edlc_codegen_cranelift 0.2.16

/*
 *     EDLc, a compiler for the EDL programming language.
 *     Copyright (C) 2026  Adrian Paskert
 *
 *     This program is free software: you can redistribute it and/or modify
 *     it under the terms of the GNU Affero General Public License as published by
 *     the Free Software Foundation, either version 3 of the License, or
 *     (at your option) any later version.
 *
 *     This program is distributed in the hope that it will be useful,
 *     but WITHOUT ANY WARRANTY; without even the implied warranty of
 *     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *     GNU Affero General Public License for more details.
 *
 *     You should have received a copy of the GNU Affero General Public License
 *     along with this program.  If not, see <http://www.gnu.org/licenses/>.
 */

#[cfg(any(target_arch="x86_64", target_arch="x86"))]
mod x86;
#[cfg(any(target_os="linux", target_os="macos", target_os="freebsd", target_os="openbsd"))]
mod unix;
mod signal_stack;
mod cfi;
mod range_vec;

use std::cell::{LazyCell, RefCell};
use std::fmt::{Display, Formatter};
use std::hash::DefaultHasher;
use std::ops::Index;
use std::sync::atomic::{AtomicBool, Ordering};
use std::sync::Mutex;
use cranelift_codegen::ir::TrapCode;
use log::{error, warn};
#[cfg(feature="serde")]
use serde::{Deserialize, Serialize};
use edlc_core::lexer::SrcPos;
use edlc_core::prelude::{AmorphusDataCopy, HirPhase, MirPhase, ModuleSrc, TrapInfo, TypeArgument, TypeArguments};
use edlc_core::prelude::mir_backend::Backend;
use edlc_core::prelude::mir_funcs::{MirFuncId, MirFuncRegistry};
use edlc_core::prelude::mir_type::MirTypeId;
#[cfg(any(target_os="linux", target_os="macos", target_os="freebsd", target_os="openbsd"))]
pub use unix::{TrapHandler, jit_panic, cause_jit_async_panic, jit_sync_panic};
pub use range_vec::{RangeVec, RangeVecIter};
use crate::compiler::{UnwindInfo, JIT};

#[cfg_attr(feature="serde", derive(Serialize, Deserialize))]
#[derive(Debug)]
pub enum PrintableBacktraceElement {
    Jit {
        pos: SrcPos,
        src: ModuleSrc,
        func: MirFuncId,
    },
    JitUnknown {
        func: MirFuncId,
        offset: u32,
    },
    Host {
        func: usize,
        offset: u32,
    },
}

#[cfg_attr(feature="serde", derive(Serialize, Deserialize))]
#[derive(Default, Debug)]
pub struct PrintableBacktrace {
    pub trace: Vec<PrintableBacktraceElement>,
}


#[derive(Default, Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Hash, Debug)]
struct BacktraceFlags(u32);

const FLAG_JIT_FRAME: u32 = 0b1;

impl BacktraceFlags {
    pub fn set_jit_frame(mut self) -> Self {
        self.0 |= FLAG_JIT_FRAME;
        self
    }

    pub fn jit_frame(&self) -> bool {
        self.0 & FLAG_JIT_FRAME != 0
    }
}

#[derive(Default, Clone, Copy, Debug)]
pub struct BacktraceEntry {
    /// Function pointer in which the location lives.
    /// This is guaranteed to always be a function generated by the JIT.
    func: usize,
    /// Code offset of the panic (from the base addr).
    loc: u32,
    flags: BacktraceFlags,
}

pub struct Backtrace {
    trace: Box<[BacktraceEntry]>,
    len: usize,
}


const BACKTRACE_SIZE_MAX: usize = 32;

impl Backtrace {
    fn new() -> Self {
        let trace = vec![const { BacktraceEntry {
            loc: 0,
            func: 0,
            flags: BacktraceFlags(0),
        } }; BACKTRACE_SIZE_MAX];
        Backtrace {
            trace: trace.into_boxed_slice(),
            len: 0,
        }
    }

    fn push(&mut self, entry: BacktraceEntry) {
        if self.len >= BACKTRACE_SIZE_MAX {
            return;
        }
        self.trace[self.len] = entry;
        self.len += 1;
    }

    fn slice(&self) -> &[BacktraceEntry] {
        &self.trace[..self.len]
    }

    fn clear(&mut self) {
        self.len = 0;
    }

    fn len(&self) -> usize {
        self.len
    }

    fn is_empty(&self) -> bool {
        self.len == 0
    }

    /// Returns true if the panic unwind was initiated by a synchronous panic.
    fn synchronous(&self) -> bool {
        if let Some(first) = self.slice().first() {
            first.func as *const u8 == jit_sync_panic as *const u8
        } else {
            false
        }
    }

