linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Dmitry Safonov <dima@arista.com>
To: linux-kernel@vger.kernel.org
Cc: Dmitry Safonov <0x7f454c46@gmail.com>,
	Dmitry Safonov <dima@arista.com>, Adrian Reber <adrian@lisas.de>,
	Alexander Viro <viro@zeniv.linux.org.uk>,
	Andrei Vagin <avagin@openvz.org>,
	Andy Lutomirski <luto@kernel.org>,
	Cyrill Gorcunov <gorcunov@openvz.org>,
	Ingo Molnar <mingo@redhat.com>, Oleg Nesterov <oleg@redhat.com>,
	Pavel Emelyanov <xemul@virtuozzo.com>,
	Thomas Gleixner <tglx@linutronix.de>,
	containers@lists.linux-foundation.org,
	linux-fsdevel@vger.kernel.org
Subject: [PATCH 5/9] select: Convert select_estimate_accuracy() to take ktime_t
Date: Mon,  9 Sep 2019 11:23:36 +0100	[thread overview]
Message-ID: <20190909102340.8592-6-dima@arista.com> (raw)
In-Reply-To: <20190909102340.8592-1-dima@arista.com>

Instead of converting the time on the first loop, the same
if (end_time) can be shared. Simplify the loop by taking time
conversion out.

Also prepare the ground for converting poll() restart_block timeout into
ktime_t - that's the only user that leaves it in timespec.
The conversion is needed to introduce an API for ptrace() to get
a timeout from restart_block.

Signed-off-by: Dmitry Safonov <dima@arista.com>
---
 fs/eventpoll.c       |  4 ++--
 fs/select.c          | 38 ++++++++++++--------------------------
 include/linux/poll.h |  2 +-
 3 files changed, 15 insertions(+), 29 deletions(-)

diff --git a/fs/eventpoll.c b/fs/eventpoll.c
index d7f1f5011fac..d5120fc49a39 100644
--- a/fs/eventpoll.c
+++ b/fs/eventpoll.c
@@ -1836,9 +1836,9 @@ static int ep_poll(struct eventpoll *ep, struct epoll_event __user *events,
 	if (timeout > 0) {
 		struct timespec64 end_time = ep_set_mstimeout(timeout);
 
-		slack = select_estimate_accuracy(&end_time);
+		expires = timespec64_to_ktime(end_time);
 		to = &expires;
-		*to = timespec64_to_ktime(end_time);
+		slack = select_estimate_accuracy(expires);
 	} else if (timeout == 0) {
 		/*
 		 * Avoid the unnecessary trip to the wait queue loop, if the
diff --git a/fs/select.c b/fs/select.c
index 2477c202631e..458f2a944318 100644
--- a/fs/select.c
+++ b/fs/select.c
@@ -66,7 +66,7 @@ static long __estimate_accuracy(ktime_t slack)
 	return slack;
 }
 
-u64 select_estimate_accuracy(struct timespec64 *timeout)
+u64 select_estimate_accuracy(ktime_t timeout)
 {
 	ktime_t now, slack;
 
@@ -77,7 +77,7 @@ u64 select_estimate_accuracy(struct timespec64 *timeout)
 		return 0;
 
 	now = ktime_get();
-	slack = now - timespec64_to_ktime(*timeout);
+	slack = now - timeout;
 
 	slack = __estimate_accuracy(slack);
 	if (slack < current->timer_slack_ns)
@@ -490,8 +490,11 @@ static int do_select(int n, fd_set_bits *fds, struct timespec64 *end_time)
 		timed_out = 1;
 	}
 
-	if (end_time && !timed_out)
-		slack = select_estimate_accuracy(end_time);
+	if (end_time && !timed_out) {
+		expire = timespec64_to_ktime(*end_time);
+		to = &expire;
+		slack = select_estimate_accuracy(expire);
+	}
 
 	retval = 0;
 	for (;;) {
@@ -582,16 +585,6 @@ static int do_select(int n, fd_set_bits *fds, struct timespec64 *end_time)
 		}
 		busy_flag = 0;
 
