devicetree.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: NeilBrown <neilb@suse.de>
To: Jiri Slaby <jslaby@suse.cz>
Cc: NeilBrown <neil@brown.name>, Mark Rutland <mark.rutland@arm.com>,
	One Thousand Gnomes <gnomes@lxorguk.ukuu.org.uk>,
	Peter Hurley <peter@hurleysoftware.com>,
	Arnd Bergmann <arnd@arndb.de>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Sebastian Reichel <sre@kernel.org>, Pavel Machek <pavel@ucw.cz>,
	Grant Likely <grant.likely@linaro.org>,
	GTA04 owners <gta04-owner@goldelico.com>,
	devicetree@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/3] TTY: add support for tty_slave devices.
Date: Tue, 31 Mar 2015 10:45:20 +1100	[thread overview]
Message-ID: <20150331104520.1002b7f0@notabene.brown> (raw)
In-Reply-To: <55113D1B.1030906@suse.cz>

[-- Attachment #1: Type: text/plain, Size: 5974 bytes --]

On Tue, 24 Mar 2015 11:31:55 +0100 Jiri Slaby <jslaby@suse.cz> wrote:

> Hi,
> 
> On 03/18/2015, 06:58 AM, NeilBrown wrote:
> > --- /dev/null
> > +++ b/drivers/tty/slave/tty_slave_core.c
> > @@ -0,0 +1,136 @@
> 
> ...
> 
> > +static int tty_slave_match(struct device *dev, struct device_driver *drv)
> > +{
> > +	return of_driver_match_device(dev, drv);
> > +}
> > +
> > +static void tty_slave_release(struct device *dev)
> > +{
> > +	kfree(dev);
> 
> This should free the slave where the dev is contained. This is never
> called IMO due to missing put_device's in the code.

Yes of course, thanks.  I've fix the 'kfree'.


> 
> > +}
> > +
> > +struct bus_type tty_slave_bus_type = {
> > +	.name		= "tty-slave",
> > +	.match		= tty_slave_match,
> > +};
> > +
> > +int tty_slave_register(struct device *parent, struct device_node *node,
> > +		       struct device *tty, struct tty_driver *drv)
> > +{
> > +	struct tty_slave *slave;
> > +	int retval;
> > +
> > +	if (!of_get_property(node, "compatible", NULL))
> > +		return -ENODEV;
> > +
> > +	slave = kzalloc(sizeof(*slave), GFP_KERNEL);
> > +	if (!slave)
> > +		return -ENOMEM;
> > +
> 
> device_initialize();

device_initialize is only needed if you call device_add()
I use device_register() which calls both.


> 
> > +	slave->dev.bus = &tty_slave_bus_type;
> > +	slave->dev.parent = parent;
> > +	slave->dev.release = tty_slave_release;
> > +	slave->dev.of_node = of_node_get(node);
> > +	dev_set_name(&slave->dev, "%s", node->name);
> > +	slave->tty_dev = tty;
> > +	slave->tty_drv = drv;
> > +	slave->ops = *drv->ops;
> > +	retval = device_register(&slave->dev);
> > +	if (retval) {
> > +		of_node_put(node);
> > +		kfree(slave);
> 
> Do device_put() instead of the two. And do the two in .release.

Done, thanks.