    /// Returns true if it is safe to resume program execution after this point.
    /// Some panic unwinds can be recovered gracefully, whereas other panics are completely unsafe
    /// to resume, as not all cleanup routines are executed.
    pub fn is_recoverable(&self) -> bool {
        let mut iter = self.slice().iter();
        let Some(first) = iter.next() else {
            return true;
        };
        if first.flags.jit_frame() {
            return true;
        }
        if first.func as *const u8 != jit_sync_panic as *const u8 {
            error!("[EDL-JIT] non-synchronous panic originating in host frames");
            return false; // non-synchronous panic originating in host frames
        }
        let Some(next) = iter.next() else {
            return true;
        };
        if next.flags.jit_frame() {
            return true; // jit_sync_panic was called from JIT -> we are recoverable
        }
        let Some(next) = iter.next() else {
            return true;
        };
        if !next.flags.jit_frame() {
            error!("[EDL-JIT] jit_sync_panic called from a host function that is nested more than \
            one function call away from JIT frame");
            return false; // jit_sync_panic was called from a host function that is not directly
                          // called from the JIT -> cleanup routine in host frame not executed
        }
        true
    }

    fn get_panic_type(&self, debug_frames: &UnwindInfo) -> Option<PanicType> {
        let Some(first) = self.slice().first() else {
            return None;
        };
        if let Some(id) = debug_frames.find_id(&first.func) {
            if let Some(source_frame) = debug_frames.get_source(&id) {
                if let Some(trap_code) = source_frame.trap_code(first.loc) {
                    match *trap_code {
                        TrapCode::BAD_CONVERSION_TO_INTEGER => Some(PanicType::BadConversionToInteger),
                        TrapCode::HEAP_OUT_OF_BOUNDS => Some(PanicType::HeapOutOfBounds),
                        TrapCode::INTEGER_DIVISION_BY_ZERO => Some(PanicType::DivideByZero),
                        TrapCode::INTEGER_OVERFLOW => Some(PanicType::IntegerOverflow),
                        TrapCode::STACK_OVERFLOW => Some(PanicType::StackOverflow),
                        _ => {
                            // assume that this is a user-defined trap code
                            if let Some(trap_info) = source_frame.trap_info(first.loc) {
                                match trap_info {
                                    TrapInfo::DivideByZero => Some(PanicType::DivideByZero),
                                    TrapInfo::ArrayIndex => Some(PanicType::ArrayIndexOutOfBounds),
                                    TrapInfo::SliceIndex => Some(PanicType::SliceIndexOutOfBounds),
                                    TrapInfo::ArrayRange => Some(PanicType::ArrayRangeOutOfBounds),
                                    TrapInfo::SliceRange => Some(PanicType::SliceRangeOutOfBounds),
                                    TrapInfo::ExplicitPanic => Some(PanicType::Explicit),
                                    TrapInfo::AssertionFailed => Some(PanicType::Assertion),
                                    TrapInfo::Other(reason) => Some(PanicType::Other(reason)),
                                }
                            } else {
                                Some(PanicType::Unknown)
                            }
                        },
                    }
                } else {
                    Some(PanicType::Unknown)
                }
            } else {
                None
            }
        } else {
            None
        }
    }

    fn printable(
        &self,
        debug_frames: &UnwindInfo,
    ) -> PrintableBacktrace {
        let mut out = PrintableBacktrace::default();
        // print stack trace
        for trace in self.slice().iter() {
            let el = if let Some(id) = debug_frames.find_id(&trace.func) {
                if let Some(source_frame) = debug_frames.get_source(&id) {
                    if let Some(loc) = source_frame.source_location(trace.loc) {
                        PrintableBacktraceElement::Jit {
                            func: id,
                            src: loc.src.clone(),
                            pos: loc.pos,
                        }
                    } else {
                        PrintableBacktraceElement::JitUnknown {
                            func: id,
                            offset: trace.loc,
                        }
                    }
                } else {
                    PrintableBacktraceElement::JitUnknown {
                        func: id,
                        offset: trace.loc,
                    }
                }
            } else {
                PrintableBacktraceElement::Host {
                    func: trace.func,
                    offset: trace.loc,
                }
            };
            out.trace.push(el);
        }
        out
    }

