Skip to contentSkip to conversation

staging: r8188eu: use completions and clean rtw_cmd_thread()

PATCH Superseded Fabio M. De Francesco 4 participants 10 messages Oct 16, 2021 – Oct 17, 2021 Staging

v1 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

cover staging: r8188eu: use completions and clean rtw_cmd_thread()

This series replaces two semaphores with three completion variables in rtw_cmd_thread(). Completions variables are better suited for the purposes that are explained in detail in the commit messages of patches 1/3 and 2/3. Furthermore, patch 3/3 removes a redundant 'if' statement from that same rtw_cmd_thread().

Tested with ASUSTek Computer, Inc. Realtek 8188EUS [USB-N10 Nano]

Many thanks to Dan Carpenter <[email protected]> who helped with his review of the RFC Patch.

Fabio M. De Francesco (3):
staging: r8188eu: Use completions for signaling start and end of
kthread
staging: r8188eu: Use completions for signaling enqueueing of commands
staging: r8188eu: Remove redundant 'if' statement

 drivers/staging/r8188eu/core/rtw_cmd.c    | 19 +++++++------------
 drivers/staging/r8188eu/include/rtw_cmd.h |  5 +++--
 drivers/staging/r8188eu/os_dep/os_intfs.c |  8 +++++---
 3 files changed, 15 insertions(+), 17 deletions(-)
Signature · 1 line
2.33.0
#1 lore raw

PATCH 1/3 staging: r8188eu: Use completions for signaling start and end kthread

rtw_cmd_thread() "up(s)" a semaphore twice, first to notify callers when its execution has started and then to notify when it is about to end.

It makes the same semaphore go "up" twice in the same kthread. This construct makes Smatch to warn of duplicate "up(s)".

This kthread uses interruptible semaphores where instead completions are more suitable. For this purpose it calls an helper (_rtw_down_sema()) that returns values that are never checked. It may lead to bugs.

To address the above-mentioned issues, use two completions variables instead of semaphores. Use the uninterruptible versions of wait_for_completion*() because the interruptible / killable versions are not necessary.

Tested with "ASUSTek Computer, Inc. Realtek 8188EUS [USB-N10 Nano]".

---
 drivers/staging/r8188eu/core/rtw_cmd.c    | 7 ++++---
 drivers/staging/r8188eu/include/rtw_cmd.h | 3 ++-
 drivers/staging/r8188eu/os_dep/os_intfs.c | 6 ++++--
 3 files changed, 10 insertions(+), 6 deletions(-)
drivers/staging/r8188eu/core/rtw_cmd.c +4 -3
index e17332677daa..195390449502 100644
@@ -23,7 +23,8 @@ static int _rtw_init_cmd_priv(struct cmd_priv *pcmdpriv)
 
 	sema_init(&pcmdpriv->cmd_queue_sema, 0);
 	/* sema_init(&(pcmdpriv->cmd_done_sema), 0); */
-	sema_init(&pcmdpriv->terminate_cmdthread_sema, 0);
+	init_completion(&pcmdpriv->start_cmd_thread);
+	init_completion(&pcmdpriv->stop_cmd_thread);
 
 	rtw_init_queue(&pcmdpriv->cmd_queue);
 
@@ -248,7 +249,7 @@ int rtw_cmd_thread(void *context)
 	pcmdbuf = pcmdpriv->cmd_buf;
 
 	pcmdpriv->cmdthd_running = true;
-	up(&pcmdpriv->terminate_cmdthread_sema);
+	complete(&pcmdpriv->start_cmd_thread);
 
 	while (1) {
 		if (_rtw_down_sema(&pcmdpriv->cmd_queue_sema) == _FAIL)
@@ -329,7 +330,7 @@ int rtw_cmd_thread(void *context)
 		rtw_free_cmd_obj(pcmd);
 	} while (1);
 
