Skip to contentSkip to conversation

cifsd: fix WARNING: convert list_for_each to entry variant in smb2pdu.c

PATCH Superseded Baokun Li 2 participants 6 messages Jun 17, 2021 – Jun 18, 2021

Subsystems (3)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore

Baokun Li

author
#0 lore raw

PATCH 1/1 cifsd: fix WARNING: convert list_for_each to entry variant in smb2pdu.c

convert list_for_each() to list_for_each_entry() where applicable.

---
 fs/cifsd/smb2pdu.c | 15 ++++-----------
 1 file changed, 4 insertions(+), 11 deletions(-)
fs/cifsd/smb2pdu.c +4 -11
index ac15a9287310..22ef1d9eed1b 100644
@@ -74,10 +74,7 @@ static inline int check_session_id(struct ksmbd_conn *conn, u64 id)
 struct channel *lookup_chann_list(struct ksmbd_session *sess)
 {
 	struct channel *chann;
-	struct list_head *t;
-
-	list_for_each(t, &sess->ksmbd_chann_list) {
-		chann = list_entry(t, struct channel, chann_list);
+	list_for_each_entry(chann, &sess->ksmbd_chann_list, chann_list) {
 		if (chann && chann->conn == sess->conn)
 			return chann;
 	}
@@ -6258,7 +6255,6 @@ int smb2_cancel(struct ksmbd_work *work)
 	struct smb2_hdr *hdr = work->request_buf;
 	struct smb2_hdr *chdr;
 	struct ksmbd_work *cancel_work = NULL;
-	struct list_head *tmp;
 	int canceled = 0;
 	struct list_head *command_list;
 
@@ -6269,9 +6265,8 @@ int smb2_cancel(struct ksmbd_work *work)
 		command_list = &conn->async_requests;
 
 		spin_lock(&conn->request_lock);
-		list_for_each(tmp, command_list) {
-			cancel_work = list_entry(tmp, struct ksmbd_work,
-						 async_request_entry);
+		list_for_each_entry(cancel_work, command_list,
+				    async_request_entry) {
 			chdr = cancel_work->request_buf;
 
 			if (cancel_work->async_id !=
@@ -6290,9 +6285,7 @@ int smb2_cancel(struct ksmbd_work *work)
 		command_list = &conn->requests;
 
 		spin_lock(&conn->request_lock);
-		list_for_each(tmp, command_list) {
-			cancel_work = list_entry(tmp, struct ksmbd_work,
-						 request_entry);
+		list_for_each_entry(cancel_work, command_list, request_entry) {
 			chdr = cancel_work->request_buf;
 
 			if (chdr->MessageId != hdr->MessageId ||
#1 lore raw
Quoted 14 lines · Baokun Li

convert list_for_each() to list_for_each_entry() where applicable.

---
 fs/cifsd/smb2pdu.c | 15 ++++-----------
 1 file changed, 4 insertions(+), 11 deletions(-)
fs/cifsd/smb2pdu.c +0 -0
index ac15a9287310..22ef1d9eed1b 100644

fs/cifsd/smb2pdu.c static inline int check_session_id(struct ksmbd_conn *conn, u64 id) · line 78

Quoted 9 lines
fs/cifsd/smb2pdu.c +1 -4
@@ -74,10 +74,7 @@ static inline int check_session_id(struct ksmbd_conn *conn, u64 id)
 struct channel *lookup_chann_list(struct ksmbd_session *sess)
 {
 	struct channel *chann;
-	struct list_head *t;
-
-	list_for_each(t, &sess->ksmbd_chann_list) {
-		chann = list_entry(t, struct channel, chann_list);
+	list_for_each_entry(chann, &sess->ksmbd_chann_list, chann_list) {
 		if (chann && chann->conn == sess->conn)

"chan" is the list iterator and it can't be NULL.

fs/cifsd/smb2pdu.c static inline int check_session_id(struct ksmbd_conn *conn, u64 id) · line 80

Quoted 2 lines
fs/cifsd/smb2pdu.c +0 -0
 			return chann;
 	}

regards,
dan carpenter

#2 lore raw

I don't know what the difference is between

list_for_each_entry() and list_for_each() for 'struct channel *chann',

but I don't think there's any difference here.

