All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jan Kiszka <jan.kiszka@siemens.com>
To: xenomai@xenomai.org
Subject: [PATCH 12/17] cobalt/tick: dovetail: improve accuracy of the host tick delay
Date: Fri, 11 Jun 2021 20:05:38 +0200	[thread overview]
Message-ID: <7b6dd52ed44d59ed87d260d7e35663dbb8cb3cd2.1623434743.git.jan.kiszka@siemens.com> (raw)
In-Reply-To: <cover.1623434743.git.jan.kiszka@siemens.com>

From: Philippe Gerum <rpm@xenomai.org>

Expiration dates of in-band timers are based on the monotonic time
base read by ktime_get(), which is ok for us to use from non-NMI
context.

Signed-off-by: Philippe Gerum <rpm@xenomai.org>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
---
 kernel/cobalt/dovetail/tick.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

diff --git a/kernel/cobalt/dovetail/tick.c b/kernel/cobalt/dovetail/tick.c
index fa041092e..ac343e4d3 100644
--- a/kernel/cobalt/dovetail/tick.c
+++ b/kernel/cobalt/dovetail/tick.c
@@ -55,19 +55,21 @@ void pipeline_set_timer_shot(unsigned long delay) /* ns */
 }
 
 static int proxy_set_next_ktime(ktime_t expires,
-				struct clock_event_device *proxy_dev)
+				struct clock_event_device *proxy_dev) /* hard irqs on/off */
 {
 	struct xnsched *sched;
-	ktime_t delta;
 	unsigned long flags;
+	ktime_t delta;
 	int ret;
 
 	/*
-	 * When Negative delta have been observed, we set delta zero.
-	 * Or else exntimer_start() will return -ETIMEDOUT and do not
-	 * trigger shot
+	 * Expiration dates of in-band timers are based on the common
+	 * monotonic time base. If the timeout date has already
+	 * elapsed, make sure xntimer_start() does not fail with
+	 * -ETIMEDOUT but programs the hardware for ticking
+	 * immediately instead.
 	 */
-	delta = ktime_sub(expires, ktime_get_mono_fast_ns());
+	delta = ktime_sub(expires, ktime_get());
 	if (delta < 0)
 		delta = 0;
 
-- 
2.26.2



  parent reply	other threads:[~2021-06-11 18:05 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-06-11 18:05 [PATCH 00/17] Dovetail integration - the finals Jan Kiszka
2021-06-11 18:05 ` [PATCH 01/17] cobalt/thread: Rework kthread check for xnthread_signal Jan Kiszka
2021-06-11 18:05 ` [PATCH 02/17] cobalt/thread: Move xnthread_signal work off the stack Jan Kiszka
2021-06-11 18:05 ` [PATCH 03/17] cobalt/thread: Privatize xnthread_map to map_kthread Jan Kiszka
2021-06-11 18:05 ` [PATCH 04/17] cobalt/thread: Pull kthread inband work onto creator stack Jan Kiszka
2021-06-11 18:05 ` [PATCH 05/17] cobalt/thread: Make sure relax inband work is on a stack outliving the wakeup Jan Kiszka
2021-06-11 18:05 ` [PATCH 06/17] cobalt/thread: dovetail: keep hard irqs off on transition to in-band Jan Kiszka
2022-01-24 16:15   ` Jan Kiszka
2022-01-24 16:55     ` Philippe Gerum
2022-01-24 17:00       ` Philippe Gerum
2022-01-24 17:05         ` Jan Kiszka
2021-06-11 18:05 ` [PATCH 07/17] cobalt/sched: dovetail: prevent early scheduling on uninit runqueues Jan Kiszka
2021-06-11 18:05 ` [PATCH 08/17] cobalt/arm: dovetail: add architecture bits Jan Kiszka
2021-06-11 18:05 ` [PATCH 09/17] lib/cobalt: Reorder low_init for having cobalt_use_legacy_tsc earlier available Jan Kiszka
2021-06-11 18:05 ` [PATCH 10/17] lib/cobalt/arm: dovetail: skip detection of KUSER_TSC support Jan Kiszka
2021-06-11 18:05 ` [PATCH 11/17] lib/cobalt/arm64: " Jan Kiszka
2021-06-11 18:05 ` Jan Kiszka [this message]
2021-06-11 18:05 ` [PATCH 13/17] cobalt/tick: dovetail: Drop pointless inline specifier from pipeline_must_force_program_tick Jan Kiszka
2021-06-11 18:05 ` [PATCH 14/17] cobalt/clock: dovetail: abstract interface to hardware TSC Jan Kiszka
2021-06-11 18:05 ` [PATCH 15/17] cobalt/arm64: dovetail: add architecture bits Jan Kiszka
2021-06-11 18:05 ` [PATCH 16/17] cobalt/syscall: Account for different syscall argument marshaling Jan Kiszka
2021-06-11 18:05 ` [PATCH 17/17] ci: Add arm, arm64 and x86 dovetail targets for kernel 5.10 Jan Kiszka
2021-06-13 17:07 ` [PATCH 00/17] Dovetail integration - the finals Philippe Gerum
2021-06-13 17:12   ` Jan Kiszka
2021-06-14  6:24     ` Philippe Gerum

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=7b6dd52ed44d59ed87d260d7e35663dbb8cb3cd2.1623434743.git.jan.kiszka@siemens.com \
    --to=jan.kiszka@siemens.com \
    --cc=xenomai@xenomai.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.