.rs
+++ b/src/parse_util.rs
@@ -31,7 +31,7 @@ use std::ops::Range;
 use std::{iter, ops};
 
 /// Handles slices: the square brackets in an expression like $foo[5..4]
-/// \return the length of the slice starting at \p in, or 0 if there is no slice, or None on error.
+/// \return the length of the slice starting at `in`, or 0 if there is no slice, or None on error.
 /// This never accepts incomplete slices.
 pub fn parse_util_slice_length(input: &wstr) -> Option<usize> {
     const openc: char = '[';
@@ -1566,8 +1566,8 @@ fn detect_errors_in_backgrounded_job(
     errored
 }
 
-/// Given a source buffer \p buff_src and decorated statement \p dst within it, return true if there
-/// is an error and false if not. \p storage may be used to reduce allocations.
+/// Given a source buffer `buff_src` and decorated statement `dst` within it, return true if there
+/// is an error and false if not. `storage` may be used to reduce allocations.
 fn detect_errors_in_decorated_statement(
     buff_src: &wstr,
     dst: &ast::DecoratedStatement,
diff --git a/src/parser.rs b/src/parser.rs
index 6ec0cf2b1..d6a43e061 100644
--- a/src/parser.rs
+++ b/src/parser.rs
@@ -807,7 +807,7 @@ impl Parser {
     }
 
     /// Update any universal variables and send event handlers.
-    /// If \p always is set, then do it even if we have no pending changes (that is, look for
+    /// If `always` is set, then do it even if we have no pending changes (that is, look for
     /// changes from other fish instances); otherwise only sync if this instance has changed uvars.
     pub fn sync_uvars_and_fire(&self, always: bool) {
         if self.syncs_uvars.load() {
@@ -1128,7 +1128,7 @@ fn print_profile(items: &[ProfileItem], out: RawFd) {
     }
 }
 
-/// Append stack trace info for the block \p b to \p trace.
+/// Append stack trace info for the block `b` to `trace`.
 fn append_block_description_to_stack_trace(parser: &Parser, b: &Block, trace: &mut WString) {
     let mut print_call_site = false;
     match b.typ() {
diff --git a/src/path.rs b/src/path.rs
index 09e3337b3..bad63a6fd 100644
--- a/src/path.rs
+++ b/src/path.rs
@@ -186,7 +186,7 @@ fn maybe_issue_path_warning(
     printf!("\n");
 }
 
-/// Finds the path of an executable named \p cmd, by looking in $PATH taken from \p vars.
+/// Finds the path of an executable named `cmd`, by looking in $PATH taken from `vars`.
 /// \returns the path if found, none if not.
 pub fn path_get_path(cmd: &wstr, vars: &dyn Environment) -> Option<WString> {
     let result = path_try_get_path(cmd, vars);
@@ -206,7 +206,7 @@ pub static DEFAULT_PATH: Lazy<[WString; 3]> = Lazy::new(|| {
     ]
 });
 
-/// Finds the path of an executable named \p cmd, by looking in $PATH taken from \p vars.
+/// Finds the path of an executable named `cmd`, by looking in $PATH taken from `vars`.
 /// On success, err will be 0 and the path is returned.
 /// On failure, we return the "best path" with err set appropriately.
 /// For example, if we find a non-executable file, we will return its path and EACCESS.
@@ -591,8 +591,8 @@ impl BaseDirectory {
     }
 }
 
-/// Attempt to get a base directory, creating it if necessary. If a variable named \p xdg_var is
-/// set, use that directory; otherwise use the path \p non_xdg_homepath rooted in $HOME. \return the
+/// Attempt to get a base directory, creating it if necessary. If a variable named `xdg_var` is
+/// set, use that directory; otherwise use the path `non_xdg_homepath` rooted in $HOME. \return the
 /// result; see the base_directory_t fields.
 #[cfg_attr(test, allow(unused_variables), allow(unreachable_code))]
 fn make_base_directory(xdg_var: &wstr, non_xdg_homepath: &wstr) -> BaseDirectory {
diff --git a/src/proc.rs b/src/proc.rs
index 0bb63eadd..aa3b88615 100644
--- a/src/proc.rs
+++ b/src/proc.rs
@@ -153,7 +153,7 @@ impl ProcStatus {
         (status, empty)
     }
 
-    /// Encode a return value \p ret and signal \p sig into a status value like waitpid() does.
+    /// Encode a return value `ret` and signal `sig` into a status value like waitpid() does.
     const fn w_exitcode(ret: i32, sig: i32) -> i32 {
         #[cfg(HAVE_WAITSTATUS_SIGNAL_RET)]
         // It's encoded signal and then status
@@ -1145,7 +1145,7 @@ pub fn set_job_control_mode(mode: JobControl) {
 static JOB_CONTROL_MODE: AtomicU8 = AtomicU8::new(JobControl::interactive as u8);
 
