[PATCH] bsg: minor bug fixes

[PATCH] bsg: minor bug fixes

[FUJITA Tomonori]

This fixes the following minor issues:

- When bsg_register_queue is called with a request queue that doesn't
have request_fn, it returns -EINVAL instead of zero.

- add EXPORT_SYMBOL_GPL for bsg_register_queue and
bsg_unregister_queue.

- shut up gcc warnings

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
---
 block/bsg.c         |    6 ++++--
 include/linux/bsg.h |    2 +-
 2 files changed, 5 insertions(+), 3 deletions(-)

diff --git a/block/bsg.c b/block/bsg.c
index 4ef3cc5..53bf6d4 100644
--- a/block/bsg.c
+++ b/block/bsg.c
@@ -961,8 +961,9 @@ void bsg_unregister_queue(struct request
 	bsg_device_nr--;
 	mutex_unlock(&bsg_mutex);
 }
+EXPORT_SYMBOL_GPL(bsg_unregister_queue);
 
-int bsg_register_queue(struct request_queue *q, char *name)
+int bsg_register_queue(struct request_queue *q, const char *name)
 {
 	struct bsg_class_device *bcd, *__bcd;
 	dev_t dev;
@@ -973,7 +974,7 @@ int bsg_register_queue(struct request_qu
 	 * we need a proper transport to send commands, not a stacked device
 	 */
 	if (!q->request_fn)
-		return 0;
+		return -EINVAL;
 
 	bcd = &q->bsg_dev;
 	memset(bcd, 0, sizeof(*bcd));
@@ -1025,6 +1026,7 @@ err:
 	mutex_unlock(&bsg_mutex);
 	return ret;
 }
+EXPORT_SYMBOL_GPL(bsg_register_queue);
 
 static int bsg_add(struct class_device *cl_dev, struct class_interface *cl_intf)
 {
diff --git a/include/linux/bsg.h b/include/linux/bsg.h
index 0475a6d..0dd01f9 100644
--- a/include/linux/bsg.h
+++ b/include/linux/bsg.h
@@ -51,7 +51,7 @@ struct bsg_class_device {
 	struct request_queue *queue;
 };
 
-extern int bsg_register_queue(struct request_queue *, char *);
+extern int bsg_register_queue(struct request_queue *, const char *);
 extern void bsg_unregister_queue(struct request_queue *);
 #else
 struct bsg_class_device { };
-- 
1.4.3.2

Re: [PATCH] bsg: minor bug fixes

[Jens Axboe]

On Fri, Mar 30 2007, FUJITA Tomonori wrote:
> This fixes the following minor issues:
> 
> - When bsg_register_queue is called with a request queue that doesn't
> have request_fn, it returns -EINVAL instead of zero.

Seems a little backwards, you wont be able to talk to the device then.
So, what is the purpose of this patch?

> - add EXPORT_SYMBOL_GPL for bsg_register_queue and
> bsg_unregister_queue.

ok

> - shut up gcc warnings

ok


-- 
Jens Axboe

Re: [PATCH] bsg: minor bug fixes

[FUJITA Tomonori]

From: Jens Axboe <jens.axboe@oracle.com>
Subject: Re: [PATCH] bsg: minor bug fixes
Date: Fri, 30 Mar 2007 08:38:04 +0200

> On Fri, Mar 30 2007, FUJITA Tomonori wrote:
> > This fixes the following minor issues:
> > 
> > - When bsg_register_queue is called with a request queue that doesn't
> > have request_fn, it returns -EINVAL instead of zero.
> 
> Seems a little backwards, you wont be able to talk to the device then.

What do you mean? All the devices that we might bind to bsg has
request_fn?


> So, what is the purpose of this patch?

When I wrote a SMP pass through interface via bsg, I thought that it
would be better to return an error instead of success when we call
bsg_register_queue with a request queue that doesn't have a
request_fn.


Anyway, it doesn't matter much. Here is a new version.

---
>From db616d88a5e180a6be4928ba66947e501da581fa Mon Sep 17 00:00:00 2001
From: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Date: Fri, 30 Mar 2007 17:24:35 +0900
Subject: [PATCH] bsg: minor bug fixes

This fixes the following minor issues:

- add EXPORT_SYMBOL_GPL for bsg_register_queue and
bsg_unregister_queue.

- shut up gcc warnings

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
---
 block/bsg.c         |    4 +++-
 include/linux/bsg.h |    2 +-
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/block/bsg.c b/block/bsg.c
index 4ef3cc5..a333c93 100644
--- a/block/bsg.c
+++ b/block/bsg.c
@@ -961,8 +961,9 @@ void bsg_unregister_queue(struct request
 	bsg_device_nr--;
 	mutex_unlock(&bsg_mutex);
 }
+EXPORT_SYMBOL_GPL(bsg_unregister_queue);
 
-int bsg_register_queue(struct request_queue *q, char *name)
+int bsg_register_queue(struct request_queue *q, const char *name)
 {
 	struct bsg_class_device *bcd, *__bcd;
 	dev_t dev;
@@ -1025,6 +1026,7 @@ err:
 	mutex_unlock(&bsg_mutex);
 	return ret;
 }
+EXPORT_SYMBOL_GPL(bsg_register_queue);
 
 static int bsg_add(struct class_device *cl_dev, struct class_interface *cl_intf)
 {
diff --git a/include/linux/bsg.h b/include/linux/bsg.h
index 0475a6d..0dd01f9 100644
--- a/include/linux/bsg.h
+++ b/include/linux/bsg.h
@@ -51,7 +51,7 @@ struct bsg_class_device {
 	struct request_queue *queue;
 };
 
-extern int bsg_register_queue(struct request_queue *, char *);
+extern int bsg_register_queue(struct request_queue *, const char *);
 extern void bsg_unregister_queue(struct request_queue *);
 #else
 struct bsg_class_device { };
-- 
1.4.3.2

Re: [PATCH] bsg: minor bug fixes

[Jens Axboe]

On Fri, Mar 30 2007, FUJITA Tomonori wrote:
> From: Jens Axboe <jens.axboe@oracle.com>
> Subject: Re: [PATCH] bsg: minor bug fixes
> Date: Fri, 30 Mar 2007 08:38:04 +0200
> 
> > On Fri, Mar 30 2007, FUJITA Tomonori wrote:
> > > This fixes the following minor issues:
> > > 
> > > - When bsg_register_queue is called with a request queue that doesn't
> > > have request_fn, it returns -EINVAL instead of zero.
> > 
> > Seems a little backwards, you wont be able to talk to the device then.
> 
> What do you mean? All the devices that we might bind to bsg has
> request_fn?

They have to, how else would you send commands to it?

> > So, what is the purpose of this patch?
> 
> When I wrote a SMP pass through interface via bsg, I thought that it
> would be better to return an error instead of success when we call
> bsg_register_queue with a request queue that doesn't have a
> request_fn.
> 
> 
> Anyway, it doesn't matter much. Here is a new version.

Aplied this one.

-- 
Jens Axboe