x86/kprobes: Fix crash when probing CS CALL instructionsgrafted

When using eBPF to probe CS CALL instructions within a function,
a crash can be triggered.

The eBPF tool probes offset 257 of the __hrtimer_run_queues()
function:

<__hrtimer_run_queues+249>:  nopl   0x0(%rax,%rax,1)
<__hrtimer_run_queues+254>:  mov    %r14,%rdi
<__hrtimer_run_queues+257>:  cs call <__x86_indirect_thunk_r12>
<__hrtimer_run_queues+263>:  mov    %eax,%r12d
<__hrtimer_run_queues+266>:  xchg   %ax,%ax
<__hrtimer_run_queues+268>:  mov    %r13,%rdi

Which triggers this crash:

  BUG: unable to handle page fault for address: 00000000000f41c9
  #PF: supervisor write access in kernel mode
  #PF: error_code(0x0002) - not-present page
  PGD 0 P4D 0
  Oops: 0002 [#1] SMP NOPTI
  CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P
  RIP: 0010:__hrtimer_run_queues+0x106/0x230

Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is
at the 6th byte of the above CS CALL instruction. Since the CS CALL
instruction occupies 6 bytes, the exception occurred in the middle of that
call instruction.

The root cause is that when using eBPF tools to probe in the middle of a
function, a kprobe with INT3 is used as the underlying implementation.

During single-step emulation of the original CALL instruction,
int3_emulate_call() assumes that the probed CALL instruction is 5 bytes
long. However, the actual CS-prefixed CALL instruction occupies 6 bytes,
so it constructs an incorrect exception return address. When the CPU
returns from the kprobe handler, the next instruction to be executed is at
the address of the last byte of that CS CALL instruction. Coincidentally,
starting from that address, the CPU fetches and decodes a completely
different instruction, which ultimately triggers a kernel crash.

Fix the issue by using the actual instruction length obtained from
the instruction decoder when constructing the exception return
address, rather than relying on the hardcoded CALL_INSN_SIZE macro.

[ mingo: Refined the changelog ]

Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step")
Suggested-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Jinke Han <jinkehan@didiglobal.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Acked-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000