-	up(&pcmdpriv->terminate_cmdthread_sema);
+	complete(&pcmdpriv->stop_cmd_thread);
 
 	thread_exit();
 }
drivers/staging/r8188eu/include/rtw_cmd.h +2 -1
index 83fbb922db2c..b6266e3e2c40 100644
@@ -34,7 +34,8 @@ struct cmd_obj {
 
 struct cmd_priv {
 	struct semaphore cmd_queue_sema;
-	struct semaphore terminate_cmdthread_sema;
+	struct completion start_cmd_thread;
+	struct completion stop_cmd_thread;
 	struct __queue cmd_queue;
 	u8	cmd_seq;
 	u8	*cmd_buf;	/* shall be non-paged, and 4 bytes aligned */
drivers/staging/r8188eu/os_dep/os_intfs.c +4 -2
index e7964a048c99..0bcea66f550b 100644
@@ -385,7 +385,8 @@ u32 rtw_start_drv_threads(struct adapter *padapter)
 	if (IS_ERR(padapter->cmdThread))
 		_status = _FAIL;
 	else
-		_rtw_down_sema(&padapter->cmdpriv.terminate_cmdthread_sema); /* wait for cmd_thread to run */
+		/* wait for rtw_cmd_thread() to start running */
+		wait_for_completion(&padapter->cmdpriv.start_cmd_thread);
 
 	return _status;
 }
@@ -395,7 +396,8 @@ void rtw_stop_drv_threads(struct adapter *padapter)
 	/* Below is to termindate rtw_cmd_thread & event_thread... */
 	up(&padapter->cmdpriv.cmd_queue_sema);
 	if (padapter->cmdThread)
-		_rtw_down_sema(&padapter->cmdpriv.terminate_cmdthread_sema);
+		/* wait for rtw_cmd_thread() to stop running */
+		wait_for_completion(&padapter->cmdpriv.stop_cmd_thread);
 }
 
 static u8 rtw_init_default_value(struct adapter *padapter)
Signature · 1 line
2.33.0
#2 lore raw

Re: [PATCH 1/3] staging: r8188eu: Use completions for signaling start and end kthread

Yeah, you're right. This is fine. It's a bit confusing how Arnd re-used complete to start the function.

regards,
dan carpenter

#3 lore raw

Re: [PATCH 1/3] staging: r8188eu: Use completions for signaling start and end kthread

Quoted 6 lines · CEST Dan Carpenter

Yeah, you're right. This is fine. It's a bit confusing how Arnd re-used complete to start the function.

regards,
dan carpenter

Thanks, Dan.

What still confuses me is why rtw_start_drv_threads() wants to be notified that the kthread has started. Can you please answer this question?

Regards,

Fabio

#4 lore raw

PATCH 2/3 staging: r8188eu: Use completions for signaling enqueueing

In reply to Fabio M. De Francesco

rtw_enqueue_cmd() uses a semaphore to notify rtw_cmd_thread() that it has enqueued commands. rtw_cmd_thread() "down(s)" in interruptible mode to wait to be notified.

Use completion variables because they are better suited for the purpose.

In rtw_cmd_thread(), wait in killable mode because there are no other reasons to interrupt waiting if not for killing.

---
 drivers/staging/r8188eu/core/rtw_cmd.c    | 6 +++---
 drivers/staging/r8188eu/include/rtw_cmd.h | 2 +-
 drivers/staging/r8188eu/os_dep/os_intfs.c | 2 +-
 3 files changed, 5 insertions(+), 5 deletions(-)
drivers/staging/r8188eu/core/rtw_cmd.c +3 -3
index 195390449502..2900c31d8ec9 100644
@@ -21,7 +21,7 @@ static int _rtw_init_cmd_priv(struct cmd_priv *pcmdpriv)
 {
 	int res = _SUCCESS;
 