    fn print<B: Backend>(
        &self,
        phase: &HirPhase,
        funcs: &MirFuncRegistry<B>,
        debug_frames: &UnwindInfo,
    ) {
        // format trap reason
        let Some(first) = self.slice().first() else {
            return;
        };
        if let Some(id) = debug_frames.find_id(&first.func) {
            if let Some(source_frame) = debug_frames.get_source(&id) {
                if let Some(trap_code) = source_frame.trap_code(first.loc) {
                    match *trap_code {
                        TrapCode::BAD_CONVERSION_TO_INTEGER => {
                            error!("bad conversion to integer");
                        },
                        TrapCode::HEAP_OUT_OF_BOUNDS => {
                            error!("heap out of bounds");
                        },
                        TrapCode::INTEGER_DIVISION_BY_ZERO => {
                            error!("integer division by zero")
                        },
                        TrapCode::INTEGER_OVERFLOW => {
                            error!("integer overflow")
                        },
                        TrapCode::STACK_OVERFLOW => {
                            error!("stack overflow")
                        },
                        _ => {
                            // assume that this is a user-defined trap code
                            if let Some(trap_info) = source_frame.trap_info(first.loc) {
                                match trap_info {
                                    TrapInfo::DivideByZero => {
                                        error!("division by zero");
                                    }
                                    TrapInfo::ArrayIndex => {
                                        error!("array index out of bounds");
                                    }
                                    TrapInfo::SliceIndex => {
                                        error!("slice index out of bounds");
                                    }
                                    TrapInfo::ArrayRange => {
                                        error!("array range out of bounds");
                                    }
                                    TrapInfo::SliceRange => {
                                        error!("slice range out of bounds");
                                    }
                                    TrapInfo::ExplicitPanic => {
                                        error!("explicit panic");
                                    }
                                    TrapInfo::AssertionFailed => {
                                        error!("assertion failed");
                                    }
                                    TrapInfo::Other(reason) => {
                                        error!("{}", reason);
                                    }
                                }
                            } else {
                                error!("<unknown error>");
                            }
                        },
                    }
                } else {
                    error!("<unknown error>");
                }
            }
        }

        // print stack trace
        for trace in self.slice().iter() {
            if let Some(id) = debug_frames.find_id(&trace.func) {
                let edl_func = funcs.get_edl_id(id)
                    .expect("failed to get EDL function id from MIR function id");
                let sig = phase.types.get_fn_signature(edl_func)
                    .expect("failed to get function signature");

                if let Some((pos, src)) = funcs.get_source_information(id) {
                    error!("    {} at {}", TypeArguments::<'_, DefaultHasher>::new(&[
                        TypeArgument::new_edl(sig)
                    ]).printable(&phase.types, &phase.vars), src.format_pos(pos));
                } else {
                    error!("    {}", TypeArguments::<'_, DefaultHasher>::new(&[
                        TypeArgument::new_edl(sig),
                    ]).printable(&phase.types, &phase.vars));
                }

                if let Some(source_frame) = debug_frames.get_source(&id) {
                    if let Some(loc) = source_frame.source_location(trace.loc) {
                        error!("        at {}", loc.src.format_pos(loc.pos));
                    } else {
                        error!("        at <unknown> {:p}", trace.loc as *const ());
                    }
                } else {
                    error!("        at <unknown> {:p}", trace.loc as *const ());
                }
            } else if trace.func as *const u8 == jit_sync_panic as *const u8 {
                error!("    jit_sync_panic (EDL runtime synchronous panic unwind hook)")
            } else {
                error!("    <unknown> {:p}", trace.func as *const ());
                error!("        at <unknown> {:p}", trace.loc as *const ());
            }
        }
    }
}

impl Index<usize> for Backtrace {
    type Output = BacktraceEntry;

    fn index(&self, index: usize) -> &Self::Output {
        self.trace.get(index).unwrap()
    }
}

struct PanicPayload {
    backtrace: Backtrace,
    reached_host: bool,
}

impl PanicPayload {
    fn clear(&mut self) {
        self.backtrace.clear();
        self.reached_host = false;
    }
}

/// Encodes information about a panic.
pub struct PanicData {
    panic: AtomicBool,
    cleaning_up: AtomicBool,
    payload: Mutex<PanicPayload>,
}

thread_local! {
    static PANIC: LazyCell<PanicData> = LazyCell::new(PanicData::new);
}

impl PanicData {
    fn new() -> PanicData {
        PanicData {
            panic: AtomicBool::new(false),
            cleaning_up: AtomicBool::new(false),
            payload: Mutex::new(PanicPayload {
                backtrace: Backtrace::new(),
                reached_host: false,
            }),
        }
    }

