On Tue, Nov 15, 2016 at 11:03:59AM +0100, Ingo Molnar wrote:
> > Should I also make a CONFIG knob that implements refcount_t with the
> > 'normal' atomic_t primitives?
>
> I'd suggest doing the saturation/safe-wrap semantics only for now (i.e. the
> current patch, split into two perhaps), and reconsider if there's any
> complaints?
>
> > And possibly another knob to toggle the BUG()s into WARN()s. With the
> > full saturation semantics WARN() is a lot safer and will not corrupt
> > kernel state as much.
>
> I'd suggest changing it to a WARN() straight away, no extra knobs.
OK, a little like so then? Note that the overflow tests went away
because increments guarantee we saturate before we overflow.
---
Subject: refcount_t: A special purpose refcount type
From: Peter Zijlstra <pet...@infradead.org>
Date: Mon Nov 14 18:06:19 CET 2016
Provide refcount_t, an atomic_t like primitive built just for
refcounting.
It provides saturation semantics such that overflow becomes impossible
and thereby 'spurious' use-after-free is avoided.
Signed-off-by: Peter Zijlstra (Intel) <pet...@infradead.org>
---
include/linux/refcount.h | 229 +++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 229 insertions(+)
--- /dev/null
+++ b/include/linux/refcount.h
@@ -0,0 +1,229 @@
+#ifndef _LINUX_REFCOUNT_H
+#define _LINUX_REFCOUNT_H
+
+/*
+ * Variant of atomic_t specialized for refcounting.
+ *
+ * The interface matches the atomic_t interface (to aid in porting) but only
+ * provides the few functions one should use for refcounting.
+ *
+ * It differs in that the counter saturates at UINT_MAX and will not move once
+ * there. This avoids wrapping the counter and causing 'spurious'
+ * use-after-free issues.
+ *
+ * Memory ordering rules are slightly relaxed wrt regular atomic_t functions
+ * and provide only what is strictly required for refcounts.
+ *
+ * The increments are fully relaxed; these will not provide ordering. The
+ * rationale is that whatever is used to obtain the object we're increasing the
+ * reference count on will provide the ordering. For locked data structures,
+ * its the lock acquire, for RCU/lockless data structures its the dependent
+ * load.
+ *
+ * Do note that inc_not_zero() provides a control dependency which will order
+ * future stores against the inc, this ensures we'll never modify the object
+ * if we did not in fact acquire a reference.
+ *
+ * The decrements will provide release order, such that all the prior loads and
+ * stores will be issued before we proceed with freeing the object.
+ *
+ * Note: the implementation hard relies on increments, bigger than 1 additions
+ * need explicit overflow -> saturation logic.
+ *
+ */
+
+#include <linux/atomic.h>
+#include <linux/bug.h>
+#include <linux/mutex.h>
+#include <linux/spinlock.h>
+
+typedef struct refcount_struct {
+ atomic_t refs;
+} refcount_t;
+
+#define REFCOUNT_INIT(n) { .refs = ATOMIC_INIT(n), }
+
+static inline void refcount_set(refcount_t *r, int n)
+{
+ atomic_set(&r->refs, n);
+}
+
+static inline unsigned int refcount_read(const refcount_t *r)
+{
+ return atomic_read(&r->refs);
+}
+
+/*
+ * Similar to atomic_inc(), will saturate at UINT_MAX and WARN.
+ *
+ * Provides no memory ordering, it is assumed the caller already has a
+ * reference on the object, will WARN when this is not so.
+ */
+static inline void refcount_inc(refcount_t *r)
+{
+ unsigned int old, new, val = atomic_read(&r->refs);
+
+ for (;;) {
+ WARN(!val, "refcount_t: increment on 0; use-after-free.\n");
+
+ if (unlikely(val == UINT_MAX))
+ return;
+
+ new = val + 1;
+ old = atomic_cmpxchg_relaxed(&r->refs, val, new);
+ if (old == val)
+ break;
+
+ val = old;
+ }
+
+ WARN(new == UINT_MAX, "refcount_t: saturated; leaking memory.\n");
+}
+
+/*
+ * Similar to atomic_inc_not_zero(), will saturate at UINT_MAX and WARN.
