1 files changed, 83 insertions, 18 deletions

diff --git a/compiler/rustc_driver_impl/src/signal_handler.rs b/compiler/rustc_driver_impl/src/signal_handler.rs
index 1bf5c5af866..5971562627c 100644
--- a/compiler/rustc_driver_impl/src/signal_handler.rs
+++ b/compiler/rustc_driver_impl/src/signal_handler.rs
@@ -2,43 +2,108 @@
 //! Primarily used to extract a backtrace from stack overflow
 
 use std::alloc::{alloc, Layout};
-use std::{mem, ptr};
+use std::{fmt, mem, ptr};
 
 extern "C" {
     fn backtrace_symbols_fd(buffer: *const *mut libc::c_void, size: libc::c_int, fd: libc::c_int);
 }
 
+fn backtrace_stderr(buffer: &[*mut libc::c_void]) {
+    let size = buffer.len().try_into().unwrap_or_default();
+    unsafe { backtrace_symbols_fd(buffer.as_ptr(), size, libc::STDERR_FILENO) };
+}
+
+/// Unbuffered, unsynchronized writer to stderr.
+///
+/// Only acceptable because everything will end soon anyways.
+struct RawStderr(());
+
+impl fmt::Write for RawStderr {
+    fn write_str(&mut self, s: &str) -> Result<(), fmt::Error> {
+        let ret = unsafe { libc::write(libc::STDERR_FILENO, s.as_ptr().cast(), s.len()) };
+        if ret == -1 { Err(fmt::Error) } else { Ok(()) }
+    }
+}
+
+/// We don't really care how many bytes we actually get out. SIGSEGV comes for our head.
+/// Splash stderr with letters of our own blood to warn our friends about the monster.
+macro raw_errln($tokens:tt) {
+    let _ = ::core::fmt::Write::write_fmt(&mut RawStderr(()), format_args!($tokens));
+    let _ = ::core::fmt::Write::write_char(&mut RawStderr(()), '\n');
+}
+
 /// Signal handler installed for SIGSEGV
 extern "C" fn print_stack_trace(_: libc::c_int) {
     const MAX_FRAMES: usize = 256;
     // Reserve data segment so we don't have to malloc in a signal handler, which might fail
     // in incredibly undesirable and unexpected ways due to e.g. the allocator deadlocking
     static mut STACK_TRACE: [*mut libc::c_void; MAX_FRAMES] = [ptr::null_mut(); MAX_FRAMES];
-    unsafe {
+    let stack = unsafe {
         // Collect return addresses
         let depth = libc::backtrace(STACK_TRACE.as_mut_ptr(), MAX_FRAMES as i32);
         if depth == 0 {
             return;
         }
-        // Just a stack trace is cryptic. Explain what we're doing.
-        write_raw_err("error: rustc interrupted by SIGSEGV, printing stack trace:\n\n");
-        // Elaborate return addrs into symbols and write them directly to stderr
-        backtrace_symbols_fd(STACK_TRACE.as_ptr(), depth, libc::STDERR_FILENO);
-        if depth > 22 {
-            // We probably just scrolled that "we got SIGSEGV" message off the terminal
-            write_raw_err("\nerror: stack trace dumped due to SIGSEGV, possible stack overflow");
+        &STACK_TRACE.as_slice()[0..(depth as _)]
+    };
+
+    // Just a stack trace is cryptic. Explain what we're doing.
+    raw_errln!("error: rustc interrupted by SIGSEGV, printing backtrace\n");
+    let mut written = 1;
+    let mut consumed = 0;
+    // Begin elaborating return addrs into symbols and writing them directly to stderr
+    // Most backtraces are stack overflow, most stack overflows are from recursion
+    // Check for cycles before writing 250 lines of the same ~5 symbols
+    let cycled = |(runner, walker)| runner == walker;
+    let mut cyclic = false;
+    if let Some(period) = stack.iter().skip(1).step_by(2).zip(stack).position(cycled) {
+        let period = period.saturating_add(1); // avoid "what if wrapped?" branches
+        let Some(offset) = stack.iter().skip(period).zip(stack).position(cycled) else {
+            // impossible.
+            return;
+        };
+
+        // Count matching trace slices, else we could miscount "biphasic cycles"
+        // with the same period + loop entry but a different inner loop
+        let next_cycle = stack[offset..].chunks_exact(period).skip(1);
+        let cycles = 1 + next_cycle
+            .zip(stack[offset..].chunks_exact(period))
+            .filter(|(next, prev)| next == prev)
+            .count();
+        backtrace_stderr(&stack[..offset]);
+        written += offset;
+        consumed += offset;
+        if cycles > 1 {
+            raw_errln!("\n### cycle encountered after {offset} frames with period {period}");
+            backtrace_stderr(&stack[consumed..consumed + period]);
+            raw_errln!("### recursed {cycles} times\n");
+            written += period + 4;
+            consumed += period * cycles;
+            cyclic = true;
         };
-        write_raw_err("\nerror: please report a bug to https://github.com/rust-lang/rust\n");
     }
-}
+    let rem = &stack[consumed..];
+    backtrace_stderr(rem);
+    raw_errln!("");
+    written += rem.len() + 1;
 
-/// Write without locking stderr.
-///
-/// Only acceptable because everything will end soon anyways.
-fn write_raw_err(input: &str) {
-    // We do not care how many bytes we actually get out. SIGSEGV comes for our head.
-    // Splash stderr with letters of our own blood to warn our friends about the monster.
-    let _ = unsafe { libc::write(libc::STDERR_FILENO, input.as_ptr().cast(), input.len()) };
+    if cyclic || stack.len() == 256 {
+        // technically speculation, but assert it with confidence anyway.
+        // rustc only arrived in this signal handler because bad things happened
+        // and this message is for explaining it's not the programmer's fault
+        raw_errln!("note: rustc unexpectedly overflowed its stack! this is a bug");
+        written += 1;
+    }
+    if stack.len() == 256 {
+        raw_errln!("note: maximum backtrace depth reached, frames may have been lost");
+        written += 1;
+    }
+    raw_errln!("note: we would appreciate a report at https://github.com/rust-lang/rust");
+    written += 1;
+    if written > 24 {
+        // We probably just scrolled the earlier "we got SIGSEGV" message off the terminal
+        raw_errln!("note: backtrace dumped due to SIGSEGV! resuming signal");
+    };
 }
 
 /// When SIGSEGV is delivered to the process, print a stack trace and then exit.