[PATCH 1/2] nbd: use list_for_each_entry_safe to make it more consolidated and readable

[PATCH 1/2] nbd: use list_for_each_entry_safe to make it more consolidated and readable

[Denis Cheng]

Thus the traverse of the loop may delete nodes, use the safe version.

Signed-off-by: Denis Cheng <crquan@gmail.com>
---
 drivers/block/nbd.c |    6 ++----
 1 files changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index c129510..86639c0 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -237,8 +237,7 @@ error_out:
 
 static struct request *nbd_find_request(struct nbd_device *lo, char *handle)
 {
-	struct request *req;
-	struct list_head *tmp;
+	struct request *req, *n;
 	struct request *xreq;
 	int err;
 
@@ -249,8 +248,7 @@ static struct request *nbd_find_request(struct nbd_device *lo, char *handle)
 		goto out;
 
 	spin_lock(&lo->queue_lock);
-	list_for_each(tmp, &lo->queue_head) {
-		req = list_entry(tmp, struct request, queuelist);
+	list_for_each_entry_safe(req, n, &lo->queue_head, queuelist) {
 		if (req != xreq)
 			continue;
 		list_del_init(&req->queuelist);
-- 
1.5.2.2

[PATCH 2/2] nbd: change a parameter's type to remove a memcpy call

[Denis Cheng]

this memcpy looks so strange, in fact it's merely a pointer dereference,
so I change the parameter's type to refer it more directly,
this could make the memcpy not needed anymore.

in the function nbd_read_stat where nbd_find_request is only once called,
the parameter served should be transformed accordingly.

Signed-off-by: Denis Cheng <crquan@gmail.com>
---
 drivers/block/nbd.c |    7 ++-----
 1 files changed, 2 insertions(+), 5 deletions(-)

diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index 86639c0..a4d8508 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -235,14 +235,11 @@ error_out:
 	return 1;
 }
 
-static struct request *nbd_find_request(struct nbd_device *lo, char *handle)
+static struct request *nbd_find_request(struct nbd_device *lo, struct request *xreq)
 {
 	struct request *req, *n;
-	struct request *xreq;
 	int err;
 
-	memcpy(&xreq, handle, sizeof(xreq));
-
 	err = wait_event_interruptible(lo->active_wq, lo->active_req != xreq);
 	if (unlikely(err))
 		goto out;
@@ -297,7 +294,7 @@ static struct request *nbd_read_stat(struct nbd_device *lo)
 		goto harderror;
 	}
 
-	req = nbd_find_request(lo, reply.handle);
+	req = nbd_find_request(lo, *(struct request **)reply.handle);
 	if (unlikely(IS_ERR(req))) {
 		result = PTR_ERR(req);
 		if (result != -ENOENT)
-- 
1.5.2.2

Re: [PATCH 1/2] nbd: use list_for_each_entry_safe to make it more consolidated and readable

[Paul Clements]

Denis Cheng wrote:
> Thus the traverse of the loop may delete nodes, use the safe version.
> 
> Signed-off-by: Denis Cheng <crquan@gmail.com>
> ---
>  drivers/block/nbd.c |    6 ++----
>  1 files changed, 2 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
> index c129510..86639c0 100644
> --- a/drivers/block/nbd.c
> +++ b/drivers/block/nbd.c
> @@ -237,8 +237,7 @@ error_out:
>  
>  static struct request *nbd_find_request(struct nbd_device *lo, char *handle)
>  {
> -	struct request *req;
> -	struct list_head *tmp;
> +	struct request *req, *n;
>  	struct request *xreq;
>  	int err;
>  
> @@ -249,8 +248,7 @@ static struct request *nbd_find_request(struct nbd_device *lo, char *handle)
>  		goto out;
>  
>  	spin_lock(&lo->queue_lock);
> -	list_for_each(tmp, &lo->queue_head) {
> -		req = list_entry(tmp, struct request, queuelist);
> +	list_for_each_entry_safe(req, n, &lo->queue_head, queuelist) {
>  		if (req != xreq)
>  			continue;
>  		list_del_init(&req->queuelist);

Could you name "n" as "tmp" (as in the previous code) so that it's clear 
that's only a temporary variable. Other than that, this looks good.

Thanks,
Paul

Re: [PATCH 1/2] nbd: use list_for_each_entry_safe to make it more consolidated and readable

[rae l]

On 7/19/07, Paul Clements <paul.clements@steeleye.com> wrote:
> Could you name "n" as "tmp" (as in the previous code) so that it's clear
> that's only a temporary variable. Other than that, this looks good.
Sure. I just use the name "n" as in the declaration of
list_for_each_entry_safe in the header file <linux/list.h>
I'll resend it a little later.

>
> Thanks,
> Paul
>

-- 
Denis Cheng
Linux Application Developer

"One of my most productive days was throwing away 1000 lines of code."
 - Ken Thompson.

