mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Arnd Bergmann <arnd@arndb.de>
To: linux-kernel@vger.kernel.org
Cc: Arnd Bergmann <arnd@arndb.de>, Jaroslav Kysela <perex@perex.cz>,
	Takashi Iwai <tiwai@suse.de>,
	alsa-devel@alsa-project.org
Subject: [PATCH, RFC 26/30] oss: dmasound: kill SLEEP() macro to avoid race
Date: Thu,  2 Jan 2014 13:07:50 +0100	[thread overview]
Message-ID: <1388664474-1710039-27-git-send-email-arnd@arndb.de> (raw)
In-Reply-To: <1388664474-1710039-1-git-send-email-arnd@arndb.de>

The use of interruptible_sleep_on_timeout in the dmasound driver
is questionable and we want to kill off all sleep_on variants.
This replaces the calls with wait_event_interruptible_timeout
where possible, to wait for a particular event instead of blocking
in a racy way. In the sq_write function, the easiest solution is
an open-coded prepare_to_wait loop.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Jaroslav Kysela <perex@perex.cz>
Cc: Takashi Iwai <tiwai@suse.de>
Cc: alsa-devel@alsa-project.org
---
 sound/oss/dmasound/dmasound.h      |  1 -
 sound/oss/dmasound/dmasound_core.c | 28 +++++++++++++++++++---------
 2 files changed, 19 insertions(+), 10 deletions(-)

diff --git a/sound/oss/dmasound/dmasound.h b/sound/oss/dmasound/dmasound.h
index 1308d8d..01019f0 100644
--- a/sound/oss/dmasound/dmasound.h
+++ b/sound/oss/dmasound/dmasound.h
@@ -239,7 +239,6 @@ struct sound_queue {
     int busy, syncing, xruns, died;
 };
 
-#define SLEEP(queue)		interruptible_sleep_on_timeout(&queue, HZ)
 #define WAKE_UP(queue)		(wake_up_interruptible(&queue))
 
 extern struct sound_queue dmasound_write_sq;
diff --git a/sound/oss/dmasound/dmasound_core.c b/sound/oss/dmasound/dmasound_core.c
index bac43b5..f4ee85a 100644
--- a/sound/oss/dmasound/dmasound_core.c
+++ b/sound/oss/dmasound/dmasound_core.c
@@ -619,15 +619,27 @@ static ssize_t sq_write(struct file *file, const char __user *src, size_t uLeft,
 	}
 
 	while (uLeft) {
+		DEFINE_WAIT(wait);
+
 		while (write_sq.count >= write_sq.max_active) {
+			prepare_to_wait(&write_sq.action_queue, &wait, TASK_INTERRUPTIBLE);
 			sq_play();
-			if (write_sq.non_blocking)
+			if (write_sq.non_blocking) {
+				finish_wait(&write_sq.action_queue, &wait);
 				return uWritten > 0 ? uWritten : -EAGAIN;
-			SLEEP(write_sq.action_queue);
-			if (signal_pending(current))
+			}
+			if (write_sq.count < write_sq.max_active)
+				break;
+
+			schedule_timeout(HZ);
+			if (signal_pending(current)) {
+				finish_wait(&write_sq.action_queue, &wait);
 				return uWritten > 0 ? uWritten : -EINTR;
+			}
 		}
 
+		finish_wait(&write_sq.action_queue, &wait);
+
 		/* Here, we can avoid disabling the interrupt by first
 		 * copying and translating the data, and then updating
 		 * the write_sq variables. Until this is done, the interrupt
@@ -707,11 +719,8 @@ static int sq_open2(struct sound_queue *sq, struct file *file, fmode_t mode,
 			if (file->f_flags & O_NONBLOCK)
 				return rc;
 			rc = -EINTR;
-			while (sq->busy) {
-				SLEEP(sq->open_queue);
-				if (signal_pending(current))
-					return rc;
-			}
+			if (wait_event_interruptible(sq->open_queue, !sq->busy))
+				return rc;
 			rc = 0;
 #else
 			/* OSS manual says we will return EBUSY regardless
@@ -844,7 +853,8 @@ static int sq_fsync(void)
 	sq_play();	/* there may be an incomplete frame waiting */
 
 	while (write_sq.active) {
-		SLEEP(write_sq.sync_queue);
+		wait_event_interruptible_timeout(write_sq.sync_queue,
+						 !write_sq.active, HZ);
 		if (signal_pending(current)) {
 			/* While waiting for audio output to drain, an
 			 * interrupt occurred.  Stop audio output immediately
-- 
1.8.3.2


  parent reply	other threads:[~2014-01-02 12:11 UTC|newest]

