Skip to contentSkip to conversation

ucount: Fix atomic_long_inc_below() argument type

PATCH RESEND Applied Uros Bizjak 3 participants 5 messages Jul 21, 2025 – Jul 22, 2025

lore
Share & tools
lore
#0 lore raw

PATCH RESEND v2 1/2 ucount: Fix atomic_long_inc_below() argument type

The type of u argument of atomic_long_inc_below() should be long to avoid unwanted truncation to int.

Fixes f9c82a4ea89c ("Increase size of ucounts to atomic_long_t")
Cc Andrew Morton <[email protected]>
Cc Sebastian Andrzej Siewior <[email protected]>
Cc "Paul E. McKenney" <[email protected]>
Cc Alexey Gladkov <[email protected]>
Cc Roman Gushchin <[email protected]>
Cc MengEn Sun <[email protected]>
Cc "Thomas Weißschuh" <[email protected]>
---
 kernel/ucount.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
kernel/ucount.c +1 -1
index 8686e329b8f2..f629db485a07 100644
@@ -199,7 +199,7 @@ void put_ucounts(struct ucounts *ucounts)
 	}
 }
 
-static inline bool atomic_long_inc_below(atomic_long_t *v, int u)
+static inline bool atomic_long_inc_below(atomic_long_t *v, long u)
 {
 	long c, old;
 	c = atomic_long_read(v);
Signature · 1 line
2.50.1
#1 lore raw

PATCH RESEND v2 2/2 ucount: Use atomic_long_try_cmpxchg() in atomic_long_inc_below()

Use atomic_long_try_cmpxchg() instead of
atomic_long_cmpxchg (*ptr, old, new) == old in atomic_long_inc_below(). x86 CMPXCHG instruction returns success in ZF flag, so this change saves a compare after cmpxchg (and related move instruction in front of cmpxchg).

Also, atomic_long_try_cmpxchg implicitly assigns old *ptr value to "old" when cmpxchg fails, enabling further code simplifications.

No functional change intended.

Cc Andrew Morton <[email protected]>
Cc Sebastian Andrzej Siewior <[email protected]>
Cc "Paul E. McKenney" <[email protected]>
Cc Alexey Gladkov <[email protected]>
Cc Roman Gushchin <[email protected]>
Cc MengEn Sun <[email protected]>
Cc "Thomas Weißschuh" <[email protected]>
---
 kernel/ucount.c | 14 ++++++--------
 1 file changed, 6 insertions(+), 8 deletions(-)
kernel/ucount.c +6 -8
index f629db485a07..586af49fc03e 100644
@@ -201,16 +201,14 @@ void put_ucounts(struct ucounts *ucounts)
 
 static inline bool atomic_long_inc_below(atomic_long_t *v, long u)
 {
-	long c, old;
-	c = atomic_long_read(v);
-	for (;;) {
+	long c = atomic_long_read(v);
+
+	do {
 		if (unlikely(c >= u))
 			return false;
-		old = atomic_long_cmpxchg(v, c, c+1);
-		if (likely(old == c))
-			return true;
-		c = old;
-	}
+	} while (!atomic_long_try_cmpxchg(v, &c, c+1));
+
+	return true;
 }
 
 struct ucounts *inc_ucount(struct user_namespace *ns, kuid_t uid,
Signature · 1 line
2.50.1
#2 lore raw
In reply to Uros Bizjak
Quoted 4 lines · On Mon, 21 Jul 2025 19:45:57 +0200 Uros Bizjak <[email protected]> wrote:

The type of u argument of atomic_long_inc_below() should be long to avoid unwanted truncation to int.

Fixes f9c82a4ea89c ("Increase size of ucounts to atomic_long_t")

Please (always!) provide a description of the userspace-visible effects of the bug. That way I (and others) can decide whether the fix should be backported. And people will be able to determine whether this patch may fix problems which they are observing. Thanks.

#3 lore raw
Quoted 6 lines · Andrew Morton
Quoted 3 lines · On Mon, 21 Jul 2025 19:45:57 +0200 Uros Bizjak <[email protected]> wrote:

The type of u argument of atomic_long_inc_below() should be long to avoid unwanted truncation to int.

Quoted 6 lines
Quoted 1 line
Fixes f9c82a4ea89c ("Increase size of ucounts to atomic_long_t")

Please (always!) provide a description of the userspace-visible effects of the bug. That way I (and others) can decide whether the fix should be backported. And people will be able to determine whether this patch may fix problems which they are observing. Thanks.

The patch fixes the wrong argument type of an internal function to prevent unwanted argument truncation. It fixes an internal locking primitive; it should not have any direct effect on userspace.

Uros.

#4 lore raw
Quoted 10 lines · Uros Bizjak
Quoted 8 lines · Andrew Morton
Quoted 4 lines · On Mon, 21 Jul 2025 19:45:57 +0200 Uros Bizjak <[email protected]> wrote:

The type of u argument of atomic_long_inc_below() should be long to avoid unwanted truncation to int.

Fixes f9c82a4ea89c ("Increase size of ucounts to atomic_long_t")
Quoted 8 lines
Quoted 4 lines

Please (always!) provide a description of the userspace-visible effects of the bug. That way I (and others) can decide whether the fix should be backported. And people will be able to determine whether this patch may fix problems which they are observing. Thanks.

The patch fixes the wrong argument type of an internal function to prevent unwanted argument truncation. It fixes an internal locking primitive; it should not have any direct effect on userspace.

AFAICT there's no problem in practice because atomic_long_inc_below() is only used by inc_ucount(), and it looks like the value is constrained between 0 and INT_MAX.

In inc_ucount() the limit value is taken from user_namespace::ucount_max[], and AFAICT that's only written by sysctls, to the table setup by setup_userns_sysctls(), where UCOUNT_ENTRY() limits the value between 0 and INT_MAX.

This is certainly a cleanup, but there might be no functional issue in practice as above.

Mark.