1 files changed, 654 insertions, 0 deletions

diff --git a/rust/macros/module.rs b/rust/macros/module.rs
new file mode 100644
index 000000000..bc7027f8d
--- /dev/null
+++ b/rust/macros/module.rs
@@ -0,0 +1,654 @@
+// SPDX-License-Identifier: GPL-2.0
+
+use std::ffi::CString;
+
+use proc_macro2::{
+    Literal,
+    TokenStream, //
+};
+use quote::{
+    format_ident,
+    quote, //
+};
+use syn::{
+    braced,
+    bracketed,
+    ext::IdentExt,
+    parse::{
+        Parse,
+        ParseStream, //
+    },
+    parse_quote,
+    punctuated::Punctuated,
+    Error,
+    Expr,
+    Ident,
+    LitStr,
+    Path,
+    Result,
+    Token,
+    Type, //
+};
+
+use crate::helpers::*;
+
+struct ModInfoBuilder<'a> {
+    module: &'a str,
+    counter: usize,
+    ts: TokenStream,
+    param_ts: TokenStream,
+}
+
+impl<'a> ModInfoBuilder<'a> {
+    fn new(module: &'a str) -> Self {
+        ModInfoBuilder {
+            module,
+            counter: 0,
+            ts: TokenStream::new(),
+            param_ts: TokenStream::new(),
+        }
+    }
+
+    fn emit_base(&mut self, field: &str, content: &str, builtin: bool, param: bool) {
+        let string = if builtin {
+            // Built-in modules prefix their modinfo strings by `module.`.
+            format!("{module}.{field}={content}\0", module = self.module)
+        } else {
+            // Loadable modules' modinfo strings go as-is.
+            format!("{field}={content}\0")
+        };
+        let length = string.len();
+        let string = Literal::byte_string(string.as_bytes());
+        let cfg = if builtin {
+            quote!(#[cfg(not(MODULE))])
+        } else {
+            quote!(#[cfg(MODULE)])
+        };
+
+        let counter = format_ident!(
+            "__{module}_{counter}",
+            module = self.module.to_uppercase(),
+            counter = self.counter
+        );
+        let item = quote! {
+            #cfg
+            #[cfg_attr(not(target_os = "macos"), link_section = ".modinfo")]
+            #[used(compiler)]
+            pub static #counter: [u8; #length] = *#string;
+        };
+
+        if param {
+            self.param_ts.extend(item);
+        } else {
+            self.ts.extend(item);
+        }
+
+        self.counter += 1;
+    }
+
+    fn emit_only_builtin(&mut self, field: &str, content: &str, param: bool) {
+        self.emit_base(field, content, true, param)
+    }
+
+    fn emit_only_loadable(&mut self, field: &str, content: &str, param: bool) {
+        self.emit_base(field, content, false, param)
+    }
+
+    fn emit(&mut self, field: &str, content: &str) {
+        self.emit_internal(field, content, false);
+    }
+
+    fn emit_internal(&mut self, field: &str, content: &str, param: bool) {
+        self.emit_only_builtin(field, content, param);
+        self.emit_only_loadable(field, content, param);
+    }
+
+    fn emit_param(&mut self, field: &str, param: &str, content: &str) {
+        let content = format!("{param}:{content}");
+        self.emit_internal(field, &content, true);
+    }
+
+    fn emit_params(&mut self, info: &ModuleInfo) {
+        let Some(params) = &info.params else {
+            return;
+        };
+
+        for param in params {
+            let param_name_str = param.name.to_string();
+            let param_type_str = param.ptype.to_string();
+
+            let ops = param_ops_path(&param_type_str);
+
+            // Note: The spelling of these fields is dictated by the user space
+            // tool `modinfo`.
+            self.emit_param("parmtype", &param_name_str, &param_type_str);
+            self.emit_param("parm", &param_name_str, &param.description.value());
+
+            let static_name = format_ident!("__{}_{}_struct", self.module, param.name);
+            let param_name_cstr =
+                CString::new(param_name_str).expect("name contains NUL-terminator");
+            let param_name_cstr_with_module =
+                CString::new(format!("{}.{}", self.module, param.name))
+                    .expect("name contains NUL-terminator");
+
+            let param_name = &param.name;
+            let param_type = &param.ptype;
+            let param_default = &param.default;
+
+            self.param_ts.extend(quote! {
+                #[allow(non_upper_case_globals)]
+                pub(crate) static #param_name:
+                    ::kernel::module_param::ModuleParamAccess<#param_type> =
+                        ::kernel::module_param::ModuleParamAccess::new(#param_default);
+
+                const _: () = {
+                    #[allow(non_upper_case_globals)]
+                    #[link_section = "__param"]
+                    #[used(compiler)]
+                    static #static_name:
+                        ::kernel::module_param::KernelParam =
+                        ::kernel::module_param::KernelParam::new(
+                            ::kernel::bindings::kernel_param {
+                                name: kernel::str::as_char_ptr_in_const_context(
+                                    if ::core::cfg!(MODULE) {
+                                        #param_name_cstr
+                                    } else {
+                                        #param_name_cstr_with_module
+                                    }
+                                ),
+                                // SAFETY: `__this_module` is constructed by the kernel at load
+                                // time and will not be freed until the module is unloaded.
+                                #[cfg(MODULE)]
+                                mod_: unsafe {
+                                    core::ptr::from_ref(&::kernel::bindings::__this_module)
+                                        .cast_mut()
+                                },
+                                #[cfg(not(MODULE))]
+                                mod_: ::core::ptr::null_mut(),
+                                ops: core::ptr::from_ref(&#ops),
+                                perm: 0, // Will not appear in sysfs
+                                level: -1,
+                                flags: 0,
+                                __bindgen_anon_1: ::kernel::bindings::kernel_param__bindgen_ty_1 {
+                                    arg: #param_name.as_void_ptr()
+                                },
+                            }
+                        );
+                };
+            });
+        }
+    }
+}
+
+fn param_ops_path(param_type: &str) -> Path {
+    match param_type {
+        "i8" => parse_quote!(::kernel::module_param::PARAM_OPS_I8),
+        "u8" => parse_quote!(::kernel::module_param::PARAM_OPS_U8),
+        "i16" => parse_quote!(::kernel::module_param::PARAM_OPS_I16),
+        "u16" => parse_quote!(::kernel::module_param::PARAM_OPS_U16),
+        "i32" => parse_quote!(::kernel::module_param::PARAM_OPS_I32),
+        "u32" => parse_quote!(::kernel::module_param::PARAM_OPS_U32),
+        "i64" => parse_quote!(::kernel::module_param::PARAM_OPS_I64),
+        "u64" => parse_quote!(::kernel::module_param::PARAM_OPS_U64),
+        "isize" => parse_quote!(::kernel::module_param::PARAM_OPS_ISIZE),
+        "usize" => parse_quote!(::kernel::module_param::PARAM_OPS_USIZE),
+        "bool" => parse_quote!(::kernel::module_param::PARAM_OPS_BOOL),
+        t => panic!("Unsupported parameter type {}", t),
+    }
+}
+
+/// Parse fields that are required to use a specific order.
+///
+/// As fields must follow a specific order, we *could* just parse fields one by one by peeking.
+/// However the error message generated when implementing that way is not very friendly.
+///
+/// So instead we parse fields in an arbitrary order, but only enforce the ordering after parsing,
+/// and if the wrong order is used, the proper order is communicated to the user with error message.
+///
+/// Usage looks like this:
+/// ```ignore
+/// parse_ordered_fields! {
+///     from input;
+///
+///     // This will extract "foo: <field>" into a variable named "foo".
+///     // The variable will have type `Option<_>`.
+///     foo => <expression that parses the field>,
+///
+///     // If you need the variable name to be different than the key name.
+///     // This extracts "baz: <field>" into a variable named "bar".
+///     // You might want this if "baz" is a keyword.
+///     baz as bar => <expression that parse the field>,
+///
+///     // You can mark a key as required, and the variable will no longer be `Option`.
+///     // foobar will be of type `Expr` instead of `Option<Expr>`.
+///     foobar [required] => input.parse::<Expr>()?,
+/// }
+/// ```
+macro_rules! parse_ordered_fields {
+    (@gen
+        [$input:expr]
+        [$([$name:ident; $key:ident; $parser:expr])*]
+        [$([$req_name:ident; $req_key:ident])*]
+    ) => {
+        $(let mut $name = None;)*
+
+        const EXPECTED_KEYS: &[&str] = &[$(stringify!($key),)*];
+        const REQUIRED_KEYS: &[&str] = &[$(stringify!($req_key),)*];
+
+        let span = $input.span();
+        let mut seen_keys = Vec::new();
+
+        while !$input.is_empty() {
+            let key = $input.call(Ident::parse_any)?;
+
+            if seen_keys.contains(&key) {
+                Err(Error::new_spanned(
+                    &key,
+                    format!(r#"duplicated key "{key}". Keys can only be specified once."#),
+                ))?
+            }
+
+            $input.parse::<Token![:]>()?;
+
+            match &*key.to_string() {