    /// Sets the panic data.
    /// If this function is called, the host's safety features will assume that the JIT code paniced
    /// and the stack was unrolled into this collection of panic data.
    fn set<R, F: FnOnce(&mut PanicPayload) -> R>(with: F, default: R) -> R {
        PANIC.with(|panic| {
            panic.panic.store(true, Ordering::Relaxed);
            if let Ok(mut backtrace) = panic.payload.lock() {
                backtrace.clear();
                with(&mut backtrace)
            } else {
                default
            }
        })
    }

    /// Fetches the local panic handling data from thread local storage.
    /// If there was no panic since the last time this function was called, it will return `None`.
    pub fn fetch<Runtime: 'static>(jit: &JIT<Runtime>, hir_phase: &HirPhase) -> Result<(), PanicError> {
        PANIC.with(|panic| if panic
            .panic
            .fetch_and(false, Ordering::Relaxed) {

            let payload = panic.payload.lock().unwrap();
            error!("encountered panic while executing JIT code in thread: {}",
                std::thread::current().name().unwrap_or("unknown"));
            error!("stack trace:");
            let t = payload.backtrace.get_panic_type(&jit.unwind_info)
                .unwrap_or(PanicType::Unknown);
            payload.backtrace.print(hir_phase, &*jit.func_reg.borrow(), &jit.unwind_info);
            error!("end of stack trace.");
            let recoverable = payload.reached_host && payload.backtrace.is_recoverable();
            if recoverable {
                error!("host function reached – attempting to recover through graceful panic protocol");
            } else {
                error!("host function was not reached during unwinding – graceful panic not possible");
                panic!("unrecoverable panic during JIT code execution");
            }

            let backtrace = payload.backtrace.printable(&jit.unwind_info);
            let msg = PanicMessage::take().map(|msg| msg.data);
            Err(PanicError {
                msg,
                graceful: recoverable,
                panic_type: t,
                backtrace,
            })
        } else {
            Ok(())
        })
    }
}

#[cfg_attr(feature="serde", derive(Serialize, Deserialize))]
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum PanicType {
    BadConversionToInteger,
    HeapOutOfBounds,
    IntegerOverflow,
    StackOverflow,
    DivideByZero,
    ArrayIndexOutOfBounds,
    SliceIndexOutOfBounds,
    ArrayRangeOutOfBounds,
    SliceRangeOutOfBounds,
    Assertion,
    Explicit,
    Segfault,
    Other(&'static str),
    Unknown,
}

impl Display for PanicType {
    fn fmt(&self, f: &mut Formatter<'_>) -> std::fmt::Result {
        match self {
            PanicType::BadConversionToInteger => write!(f, "bad conversion to integer"),
            PanicType::HeapOutOfBounds => write!(f, "heap out of bounds"),
            PanicType::IntegerOverflow => write!(f, "integer overflow"),
            PanicType::StackOverflow => write!(f, "stack overflow"),
            PanicType::DivideByZero => write!(f, "divide by zero"),
            PanicType::ArrayIndexOutOfBounds => write!(f, "array index out of bounds"),
            PanicType::SliceIndexOutOfBounds => write!(f, "slice index out of bounds"),
            PanicType::ArrayRangeOutOfBounds => write!(f, "array range out of bounds"),
            PanicType::SliceRangeOutOfBounds => write!(f, "slice range out of bounds"),
            PanicType::Assertion => write!(f, "assertion failed"),
            PanicType::Explicit => write!(f, "explicit panic"),
            PanicType::Segfault => write!(f, "segfault panic"),
            PanicType::Other(s) => write!(f, "{s}"),
            PanicType::Unknown => write!(f, "unknown"),
        }
    }
}

#[cfg_attr(feature="serde", derive(Serialize, Deserialize))]
#[derive(Debug)]
pub struct PanicError {
    pub msg: Option<String>,
    pub graceful: bool,
    pub panic_type: PanicType,
    pub backtrace: PrintableBacktrace,
}

impl Display for PanicError {
    fn fmt(&self, f: &mut Formatter<'_>) -> std::fmt::Result {
        write!(f, "{}", self.panic_type)?;
        if let Some(msg) = self.msg.as_ref() {
            write!(f, ": {msg}")?;
        }
        if !self.graceful {
            write!(f, " !non recoverable - aborting!")?;
        }
        Ok(())
    }
}

/// User-defined panic message, that can be passed to a manual call to core::panic.
pub struct PanicMessage {
    pub data: String,
}

thread_local! {
    static PANIC_MSG: RefCell<Option<PanicMessage>> = const { RefCell::new(None) };
}

impl PanicMessage {
    pub fn set(msg: PanicMessage) {
        PANIC_MSG.set(Some(msg));
    }

    fn take() -> Option<PanicMessage> {
        PANIC_MSG.take()
    }
}