Skip to contentSkip to conversation

dlm: remove unnecessary error check

PATCH In review Guoqing Jiang 3 participants 11 messages Jun 9, 2015 – Jun 17, 2015 Distributed lock manager (DLM)

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 dlm: remove unnecessary error check

We don't need the redundant logic since send_message always returns 0.

---
 fs/dlm/lock.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)
fs/dlm/lock.c +2 -8
index 35502d4..6fc3de9 100644
@@ -3656,10 +3656,7 @@ static int send_common(struct dlm_rsb *r, struct dlm_lkb *lkb, int mstype)
 
 	send_args(r, lkb, ms);
 
-	error = send_message(mh, ms);
-	if (error)
-		goto fail;
-	return 0;
+	return send_message(mh, ms);
 
  fail:
 	remove_from_waiters(lkb, msg_reply_type(mstype));
@@ -3763,10 +3760,7 @@ static int send_lookup(struct dlm_rsb *r, struct dlm_lkb *lkb)
 
 	send_args(r, lkb, ms);
 
-	error = send_message(mh, ms);
-	if (error)
-		goto fail;
-	return 0;
+	return send_message(mh, ms);
 
  fail:
 	remove_from_waiters(lkb, DLM_MSG_LOOKUP_REPLY);
Signature · 1 line
1.7.12.4
#1 lore raw

----- Original Message -----

Quoted 36 lines

We don't need the redundant logic since send_message always returns 0.

---
 fs/dlm/lock.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)
fs/dlm/lock.c +1 -4
index 35502d4..6fc3de9 100644
@@ -3656,10 +3656,7 @@ static int send_common(struct dlm_rsb *r, struct
 dlm_lkb *lkb, int mstype)
 
 	send_args(r, lkb, ms);
 
-	error = send_message(mh, ms);
-	if (error)
-		goto fail;
-	return 0;
+	return send_message(mh, ms);
 
  fail:
fs/dlm/lock.c +1 -4
 	remove_from_waiters(lkb, msg_reply_type(mstype));
@@ -3763,10 +3760,7 @@ static int send_lookup(struct dlm_rsb *r, struct
 dlm_lkb *lkb)
 
 	send_args(r, lkb, ms);
 
-	error = send_message(mh, ms);
-	if (error)
-		goto fail;
-	return 0;
+	return send_message(mh, ms);
 
  fail:
Quoted 3 lines

remove_from_waiters(lkb, DLM_MSG_LOOKUP_REPLY);
--
1.7.12.4

Hi,

The patch looks okay, but if remove_from_waiters() always returns 0, wouldn't it be better to change the function from int to void and return 0 here? The advantage is that code spelunkers wouldn't need to back-track one more level (not to mention the instruction or two it might save).

Regards,

Bob Peterson
Red Hat File Systems

#2 lore raw

Hi Bob,

Quoted 45 lines · Bob Peterson

----- Original Message -----

Quoted 36 lines

We don't need the redundant logic since send_message always returns 0.

---
 fs/dlm/lock.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)
fs/dlm/lock.c +1 -4
index 35502d4..6fc3de9 100644
@@ -3656,10 +3656,7 @@ static int send_common(struct dlm_rsb *r, struct
 dlm_lkb *lkb, int mstype)
 
 	send_args(r, lkb, ms);
 
-	error = send_message(mh, ms);
-	if (error)
-		goto fail;
-	return 0;
+	return send_message(mh, ms);
 
  fail:
fs/dlm/lock.c +1 -4
 	remove_from_waiters(lkb, msg_reply_type(mstype));
@@ -3763,10 +3760,7 @@ static int send_lookup(struct dlm_rsb *r, struct
 dlm_lkb *lkb)
 
 	send_args(r, lkb, ms);
 
-	error = send_message(mh, ms);
-	if (error)
-		goto fail;
-	return 0;
+	return send_message(mh, ms);
 
  fail:
Quoted 4 lines

remove_from_waiters(lkb, DLM_MSG_LOOKUP_REPLY);
--
1.7.12.4

Hi,

Quoted 7 lines

The patch looks okay, but if remove_from_waiters() always returns 0, wouldn't it be better to change the function from int to void and return 0 here? The advantage is that code spelunkers wouldn't need to back-track one more level (not to mention the instruction or two it might save).