+                $(
+                    stringify!($key) => $name = Some($parser),
+                )*
+                _ => {
+                    Err(Error::new_spanned(
+                        &key,
+                        format!(r#"unknown key "{key}". Valid keys are: {EXPECTED_KEYS:?}."#),
+                    ))?
+                }
+            }
+
+            $input.parse::<Token![,]>()?;
+            seen_keys.push(key);
+        }
+
+        for key in REQUIRED_KEYS {
+            if !seen_keys.iter().any(|e| e == key) {
+                Err(Error::new(span, format!(r#"missing required key "{key}""#)))?
+            }
+        }
+
+        let mut ordered_keys: Vec<&str> = Vec::new();
+        for key in EXPECTED_KEYS {
+            if seen_keys.iter().any(|e| e == key) {
+                ordered_keys.push(key);
+            }
+        }
+
+        if seen_keys != ordered_keys {
+            Err(Error::new(
+                span,
+                format!(r#"keys are not ordered as expected. Order them like: {ordered_keys:?}."#),
+            ))?
+        }
+
+        $(let $req_name = $req_name.expect("required field");)*
+    };
+
+    // Handle required fields.
+    (@gen
+        [$input:expr] [$($tok:tt)*] [$($req:tt)*]
+        $key:ident as $name:ident [required] => $parser:expr,
+        $($rest:tt)*
+    ) => {
+        parse_ordered_fields!(
+            @gen [$input] [$($tok)* [$name; $key; $parser]] [$($req)* [$name; $key]] $($rest)*
+        )
+    };
+    (@gen
+        [$input:expr] [$($tok:tt)*] [$($req:tt)*]
+        $name:ident [required] => $parser:expr,
+        $($rest:tt)*
+    ) => {
+        parse_ordered_fields!(
+            @gen [$input] [$($tok)* [$name; $name; $parser]] [$($req)* [$name; $name]] $($rest)*
+        )
+    };
+
+    // Handle optional fields.
+    (@gen
+        [$input:expr] [$($tok:tt)*] [$($req:tt)*]
+        $key:ident as $name:ident => $parser:expr,
+        $($rest:tt)*
+    ) => {
+        parse_ordered_fields!(
+            @gen [$input] [$($tok)* [$name; $key; $parser]] [$($req)*] $($rest)*
+        )
+    };
+    (@gen
+        [$input:expr] [$($tok:tt)*] [$($req:tt)*]
+        $name:ident => $parser:expr,
+        $($rest:tt)*
+    ) => {
+        parse_ordered_fields!(
+            @gen [$input] [$($tok)* [$name; $name; $parser]] [$($req)*] $($rest)*
+        )
+    };
+
+    (from $input:expr; $($tok:tt)*) => {
+        parse_ordered_fields!(@gen [$input] [] [] $($tok)*)
+    }
+}
+
+struct Parameter {
+    name: Ident,
+    ptype: Ident,
+    default: Expr,
+    description: LitStr,
+}
+
+impl Parse for Parameter {
+    fn parse(input: ParseStream<'_>) -> Result<Self> {
+        let name = input.parse()?;
+        input.parse::<Token![:]>()?;
+        let ptype = input.parse()?;
+
+        let fields;
+        braced!(fields in input);
+
+        parse_ordered_fields! {
+            from fields;
+            default [required] => fields.parse()?,
+            description [required] => fields.parse()?,
+        }
+
+        Ok(Self {
+            name,
+            ptype,
+            default,
+            description,
+        })
+    }
+}
+
+pub(crate) struct ModuleInfo {
+    type_: Type,
+    license: AsciiLitStr,
+    name: AsciiLitStr,
+    authors: Option<Punctuated<AsciiLitStr, Token![,]>>,
+    description: Option<LitStr>,
+    alias: Option<Punctuated<AsciiLitStr, Token![,]>>,
+    firmware: Option<Punctuated<AsciiLitStr, Token![,]>>,
+    imports_ns: Option<Punctuated<AsciiLitStr, Token![,]>>,
+    params: Option<Punctuated<Parameter, Token![,]>>,
+}
+
+impl Parse for ModuleInfo {
+    fn parse(input: ParseStream<'_>) -> Result<Self> {
+        parse_ordered_fields!(
+            from input;
+            type as type_ [required] => input.parse()?,
+            name [required] => input.parse()?,
+            authors => {
+                let list;
+                bracketed!(list in input);
+                Punctuated::parse_terminated(&list)?
+            },
+            description => input.parse()?,
+            license [required] => input.parse()?,
+            alias => {
+                let list;
+                bracketed!(list in input);
+                Punctuated::parse_terminated(&list)?
+            },
+            firmware => {
+                let list;
+                bracketed!(list in input);
+                Punctuated::parse_terminated(&list)?
+            },
+            imports_ns => {
+                let list;
+                bracketed!(list in input);
+                Punctuated::parse_terminated(&list)?
+            },
+            params => {
+                let list;
+                braced!(list in input);
+                Punctuated::parse_terminated(&list)?