-	sema_init(&pcmdpriv->cmd_queue_sema, 0);
+	init_completion(&pcmdpriv->enqueue_cmd);
 	/* sema_init(&(pcmdpriv->cmd_done_sema), 0); */
 	init_completion(&pcmdpriv->start_cmd_thread);
 	init_completion(&pcmdpriv->stop_cmd_thread);
@@ -198,7 +198,7 @@ u32 rtw_enqueue_cmd(struct cmd_priv *pcmdpriv, struct cmd_obj *cmd_obj)
 	res = _rtw_enqueue_cmd(&pcmdpriv->cmd_queue, cmd_obj);
 
 	if (res == _SUCCESS)
-		up(&pcmdpriv->cmd_queue_sema);
+		complete(&pcmdpriv->enqueue_cmd);
 
 exit:
 
@@ -252,7 +252,7 @@ int rtw_cmd_thread(void *context)
 	complete(&pcmdpriv->start_cmd_thread);
 
 	while (1) {
-		if (_rtw_down_sema(&pcmdpriv->cmd_queue_sema) == _FAIL)
+		if (wait_for_completion_killable(&pcmdpriv->enqueue_cmd))
 			break;
 
 		if (padapter->bDriverStopped ||
drivers/staging/r8188eu/include/rtw_cmd.h +1 -1
index b6266e3e2c40..47c3c80cc24a 100644
@@ -33,7 +33,7 @@ struct cmd_obj {
 };
 
 struct cmd_priv {
-	struct semaphore cmd_queue_sema;
+	struct completion enqueue_cmd;
 	struct completion start_cmd_thread;
 	struct completion stop_cmd_thread;
 	struct __queue cmd_queue;
drivers/staging/r8188eu/os_dep/os_intfs.c +1 -1
index 0bcea66f550b..96e9897085fe 100644
@@ -394,7 +394,7 @@ u32 rtw_start_drv_threads(struct adapter *padapter)
 void rtw_stop_drv_threads(struct adapter *padapter)
 {
 	/* Below is to termindate rtw_cmd_thread & event_thread... */
-	up(&padapter->cmdpriv.cmd_queue_sema);
+	complete(&padapter->cmdpriv.enqueue_cmd);
 	if (padapter->cmdThread)
 		/* wait for rtw_cmd_thread() to stop running */
 		wait_for_completion(&padapter->cmdpriv.stop_cmd_thread);
Signature · 1 line
2.33.0
#5 lore raw

PATCH 3/3 staging: r8188eu: Remove redundant 'if' statement

In reply to Fabio M. De Francesco

Remove a redundant 'if' statement.

---
 drivers/staging/r8188eu/core/rtw_cmd.c | 6 ------
 1 file changed, 6 deletions(-)
drivers/staging/r8188eu/core/rtw_cmd.c +0 -6
index 2900c31d8ec9..2263e35b45cb 100644
@@ -255,12 +255,6 @@ int rtw_cmd_thread(void *context)
 		if (wait_for_completion_killable(&pcmdpriv->enqueue_cmd))
 			break;
 
-		if (padapter->bDriverStopped ||
-		    padapter->bSurpriseRemoved) {
-			DBG_88E("%s: DriverStopped(%d) SurpriseRemoved(%d) break at line %d\n",
-				__func__, padapter->bDriverStopped, padapter->bSurpriseRemoved, __LINE__);
-			break;
-		}
 _next:
 		if (padapter->bDriverStopped ||
 		    padapter->bSurpriseRemoved) {
Signature · 1 line
2.33.0
Acked-by on v1 3/3
#6 lore raw

Re: [PATCH 3/3] staging: r8188eu: Remove redundant 'if' statement

Thus wrote Fabio M. De Francesco ([email protected]):

Quoted 26 lines

Remove a redundant 'if' statement.