Seems remove_from_waiters is not always returns 0, the return value could be -1 or 0 which depends on _remove_from_waiters.

BTW, I found that there are no big difference between send_common and send_lookup, since the send_common can also be use to send lookup message, I guess send_lookup can be removed as well.

Thanks,
Guoqing

#3 lore raw

----- Original Message -----

Quoted 58 lines

Hi Bob,

Quoted 45 lines · Bob Peterson

----- Original Message -----

Quoted 36 lines

We don't need the redundant logic since send_message always returns 0.

---
 fs/dlm/lock.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)
fs/dlm/lock.c +1 -4
index 35502d4..6fc3de9 100644
@@ -3656,10 +3656,7 @@ static int send_common(struct dlm_rsb *r, struct
 dlm_lkb *lkb, int mstype)
 
 	send_args(r, lkb, ms);
 
-	error = send_message(mh, ms);
-	if (error)
-		goto fail;
-	return 0;
+	return send_message(mh, ms);
 
  fail:
fs/dlm/lock.c +1 -4
 	remove_from_waiters(lkb, msg_reply_type(mstype));
@@ -3763,10 +3760,7 @@ static int send_lookup(struct dlm_rsb *r, struct
 dlm_lkb *lkb)
 
 	send_args(r, lkb, ms);
 
-	error = send_message(mh, ms);
-	if (error)
-		goto fail;
-	return 0;
+	return send_message(mh, ms);
 
  fail:
Quoted 4 lines

remove_from_waiters(lkb, DLM_MSG_LOOKUP_REPLY);
--
1.7.12.4

Hi,

Quoted 7 lines

The patch looks okay, but if remove_from_waiters() always returns 0, wouldn't it be better to change the function from int to void and return 0 here? The advantage is that code spelunkers wouldn't need to back-track one more level (not to mention the instruction or two it might save).

Seems remove_from_waiters is not always returns 0, the return value could be -1 or 0 which depends on _remove_from_waiters.

Quoted 6 lines

BTW, I found that there are no big difference between send_common and send_lookup, since the send_common can also be use to send lookup message, I guess send_lookup can be removed as well.

Thanks,
Guoqing

Hi Guoqing,

If remove_from_waiters can return -1, then the patch would prevent the code from calling remove_from_waiters. So the patch still doesn't look right to me.

Regards,

Bob Peterson
Red Hat File Systems

#4 lore raw
Quoted 71 lines · Bob Peterson

----- Original Message -----

Quoted 61 lines

Hi Bob,

Quoted 46 lines · Bob Peterson

----- Original Message -----

Quoted 36 lines

We don't need the redundant logic since send_message always returns 0.

---
 fs/dlm/lock.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)
fs/dlm/lock.c +1 -4
index 35502d4..6fc3de9 100644
@@ -3656,10 +3656,7 @@ static int send_common(struct dlm_rsb *r, struct
 dlm_lkb *lkb, int mstype)
 
 	send_args(r, lkb, ms);
 
-	error = send_message(mh, ms);
-	if (error)
-		goto fail;
-	return 0;
+	return send_message(mh, ms);
 
  fail:
fs/dlm/lock.c +1 -4
 	remove_from_waiters(lkb, msg_reply_type(mstype));
@@ -3763,10 +3760,7 @@ static int send_lookup(struct dlm_rsb *r, struct
 dlm_lkb *lkb)
 
 	send_args(r, lkb, ms);
 
-	error = send_message(mh, ms);
-	if (error)
-		goto fail;
-	return 0;
+	return send_message(mh, ms);
 
  fail:
Quoted 5 lines

remove_from_waiters(lkb, DLM_MSG_LOOKUP_REPLY);
--
1.7.12.4

Hi,

Quoted 8 lines

The patch looks okay, but if remove_from_waiters() always returns 0, wouldn't it be better to change the function from int to void and return 0 here? The advantage is that code spelunkers wouldn't need to back-track one more level (not to mention the instruction or two it might save).