 /// Notify the user about stopped or terminated jobs, and delete completed jobs from the job list.
-/// If \p interactive is set, allow removing interactive jobs; otherwise skip them.
+/// If `interactive` is set, allow removing interactive jobs; otherwise skip them.
 /// \return whether text was printed to stdout.
 pub fn job_reap(parser: &Parser, allow_interactive: bool) -> bool {
     parser.assert_can_execute();
@@ -1243,7 +1243,7 @@ pub fn proc_init() {
     signal_set_handlers_once(false);
 }
 
-/// Set the status of \p proc to \p status.
+/// Set the status of `proc` to `status`.
 fn handle_child_status(job: &Job, proc: &Process, status: &ProcStatus) {
     proc.status.update(status);
     if status.stopped() {
@@ -1288,7 +1288,7 @@ pub fn proc_wait_any(parser: &Parser) {
     process_clean_after_marking(parser, is_interactive);
 }
 
-/// Send SIGHUP to the list \p jobs, excepting those which are in fish's pgroup.
+/// Send SIGHUP to the list `jobs`, excepting those which are in fish's pgroup.
 pub fn hup_jobs(jobs: &JobList) {
     let fish_pgrp = unsafe { libc::getpgrp() };
     for j in jobs {
@@ -1476,7 +1476,7 @@ fn process_mark_finished_children(parser: &Parser, block_ok: bool) {
     reap_disowned_pids();
 }
 
-/// Generate process_exit events for any completed processes in \p j.
+/// Generate process_exit events for any completed processes in `j`.
 fn generate_process_exit_events(j: &Job, out_evts: &mut Vec<Event>) {
     // Historically we have avoided generating events for foreground jobs from event handlers, as an
     // event handler may itself produce a new event.
diff --git a/src/reader.rs b/src/reader.rs
index 30acab14f..c65d38b20 100644
--- a/src/reader.rs
+++ b/src/reader.rs
@@ -240,8 +240,8 @@ fn reader_push_ret(
     data
 }
 
-/// Push a new reader environment controlled by \p conf, using the given history name.
-/// If \p history_name is empty, then save history in-memory only; do not write it to disk.
+/// Push a new reader environment controlled by `conf`, using the given history name.
+/// If `history_name` is empty, then save history in-memory only; do not write it to disk.
 pub fn reader_push(parser: &Parser, history_name: &wstr, conf: ReaderConfig) {
     // Public variant which discards the return value.
     reader_push_ret(parser, history_name, conf);
@@ -1342,7 +1342,7 @@ pub fn combine_command_and_autosuggestion(cmdline: &wstr, autosuggestion: &wstr)
 }
 
 impl ReaderData {
-    /// \return true if the command line has changed and repainting is needed. If \p colors is not
+    /// \return true if the command line has changed and repainting is needed. If `colors` is not
     /// null, then also return true if the colors have changed.
     fn is_repaint_needed(&self, mcolors: Option<&[HighlightSpec]>) -> bool {
         // Note: this function is responsible for detecting all of the ways that the command line may
@@ -1394,7 +1394,7 @@ impl ReaderData {
     }
 
     /// Generate a new layout data from the current state of the world.
-    /// If \p mcolors has a value, then apply it; otherwise extend existing colors.
+    /// If `mcolors` has a value, then apply it; otherwise extend existing colors.
     fn make_layout_data(&self) -> LayoutData {
         let mut result = LayoutData::default();
         let focused_on_pager = self.active_edit_line_tag() == EditableLineTag::SearchField;
@@ -1417,14 +1417,14 @@ impl ReaderData {
     }
 
     /// Generate a new layout data from the current state of the world, and paint with it.
-    /// If \p mcolors has a value, then apply it; otherwise extend existing colors.
+    /// If `mcolors` has a value, then apply it; otherwise extend existing colors.
     fn layout_and_repaint(&mut self, reason: &wstr) {
         self.rendered_layout = self.make_layout_data();
         self.paint_layout(reason);
     }
 
     /// Paint the last rendered layout.
-    /// \p reason is used in FLOG to explain why.
+    /// `reason` is used in FLOG to explain why.
     fn paint_layout(&mut self, reason: &wstr) {
         FLOGF!(reader_render, "Repainting from %ls", reason);
         let data = &self.rendered_layout;
@@ -4493,7 +4493,7 @@ fn expand_replacer(
     ))
 }
 