> 
> > +	}
> > +	return retval;
> > +}
> > +EXPORT_SYMBOL(tty_slave_register);
> ...
> > --- a/drivers/tty/tty_io.c
> > +++ b/drivers/tty/tty_io.c
> ...
> > @@ -3205,6 +3208,29 @@ static void tty_device_create_release(struct device *dev)
> >  	kfree(dev);
> >  }
> >  
> > +int tty_register_finalize(struct tty_driver *driver, struct device *dev)
> > +{
> > +	int retval;
> > +	bool cdev = false;
> > +	int index = dev->devt - MKDEV(driver->major,
> > +				      driver->minor_start);
> > +	printk("REGISTER %d %d 0x%x %d\n", driver->major, driver->minor_start, dev->devt, index);
> > +	if (!(driver->flags & TTY_DRIVER_DYNAMIC_ALLOC)) {
> > +		retval = tty_cdev_add(driver,
> > +				      dev->devt,
> > +				      index, 1);
> > +		if (retval)
> > +			return retval;
> > +		cdev = true;
> > +	}
> > +	retval = device_register(dev);
> > +	if (retval == 0)
> > +		return 0;
> > +	if (cdev)
> > +		cdev_del(&driver->cdevs[index]);
> > +	return retval;
> > +}
> > +EXPORT_SYMBOL(tty_register_finalize);
> >  /**
> >   *	tty_register_device_attr - register a tty device
> >   *	@driver: the tty driver that describes the tty device
> > @@ -3234,7 +3260,8 @@ struct device *tty_register_device_attr(struct tty_driver *driver,
> >  	dev_t devt = MKDEV(driver->major, driver->minor_start) + index;
> >  	struct device *dev = NULL;
> >  	int retval = -ENODEV;
> > -	bool cdev = false;
> > +	struct device_node *node;
> > +	bool slave_registered = false;
> >  
> >  	if (index >= driver->num) {
> >  		printk(KERN_ERR "Attempt to register invalid tty line number "
> > @@ -3247,13 +3274,6 @@ struct device *tty_register_device_attr(struct tty_driver *driver,
> >  	else
> >  		tty_line_name(driver, index, name);
> >  
> > -	if (!(driver->flags & TTY_DRIVER_DYNAMIC_ALLOC)) {
> > -		retval = tty_cdev_add(driver, devt, index, 1);
> > -		if (retval)
> > -			goto error;
> > -		cdev = true;
> > -	}
> > -
> >  	dev = kzalloc(sizeof(*dev), GFP_KERNEL);
> >  	if (!dev) {
> >  		retval = -ENOMEM;
> > @@ -3268,16 +3288,24 @@ struct device *tty_register_device_attr(struct tty_driver *driver,
> >  	dev->groups = attr_grp;
> >  	dev_set_drvdata(dev, drvdata);
> >  
> > -	retval = device_register(dev);
> > -	if (retval)
> > -		goto error;
> > +	if (device && device->of_node)
> > +		for_each_available_child_of_node(device->of_node, node) {
> > +			if (tty_slave_register(device, node, dev, driver) == 0)
> > +				slave_registered = true;
> > +			if (slave_registered)
> > +				break;
> > +		}
> > +
> > +	if (!slave_registered) {
> > +		retval = tty_register_finalize(driver, dev);
> > +		if (retval)
> > +			goto error;
> > +	}
> >  
> >  	return dev;
> 
> And what about ttys not using the tty_register_device* helpers?

I guess they are on their own - they don't get any help...
Is there a problem that I'm not seeing here?

I think this only applies to code that calls device_create(tty_class, ...)
which is pty.c, vt.c, and "/dev/tty".  These are all virtual devices so
having a slave doesn't make much sense... does it?

> 
> What happens when the tty is unregistered?

Good question.  I hadn't thought that through.

When the tty is unregistered, it will drop the reference it has on ->parent
which the tty_slave.  So we want that to be the last reference.
So in tty_slave_finalize() I need a put_dev() on the slave.
That must be the "put_dev" that you thought was missing earlier.

However that leaves a loose end. destruct_tty_driver() calls
tty_unregister_device() on all registered ttys for that driver.
If one had a tty_slave which was finalized, that will drop the reference on
the slave so it will disappear, which is all good.
But if one has a slave for which the driver hasn't been found yet, then there
will be no tty to unregister so the tty_slave cannot be dropped.

I guess I need to either take a reference to the driver - which doesn't seem
like a good idea - or use bus_find_device() to find any tty_slaves with that
driver, and drop them.
I'll see how that works out.


Thanks a lot!

NeilBrown

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 811 bytes --]

  reply	other threads:[~2015-03-30 23:45 UTC|newest]

