[dev.typeparams] runtime: simplify defer record allocation

Now that deferred functions are always argumentless and defer
records are no longer with arguments, defer record can be fixed
size (just the _defer struct). This allows us to simplify the
allocation of defer records, specifically, remove the defer
classes and the pools of different sized defers.

Change-Id: Icc4b16afc23b38262ca9dd1f7369ad40874cf701
Reviewed-on: https://go-review.googlesource.com/c/go/+/326062
Trust: Cherry Mui <cherryyz@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Go Bot <gobot@golang.org>

6 files changed, 50 insertions, 142 deletions

diff --git a/src/cmd/compile/internal/test/inl_test.go b/src/cmd/compile/internal/test/inl_test.go
index 5b0db83301..bbdbe0c37c 100644
--- a/src/cmd/compile/internal/test/inl_test.go
+++ b/src/cmd/compile/internal/test/inl_test.go
@@ -42,7 +42,6 @@ func TestIntendedInlining(t *testing.T) {
 			"bucketMask",
 			"bucketShift",
 			"chanbuf",
-			"deferclass",
 			"evacuated",
 			"fastlog2",
 			"fastrand",
@@ -63,7 +62,6 @@ func TestIntendedInlining(t *testing.T) {
 			"subtract1",
 			"subtractb",
 			"tophash",
-			"totaldefersize",
 			"(*bmap).keys",
 			"(*bmap).overflow",
 			"(*waitq).enqueue",
diff --git a/src/runtime/malloc.go b/src/runtime/malloc.go
index 2759bbdaf9..c5f62483ff 100644
--- a/src/runtime/malloc.go
+++ b/src/runtime/malloc.go
@@ -420,8 +420,6 @@ func mallocinit() {
 		throw("bad TinySizeClass")
 	}
 
-	testdefersizes()
-
 	if heapArenaBitmapBytes&(heapArenaBitmapBytes-1) != 0 {
 		// heapBits expects modular arithmetic on bitmap
 		// addresses to work.
@@ -1088,15 +1086,6 @@ func mallocgc(size uintptr, typ *_type, needzero bool) unsafe.Pointer {
 
 	var scanSize uintptr
 	if !noscan {
-		// If allocating a defer+arg block, now that we've picked a malloc size
-		// large enough to hold everything, cut the "asked for" size down to
-		// just the defer header, so that the GC bitmap will record the arg block
-		// as containing nothing at all (as if it were unused space at the end of
-		// a malloc block caused by size rounding).
-		// The defer arg areas are scanned as part of scanstack.
-		if typ == deferType {
-			dataSize = unsafe.Sizeof(_defer{})
-		}
 		heapBitsSetType(uintptr(x), size, dataSize, typ)
 		if dataSize > typ.size {
 			// Array allocation. If there are any
diff --git a/src/runtime/mgc.go b/src/runtime/mgc.go
index c239fa0f63..34b5b482a3 100644
--- a/src/runtime/mgc.go
+++ b/src/runtime/mgc.go
@@ -1558,19 +1558,17 @@ func clearpools() {
 	sched.sudogcache = nil
 	unlock(&sched.sudoglock)
 
-	// Clear central defer pools.
+	// Clear central defer pool.
 	// Leave per-P pools alone, they have strictly bounded size.
 	lock(&sched.deferlock)
-	for i := range sched.deferpool {
-		// disconnect cached list before dropping it on the floor,
-		// so that a dangling ref to one entry does not pin all of them.
-		var d, dlink *_defer
-		for d = sched.deferpool[i]; d != nil; d = dlink {
-			dlink = d.link
-			d.link = nil
-		}
-		sched.deferpool[i] = nil
+	// disconnect cached list before dropping it on the floor,
+	// so that a dangling ref to one entry does not pin all of them.
+	var d, dlink *_defer
+	for d = sched.deferpool; d != nil; d = dlink {
+		dlink = d.link
+		d.link = nil
 	}
+	sched.deferpool = nil
 	unlock(&sched.deferlock)
 }
 
diff --git a/src/runtime/panic.go b/src/runtime/panic.go
index 39013163b6..86d41c4e1c 100644
--- a/src/runtime/panic.go
+++ b/src/runtime/panic.go
@@ -236,7 +236,7 @@ func deferproc(fn func()) {
 	sp := getcallersp()
 	callerpc := getcallerpc()
 
-	d := newdefer(0)
+	d := newdefer()
 	if d._panic != nil {
 		throw("deferproc: d.panic != nil after newdefer")
 	}
@@ -302,107 +302,38 @@ func deferprocStack(d *_defer) {
 	// been set and must not be clobbered.
 }
 