+ *
+ * Provides no memory ordering, it is assumed the caller has guaranteed the
+ * object memory to be stable (RCU, etc.). It does provide a control dependency
+ * and thereby orders future stores.
+ */
+static inline __must_check
+bool refcount_inc_not_zero(refcount_t *r)
+{
+ unsigned int old, new, val = atomic_read(&r->refs);
+
+ for (;;) {
+ if (!val)
+ return false;
+
+ if (unlikely(val == UINT_MAX))
+ return true;
+
+ new = val + 1;
+ old = atomic_cmpxchg_relaxed(&r->refs, val, new);
+ if (old == val)
+ break;
+
+ val = old;
+ }
+
+ WARN(new == UINT_MAX, "refcount_t: saturated; leaking memory.\n");
+
+ return true;
+}
+
+/*
+ * Similar to atomic_dec_and_test(), it will WARN on underflow and fail to
+ * decrement when saturated at UINT_MAX.
+ *
+ * Provides release memory ordering, such that prior loads and stores are done
+ * before a subsequent free.
+ */
+static inline __must_check
+bool refcount_dec_and_test(refcount_t *r)
+{
+ unsigned int old, new, val = atomic_read(&r->refs);
+
+ for (;;) {
+ if (val == UINT_MAX)
+ return false;
+
+ new = val - 1;
+ if (WARN(new > val, "refcount_t: underflow; use-after-free.\n"))
+ return false;
+
+ old = atomic_cmpxchg_release(&r->refs, val, new);
+ if (old == val)
+ break;
+
+ val = old;
+ }
+
+ return !new;
+}
+
+/*
+ * Similar to atomic_dec_and_mutex_lock(), it will WARN on underflow and fail
+ * to decrement when saturated at UINT_MAX.
+ *
+ * Provides release memory ordering, such that prior loads and stores are done
+ * before a subsequent free. This allows free() while holding the mutex.
+ */
+static inline __must_check
+bool refcount_dec_and_mutex_lock(refcount_t *r, struct mutex *lock)
+{
+ unsigned int old, new, val = atomic_read(&r->refs);
+ bool locked = false;
+
+ for (;;) {
+ if (val == UINT_MAX)
+ return false;
+
+ if (val == 1 && !locked) {
+ locked = true;
+ mutex_lock(lock);
+ }
+
+ new = val - 1;
+ if (WARN(new > val, "refcount_t: underflow;
use-after-free.\n")) {
+ if (locked)
+ mutex_unlock(lock);
+ return false;
+ }
+
+ old = atomic_cmpxchg_release(&r->refs, val, new);
+ if (old == val)
+ break;
+
+ val = old;
+ }
+
+ if (new && locked)
+ mutex_unlock(lock);
+
+ return !new;
+}
+
+/*
+ * Similar to atomic_dec_and_lock(), it will WARN on underflow and fail to
+ * decrement when saturated at UINT_MAX.
+ *
+ * Provides release memory ordering, such that prior loads and stores are done
+ * before a subsequent free. This allows free() while holding the lock.
+ */
+static inline __must_check
+bool refcount_dec_and_lock(refcount_t *r, spinlock_t *lock)
+{
+ unsigned int old, new, val = atomic_read(&r->refs);
+ bool locked = false;
+
+ for (;;) {
+ if (val == UINT_MAX)
+ return false;
+
+ if (val == 1 && !locked) {
+ locked = true;
+ spin_lock(lock);
+ }
+
+ new = val - 1;
+ if (WARN(new > val, "refcount_t: underflow;
use-after-free.\n")) {
+ if (locked)
+ mutex_unlock(lock);
+ return false;
+ }
+
+ old = atomic_cmpxchg_release(&r->refs, val, new);
+ if (old == val)
+ break;
+
+ val = old;
+ }
+
+ if (new && locked)
+ spin_unlock(lock);
+
+ return !new;
+}
+
+#endif /* _LINUX_REFCOUNT_H */