+            },
+        );
+
+        Ok(ModuleInfo {
+            type_,
+            license,
+            name,
+            authors,
+            description,
+            alias,
+            firmware,
+            imports_ns,
+            params,
+        })
+    }
+}
+
+pub(crate) fn module(info: ModuleInfo) -> Result<TokenStream> {
+    let ModuleInfo {
+        type_,
+        license,
+        name,
+        authors,
+        description,
+        alias,
+        firmware,
+        imports_ns,
+        params: _,
+    } = &info;
+
+    // Rust does not allow hyphens in identifiers, use underscore instead.
+    let ident = name.value().replace('-', "_");
+    let mut modinfo = ModInfoBuilder::new(ident.as_ref());
+    if let Some(authors) = authors {
+        for author in authors {
+            modinfo.emit("author", &author.value());
+        }
+    }
+    if let Some(description) = description {
+        modinfo.emit("description", &description.value());
+    }
+    modinfo.emit("license", &license.value());
+    if let Some(aliases) = alias {
+        for alias in aliases {
+            modinfo.emit("alias", &alias.value());
+        }
+    }
+    if let Some(firmware) = firmware {
+        for fw in firmware {
+            modinfo.emit("firmware", &fw.value());
+        }
+    }
+    if let Some(imports) = imports_ns {
+        for ns in imports {
+            modinfo.emit("import_ns", &ns.value());
+        }
+    }
+
+    // Built-in modules also export the `file` modinfo string.
+    let file =
+        std::env::var("RUST_MODFILE").expect("Unable to fetch RUST_MODFILE environmental variable");
+    modinfo.emit_only_builtin("file", &file, false);
+
+    modinfo.emit_params(&info);
+
+    let modinfo_ts = modinfo.ts;
+    let params_ts = modinfo.param_ts;
+
+    let ident_init = format_ident!("__{ident}_init");
+    let ident_exit = format_ident!("__{ident}_exit");
+    let ident_initcall = format_ident!("__{ident}_initcall");
+    let initcall_section = ".initcall6.init";
+
+    let global_asm = format!(
+        r#".section "{initcall_section}", "a"
+        __{ident}_initcall:
+            .long   __{ident}_init - .
+            .previous
+        "#
+    );
+
+    let name_cstr = CString::new(name.value()).expect("name contains NUL-terminator");
+
+    Ok(quote! {
+        /// The module name.
+        ///
+        /// Used by the printing macros, e.g. [`info!`].
+        const __LOG_PREFIX: &[u8] = #name_cstr.to_bytes_with_nul();
+
+        /// The `LocalModule` type is the type of the module created by `module!`,
+        /// `module_pci_driver!`, `module_platform_driver!`, etc.
+        type LocalModule = #type_;
+
+        impl ::kernel::ModuleMetadata for #type_ {
+            const NAME: &'static ::kernel::str::CStr = #name_cstr;
+
+            #[cfg(MODULE)]
+            const THIS_MODULE: ::kernel::ThisModule = {
+                extern "C" {
+                    static __this_module: ::kernel::types::Opaque<::kernel::bindings::module>;
+                }
+
+                // SAFETY: `__this_module` is constructed by the kernel at load time
+                // and lives until the module is unloaded.
+                unsafe { ::kernel::ThisModule::from_ptr(__this_module.get()) }
+            };
+
+            #[cfg(not(MODULE))]
+            const THIS_MODULE: ::kernel::ThisModule = unsafe {
+                ::kernel::ThisModule::from_ptr(::core::ptr::null_mut())
+            };
+        }
+
+        // Double nested modules, since then nobody can access the public items inside.
+        #[doc(hidden)]
+        mod __module_init {
+            mod __module_init {
+                use pin_init::PinInit;
+
+                /// The "Rust loadable module" mark.
+                //
+                // This may be best done another way later on, e.g. as a new modinfo
+                // key or a new section. For the moment, keep it simple.
+                #[cfg(MODULE)]
+                #[used(compiler)]
+                static __IS_RUST_MODULE: () = ();
+
+                static mut __MOD: ::core::mem::MaybeUninit<super::super::LocalModule> =
+                    ::core::mem::MaybeUninit::uninit();
+
+                // Loadable modules need to export the `{init,cleanup}_module` identifiers.
+                /// # Safety
+                ///
+                /// This function must not be called after module initialization, because it may be
+                /// freed after that completes.
+                #[cfg(MODULE)]
+                #[no_mangle]
+                #[link_section = ".init.text"]
+                pub unsafe extern "C" fn init_module() -> ::kernel::ffi::c_int {
