All of lore.kernel.org
 help / color / mirror / Atom feed
From: Adrian Hunter <adrian.hunter@intel.com>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>,
	Dave Hansen <dave.hansen@linux.intel.com>,
	John Stultz <jstultz@google.com>,
	"H. Peter Anvin" <hpa@zytor.com>,
	Alexander Gordeev <agordeev@linux.ibm.com>,
	Vincenzo Frascino <vincenzo.frascino@arm.com>,
	linux-s390@vger.kernel.org, Arnd Bergmann <arnd@arndb.de>,
	x86@kernel.org, "Aneesh Kumar K.V" <aneesh.kumar@kernel.org>,
	Ingo Molnar <mingo@redhat.com>,
	"Naveen N. Rao" <naveen.n.rao@linux.ibm.com>,
	Christian Borntraeger <borntraeger@linux.ibm.com>,
	Vasily Gorbik <gor@linux.ibm.com>,
	Heiko Carstens <hca@linux.ibm.com>,
	Nicholas Piggin <npiggin@gmail.com>,
	Borislav Petkov <bp@alien8.de>, Andy Lutomirski <luto@kernel.org>,
	Bjorn Helgaas <bhelgaas@google.com>,
	Anna-Maria Behnsen <anna-maria@linutronix.de>,
	Stephen Boyd <sboyd@kernel.org>,
	Randy Dunlap <rdunlap@infradead.org>,
	linux-kernel@vger.kernel.org, Sven Schnelle <svens@linux.ibm.com>,
	linuxppc-dev@lists.ozlabs.org
Subject: [PATCH V2 18/19] timekeeping: Let timekeeping_cycles_to_ns() handle both under and overflow
Date: Mon, 25 Mar 2024 08:40:22 +0200	[thread overview]
Message-ID: <20240325064023.2997-19-adrian.hunter@intel.com> (raw)
In-Reply-To: <20240325064023.2997-1-adrian.hunter@intel.com>

For the case !CONFIG_CLOCKSOURCE_VALIDATE_LAST_CYCLE, forego overflow
protection in the range (mask << 1) < delta <= mask, and interpret it
always as an inconsistency between CPU clock values. That allows
slightly neater code, and it is on a slow path so has no effect on
performance.

Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 kernel/time/timekeeping.c | 31 +++++++++++++------------------
 1 file changed, 13 insertions(+), 18 deletions(-)

diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c
index 111dfdbd488f..4e18db1819f8 100644
--- a/kernel/time/timekeeping.c
+++ b/kernel/time/timekeeping.c
@@ -266,17 +266,14 @@ static inline u64 timekeeping_debug_get_ns(const struct tk_read_base *tkr)
 	 * Try to catch underflows by checking if we are seeing small
 	 * mask-relative negative values.
 	 */
-	if (unlikely((~delta & mask) < (mask >> 3))) {
+	if (unlikely((~delta & mask) < (mask >> 3)))
 		tk->underflow_seen = 1;
-		now = last;
-	}
 
-	/* Cap delta value to the max_cycles values to avoid mult overflows */
-	if (unlikely(delta > max)) {
+	/* Check for multiplication overflows */
+	if (unlikely(delta > max))
 		tk->overflow_seen = 1;
-		now = last + max;
-	}
 
+	/* timekeeping_cycles_to_ns() handles both under and overflow */
 	return timekeeping_cycles_to_ns(tkr, now);
 }
 #else
@@ -375,19 +372,17 @@ static inline u64 timekeeping_cycles_to_ns(const struct tk_read_base *tkr, u64 c
 	u64 mask = tkr->mask, delta = (cycles - tkr->cycle_last) & mask;
 
 	/*
-	 * This detects the case where the delta overflows the multiplication
-	 * with tkr->mult.
+	 * This detects both negative motion and the case where the delta
+	 * overflows the multiplication with tkr->mult.
 	 */
 	if (unlikely(delta > tkr->clock->max_cycles)) {
-		if (IS_ENABLED(CONFIG_CLOCKSOURCE_VALIDATE_LAST_CYCLE)) {
-			/*
-			 * Handle clocksource inconsistency between CPUs to prevent
-			 * time from going backwards by checking for the MSB of the
-			 * mask being set in the delta.
-			 */
-			if (unlikely(delta & ~(mask >> 1)))
-				return tkr->xtime_nsec >> tkr->shift;
-		}
+		/*
+		 * Handle clocksource inconsistency between CPUs to prevent
+		 * time from going backwards by checking for the MSB of the
+		 * mask being set in the delta.
+		 */
+		if (delta & ~(mask >> 1))
+			return tkr->xtime_nsec >> tkr->shift;
 