-// Small malloc size classes >= 16 are the multiples of 16: 16, 32, 48, 64, 80, 96, 112, 128, 144, ...
-// Each P holds a pool for defers with small arg sizes.
-// Assign defer allocations to pools by rounding to 16, to match malloc size classes.
-
-const (
-	deferHeaderSize = unsafe.Sizeof(_defer{})
-	minDeferAlloc   = (deferHeaderSize + 15) &^ 15
-	minDeferArgs    = minDeferAlloc - deferHeaderSize
-)
-
-// defer size class for arg size sz
-//go:nosplit
-func deferclass(siz uintptr) uintptr {
-	if siz <= minDeferArgs {
-		return 0
-	}
-	return (siz - minDeferArgs + 15) / 16
-}
-
-// total size of memory block for defer with arg size sz
-func totaldefersize(siz uintptr) uintptr {
-	if siz <= minDeferArgs {
-		return minDeferAlloc
-	}
-	return deferHeaderSize + siz
-}
-
-// Ensure that defer arg sizes that map to the same defer size class
-// also map to the same malloc size class.
-func testdefersizes() {
-	var m [len(p{}.deferpool)]int32
-
-	for i := range m {
-		m[i] = -1
-	}
-	for i := uintptr(0); ; i++ {
-		defersc := deferclass(i)
-		if defersc >= uintptr(len(m)) {
-			break
-		}
-		siz := roundupsize(totaldefersize(i))
-		if m[defersc] < 0 {
-			m[defersc] = int32(siz)
-			continue
-		}
-		if m[defersc] != int32(siz) {
-			print("bad defer size class: i=", i, " siz=", siz, " defersc=", defersc, "\n")
-			throw("bad defer size class")
-		}
-	}
-}
-
-var deferType *_type // type of _defer struct
-
-func init() {
-	var x interface{}
-	x = (*_defer)(nil)
-	deferType = (*(**ptrtype)(unsafe.Pointer(&x))).elem
-}
+// Each P holds a pool for defers.
 
 // Allocate a Defer, usually using per-P pool.
 // Each defer must be released with freedefer.  The defer is not
 // added to any defer chain yet.
-//
-// This must not grow the stack because there may be a frame without
-// stack map information when this is called.
-//
-//go:nosplit
-func newdefer(siz int32) *_defer {
+func newdefer() *_defer {
 	var d *_defer
-	sc := deferclass(uintptr(siz))
 	gp := getg()
-	if sc < uintptr(len(p{}.deferpool)) {
-		pp := gp.m.p.ptr()
-		if len(pp.deferpool[sc]) == 0 && sched.deferpool[sc] != nil {
-			// Take the slow path on the system stack so
-			// we don't grow newdefer's stack.
-			systemstack(func() {
-				lock(&sched.deferlock)
-				for len(pp.deferpool[sc]) < cap(pp.deferpool[sc])/2 && sched.deferpool[sc] != nil {
-					d := sched.deferpool[sc]
-					sched.deferpool[sc] = d.link
-					d.link = nil
-					pp.deferpool[sc] = append(pp.deferpool[sc], d)
-				}
-				unlock(&sched.deferlock)
-			})
-		}
-		if n := len(pp.deferpool[sc]); n > 0 {
-			d = pp.deferpool[sc][n-1]
-			pp.deferpool[sc][n-1] = nil
-			pp.deferpool[sc] = pp.deferpool[sc][:n-1]
-		}
-	}
-	if d == nil {
-		// Allocate new defer+args.
+	pp := gp.m.p.ptr()
+	if len(pp.deferpool) == 0 && sched.deferpool != nil {
+		// Take the slow path on the system stack so
+		// we don't grow newdefer's stack.
 		systemstack(func() {
-			total := roundupsize(totaldefersize(uintptr(siz)))
-			d = (*_defer)(mallocgc(total, deferType, true))
+			lock(&sched.deferlock)
+			for len(pp.deferpool) < cap(pp.deferpool)/2 && sched.deferpool != nil {
+				d := sched.deferpool
+				sched.deferpool = d.link
+				d.link = nil
+				pp.deferpool = append(pp.deferpool, d)
+			}
+			unlock(&sched.deferlock)
 		})
 	}
+	if n := len(pp.deferpool); n > 0 {
+		d = pp.deferpool[n-1]
+		pp.deferpool[n-1] = nil
+		pp.deferpool = pp.deferpool[:n-1]
+	}
+	if d == nil {
+		// Allocate new defer.
+		d = new(_defer)
+	}
 	d.heap = true
 	return d
 }
@@ -424,23 +355,19 @@ func freedefer(d *_defer) {
 	if !d.heap {
 		return
 	}
-	sc := deferclass(0)
-	if sc >= uintptr(len(p{}.deferpool)) {
-		return
-	}
 	pp := getg().m.p.ptr()
-	if len(pp.deferpool[sc]) == cap(pp.deferpool[sc]) {
+	if len(pp.deferpool) == cap(pp.deferpool) {
 		// Transfer half of local cache to the central cache.
 		//
 		// Take this slow path on the system stack so
 		// we don't grow freedefer's stack.
 		systemstack(func() {
 			var first, last *_defer
-			for len(pp.deferpool[sc]) > cap(pp.deferpool[sc])/2 {
-				n := len(pp.deferpool[sc])
-				d := pp.deferpool[sc][n-1]
-				pp.deferpool[sc][n-1] = nil
-				pp.deferpool[sc] = pp.deferpool[sc][:n-1]
+			for len(pp.deferpool) > cap(pp.deferpool)/2 {
+				n := len(pp.deferpool)
+				d := pp.deferpool[n-1]
+				pp.deferpool[n-1] = nil
+				pp.deferpool = pp.deferpool[:n-1]
 				if first == nil {
 					first = d
 				} else {
@@ -449,8 +376,8 @@ func freedefer(d *_defer) {
 				last = d
 			}
 			lock(&sched.deferlock)
-			last.link = sched.deferpool[sc]
-			sched.deferpool[sc] = first
+			last.link = sched.deferpool
+			sched.deferpool = first
 			unlock(&sched.deferlock)
 		})
 	}
@@ -469,7 +396,7 @@ func freedefer(d *_defer) {
 	// both of which throw.
 	d.link = nil
 