Seems remove_from_waiters is not always returns 0, the return value could be -1 or 0 which depends on _remove_from_waiters.

Quoted 7 lines

BTW, I found that there are no big difference between send_common and send_lookup, since the send_common can also be use to send lookup message, I guess send_lookup can be removed as well.

Thanks,
Guoqing

Quoted 7 lines

Hi Guoqing,

If remove_from_waiters can return -1, then the patch would prevent the code from calling remove_from_waiters. So the patch still doesn't look right to me.

Hi Bob,

The remove_from_waiters could only be invoked after failed to create_message, right?
Since send_message always returns 0, this patch doesn't touch anything about the failure
path, and it also doesn't change the original semantic.

Thanks,
Guoqing

#5 lore raw

----- Original Message -----

Quoted 24 lines
Quoted 23 lines · Bob Peterson

----- Original Message -----

Quoted 21 lines

Hi Bob,

Quoted 17 lines · Bob Peterson

----- Original Message -----

Quoted 14 lines

We don't need the redundant logic since send_message always returns 0.

---
 fs/dlm/lock.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)
fs/dlm/lock.c +0 -0
index 35502d4..6fc3de9 100644
@@ -3656,10 +3656,7 @@ static int send_common(struct dlm_rsb *r, struct
 dlm_lkb *lkb, int mstype)
 
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

fs/dlm/lock.c static int send_common(struct dlm_rsb *r, struct dlm_lkb *lkb, int mstype) · line 3659

Quoted 7 lines
fs/dlm/lock.c +1 -4
 	send_args(r, lkb, ms);
 
-	error = send_message(mh, ms);
-	if (error)
-		goto fail;
-	return 0;
+	return send_message(mh, ms);

Hi Guoqing,

Sorry, I was momentarily confused. I think you misunderstood what I was saying. What I meant was: Instead of doing:

+ return send_message(mh, ms);
...where send_message returns 0, it might be better to have:

static void send_message(struct dlm_mhandle *mh, struct dlm_message *ms)

{
	dlm_message_out(ms);
	dlm_lowcomms_commit_buffer(mh);
}

...And in send_common, do (in both places):
+ send_message(mh, ms);
+ return 0;

Since it's so short, it might even be better to code send_message as a macro, or at least an "inline" function.

Regards,

Bob Peterson
Red Hat File Systems

#6 lore raw
Quoted 48 lines · Bob Peterson
Quoted 21 lines
Quoted 21 lines
Quoted 21 lines
Quoted 21 lines

----- Original Message -----

Quoted 14 lines

We don't need the redundant logic since send_message always returns 0.

---
 fs/dlm/lock.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)
fs/dlm/lock.c +0 -0
index 35502d4..6fc3de9 100644
@@ -3656,10 +3656,7 @@ static int send_common(struct dlm_rsb *r, struct
 dlm_lkb *lkb, int mstype)
 
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

fs/dlm/lock.c static int send_common(struct dlm_rsb *r, struct dlm_lkb *lkb, int mstype) · line 3660

Quoted 8 lines
fs/dlm/lock.c +1 -4
 	send_args(r, lkb, ms);
 
-	error = send_message(mh, ms);
-	if (error)
-		goto fail;
-	return 0;
+	return send_message(mh, ms);
 

Hi Guoqing,

Sorry, I was momentarily confused. I think you misunderstood what I was saying. What I meant was: Instead of doing:

(quoted diff) +1 -0
+	return send_message(mh, ms);

...where send_message returns 0, it might be better to have:

static void send_message(struct dlm_mhandle *mh, struct dlm_message *ms)

{
	dlm_message_out(ms);
	dlm_lowcomms_commit_buffer(mh);
}

...And in send_common, do (in both places):

(quoted diff) +2 -0
+	send_message(mh, ms);
+	return 0;
Quoted 4 lines

Since it's so short, it might even be better to code send_message as a macro, or at least an "inline" function.