---
 drivers/staging/r8188eu/core/rtw_cmd.c | 6 ------
 1 file changed, 6 deletions(-)
drivers/staging/r8188eu/core/rtw_cmd.c +0 -6
index 2900c31d8ec9..2263e35b45cb 100644
@@ -255,12 +255,6 @@ int rtw_cmd_thread(void *context)
 		if (wait_for_completion_killable(&pcmdpriv->enqueue_cmd))
 			break;
 
-		if (padapter->bDriverStopped ||
-		    padapter->bSurpriseRemoved) {
-			DBG_88E("%s: DriverStopped(%d) SurpriseRemoved(%d) break at line %d\n",
-				__func__, padapter->bDriverStopped, padapter->bSurpriseRemoved, __LINE__);
-			break;
-		}
 _next:
 		if (padapter->bDriverStopped ||
 		    padapter->bSurpriseRemoved) {
Signature · 1 line
2.33.0
Acked-by on the v1 series
#7 lore raw
In reply to Fabio M. De Francesco
Quoted 17 lines · Fabio M. De Francesco

This series replaces two semaphores with three completion variables in rtw_cmd_thread(). Completions variables are better suited for the purposes that are explained in detail in the commit messages of patches 1/3 and 2/3. Furthermore, patch 3/3 removes a redundant 'if' statement from that same rtw_cmd_thread().

Tested with ASUSTek Computer, Inc. Realtek 8188EUS [USB-N10 Nano]

Many thanks to Dan Carpenter <[email protected]> who helped with his review of the RFC Patch.

Fabio M. De Francesco (3):
staging: r8188eu: Use completions for signaling start and end of
kthread
staging: r8188eu: Use completions for signaling enqueueing of commands
staging: r8188eu: Remove redundant 'if' statement

Quoted 8 lines
 drivers/staging/r8188eu/core/rtw_cmd.c    | 19 +++++++------------
 drivers/staging/r8188eu/include/rtw_cmd.h |  5 +++--
 drivers/staging/r8188eu/os_dep/os_intfs.c |  8 +++++---
 3 files changed, 15 insertions(+), 17 deletions(-)
Signature · 1 line
2.33.0

Dear Fabio,

Built and tested on my USB-N10 Nano, working well. In terms of how you've split the patches out, I have no problem with it personally, given that one semaphore was there for kthread start/stop and the other for queuing. Looks good to me anyway based on what I know of completion variables :-) I assume you've not made the waits killable or interruptible in patch 1 due to the fact they are specifically related to kthread start/stop? Anyhow:

For whole series:

Regards,
Phil

#8 lore raw
Quoted 7 lines · CEST Phillip Potter
Quoted 6 lines · Fabio M. De Francesco

This series replaces two semaphores with three completion variables in rtw_cmd_thread(). Completions variables are better suited for the purposes that are explained in detail in the commit messages of patches 1/3 and 2/3. Furthermore, patch 3/3 removes a redundant 'if' statement from that same rtw_cmd_thread().

Quoted 7 lines
Quoted 7 lines

Tested with ASUSTek Computer, Inc. Realtek 8188EUS [USB-N10 Nano]

Many thanks to Dan Carpenter <[email protected]> who helped with his review of the RFC Patch.

[...]

Dear Phil,

Quoted 5 lines

Dear Fabio,

Built and tested on my USB-N10 Nano, working well. In terms of how you've split the patches out, I have no problem with it personally,

I guess that Dan will disagree with us :) Did you read his last message?

I hope that he has time to review these patches. He expressed some doubts about splitting the changes in two separate patches. As far as I know, since Dan is a very experienced engineer (I am not even graduated and everything I know of Computer Science is self-taught), I could have been wrong in doing this work the way I did.

Quoted 5 lines

given that one semaphore was there for kthread start/stop and the other for queuing. Looks good to me anyway based on what I know of completion variables :-) I assume you've not made the waits killable or interruptible in patch 1 due to the fact they are specifically related to kthread start/stop?

Good question! :)

Let me explain how I chose to make one wait killable and the other uninterruptible.

As far as I know, waiters may spin or sleep while waiting to acquire a lock (see spinlocks or mutexes for instance) or to be awakened (completions and condition variables for instance).

These were the cases of sleeping waiters. Sleeping can be done in uninterruptible, interruptible / killable, and timed-out states.