 		return delta_to_ns_safe(tkr, delta);
 	}
-- 
2.34.1


WARNING: multiple messages have this Message-ID (diff)
From: Adrian Hunter <adrian.hunter@intel.com>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: Michael Ellerman <mpe@ellerman.id.au>,
	Nicholas Piggin <npiggin@gmail.com>,
	Christophe Leroy <christophe.leroy@csgroup.eu>,
	"Aneesh Kumar K.V" <aneesh.kumar@kernel.org>,
	"Naveen N. Rao" <naveen.n.rao@linux.ibm.com>,
	Heiko Carstens <hca@linux.ibm.com>,
	Vasily Gorbik <gor@linux.ibm.com>,
	Alexander Gordeev <agordeev@linux.ibm.com>,
	Christian Borntraeger <borntraeger@linux.ibm.com>,
	Sven Schnelle <svens@linux.ibm.com>,
	Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
	Dave Hansen <dave.hansen@linux.intel.com>,
	x86@kernel.org, "H. Peter Anvin" <hpa@zytor.com>,
	Andy Lutomirski <luto@kernel.org>,
	Vincenzo Frascino <vincenzo.frascino@arm.com>,
	John Stultz <jstultz@google.com>, Stephen Boyd <sboyd@kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Randy Dunlap <rdunlap@infradead.org>,
	Bjorn Helgaas <bhelgaas@google.com>,
	Arnd Bergmann <arnd@arndb.de>,
	Anna-Maria Behnsen <anna-maria@linutronix.de>,
	linuxppc-dev@lists.ozlabs.org, linux-kernel@vger.kernel.org,
	linux-s390@vger.kernel.org
Subject: [PATCH V2 18/19] timekeeping: Let timekeeping_cycles_to_ns() handle both under and overflow
Date: Mon, 25 Mar 2024 08:40:22 +0200	[thread overview]
Message-ID: <20240325064023.2997-19-adrian.hunter@intel.com> (raw)
In-Reply-To: <20240325064023.2997-1-adrian.hunter@intel.com>

For the case !CONFIG_CLOCKSOURCE_VALIDATE_LAST_CYCLE, forego overflow
protection in the range (mask << 1) < delta <= mask, and interpret it
always as an inconsistency between CPU clock values. That allows
slightly neater code, and it is on a slow path so has no effect on
performance.

Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 kernel/time/timekeeping.c | 31 +++++++++++++------------------
 1 file changed, 13 insertions(+), 18 deletions(-)

diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c
index 111dfdbd488f..4e18db1819f8 100644
--- a/kernel/time/timekeeping.c
+++ b/kernel/time/timekeeping.c
@@ -266,17 +266,14 @@ static inline u64 timekeeping_debug_get_ns(const struct tk_read_base *tkr)
 	 * Try to catch underflows by checking if we are seeing small
 	 * mask-relative negative values.
 	 */
-	if (unlikely((~delta & mask) < (mask >> 3))) {
+	if (unlikely((~delta & mask) < (mask >> 3)))
 		tk->underflow_seen = 1;
-		now = last;
-	}
 
-	/* Cap delta value to the max_cycles values to avoid mult overflows */
-	if (unlikely(delta > max)) {
+	/* Check for multiplication overflows */
+	if (unlikely(delta > max))
 		tk->overflow_seen = 1;
-		now = last + max;
-	}
 
+	/* timekeeping_cycles_to_ns() handles both under and overflow */
 	return timekeeping_cycles_to_ns(tkr, now);
 }
 #else
@@ -375,19 +372,17 @@ static inline u64 timekeeping_cycles_to_ns(const struct tk_read_base *tkr, u64 c
 	u64 mask = tkr->mask, delta = (cycles - tkr->cycle_last) & mask;
 