-// Extract all the token ranges in \p str, along with whether they are a command.
+// Extract all the token ranges in `str`, along with whether they are a command.
 // Tokens containing command substitutions are skipped; this ensures tokens are non-overlapping.
 struct PositionedToken {
     range: SourceRange,
@@ -4997,8 +4997,8 @@ pub fn check_exit_loop_maybe_warning(data: Option<&mut ReaderData>) -> bool {
     true
 }
 
-/// Given that the user is tab-completing a token \p wc whose cursor is at \p pos in the token,
-/// try expanding it as a wildcard, populating \p result with the expanded string.
+/// Given that the user is tab-completing a token `wc` whose cursor is at `pos` in the token,
+/// try expanding it as a wildcard, populating `result` with the expanded string.
 fn try_expand_wildcard(
     parser: &Parser,
     wc: WString,
diff --git a/src/reader_history_search.rs b/src/reader_history_search.rs
index ff2abe365..346626caf 100644
--- a/src/reader_history_search.rs
+++ b/src/reader_history_search.rs
@@ -79,7 +79,7 @@ impl ReaderHistorySearch {
         self.mode == SearchMode::Prefix
     }
 
-    /// Move the history search in the given direction \p dir.
+    /// Move the history search in the given direction `dir`.
     pub fn move_in_direction(&mut self, dir: SearchDirection) -> bool {
         if dir == SearchDirection::Forward {
             self.move_forwards()
diff --git a/src/screen.rs b/src/screen.rs
index ba235b5b3..fad9c8fd5 100644
--- a/src/screen.rs
+++ b/src/screen.rs
@@ -61,7 +61,7 @@ impl Line {
         self.text.clear();
     }
 
-    /// Append a single character \p txt to the line with color \p c.
+    /// Append a single character `txt` to the line with color `c`.
     pub fn append(&mut self, character: char, highlight: HighlightSpec) {
         self.text.push(HighlightedChar {
             highlight,
@@ -69,7 +69,7 @@ impl Line {
         })
     }
 
-    /// Append a nul-terminated string \p txt to the line, giving each character \p color.
+    /// Append a nul-terminated string `txt` to the line, giving each character `color`.
     pub fn append_str(&mut self, txt: &wstr, highlight: HighlightSpec) {
         for c in txt.chars() {
             self.append(c, highlight);
@@ -91,12 +91,12 @@ impl Line {
         self.text[idx].highlight
     }
 
-    /// Append the contents of \p line to this line.
+    /// Append the contents of `line` to this line.
     pub fn append_line(&mut self, line: &Line) {
         self.text.extend_from_slice(&line.text);
     }
 
-    /// \return the width of this line, counting up to no more than \p max characters.
+    /// \return the width of this line, counting up to no more than `max` characters.
     /// This follows fish_wcswidth() semantics, except that characters whose width would be -1 are
     /// treated as 0.
     pub fn wcswidth_min_0(&self, max: usize /* = usize::MAX */) -> usize {
@@ -1133,7 +1133,7 @@ impl LayoutCache {
         self.esc_cache.len()
     }
 
-    /// Insert the entry \p str in its sorted position, if it is not already present in the cache.
+    /// Insert the entry `str` in its sorted position, if it is not already present in the cache.
     pub fn add_escape_code(&mut self, s: WString) {
         if let Err(pos) = self.esc_cache.binary_search(&s) {
             self.esc_cache.insert(pos, s);
@@ -1158,7 +1158,7 @@ impl LayoutCache {
         esc_seq_len
     }
 
-    /// \return the length of a string that matches a prefix of \p entry.
+    /// \return the length of a string that matches a prefix of `entry`.
     pub fn find_escape_code(&self, entry: &wstr) -> usize {
         // Do a binary search and see if the escape code right before our entry is a prefix of our
         // entry. Note this assumes that escape codes are prefix-free: no escape code is a prefix of
@@ -1177,7 +1177,7 @@ impl LayoutCache {
         0
     }
 
-    /// Computes a prompt layout for \p prompt_str, perhaps truncating it to \p max_line_width.
+    /// Computes a prompt layout for `prompt_str`, perhaps truncating it to `max_line_width`.
     /// \return the layout, and optionally the truncated prompt itself, by reference.
     pub fn calc_prompt_layout(
         &mut self,
@@ -1494,13 +1494,13 @@ fn is_visual_escape_seq(code: &wstr) -> Option<usize> {
     None
 }
 