Would you give me more detial about this, please?

Thank you.

Best Regards.

Quoted 32 lines · Dan Carpenter
Quoted 14 lines · Baokun Li

convert list_for_each() to list_for_each_entry() where applicable.

---
  fs/cifsd/smb2pdu.c | 15 ++++-----------
  1 file changed, 4 insertions(+), 11 deletions(-)
fs/cifsd/smb2pdu.c +0 -0
index ac15a9287310..22ef1d9eed1b 100644

fs/cifsd/smb2pdu.c static inline int check_session_id(struct ksmbd_conn *conn, u64 id) · line 78

Quoted 9 lines
fs/cifsd/smb2pdu.c +1 -4
@@ -74,10 +74,7 @@ static inline int check_session_id(struct ksmbd_conn *conn, u64 id)
  struct channel *lookup_chann_list(struct ksmbd_session *sess)
  {
  	struct channel *chann;
-	struct list_head *t;
-
-	list_for_each(t, &sess->ksmbd_chann_list) {
-		chann = list_entry(t, struct channel, chann_list);
+	list_for_each_entry(chann, &sess->ksmbd_chann_list, chann_list) {
  		if (chann && chann->conn == sess->conn)

"chan" is the list iterator and it can't be NULL.

fs/cifsd/smb2pdu.c static inline int check_session_id(struct ksmbd_conn *conn, u64 id) · line 80

Quoted 2 lines
fs/cifsd/smb2pdu.c +0 -0
  			return chann;
  	}

regards,
dan carpenter

.

#3 lore raw
Quoted 5 lines · libaokun (A)

I don't know what the difference is between

list_for_each_entry() and list_for_each() for 'struct channel *chann',

but I don't think there's any difference here.

Correct. There is no difference, but Coccinelle is smart enough to parse list_for_each_entry() and it's not smart enough to parse list_for_each().

Quoted 2 lines

Would you give me more detial about this, please?

There is a Coccinelle script scripts/coccinelle/iterators/itnull.cocci which will complain about the NULL check in the new code so this patch will introduce a new warning. We may as well remove the unnecessary NULL check and avoid the warning.

regards,
dan carpenter

#4 lore raw
Quoted 16 lines · Dan Carpenter
Quoted 5 lines · libaokun (A)

I don't know what the difference is between

list_for_each_entry() and list_for_each() for 'struct channel *chann',

but I don't think there's any difference here.

Correct. There is no difference, but Coccinelle is smart enough to parse list_for_each_entry() and it's not smart enough to parse list_for_each().

Quoted 1 line

Would you give me more detial about this, please?

There is a Coccinelle script scripts/coccinelle/iterators/itnull.cocci which will complain about the NULL check in the new code so this patch will introduce a new warning. We may as well remove the unnecessary NULL check and avoid the warning.

Quoted 4 lines

regards,
dan carpenter

.

I get your point, but this bug has nothing to do with my patch.

It's  from e2f34481b24d(cifsd: add server-side procedures for SMB3).

Thank you.

Signature · 2 lines
With Best Regards,
Baokun Li
#5 lore raw
Quoted 17 lines · libaokun (A)
Quoted 16 lines · Dan Carpenter
Quoted 5 lines · libaokun (A)

I don't know what the difference is between

list_for_each_entry() and list_for_each() for 'struct channel *chann',

but I don't think there's any difference here.

Correct. There is no difference, but Coccinelle is smart enough to parse list_for_each_entry() and it's not smart enough to parse list_for_each().

Quoted 1 line

Would you give me more detial about this, please?

There is a Coccinelle script scripts/coccinelle/iterators/itnull.cocci which will complain about the NULL check in the new code so this patch will introduce a new warning. We may as well remove the unnecessary NULL check and avoid the warning.

Quoted 9 lines
Quoted 4 lines

regards,
dan carpenter

.

I get your point, but this bug has nothing to do with my patch.

It's  from e2f34481b24d(cifsd: add server-side procedures for SMB3).

Fine, but when Hulk Robot reports the new warning and someone fixes it, I'm going to insist that the Fixes tag points to your patch. ;)

regards,
dan carpenter