Thread overview: 53+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-01-02 12:07 [PATCH, RFC 00/30] sleep_on removal Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 01/30] ataflop: fix sleep_on races Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 02/30] scsi: atari_scsi: fix sleep_on race Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 03/30] DAC960: remove sleep_on usage Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 04/30] swim3: fix interruptible_sleep_on race Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 05/30] [media] omap_vout: avoid sleep_on race Arnd Bergmann
2014-01-17 10:23   ` Hans Verkuil
2014-02-26  9:03     ` Arnd Bergmann
2014-02-26  9:56       ` Hans Verkuil
2014-01-02 12:07 ` [PATCH, RFC 06/30] [media] usbvision: remove bogus sleep_on_timeout Arnd Bergmann
2014-01-17 10:26   ` Hans Verkuil
2014-01-02 12:07 ` [PATCH, RFC 07/30] [media] radio-cadet: avoid interruptible_sleep_on race Arnd Bergmann
2014-01-17 10:47   ` Hans Verkuil
2014-01-17 14:28     ` Arnd Bergmann
2014-02-07  9:32       ` Hans Verkuil
2014-02-07  9:47         ` Hans Verkuil
2014-02-07 10:17         ` Arnd Bergmann
2014-02-07 11:35           ` Hans Verkuil
2014-02-09 20:53             ` Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 08/30] [media] arv: fix sleep_on race Arnd Bergmann
2014-01-17 10:51   ` Hans Verkuil
2014-02-07  9:16     ` Hans Verkuil
2014-02-26  8:57       ` Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 09/30] staging: serqt_usb2: don't use sleep_on Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 10/30] staging: gdm72xx: fix interruptible_sleep_on race Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 11/30] staging: panel: " Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 12/30] parport: " Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 13/30] cris: sync_serial: remove interruptible_sleep_on Arnd Bergmann
2014-01-09  9:52   ` Jesper Nilsson
2014-01-02 12:07 ` [PATCH, RFC 14/30] tty/amiserial: avoid interruptible_sleep_on Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 15/30] usbserial: stop using interruptible_sleep_on Arnd Bergmann
2014-01-02 21:36   ` Johan Hovold
2014-01-03 14:01     ` Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 16/30] tty: synclink: avoid sleep_on race Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 17/30] atm: nicstar: remove interruptible_sleep_on_timeout Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 18/30] atm: firestream: fix interruptible_sleep_on race Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 19/30] isdn: pcbit: " Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 20/30] isdn: hisax/elsa: fix sleep_on race in elsa FSM Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 21/30] isdn: divert, hysdn: fix interruptible_sleep_on race Arnd Bergmann
2014-01-02 15:01   ` Sergei Shtylyov
2014-01-02 16:48     ` Arnd Bergmann
2014-01-02 23:00       ` Sergei Shtylyov
2014-01-02 12:07 ` [PATCH, RFC 22/30] isdn: fix multiple sleep_on races Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 23/30] oss: msnd_pinnacle: avoid interruptible_sleep_on_timeout Arnd Bergmann
2014-01-14 15:16   ` Takashi Iwai
2014-01-02 12:07 ` [PATCH, RFC 24/30] oss: midibuf: fix sleep_on races Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 25/30] oss: vwsnd: avoid interruptible_sleep_on Arnd Bergmann
2014-01-02 12:07 ` Arnd Bergmann [this message]
2014-01-02 12:07 ` [PATCH, RFC 27/30] oss: remove last sleep_on users Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 28/30] sgi-xp: open-code interruptible_sleep_on_timeout Arnd Bergmann
2014-01-02 16:04   ` Robin Holt
2014-01-02 12:07 ` [PATCH, RFC 29/30] char: nwbutton: open-code interruptible_sleep_on Arnd Bergmann
2014-01-02 12:07 ` [PATCH, RFC 30/30] sched: remove sleep_on() and friends 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=1388664474-1710039-27-git-send-email-arnd@arndb.de \
    --to=arnd@arndb.de \
    --cc=alsa-devel@alsa-project.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=perex@perex.cz \
    --cc=tiwai@suse.de \
    /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

all inboxes | Powered by JetHome®