Where I'm 100% sure that the code doesn't require / want to be interrupted for whatever reason I prefer to use uninterruptible variants (and so I did in 1/3).

When I'm not sure of the requirement above, I prefer to avoid that the process or the entire system hangs while waiting to acquire a mutex or to be awakened by a complete() (and so on).

Conversely, using interruptible versions without proper checking of return codes and without proper managing of errors may lead to serious bugs.

Kernel threads (kthreads) are like user processes / threads and are scheduled the same way the former are. One noteworthy difference is that their mm pointer is NULL (they have not an userspace address spaces). However they are still threads that have a PID in userspace and they can be killed by root.

This is the output of the "ps -ef" command after "modprobe r8188eu":

localhost:~ # ps -ef | grep RTW
root 1726 2 0 19:06 ? 00:00:00 [RTW_CMD_THREAD]

Since the developers who wrote the original code thought that that thread must be interrupted I thought that restricting interruptions to kills was the wiser choice in 2/3. Conversely, I cannot see reasons to interrupt the core part of a driver, so I chose to use an uninterruptible version of wait_for_completion*() in the other parts of the code.

I warned you that I'm not an engineer, so please double check my argument :)

Quoted 4 lines

Anyhow:

For whole series:

Thanks for you ack. I really appreciated it.

Best regards,

Fabio

Quoted 4 lines

Regards,
Phil

#9 lore raw
Quoted 8 lines · Fabio M. De Francesco

I guess that Dan will disagree with us :) Did you read his last message?

I hope that he has time to review these patches. He expressed some doubts about splitting the changes in two separate patches. As far as I know, since Dan is a very experienced engineer (I am not even graduated and everything I know of Computer Science is self-taught), I could have been wrong in doing this work the way I did.

I did read it yes, he makes good points, and my motivation is simply that the patches look fine as they are to me personally :-)

Quoted 39 lines
Quoted 5 lines

given that one semaphore was there for kthread start/stop and the other for queuing. Looks good to me anyway based on what I know of completion variables :-) I assume you've not made the waits killable or interruptible in patch 1 due to the fact they are specifically related to kthread start/stop?

Good question! :)

Let me explain how I chose to make one wait killable and the other uninterruptible.

As far as I know, waiters may spin or sleep while waiting to acquire a lock (see spinlocks or mutexes for instance) or to be awakened (completions and condition variables for instance).

These were the cases of sleeping waiters. Sleeping can be done in uninterruptible, interruptible / killable, and timed-out states.

Where I'm 100% sure that the code doesn't require / want to be interrupted for whatever reason I prefer to use uninterruptible variants (and so I did in 1/3).

When I'm not sure of the requirement above, I prefer to avoid that the process or the entire system hangs while waiting to acquire a mutex or to be awakened by a complete() (and so on).

Conversely, using interruptible versions without proper checking of return codes and without proper managing of errors may lead to serious bugs.

Kernel threads (kthreads) are like user processes / threads and are scheduled the same way the former are. One noteworthy difference is that their mm pointer is NULL (they have not an userspace address spaces). However they are still threads that have a PID in userspace and they can be killed by root.

This is the output of the "ps -ef" command after "modprobe r8188eu":

localhost:~ # ps -ef | grep RTW
root 1726 2 0 19:06 ? 00:00:00 [RTW_CMD_THREAD]

Quoted 8 lines

Since the developers who wrote the original code thought that that thread must be interrupted I thought that restricting interruptions to kills was the wiser choice in 2/3. Conversely, I cannot see reasons to interrupt the core part of a driver, so I chose to use an uninterruptible version of wait_for_completion*() in the other parts of the code.

I warned you that I'm not an engineer, so please double check my argument :)

Sounds good to me, just wanted to know your reasoning.

Quoted 7 lines
Quoted 4 lines

Anyhow:

For whole series:

Thanks for you ack. I really appreciated it.

You're welcome :-)

Regards,
Phil