Thread overview: 80+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-03-18  5:58 [PATCH 0/3] tty slave device support - version 3 NeilBrown
2015-03-18  5:58 ` [PATCH 3/3] tty/slaves: add a driver to power on/off UART attached devices NeilBrown
     [not found]   ` <20150318055831.21025.33670.stgit-wvvUuzkyo1EYVZTmpyfIwg@public.gmane.org>
2015-03-20  7:54     ` [Gta04-owner] " Dr. H. Nikolaus Schaller
     [not found]       ` <C53A3E5D-5AEA-4F0F-88A9-F9BC8CB0D0D6-xXXSsgcRVICgSpxsJD1C4w@public.gmane.org>
2015-03-20  8:54         ` NeilBrown
2015-03-20  9:34           ` Dr. H. Nikolaus Schaller
     [not found]             ` <14FB51CF-9568-4BF4-B917-2C019D992FDB-xXXSsgcRVICgSpxsJD1C4w@public.gmane.org>
2015-03-20 19:50               ` Pavel Machek
2015-03-20 23:31             ` NeilBrown
2015-03-24 17:58               ` Dr. H. Nikolaus Schaller
2015-03-25  1:45                 ` Sebastian Reichel
2015-03-25  7:59                   ` Dr. H. Nikolaus Schaller
2015-03-25 15:21                     ` Sebastian Reichel
2015-03-25 16:44                       ` Dr. H. Nikolaus Schaller
2015-03-26 18:08                         ` Sebastian Reichel
2015-03-27  9:22                           ` Dr. H. Nikolaus Schaller
2015-03-27 16:31                             ` Sebastian Reichel
2015-03-27 17:11                               ` Dr. H. Nikolaus Schaller
2015-04-27 20:35                                 ` Pavel Machek
2015-04-29  6:56                                   ` Dr. H. Nikolaus Schaller
2015-03-25 20:53                     ` Pavel Machek
2015-03-25 21:25                       ` Dr. H. Nikolaus Schaller
     [not found]                         ` <1F3B5E32-1330-457C-AE25-791319293C38-xXXSsgcRVICgSpxsJD1C4w@public.gmane.org>
2015-03-26  5:56                           ` Pavel Machek
2015-03-26  6:44                             ` Dr. H. Nikolaus Schaller
     [not found]                               ` <365B2C25-1782-4ADE-B620-190A4CC5E8E3-xXXSsgcRVICgSpxsJD1C4w@public.gmane.org>
2015-04-04  7:52                                 ` Pavel Machek
2015-03-25 20:42                 ` Pavel Machek
2015-03-25 21:22                   ` Dr. H. Nikolaus Schaller
2015-03-18  5:58 ` [PATCH 1/3] TTY: use class_find_device to find port in uart_suspend/resume NeilBrown
     [not found]   ` <20150318055831.21025.27864.stgit-wvvUuzkyo1EYVZTmpyfIwg@public.gmane.org>
2015-03-25 16:20     ` Peter Hurley
2015-03-29 21:49       ` [Gta04-owner] " NeilBrown
2015-03-18  5:58 ` [PATCH 2/3] TTY: add support for tty_slave devices NeilBrown
2015-03-18  9:11   ` Paul Bolle
2015-03-22  3:32     ` NeilBrown
2015-03-20 19:41   ` Pavel Machek
2015-03-22  3:42     ` [Gta04-owner] " NeilBrown
     [not found]       ` <20150322144209.14abc603-wvvUuzkyo1EYVZTmpyfIwg@public.gmane.org>
2015-03-22  7:58         ` Pavel Machek
2015-03-24 10:31   ` Jiri Slaby
2015-03-30 23:45     ` NeilBrown [this message]
2015-03-25 16:30   ` Peter Hurley
2015-03-25 21:17     ` [Gta04-owner] " NeilBrown
2015-03-27 11:09       ` Peter Hurley
2015-03-31  0:33         ` NeilBrown
2015-03-20  7:54 ` [Gta04-owner] [PATCH 0/3] tty slave device support - version 3 Dr. H. Nikolaus Schaller
     [not found]   ` <80D7E742-4633-4CCD-B754-221387D82922-xXXSsgcRVICgSpxsJD1C4w@public.gmane.org>
2015-03-20  8:43     ` NeilBrown
2015-03-20  8:54       ` Dr. H. Nikolaus Schaller
2015-03-20 13:08         ` Sebastian Reichel
2015-03-20 13:57           ` Dr. H. Nikolaus Schaller
     [not found]             ` <72268D51-F993-497A-B4F9-707C8DB7155C-xXXSsgcRVICgSpxsJD1C4w@public.gmane.org>
2015-03-20 17:14               ` Sebastian Reichel
2015-03-20 19:31 ` Pavel Machek
     [not found] ` <20150318055437.21025.13990.stgit-wvvUuzkyo1EYVZTmpyfIwg@public.gmane.org>
2015-05-05 19:54   ` Peter Hurley