 	/*
-	 * This detects the case where the delta overflows the multiplication
-	 * with tkr->mult.
+	 * This detects both negative motion and the case where the delta
+	 * overflows the multiplication with tkr->mult.
 	 */
 	if (unlikely(delta > tkr->clock->max_cycles)) {
-		if (IS_ENABLED(CONFIG_CLOCKSOURCE_VALIDATE_LAST_CYCLE)) {
-			/*
-			 * Handle clocksource inconsistency between CPUs to prevent
-			 * time from going backwards by checking for the MSB of the
-			 * mask being set in the delta.
-			 */
-			if (unlikely(delta & ~(mask >> 1)))
-				return tkr->xtime_nsec >> tkr->shift;
-		}
+		/*
+		 * Handle clocksource inconsistency between CPUs to prevent
+		 * time from going backwards by checking for the MSB of the
+		 * mask being set in the delta.
+		 */
+		if (delta & ~(mask >> 1))
+			return tkr->xtime_nsec >> tkr->shift;
 
 		return delta_to_ns_safe(tkr, delta);
 	}
-- 
2.34.1


  parent reply	other threads:[~2024-03-25  6:54 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-03-25  6:40 [PATCH V2 00/19] timekeeping: Handle potential multiplication overflow Adrian Hunter
2024-03-25  6:40 ` Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 01/19] vdso: Consolidate vdso_calc_delta() Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 02/19] vdso: Consolidate nanoseconds calculation Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 03/19] vdso: Add CONFIG_GENERIC_VDSO_OVERFLOW_PROTECT Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 04/19] math64: Tidy mul_u64_u32_shr() Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] math64: Tidy up mul_u64_u32_shr() tip-bot2 for Adrian Hunter
2024-04-24 15:11   ` [PATCH V2 04/19] math64: Tidy mul_u64_u32_shr() Peter Zijlstra
2024-04-24 15:11     ` Peter Zijlstra
2024-03-25  6:40 ` [PATCH V2 05/19] vdso: math64: Provide mul_u64_u32_add_u64_shr() Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] vdso, " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 06/19] vdso: Add vdso_data::max_cycles Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] vdso: Add vdso_data:: Max_cycles tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 07/19] vdso: Make delta calculation overflow safe Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 08/19] x86/vdso: " Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 09/19] timekeeping: Move timekeeping helper functions Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 10/19] timekeeping: Rename fast_tk_get_delta_ns() to __timekeeping_get_ns() Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 11/19] timekeeping: Tidy timekeeping_cycles_to_ns() slightly Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 12/19] timekeeping: Reuse timekeeping_cycles_to_ns() Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 13/19] timekeeping: Refactor timekeeping helpers Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 14/19] timekeeping: Consolidate " Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 15/19] timekeeping: Fold in timekeeping_delta_to_ns() Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 16/19] timekeeping: Prepare timekeeping_cycles_to_ns() for overflow safety Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 17/19] timekeeping: Make delta calculation overflow safe Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` Adrian Hunter [this message]
2024-03-25  6:40   ` [PATCH V2 18/19] timekeeping: Let timekeeping_cycles_to_ns() handle both under and overflow Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25  6:40 ` [PATCH V2 19/19] clocksource: Make watchdog and suspend-timing multiplication overflow safe Adrian Hunter
2024-03-25  6:40   ` Adrian Hunter
2024-04-08 13:10   ` [tip: timers/core] " tip-bot2 for Adrian Hunter
2024-03-25 18:11 ` [PATCH V2 00/19] timekeeping: Handle potential multiplication overflow Arnd Bergmann

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20240325064023.2997-19-adrian.hunter@intel.com \
    --to=adrian.hunter@intel.com \
    --cc=agordeev@linux.ibm.com \
    --cc=aneesh.kumar@kernel.org \
    --cc=anna-maria@linutronix.de \
    --cc=arnd@arndb.de \
    --cc=bhelgaas@google.com \
    --cc=borntraeger@linux.ibm.com \
    --cc=bp@alien8.de \
    --cc=dave.hansen@linux.intel.com \
    --cc=gor@linux.ibm.com \
    --cc=hca@linux.ibm.com \
    --cc=hpa@zytor.com \
    --cc=jstultz@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-s390@vger.kernel.org \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=luto@kernel.org \
    --cc=mingo@redhat.com \
    --cc=naveen.n.rao@linux.ibm.com \
    --cc=npiggin@gmail.com \
    --cc=peterz@infradead.org \
    --cc=rdunlap@infradead.org \
    --cc=sboyd@kernel.org \
    --cc=svens@linux.ibm.com \
    --cc=tglx@linutronix.de \
    --cc=vincenzo.frascino@arm.com \
    --cc=x86@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.