From: David Daney <ddaney@caviumnetworks.com>
To: linux-serial@vger.kernel.org, gregkh@suse.de,
linux-kernel@vger.kernel.org,
devicetree-discuss@lists.ozlabs.org, grant.likely@secretlab.ca
Cc: David Daney <ddaney@caviumnetworks.com>
Subject: [RFC PATCH 1/2] serial: 8250: Add a notifier chain for driver registration.
Date: Wed, 16 Mar 2011 18:26:06 -0700 [thread overview]
Message-ID: <1300325167-26433-2-git-send-email-ddaney@caviumnetworks.com> (raw)
In-Reply-To: <1300325167-26433-1-git-send-email-ddaney@caviumnetworks.com>
The 8250 driver is a bit weird in that in addition to supporting
platform devices, extra devices can be added by calling
serial8250_register_port().
The problem is that if we call serial8250_register_port() before the
driver is initialized Bad Things happen (we dereference NULL
pointers).
There doesn't seem to be a general way to know if a driver has been
initialized, so we add a notifier chain just for this driver. When the
SERIAL8250_DRIVER_ADD notifier is called, we know it is safe to call
serial8250_register_port(), and we can add our devices.
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
---
drivers/tty/serial/8250.c | 20 ++++++++++++++++++++
include/linux/serial_8250.h | 21 +++++++++++++++++++++
2 files changed, 41 insertions(+), 0 deletions(-)
diff --git a/drivers/tty/serial/8250.c b/drivers/tty/serial/8250.c
index 3975df6..229a1d2 100644
--- a/drivers/tty/serial/8250.c
+++ b/drivers/tty/serial/8250.c
@@ -3281,6 +3281,20 @@ void serial8250_unregister_port(int line)
}
EXPORT_SYMBOL(serial8250_unregister_port);
+static BLOCKING_NOTIFIER_HEAD(serial8250_notifier);
+
+int serial8250_register_notifier(struct notifier_block *nb)
+{
+ return blocking_notifier_chain_register(&serial8250_notifier, nb);
+}
+EXPORT_SYMBOL(serial8250_register_notifier);
+
+int serial8250_unregister_notifier(struct notifier_block *nb)
+{
+ return blocking_notifier_chain_unregister(&serial8250_notifier, nb);
+}
+EXPORT_SYMBOL(serial8250_unregister_notifier);
+
static int __init serial8250_init(void)
{
int ret;
@@ -3328,6 +3342,10 @@ unreg_uart_drv:
uart_unregister_driver(&serial8250_reg);
#endif
out:
+ if (ret == 0)
+ blocking_notifier_call_chain(&serial8250_notifier,
+ SERIAL8250_DRIVER_ADD, NULL);
+
return ret;
}
@@ -3335,6 +3353,8 @@ static void __exit serial8250_exit(void)
{
struct platform_device *isa_dev = serial8250_isa_devs;
+ blocking_notifier_call_chain(&serial8250_notifier,
+ SERIAL8250_DRIVER_REMOVE, NULL);
/*
* This tells serial8250_unregister_port() not to re-register
* the ports (thereby making serial8250_isa_driver permanently
diff --git a/include/linux/serial_8250.h b/include/linux/serial_8250.h
index 97f5b45..2155c28 100644
--- a/include/linux/serial_8250.h
+++ b/include/linux/serial_8250.h
@@ -11,6 +11,7 @@
#ifndef _LINUX_SERIAL_8250_H
#define _LINUX_SERIAL_8250_H
+#include <linux/notifier.h>
#include <linux/serial_core.h>
#include <linux/platform_device.h>
@@ -85,4 +86,24 @@ extern void serial8250_set_isa_configurator(void (*v)
(int port, struct uart_port *up,
unsigned short *capabilities));
+extern int serial8250_register_notifier(struct notifier_block *nb);
+extern int serial8250_unregister_notifier(struct notifier_block *nb);
+/*
+ * Notifiers get called to allow serial8250_register_port() after the
+ * driver is registered, and serial8250_unregister_port() to be called
+ * before it is unregistered.
+ */
+#define SERIAL8250_DRIVER_ADD 1
+#define SERIAL8250_DRIVER_REMOVE 2
+
+#define serial8250_notifier(fn, pri) \
+({ \
+ static struct notifier_block fn##_nb = { \
+ .notifier_call = fn, \
+ .priority = pri \
+ }; \
+ \
+ serial8250_register_notifier(&fn##_nb); \
+})
+
#endif
--
1.7.2.3
next prev parent reply other threads:[~2011-03-17 1:26 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-03-17 1:26 [RFC PATCH 0/2] serial: Hack up 8520.c for evil device tree hookin David Daney
2011-03-17 1:26 ` David Daney [this message]
2011-03-17 12:18 ` [RFC PATCH 1/2] serial: 8250: Add a notifier chain for driver registration Alan Cox
2011-03-17 16:42 ` David Daney
2011-03-17 18:25 ` Grant Likely
2011-03-17 18:42 ` David Daney
2011-03-17 18:47 ` Grant Likely
2011-03-17 19:24 ` Alan Cox
2011-03-17 19:31 ` Grant Likely
2011-03-17 20:13 ` David Daney
2011-03-17 20:31 ` Grant Likely
2011-03-17 23:48 ` Alan Cox
2011-03-18 5:18 ` Grant Likely
2011-03-17 1:26 ` [RFC PATCH 2/2] MIPS: Octeon: Use device tree to register serial ports David Daney
2011-03-17 18:28 ` Grant Likely
2011-03-17 18:35 ` David Daney
2011-03-17 1:46 ` [RFC PATCH 0/2] serial: Hack up 8520.c for evil device tree hookin Alan Cox
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=1300325167-26433-2-git-send-email-ddaney@caviumnetworks.com \
--to=ddaney@caviumnetworks.com \
--cc=devicetree-discuss@lists.ozlabs.org \
--cc=grant.likely@secretlab.ca \
--cc=gregkh@suse.de \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-serial@vger.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
all inboxes | Powered by JetHome®