Re: [PATCH 1/2] nbd: use list_for_each_entry_safe to make it more consolidated and readable

[rae l]

On 7/19/07, rae l <crquan@gmail.com> wrote:
> On 7/19/07, Paul Clements <paul.clements@steeleye.com> wrote:
> > Could you name "n" as "tmp" (as in the previous code) so that it's clear
> > that's only a temporary variable. Other than that, this looks good.
> Sure. I just use the name "n" as in the declaration of
> list_for_each_entry_safe in the header file <linux/list.h>
> I'll resend it a little later.

Signed-off-by: Denis Cheng <crquan@gmail.com>
---
 drivers/block/nbd.c |    6 ++----
 1 files changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index c129510..86639c0 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -237,8 +237,7 @@ error_out:

 static struct request *nbd_find_request(struct nbd_device *lo, char *handle)
 {
-	struct request *req;
-	struct list_head *tmp;
+	struct request *req, *tmp;
 	struct request *xreq;
 	int err;

@@ -249,8 +248,7 @@ static struct request *nbd_find_request(struct
nbd_device *lo, char *handle)
 		goto out;

 	spin_lock(&lo->queue_lock);
-	list_for_each(tmp, &lo->queue_head) {
-		req = list_entry(tmp, struct request, queuelist);
+	list_for_each_entry_safe(req, tmp, &lo->queue_head, queuelist) {
 		if (req != xreq)
 			continue;
 		list_del_init(&req->queuelist);

-- 
Denis Cheng
Linux Application Developer

"One of my most productive days was throwing away 1000 lines of code."
 - Ken Thompson.

Re: [PATCH 2/2] nbd: change a parameter's type to remove a memcpy call

[Paul Clements]

Denis Cheng wrote:
> this memcpy looks so strange, in fact it's merely a pointer dereference,
> so I change the parameter's type to refer it more directly,
> this could make the memcpy not needed anymore.
> 
> in the function nbd_read_stat where nbd_find_request is only once called,
> the parameter served should be transformed accordingly.

This is really a matter of preference. The generated code ends up being 
about the same, I think, while your patch makes the call to 
nbd_find_request kind of obtuse. Also, the memcpy's are balanced between 
send_req and find_request, so you can quickly see how the data is being 
transferred (from req into handle, and then back again). Your patch 
makes this less clear, at least to me.

--
Paul


> Signed-off-by: Denis Cheng <crquan@gmail.com>
> ---
>  drivers/block/nbd.c |    7 ++-----
>  1 files changed, 2 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
> index 86639c0..a4d8508 100644
> --- a/drivers/block/nbd.c
> +++ b/drivers/block/nbd.c
> @@ -235,14 +235,11 @@ error_out:
>  	return 1;
>  }
>  
> -static struct request *nbd_find_request(struct nbd_device *lo, char *handle)
> +static struct request *nbd_find_request(struct nbd_device *lo, struct request *xreq)
>  {
>  	struct request *req, *n;
> -	struct request *xreq;
>  	int err;
>  
> -	memcpy(&xreq, handle, sizeof(xreq));
> -
>  	err = wait_event_interruptible(lo->active_wq, lo->active_req != xreq);
>  	if (unlikely(err))
>  		goto out;
> @@ -297,7 +294,7 @@ static struct request *nbd_read_stat(struct nbd_device *lo)
>  		goto harderror;
>  	}
>  
> -	req = nbd_find_request(lo, reply.handle);
> +	req = nbd_find_request(lo, *(struct request **)reply.handle);
>  	if (unlikely(IS_ERR(req))) {
>  		result = PTR_ERR(req);
>  		if (result != -ENOENT)

Re: [PATCH 2/2] nbd: change a parameter's type to remove a memcpy call

[rae l]

On 7/20/07, Paul Clements <paul.clements@steeleye.com> wrote:
> Denis Cheng wrote:
> > this memcpy looks so strange, in fact it's merely a pointer dereference,
> > so I change the parameter's type to refer it more directly,
> > this could make the memcpy not needed anymore.
> >
> > in the function nbd_read_stat where nbd_find_request is only once called,
> > the parameter served should be transformed accordingly.
>
> This is really a matter of preference. The generated code ends up being
> about the same, I think, while your patch makes the call to
> nbd_find_request kind of obtuse. Also, the memcpy's are balanced between
> send_req and find_request, so you can quickly see how the data is being
> transferred (from req into handle, and then back again). Your patch
> makes this less clear, at least to me.
With one explicit memcpy stripped out, I think it's more clear to
nbd_find_request.
In nbd_read_stat, the cast to (struct request **) is not apparent, I
must admit; but I think the best solution is declaring other few
structs to make it clear, it's due to the lack of description of nbd
client and server communication protocol.

BTW, I think the nbd driver needs a clear documentation (its main site
http://nbd.sourceforge.net/ does not give it):

1. When nbd_find_request is needed to call, the 8 byte memory of char
handle[8] field in struct nbd_reply actually stores a pointer (struct
request *), that pointer is received from the network. Since a pointer
is only meaningful to the host, transfering it over the network will
be unreliable, I don't think it's a good design,

-- 
Denis Cheng
Linux Application Developer

"One of my most productive days was throwing away 1000 lines of code."
 - Ken Thompson.