2015-05-05 20:46     ` NeilBrown
     [not found]     ` <55492001.30806-WaGBZJeGNqdsbIuE7sb01tBPR1lH4CV8@public.gmane.org>
2015-05-06  5:19       ` [Gta04-owner] " Dr. H. Nikolaus Schaller
2015-05-06 11:10         ` Peter Hurley
     [not found]         ` <DE60E7AA-1BF4-4A68-9638-E09F85FD5C72-xXXSsgcRVICgSpxsJD1C4w@public.gmane.org>
2015-05-06  9:27           ` Pavel Machek
2015-05-06 11:50             ` Dr. H. Nikolaus Schaller
2015-05-06 12:05               ` Peter Hurley
2015-05-06 12:27                 ` Dr. H. Nikolaus Schaller
2015-05-06 12:36                   ` Mark Rutland
2015-05-06 13:28                     ` Dr. H. Nikolaus Schaller
2015-05-06 14:15                       ` Mark Rutland
2015-05-06 16:09                         ` Dr. H. Nikolaus Schaller
2015-05-06 17:18                           ` Mark Rutland
2015-05-07 12:46                             ` Dr. H. Nikolaus Schaller
2015-05-07 14:30                               ` Peter Hurley
2015-05-07 15:11                                 ` Dr. H. Nikolaus Schaller
2015-05-07 16:18                                   ` Peter Hurley
2015-05-07 16:57                                     ` Dr. H. Nikolaus Schaller
     [not found]                               ` <B4D487C5-C260-497B-A441-6C381D53616C-xXXSsgcRVICgSpxsJD1C4w@public.gmane.org>
2015-05-07 14:56                                 ` Peter Hurley
     [not found]                                   ` <554B7D33.602-WaGBZJeGNqdsbIuE7sb01tBPR1lH4CV8@public.gmane.org>
2015-05-07 15:34                                     ` Dr. H. Nikolaus Schaller
2015-05-07 15:51                                       ` Peter Hurley
2015-05-07 16:46                                         ` Dr. H. Nikolaus Schaller
2015-05-13  8:09                                           ` Dr. H. Nikolaus Schaller
2015-06-03 11:49                                             ` [PATCH RFC 0/3] UART slave device support Dr. H. Nikolaus Schaller
     [not found]                                               ` <56C579D3-E8F8-4B5A-B6E8-993B113F3461-xXXSsgcRVICgSpxsJD1C4w@public.gmane.org>
2015-06-06 13:09                                                 ` Pavel Machek
     [not found]                                             ` <463356C5-E3C6-432C-A1C5-71F0287F1FEE@goldelico.com>
2015-06-03 12:09                                               ` [Gta04-owner] [PATCH RFC 3/3] misc: Add w2g0004 gps receiver driver Christ van Willegen
2015-05-07 15:37                       ` [Gta04-owner] [PATCH 0/3] tty slave device support - version 3 Peter Hurley
     [not found]               ` <A2594559-CA92-40C2-A06A-AC2483E85CB1-xXXSsgcRVICgSpxsJD1C4w@public.gmane.org>
2015-05-06 14:28                 ` Pavel Machek
2015-05-06 16:12                   ` Dr. H. Nikolaus Schaller
     [not found]                     ` <C68A105A-7147-4143-9B91-1A239726A780-xXXSsgcRVICgSpxsJD1C4w@public.gmane.org>
2015-06-06 13:09                       ` Pavel Machek
2015-06-06 18:53                         ` Belisko Marek
2015-06-06 18:55                           ` Belisko Marek
2015-05-07 15:48           ` Rob Herring

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=20150331104520.1002b7f0@notabene.brown \
    --to=neilb@suse.de \
    --cc=arnd@arndb.de \
    --cc=devicetree@vger.kernel.org \
    --cc=gnomes@lxorguk.ukuu.org.uk \
    --cc=grant.likely@linaro.org \
    --cc=gregkh@linuxfoundation.org \
    --cc=gta04-owner@goldelico.com \
    --cc=jslaby@suse.cz \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mark.rutland@arm.com \
    --cc=neil@brown.name \
    --cc=pavel@ucw.cz \
    --cc=peter@hurleysoftware.com \
    --cc=sre@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 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).