[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20240308131512.44324-3-adrian.hunter@intel.com>
Date: Fri, 8 Mar 2024 15:14:55 +0200
From: Adrian Hunter <adrian.hunter@...el.com>
To: Thomas Gleixner <tglx@...utronix.de>
Cc: Michael Ellerman <mpe@...erman.id.au>,
Nicholas Piggin <npiggin@...il.com>,
Christophe Leroy <christophe.leroy@...roup.eu>,
"Aneesh Kumar K.V" <aneesh.kumar@...nel.org>,
"Naveen N. Rao" <naveen.n.rao@...ux.ibm.com>,
Heiko Carstens <hca@...ux.ibm.com>,
Vasily Gorbik <gor@...ux.ibm.com>,
Alexander Gordeev <agordeev@...ux.ibm.com>,
Christian Borntraeger <borntraeger@...ux.ibm.com>,
Sven Schnelle <svens@...ux.ibm.com>,
Ingo Molnar <mingo@...hat.com>,
Borislav Petkov <bp@...en8.de>,
Dave Hansen <dave.hansen@...ux.intel.com>,
x86@...nel.org,
"H. Peter Anvin" <hpa@...or.com>,
Andy Lutomirski <luto@...nel.org>,
Vincenzo Frascino <vincenzo.frascino@....com>,
John Stultz <jstultz@...gle.com>,
Stephen Boyd <sboyd@...nel.org>,
Peter Zijlstra <peterz@...radead.org>,
Randy Dunlap <rdunlap@...radead.org>,
Bjorn Helgaas <bhelgaas@...gle.com>,
Arnd Bergmann <arnd@...db.de>,
Anna-Maria Behnsen <anna-maria@...utronix.de>,
linuxppc-dev@...ts.ozlabs.org,
linux-kernel@...r.kernel.org,
linux-s390@...r.kernel.org
Subject: [PATCH 02/19] vdso: Consolidate nanoseconds calculation
Consolidate nanoseconds calculation to simplify and reduce code
duplication.
Suggested-by: Thomas Gleixner <tglx@...utronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@...el.com>
---
arch/x86/include/asm/vdso/gettimeofday.h | 17 +++++----
lib/vdso/gettimeofday.c | 44 +++++++++++-------------
2 files changed, 29 insertions(+), 32 deletions(-)
diff --git a/arch/x86/include/asm/vdso/gettimeofday.h b/arch/x86/include/asm/vdso/gettimeofday.h
index 8e048ca980df..5727dedd3549 100644
--- a/arch/x86/include/asm/vdso/gettimeofday.h
+++ b/arch/x86/include/asm/vdso/gettimeofday.h
@@ -300,7 +300,7 @@ static inline bool arch_vdso_cycles_ok(u64 cycles)
#define vdso_cycles_ok arch_vdso_cycles_ok
/*
- * x86 specific delta calculation.
+ * x86 specific calculation of nanoseconds for the current cycle count
*
* The regular implementation assumes that clocksource reads are globally
* monotonic. The TSC can be slightly off across sockets which can cause
@@ -308,8 +308,8 @@ static inline bool arch_vdso_cycles_ok(u64 cycles)
* jump.
*
* Therefore it needs to be verified that @cycles are greater than
- * @last. If not then use @last, which is the base time of the current
- * conversion period.
+ * @vd->cycles_last. If not then use @vd->cycles_last, which is the base
+ * time of the current conversion period.
*
* This variant also uses a custom mask because while the clocksource mask of
* all the VDSO capable clocksources on x86 is U64_MAX, the above code uses
@@ -317,25 +317,24 @@ static inline bool arch_vdso_cycles_ok(u64 cycles)
* declares everything with the MSB/Sign-bit set as invalid. Therefore the
* effective mask is S64_MAX.
*/
-static __always_inline
-u64 vdso_calc_delta(u64 cycles, u64 last, u64 mask, u32 mult)
+static __always_inline u64 vdso_calc_ns(const struct vdso_data *vd, u64 cycles, u64 base)
{
/*
* Due to the MSB/Sign-bit being used as invalid marker (see
* arch_vdso_cycles_valid() above), the effective mask is S64_MAX.
*/
- u64 delta = (cycles - last) & S64_MAX;
+ u64 delta = (cycles - vd->cycle_last) & S64_MAX;
/*
* Due to the above mentioned TSC wobbles, filter out negative motion.
* Per the above masking, the effective sign bit is now bit 62.