Hi Bob,

Got it, thanks. It is a better solution but it is not a bug fix or similar thing, so maybe just leave it as it is.

Regards,
Guoqing

maintainer
#7 lore raw
In reply to Guoqing Jiang
Quoted 5 lines · Guoqing Jiang

The remove_from_waiters could only be invoked after failed to create_message, right?
Since send_message always returns 0, this patch doesn't touch anything about the failure
path, and it also doesn't change the original semantic.

I'm not inclined to take any patches unless there's a problem identified.

#8 lore raw

Hi David,

Quoted 9 lines · David Teigland
Quoted 6 lines · Guoqing Jiang

The remove_from_waiters could only be invoked after failed to create_message, right?
Since send_message always returns 0, this patch doesn't touch anything about the failure
path, and it also doesn't change the original semantic.

Quoted 4 lines

I'm not inclined to take any patches unless there's a problem identified.

.

Do you consider take the following clean up? If yes, I will send a formal patch,
otherwise pls ignore it.

fs/dlm/lock.c +0 -0
index 35502d4..7c822f7 100644
@@ -3747,30 +3747,7 @@ static int send_bast(struct dlm_rsb *r, struct

dlm_lkb *lkb, int mode)

static int send_lookup(struct dlm_rsb *r, struct dlm_lkb *lkb) {

-       struct dlm_message *ms;
-       struct dlm_mhandle *mh;
-       int to_nodeid, error;

-
- to_nodeid = dlm_dir_nodeid(r);
-
- error = add_to_waiters(lkb, DLM_MSG_LOOKUP, to_nodeid);
- if (error)
- return error;
-
- error = create_message(r, NULL, to_nodeid, DLM_MSG_LOOKUP, &ms,
&mh);
- if (error)
- goto fail;
-
- send_args(r, lkb, ms);
-
- error = send_message(mh, ms);
- if (error)
- goto fail;
- return 0;
-
- fail:

-       remove_from_waiters(lkb, DLM_MSG_LOOKUP_REPLY);
-       return error;
+       return send_common(r, lkb, DLM_MSG_LOOKUP);

}

Thanks,
Guoqing

maintainer
#9 lore raw
Quoted 2 lines · Guoqing Jiang

Do you consider take the following clean up? If yes, I will send a formal patch, otherwise pls ignore it.

On first glance, the old and new code do not appear to do the same thing, so let's leave it as it is.

Quoted 1 line

- to_nodeid = dlm_dir_nodeid(r);

#10 lore raw

Hi David,

Quoted 12 lines · David Teigland
Quoted 3 lines · Guoqing Jiang

Do you consider take the following clean up? If yes, I will send a formal patch, otherwise pls ignore it.

On first glance, the old and new code do not appear to do the same thing, so let's leave it as it is.

Quoted 2 lines

- to_nodeid = dlm_dir_nodeid(r);

Sorry, seems it is the only different thing, if combines previous change with below modification, then the behavior is same.

@@ -3644,7 +3644,10 @@ static int send_common(struct dlm_rsb *r, struct dlm_lkb *lkb, int mstype)

        struct dlm_mhandle *mh;
        int to_nodeid, error;

- to_nodeid = r->res_nodeid;
+ if (mstype == DLM_MSG_LOOKUP)
+ to_nodeid = dlm_dir_nodeid(r);
+ else
+ to_nodeid = r->res_nodeid;

And for create_message, the second parameter (lkb) is not effective to create three type msgs (REQUEST/LOOKUP/REMOVE).

Thanks,
Guoqing