-		/*
-		 * If this is the first loop and we have a timeout
-		 * given, then we convert to ktime_t and set the to
-		 * pointer to the expiry value.
-		 */
-		if (end_time && !to) {
-			expire = timespec64_to_ktime(*end_time);
-			to = &expire;
-		}
-
 		if (!poll_schedule_timeout(&table, TASK_INTERRUPTIBLE,
 					   to, slack))
 			timed_out = 1;
@@ -876,8 +869,11 @@ static int do_poll(struct poll_list *list, struct poll_wqueues *wait,
 		timed_out = 1;
 	}
 
-	if (end_time && !timed_out)
-		slack = select_estimate_accuracy(end_time);
+	if (end_time && !timed_out) {
+		expire = timespec64_to_ktime(*end_time);
+		to = &expire;
+		slack = select_estimate_accuracy(expire);
+	}
 
 	for (;;) {
 		struct poll_list *walk;
@@ -930,16 +926,6 @@ static int do_poll(struct poll_list *list, struct poll_wqueues *wait,
 		}
 		busy_flag = 0;
 
-		/*
-		 * If this is the first loop and we have a timeout
-		 * given, then we convert to ktime_t and set the to
-		 * pointer to the expiry value.
-		 */
-		if (end_time && !to) {
-			expire = timespec64_to_ktime(*end_time);
-			to = &expire;
-		}
-
 		if (!poll_schedule_timeout(wait, TASK_INTERRUPTIBLE, to, slack))
 			timed_out = 1;
 	}
diff --git a/include/linux/poll.h b/include/linux/poll.h
index 1cdc32b1f1b0..d0f21eb19257 100644
--- a/include/linux/poll.h
+++ b/include/linux/poll.h
@@ -112,7 +112,7 @@ struct poll_wqueues {
 
 extern void poll_initwait(struct poll_wqueues *pwq);
 extern void poll_freewait(struct poll_wqueues *pwq);
-extern u64 select_estimate_accuracy(struct timespec64 *tv);
+extern u64 select_estimate_accuracy(ktime_t timeout);
 
 #define MAX_INT64_SECONDS (((s64)(~((u64)0)>>1)/HZ)-1)
 
-- 
2.23.0


  parent reply	other threads:[~2019-09-09 10:24 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-09-09 10:23 [PATCH 0/9] restart_block: Prepare the ground for dumping timeout Dmitry Safonov
2019-09-09 10:23 ` [PATCH 1/9] futex: Remove unused uaddr2 in restart_block Dmitry Safonov
2019-09-09 10:23 ` [PATCH 2/9] restart_block: Prevent userspace set part of the block Dmitry Safonov
2019-09-09 10:23 ` [PATCH 3/9] select: Convert __esimate_accuracy() to ktime_t Dmitry Safonov
2019-09-09 10:23 ` [PATCH 4/9] select: Micro-optimise __estimate_accuracy() Dmitry Safonov
2019-09-09 11:18   ` Cyrill Gorcunov
2019-09-09 11:50     ` Dmitry Safonov
2019-09-09 12:14       ` Cyrill Gorcunov
2019-09-19 14:05   ` Cyrill Gorcunov
2019-09-19 14:25     ` Dmitry Safonov
2019-09-09 10:23 ` Dmitry Safonov [this message]
2019-09-09 10:23 ` [PATCH 6/9] select: Extract common code into do_sys_ppoll() Dmitry Safonov
2019-09-09 11:15   ` kbuild test robot
2019-09-09 19:48   ` kbuild test robot
2019-09-09 10:23 ` [PATCH 7/9] select: Use ktime_t in do_sys_poll() and do_poll() Dmitry Safonov
2019-09-09 10:23 ` [PATCH 8/9] select/restart_block: Convert poll's timeout to u64 Dmitry Safonov
2019-09-09 13:07   ` David Laight
2019-09-16 15:19     ` Dmitry Safonov
2019-09-09 10:23 ` [PATCH 9/9] restart_block: Make common timeout Dmitry Safonov

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=20190909102340.8592-6-dima@arista.com \
    --to=dima@arista.com \
    --cc=0x7f454c46@gmail.com \
    --cc=adrian@lisas.de \
    --cc=avagin@openvz.org \
    --cc=containers@lists.linux-foundation.org \
    --cc=gorcunov@openvz.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=luto@kernel.org \
    --cc=mingo@redhat.com \
    --cc=oleg@redhat.com \
    --cc=tglx@linutronix.de \
    --cc=viro@zeniv.linux.org.uk \
    --cc=xemul@virtuozzo.com \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).