*/
if (unlikely(delta & (1ULL << 62)))
- return 0;
+ return base >> vd->shift;
- return delta * mult;
+ return ((delta * vd->mult) + base) >> vd->shift;
}
-#define vdso_calc_delta vdso_calc_delta
+#define vdso_calc_ns vdso_calc_ns
#endif /* !__ASSEMBLY__ */
diff --git a/lib/vdso/gettimeofday.c b/lib/vdso/gettimeofday.c
index 042b95e8164d..9fa90e0794c9 100644
--- a/lib/vdso/gettimeofday.c
+++ b/lib/vdso/gettimeofday.c
@@ -5,20 +5,12 @@
#include <vdso/datapage.h>
#include <vdso/helpers.h>
-#ifndef vdso_calc_delta
-/*
- * Default implementation which works for all sane clocksources. That
- * obviously excludes x86/TSC.
- */
-static __always_inline
-u64 vdso_calc_delta(u64 cycles, u64 last, u64 mask, u32 mult)
-{
+#ifndef vdso_calc_ns
+
#ifdef VDSO_DELTA_NOMASK
- return (cycles - last) * mult;
+# define VDSO_DELTA_MASK(vd) U64_MAX
#else
- return ((cycles - last) & mask) * mult;
-#endif
-}
+# define VDSO_DELTA_MASK(vd) (vd->mask)
#endif
#ifndef vdso_shift_ns
@@ -28,6 +20,18 @@ static __always_inline u64 vdso_shift_ns(u64 ns, u32 shift)
}
#endif
+/*
+ * Default implementation which works for all sane clocksources. That
+ * obviously excludes x86/TSC.
+ */
+static __always_inline u64 vdso_calc_ns(const struct vdso_data *vd, u64 cycles, u64 base)
+{
+ u64 delta = (cycles - vd->cycle_last) & VDSO_DELTA_MASK(vd);
+
+ return vdso_shift_ns((delta * vd->mult) + base, vd->shift);
+}
+#endif /* vdso_calc_ns */
+
#ifndef __arch_vdso_hres_capable
static inline bool __arch_vdso_hres_capable(void)
{
@@ -53,10 +57,10 @@ static inline bool vdso_cycles_ok(u64 cycles)
static __always_inline int do_hres_timens(const struct vdso_data *vdns, clockid_t clk,
struct __kernel_timespec *ts)
{
- const struct vdso_data *vd;
const struct timens_offset *offs = &vdns->offset[clk];
const struct vdso_timestamp *vdso_ts;
- u64 cycles, last, ns;
+ const struct vdso_data *vd;
+ u64 cycles, ns;
u32 seq;
s64 sec;
@@ -77,10 +81,7 @@ static __always_inline int do_hres_timens(const struct vdso_data *vdns, clockid_
cycles = __arch_get_hw_counter(vd->clock_mode, vd);
if (unlikely(!vdso_cycles_ok(cycles)))
return -1;
- ns = vdso_ts->nsec;
- last = vd->cycle_last;
- ns += vdso_calc_delta(cycles, last, vd->mask, vd->mult);
- ns = vdso_shift_ns(ns, vd->shift);
+ ns = vdso_calc_ns(vd, cycles, vdso_ts->nsec);
sec = vdso_ts->sec;
} while (unlikely(vdso_read_retry(vd, seq)));
@@ -115,7 +116,7 @@ static __always_inline int do_hres(const struct vdso_data *vd, clockid_t clk,
struct __kernel_timespec *ts)
{
const struct vdso_timestamp *vdso_ts = &vd->basetime[clk];
- u64 cycles, last, sec, ns;
+ u64 cycles, sec, ns;
u32 seq;
/* Allows to compile the high resolution parts out */
@@ -148,10 +149,7 @@ static __always_inline int do_hres(const struct vdso_data *vd, clockid_t clk,
cycles = __arch_get_hw_counter(vd->clock_mode, vd);
if (unlikely(!vdso_cycles_ok(cycles)))
return -1;
- ns = vdso_ts->nsec;
- last = vd->cycle_last;
- ns += vdso_calc_delta(cycles, last, vd->mask, vd->mult);
- ns = vdso_shift_ns(ns, vd->shift);
+ ns = vdso_calc_ns(vd, cycles, vdso_ts->nsec);
sec = vdso_ts->sec;
} while (unlikely(vdso_read_retry(vd, seq)));
--
2.34.1
Powered by blists - more mailing lists