+                    // SAFETY: This function is inaccessible to the outside due to the double
+                    // module wrapping it. It is called exactly once by the C side via its
+                    // unique name.
+                    unsafe { __init() }
+                }
+
+                #[cfg(MODULE)]
+                #[used(compiler)]
+                #[link_section = ".init.data"]
+                static __UNIQUE_ID___addressable_init_module: unsafe extern "C" fn() -> i32 =
+                    init_module;
+
+                #[cfg(MODULE)]
+                #[no_mangle]
+                #[link_section = ".exit.text"]
+                pub extern "C" fn cleanup_module() {
+                    // SAFETY:
+                    // - This function is inaccessible to the outside due to the double
+                    //   module wrapping it. It is called exactly once by the C side via its
+                    //   unique name,
+                    // - furthermore it is only called after `init_module` has returned `0`
+                    //   (which delegates to `__init`).
+                    unsafe { __exit() }
+                }
+
+                #[cfg(MODULE)]
+                #[used(compiler)]
+                #[link_section = ".exit.data"]
+                static __UNIQUE_ID___addressable_cleanup_module: extern "C" fn() = cleanup_module;
+
+                // Built-in modules are initialized through an initcall pointer
+                // and the identifiers need to be unique.
+                #[cfg(not(MODULE))]
+                #[cfg(not(CONFIG_HAVE_ARCH_PREL32_RELOCATIONS))]
+                #[link_section = #initcall_section]
+                #[used(compiler)]
+                pub static #ident_initcall: extern "C" fn() ->
+                    ::kernel::ffi::c_int = #ident_init;
+
+                #[cfg(not(MODULE))]
+                #[cfg(CONFIG_HAVE_ARCH_PREL32_RELOCATIONS)]
+                ::core::arch::global_asm!(#global_asm);
+
+                #[cfg(not(MODULE))]
+                #[no_mangle]
+                pub extern "C" fn #ident_init() -> ::kernel::ffi::c_int {
+                    // SAFETY: This function is inaccessible to the outside due to the double
+                    // module wrapping it. It is called exactly once by the C side via its
+                    // placement above in the initcall section.
+                    unsafe { __init() }
+                }
+
+                #[cfg(not(MODULE))]
+                #[no_mangle]
+                pub extern "C" fn #ident_exit() {
+                    // SAFETY:
+                    // - This function is inaccessible to the outside due to the double
+                    //   module wrapping it. It is called exactly once by the C side via its
+                    //   unique name,
+                    // - furthermore it is only called after `#ident_init` has
+                    //   returned `0` (which delegates to `__init`).
+                    unsafe { __exit() }
+                }
+
+                /// # Safety
+                ///
+                /// This function must only be called once.
+                unsafe fn __init() -> ::kernel::ffi::c_int {
+                    let initer = <super::super::LocalModule as ::kernel::InPlaceModule>::init(
+                        ::kernel::module::this_module::<super::super::LocalModule>()
+                    );
+                    // SAFETY: No data race, since `__MOD` can only be accessed by this module
+                    // and there only `__init` and `__exit` access it. These functions are only
+                    // called once and `__exit` cannot be called before or during `__init`.
+                    match unsafe { ::pin_init::raw_try_init(__MOD.as_mut_ptr(), initer) } {
+                        Ok(m) => 0,
+                        Err(e) => e.to_errno(),
+                    }
+                }
+
+                /// # Safety
+                ///
+                /// This function must
+                /// - only be called once,
+                /// - be called after `__init` has been called and returned `0`.
+                unsafe fn __exit() {
+                    // SAFETY: No data race, since `__MOD` can only be accessed by this module
+                    // and there only `__init` and `__exit` access it. These functions are only
+                    // called once and `__init` was already called.
+                    unsafe {
+                        // Invokes `drop()` on `__MOD`, which should be used for cleanup.
+                        __MOD.assume_init_drop();
+                    }
+                }
+
+                #modinfo_ts
+            }
+        }
+
+        mod module_parameters {
+            #params_ts
+        }
+    })
+}