-	pp.deferpool[sc] = append(pp.deferpool[sc], d)
+	pp.deferpool = append(pp.deferpool, d)
 }
 
 // Separate function so that it can split stack.
@@ -720,7 +647,7 @@ func addOneOpenDeferFrame(gp *g, pc uintptr, sp unsafe.Pointer) {
 					throw("missing deferreturn")
 				}
 
-				d1 := newdefer(0)
+				d1 := newdefer()
 				d1.openDefer = true
 				d1._panic = nil
 				// These are the pc/sp to set after we've
diff --git a/src/runtime/proc.go b/src/runtime/proc.go
index d6f3af690b..4a116130a5 100644
--- a/src/runtime/proc.go
+++ b/src/runtime/proc.go
@@ -4784,9 +4784,7 @@ func (pp *p) init(id int32) {
 	pp.id = id
 	pp.status = _Pgcstop
 	pp.sudogcache = pp.sudogbuf[:0]
-	for i := range pp.deferpool {
-		pp.deferpool[i] = pp.deferpoolbuf[i][:0]
-	}
+	pp.deferpool = pp.deferpoolbuf[:0]
 	pp.wbBuf.reset()
 	if pp.mcache == nil {
 		if id == 0 {
@@ -4864,12 +4862,10 @@ func (pp *p) destroy() {
 		pp.sudogbuf[i] = nil
 	}
 	pp.sudogcache = pp.sudogbuf[:0]
-	for i := range pp.deferpool {
-		for j := range pp.deferpoolbuf[i] {
-			pp.deferpoolbuf[i][j] = nil
-		}
-		pp.deferpool[i] = pp.deferpoolbuf[i][:0]
+	for j := range pp.deferpoolbuf {
+		pp.deferpoolbuf[j] = nil
 	}
+	pp.deferpool = pp.deferpoolbuf[:0]
 	systemstack(func() {
 		for i := 0; i < pp.mspancache.len; i++ {
 			// Safe to call since the world is stopped.
diff --git a/src/runtime/runtime2.go b/src/runtime/runtime2.go
index cf4b0bff43..75c4818599 100644
--- a/src/runtime/runtime2.go
+++ b/src/runtime/runtime2.go
@@ -613,8 +613,8 @@ type p struct {
 	pcache      pageCache
 	raceprocctx uintptr
 
-	deferpool    [5][]*_defer // pool of available defer structs of different sizes (see panic.go)
-	deferpoolbuf [5][32]*_defer
+	deferpool    []*_defer // pool of available defer structs (see panic.go)
+	deferpoolbuf [32]*_defer
 
 	// Cache of goroutine ids, amortizes accesses to runtime·sched.goidgen.
 	goidcache    uint64
@@ -801,9 +801,9 @@ type schedt struct {
 	sudoglock  mutex
 	sudogcache *sudog
 
-	// Central pool of available defer structs of different sizes.
+	// Central pool of available defer structs.
 	deferlock mutex
-	deferpool [5]*_defer
+	deferpool *_defer
 
 	// freem is the list of m's waiting to be freed when their
 	// m.exited is set. Linked through m.freelink.