From 86101b083ad14bb0c6ca9c55d2869cba57760046 Mon Sep 17 00:00:00 2001
From: Damien Neil <dneil@google.com>
Date: Fri, 21 Feb 2025 10:55:32 -0800
Subject: [PATCH] runtime: print stack traces for bubbled goroutines on
 synctest deadlock

When synctest.Run panics due to every goroutine in the bubble being
blocked, print a stack trace for every goroutine in the bubble.

For #67434

Change-Id: Ie751c2ee6fa136930b18f4bee0277ff30da46905
Reviewed-on: https://go-review.googlesource.com/c/go/+/645719
Auto-Submit: Damien Neil <dneil@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
---
 src/runtime/panic.go     | 29 +++++++++++++++++++++++------
 src/runtime/synctest.go  | 10 +++++++++-
 src/runtime/traceback.go |  6 +++++-
 3 files changed, 37 insertions(+), 8 deletions(-)

diff --git a/src/runtime/panic.go b/src/runtime/panic.go
index 281fe04bca..706f9879dc 100644
--- a/src/runtime/panic.go
+++ b/src/runtime/panic.go
@@ -1272,7 +1272,7 @@ func fatalthrow(t throwType) {
 
 		startpanic_m()
 
-		if dopanic_m(gp, pc, sp) {
+		if dopanic_m(gp, pc, sp, nil) {
 			// crash uses a decent amount of nosplit stack and we're already
 			// low on stack in throw, so crash on the system stack (unlike
 			// fatalpanic).
@@ -1310,7 +1310,14 @@ func fatalpanic(msgs *_panic) {
 			printpanics(msgs)
 		}
 
-		docrash = dopanic_m(gp, pc, sp)
+		// If this panic is the result of a synctest bubble deadlock,
+		// print stacks for the goroutines in the bubble.
+		var sg *synctestGroup
+		if de, ok := msgs.arg.(synctestDeadlockError); ok {
+			sg = de.sg
+		}
+
+		docrash = dopanic_m(gp, pc, sp, sg)
 	})
 
 	if docrash {
@@ -1392,7 +1399,8 @@ var deadlock mutex
 
 // gp is the crashing g running on this M, but may be a user G, while getg() is
 // always g0.
-func dopanic_m(gp *g, pc, sp uintptr) bool {
+// If sg is non-nil, print the stacks for goroutines in this group as well.
+func dopanic_m(gp *g, pc, sp uintptr, sg *synctestGroup) bool {
 	if gp.sig != 0 {
 		signame := signame(gp.sig)
 		if signame != "" {
@@ -1416,10 +1424,19 @@ func dopanic_m(gp *g, pc, sp uintptr) bool {
 			print("\nruntime stack:\n")
 			traceback(pc, sp, 0, gp)
 		}
-		if !didothers && all {
-			didothers = true
-			tracebackothers(gp)
+		if !didothers {
+			if all {
+				didothers = true
+				tracebackothers(gp)
+			} else if sg != nil {
+				// This panic is caused by a synctest bubble deadlock.
+				// Print stacks for goroutines in the deadlocked bubble.
+				tracebacksomeothers(gp, func(other *g) bool {
+					return sg == other.syncGroup
+				})
+			}
 		}
+
 	}
 	unlock(&paniclk)
 
diff --git a/src/runtime/synctest.go b/src/runtime/synctest.go
index f2ac6ab5c7..b197758ad9 100644
--- a/src/runtime/synctest.go
+++ b/src/runtime/synctest.go
@@ -220,7 +220,7 @@ func synctestRun(f func()) {
 		raceacquireg(gp, gp.syncGroup.raceaddr())
 	}
 	if total != 1 {
-		panic("deadlock: all goroutines in bubble are blocked")
+		panic(synctestDeadlockError{sg})
 	}
 	if gp.timer != nil && gp.timer.isFake {
 		// Verify that we haven't marked this goroutine's sleep timer as fake.
@@ -229,6 +229,14 @@ func synctestRun(f func()) {
 	}
 }
 
+type synctestDeadlockError struct {
+	sg *synctestGroup
+}
+
+func (synctestDeadlockError) Error() string {
+	return "deadlock: all goroutines in bubble are blocked"
+}
+
 func synctestidle_c(gp *g, _ unsafe.Pointer) bool {
 	lock(&gp.syncGroup.mu)
 	canIdle := true
diff --git a/src/runtime/traceback.go b/src/runtime/traceback.go
index 2ba05395d2..1390e8b1bd 100644
--- a/src/runtime/traceback.go
+++ b/src/runtime/traceback.go
@@ -1261,6 +1261,10 @@ func goroutineheader(gp *g) {
 }
 
 func tracebackothers(me *g) {
+	tracebacksomeothers(me, func(*g) bool { return true })
+}
+
+func tracebacksomeothers(me *g, showf func(*g) bool) {
 	level, _, _ := gotraceback()
 
 	// Show the current goroutine first, if we haven't already.
@@ -1279,7 +1283,7 @@ func tracebackothers(me *g) {
 	// against concurrent creation of new Gs, but even with allglock we may
 	// miss Gs created after this loop.
 	forEachGRace(func(gp *g) {
-		if gp == me || gp == curgp || readgstatus(gp) == _Gdead || isSystemGoroutine(gp, false) && level < 2 {
+		if gp == me || gp == curgp || readgstatus(gp) == _Gdead || !showf(gp) || (isSystemGoroutine(gp, false) && level < 2) {
 			return
 		}
 		print("\n")
-- 
2.51.0