[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <171258184044.10875.9117228310710193606.tip-bot2@tip-bot2>
Date: Mon, 08 Apr 2024 13:10:40 -0000
From: "tip-bot2 for Adrian Hunter" <tip-bot2@...utronix.de>
To: linux-tip-commits@...r.kernel.org
Cc: Thomas Gleixner <tglx@...utronix.de>,
Adrian Hunter <adrian.hunter@...el.com>, x86@...nel.org,
linux-kernel@...r.kernel.org
Subject: [tip: timers/core] vdso: Make delta calculation overflow safe
The following commit has been merged into the timers/core branch of tip:
Commit-ID: 456e3788bc7164c1c8298045e04068b8e3d8e413
Gitweb: https://git.kernel.org/tip/456e3788bc7164c1c8298045e04068b8e3d8e413
Author: Adrian Hunter <adrian.hunter@...el.com>
AuthorDate: Mon, 25 Mar 2024 08:40:11 +02:00
Committer: Thomas Gleixner <tglx@...utronix.de>
CommitterDate: Mon, 08 Apr 2024 15:03:07 +02:00
vdso: Make delta calculation overflow safe
Kernel timekeeping is designed to keep the change in cycles (since the last
timer interrupt) below max_cycles, which prevents multiplication overflow
when converting cycles to nanoseconds. However, if timer interrupts stop,
the calculation will eventually overflow.
Add protection against that, enabled by config option
CONFIG_GENERIC_VDSO_OVERFLOW_PROTECT. Check against max_cycles, falling
back to a slower higher precision calculation.
Suggested-by: Thomas Gleixner <tglx@...utronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@...el.com>
Signed-off-by: Thomas Gleixner <tglx@...utronix.de>
Link: https://lore.kernel.org/r/20240325064023.2997-8-adrian.hunter@intel.com
---
lib/vdso/gettimeofday.c | 17 ++++++++++++++++-
1 file changed, 16 insertions(+), 1 deletion(-)
diff --git a/lib/vdso/gettimeofday.c b/lib/vdso/gettimeofday.c
index 9fa90e0..9c3a8d2 100644
--- a/lib/vdso/gettimeofday.c
+++ b/lib/vdso/gettimeofday.c
@@ -13,6 +13,18 @@
# define VDSO_DELTA_MASK(vd) (vd->mask)
#endif
+#ifdef CONFIG_GENERIC_VDSO_OVERFLOW_PROTECT
+static __always_inline bool vdso_delta_ok(const struct vdso_data *vd, u64 delta)
+{
+ return delta < vd->max_cycles;
+}
+#else
+static __always_inline bool vdso_delta_ok(const struct vdso_data *vd, u64 delta)
+{
+ return true;
+}
+#endif
+
#ifndef vdso_shift_ns
static __always_inline u64 vdso_shift_ns(u64 ns, u32 shift)
{
@@ -28,7 +40,10 @@ static __always_inline u64 vdso_calc_ns(const struct vdso_data *vd, u64 cycles,
{
u64 delta = (cycles - vd->cycle_last) & VDSO_DELTA_MASK(vd);
- return vdso_shift_ns((delta * vd->mult) + base, vd->shift);
+ if (likely(vdso_delta_ok(vd, delta)))
+ return vdso_shift_ns((delta * vd->mult) + base, vd->shift);
+
+ return mul_u64_u32_add_u64_shr(delta, vd->mult, base, vd->shift);
}
#endif /* vdso_calc_ns */
Powered by blists - more mailing lists