-/// \return whether \p c ends a measuring run.
+/// \return whether `c` ends a measuring run.
 fn is_run_terminator(c: char) -> bool {
     matches!(c, '\0' | '\n' | '\r' | '\x0C')
 }
 
-/// Measure a run of characters in \p input starting at \p start.
-/// Stop when we reach a run terminator, and return its index in \p out_end (if not null).
+/// Measure a run of characters in `input` starting at `start`.
+/// Stop when we reach a run terminator, and return its index in `out_end` (if not null).
 /// Note \0 is a run terminator so there will always be one.
 /// We permit escape sequences to have run terminators other than \0. That is, escape sequences may
 /// have embedded newlines, etc.; it's unclear if this is possible but we allow it.
@@ -1532,7 +1532,7 @@ fn measure_run_from(
     width
 }
 
-/// Attempt to truncate the prompt run \p run, which has width \p width, to \p no more than
+/// Attempt to truncate the prompt run `run`, which has width `width`, to `no` more than
 /// desired_width. \return the resulting width and run by reference.
 fn truncate_run(
     run: &mut WString,
diff --git a/src/termsize.rs b/src/termsize.rs
index 5bb9a4340..b7e5bebc1 100644
--- a/src/termsize.rs
+++ b/src/termsize.rs
@@ -164,7 +164,7 @@ impl TermsizeContainer {
         data.current()
     }
 
-    /// If our termsize is stale, update it, using \p parser to fire any events that may be
+    /// If our termsize is stale, update it, using `parser` to fire any events that may be
     /// registered for COLUMNS and LINES.
     /// This requires a shared reference so it can work from a static.
     /// \return the updated termsize.
diff --git a/src/topic_monitor.rs b/src/topic_monitor.rs
index c06234688..06981e033 100644
--- a/src/topic_monitor.rs
+++ b/src/topic_monitor.rs
@@ -129,7 +129,7 @@ impl GenerationsList {
         ]
     }
 
-    /// Set the value of \p topic to the smaller of our value and the value in \p other.
+    /// Set the value of `topic` to the smaller of our value and the value in `other`.
     pub fn set_min_from(&mut self, topic: topic_t, other: &Self) {
         if self.get(topic) > other.get(topic) {
             self.set(topic, other.get(topic));
@@ -456,11 +456,11 @@ impl topic_monitor_t {
     }
 
     /// Given a list of input generations, attempt to update them to something newer.
-    /// If \p gens is older, then just return those by reference, and directly return false (not
+    /// If `gens` is older, then just return those by reference, and directly return false (not
     /// becoming the reader).
-    /// If \p gens is current and there is not a reader, then do not update \p gens and return true,
+    /// If `gens` is current and there is not a reader, then do not update `gens` and return true,
     /// indicating we should become the reader. Now it is our responsibility to wait on the
-    /// semaphore and notify on a change via the condition variable. If \p gens is current, and
+    /// semaphore and notify on a change via the condition variable. If `gens` is current, and
     /// there is already a reader, then wait until the reader notifies us and try again.
     fn try_update_gens_maybe_becoming_reader(&self, gens: &mut GenerationsList) -> bool {
         let mut become_reader = false;
@@ -547,11 +547,11 @@ impl topic_monitor_t {
         return gens;
     }
 
-    /// For each valid topic in \p gens, check to see if the current topic is larger than
-    /// the value in \p gens.
-    /// If \p wait is set, then wait if there are no changes; otherwise return immediately.
+    /// For each valid topic in `gens`, check to see if the current topic is larger than
+    /// the value in `gens`.
+    /// If `wait` is set, then wait if there are no changes; otherwise return immediately.
     /// \return true if some topic changed, false if none did.
-    /// On a true return, this updates the generation list \p gens.
+    /// On a true return, this updates the generation list `gens`.
     pub fn check(&self, gens: &GenerationsList, wait: bool) -> bool {
         if !gens.any_valid() {
             return false;
diff --git a/src/wchar_ext.rs b/src/wchar_ext.rs
index b07b14cf8..d8d851c4f 100644
--- a/src/wchar_ext.rs
+++ b/src/wchar_ext.rs
@@ -148,7 +148,7 @@ impl<'a> IntoCharIter for CharsUtf32<'a> {
     }
 }
 
-/// \return true if \p prefix is a prefix of \p contents.
+/// \return true if `prefix` is a prefix of `contents`.
 fn iter_prefixes_iter<Prefix, Contents>(prefix: Prefix, mut contents: Contents) -> bool
 where
     Prefix: Iterator,
diff --git a/src/wcstringutil.rs b/src/wcstringutil.rs
index 02b40f9ff..9a35611b7 100644
--- a/src/wcstringutil.rs
+++ b/src/wcstringutil.rs
@@ -157,9 +157,9 @@ impl StringFuzzyMatch {
         matches!(self.typ, ContainType::substr | ContainType::subseq)
     }
 
-    /// Try creating a fuzzy match for \p string against \p match_against.
-    /// \p string is something like "foo" and \p match_against is like "FooBar".
-    /// If \p anchor_start is set, then only exact and prefix matches are permitted.
+    /// Try creating a fuzzy match for `string` against `match_against`.
+    /// `string` is something like "foo" and `match_against` is like "FooBar".
+    /// If `anchor_start` is set, then only exact and prefix matches are permitted.
     pub fn try_create(
         string: &wstr,
         match_against: &wstr,
@@ -272,8 +272,8 @@ pub fn string_fuzzy_match_string(
 }
 
 /// Implementation of wcs2string that accepts a callback.
-/// This invokes \p func with (const char*, size_t) pairs.
-/// If \p func returns false, it stops; otherwise it continues.
+/// This invokes `func` with (const char*, size_t) pairs.
+/// If `func` returns false, it stops; otherwise it continues.
 /// \return false if the callback returned false, otherwise true.
 pub fn wcs2string_callback(input: &wstr, mut func: impl FnMut(&[u8]) -> bool) -> bool {
     let mut state = zero_mbstate();
@@ -330,10 +330,10 @@ pub fn split_string(val: &wstr, sep: char) -> Vec<WString> {
     val.split(sep).map(wstr::to_owned).collect()
 }
 
-/// Split a string by runs of any of the separator characters provided in \p seps.
-/// Note the delimiters are the characters in \p seps, not \p seps itself.
-/// \p seps may contain the NUL character.
-/// Do not output more than \p max_results results. If we are to output exactly that much,
+/// Split a string by runs of any of the separator characters provided in `seps`.
+/// Note the delimiters are the characters in `seps`, not `seps` itself.
+/// `seps` may contain the NUL character.
+/// Do not output more than `max_results` results. If we are to output exactly that much,
 /// the last output is the remainder of the input, including leading delimiters,
 /// except for the first. This is historical behavior.
 /// Example: split_string_tok(" a  b   c ", " ", 3) -> {"a", "b", "  c  "}
@@ -489,7 +489,7 @@ pub fn trim(input: WString, any_of: Option<&wstr>) -> WString {
 }
 
 /// \return the number of escaping backslashes before a character.
-/// \p idx may be "one past the end."
+/// `idx` may be "one past the end."
 pub fn count_preceding_backslashes(text: &wstr, idx: usize) -> usize {
     assert!(idx <= text.len(), "Out of bounds");
     text.chars()
diff --git a/src/wutil/mod.rs b/src/wutil/mod.rs
index 631a41db6..7873d07e8 100644
--- a/src/wutil/mod.rs
+++ b/src/wutil/mod.rs
@@ -269,7 +269,7 @@ fn test_normalize_path() {
     assert_eq!(norm_path(L!("foo/././bar/.././baz")), "foo/baz");
 }
 
-/// Given an input path \p path and a working directory \p wd, do a "normalizing join" in a way
+/// Given an input path `path` and a working directory `wd`, do a "normalizing join" in a way
 /// appropriate for cd. That is, return effectively wd + path while resolving leading ../s from
 /// path. The intent here is to allow 'cd' out of a directory which may no longer exist, without
 /// allowing 'cd' into a directory that may not exist; see #5341.
@@ -586,7 +586,7 @@ pub fn file_id_for_path_narrow(path: &CStr) -> FileId {
     }
     result
 }
-/// Given that \p cursor is a pointer into \p base, return the offset in characters.
+/// Given that `cursor` is a pointer into `base`, return the offset in characters.
 /// This emulates C pointer arithmetic:
 ///    `wstr_offset_in(cursor, base)` is equivalent to C++ `cursor - base`.
 pub fn wstr_offset_in(cursor: &wstr, base: &wstr) -> usize {
diff --git a/src/wutil/wcstoi.rs b/src/wutil/wcstoi.rs
index 15daf2594..e156e44ab 100644
--- a/src/wutil/wcstoi.rs
+++ b/src/wutil/wcstoi.rs
@@ -47,7 +47,7 @@ impl<Iter: Iterator<Item = char>> CharsIterator<Iter> {
     }
 }
 
-/// Parse the given \p src as an integer.
+/// Parse the given `src` as an integer.
 /// If mradix is not None, it is used as the radix; otherwise the radix is inferred:
 ///   - Leading 0x or 0X means 16.
 ///   - Leading 0 means 8.