mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
* [PATCH] usb: typec: anx7411 - stop the IRQ before destroying the workqueue
@ 2026-09-03  8:22 Linkai Gong
  2026-09-07 10:53 ` Heikki Krogerus
  2026-09-08  6:03 ` [PATCH v2] " Linkai Gong
  0 siblings, 2 replies; 8+ messages in thread
From: Linkai Gong @ 2026-09-03  8:22 UTC (permalink / raw)
  To: Heikki Krogerus, Greg Kroah-Hartman
  Cc: Xin Ji, linux-usb, linux-kernel, gonglinkai

The IRQ is devm-managed, so it can still queue plat->work while
remove() is tearing the queue down. Disable it first.

Fixes: fe6d8a9c8e64 ("usb: typec: anx7411: Add Analogix PD ANX7411 support")
Signed-off-by: Linkai Gong <gonglinkai@kylinos.cn>
---
 drivers/usb/typec/anx7411.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/typec/anx7411.c b/drivers/usb/typec/anx7411.c
index 41df115912b9..c3d36da692e4 100644
--- a/drivers/usb/typec/anx7411.c
+++ b/drivers/usb/typec/anx7411.c
@@ -1566,8 +1566,11 @@ static void anx7411_i2c_remove(struct i2c_client *client)
 	anx7411_partner_unregister_altmode(plat);
 	anx7411_unregister_partner(plat);
 
-	if (plat->workqueue)
+	if (plat->workqueue) {
+		disable_irq(plat->intp_irq);
+		cancel_work_sync(&plat->work);
 		destroy_workqueue(plat->workqueue);
+	}
 
 	i2c_unregister_device(plat->spi_client);
 
-- 
2.25.1


^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH] usb: typec: anx7411 - stop the IRQ before destroying the workqueue
  2026-09-03  8:22 [PATCH] usb: typec: anx7411 - stop the IRQ before destroying the workqueue Linkai Gong
@ 2026-09-07 10:53 ` Heikki Krogerus
  2026-09-08  3:16   ` Linkai Gong
  2026-09-08  6:02   ` Linkai Gong
  2026-09-08  6:03 ` [PATCH v2] " Linkai Gong
  1 sibling, 2 replies; 8+ messages in thread
From: Heikki Krogerus @ 2026-09-07 10:53 UTC (permalink / raw)
  To: Linkai Gong; +Cc: Greg Kroah-Hartman, Xin Ji, linux-usb, linux-kernel

On Thu, Sep 03, 2026 at 04:22:54PM +0800, Linkai Gong wrote:
> The IRQ is devm-managed, so it can still queue plat->work while
> remove() is tearing the queue down. Disable it first.
> 
> Fixes: fe6d8a9c8e64 ("usb: typec: anx7411: Add Analogix PD ANX7411 support")
> Signed-off-by: Linkai Gong <gonglinkai@kylinos.cn>
> ---
>  drivers/usb/typec/anx7411.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/usb/typec/anx7411.c b/drivers/usb/typec/anx7411.c
> index 41df115912b9..c3d36da692e4 100644
> --- a/drivers/usb/typec/anx7411.c
> +++ b/drivers/usb/typec/anx7411.c
> @@ -1566,8 +1566,11 @@ static void anx7411_i2c_remove(struct i2c_client *client)
>  	anx7411_partner_unregister_altmode(plat);
>  	anx7411_unregister_partner(plat);
>  
> -	if (plat->workqueue)
> +	if (plat->workqueue) {
> +		disable_irq(plat->intp_irq);
> +		cancel_work_sync(&plat->work);
>  		destroy_workqueue(plat->workqueue);
> +	}

Is the workqueu necessary at all? It's threaded irq.

Thanks,


>  	i2c_unregister_device(plat->spi_client);
>  
> -- 
> 2.25.1

-- 
heikki

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH] usb: typec: anx7411 - stop the IRQ before destroying the workqueue
  2026-09-07 10:53 ` Heikki Krogerus
@ 2026-09-08  3:16   ` Linkai Gong
  2026-09-08  6:02   ` Linkai Gong
  1 sibling, 0 replies; 8+ messages in thread
From: Linkai Gong @ 2026-09-08  3:16 UTC (permalink / raw)
  To: Heikki Krogerus; +Cc: Greg Kroah-Hartman, Xin Ji, linux-usb, linux-kernel

On Mon, Sep 07, 2026 at 12:53:35PM +0200, Heikki Krogerus wrote:
> Is the workqueu necessary at all? It's threaded irq.

You are right. The threaded handler only queues work, so the
extra workqueue does not buy much.

I will send a v2 that runs the handler body in the threaded IRQ
and drops the workqueue. That also removes the remove() race.

Thanks,
Linkai

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH] usb: typec: anx7411 - stop the IRQ before destroying the workqueue
  2026-09-07 10:53 ` Heikki Krogerus
  2026-09-08  3:16   ` Linkai Gong
@ 2026-09-08  6:02   ` Linkai Gong
  1 sibling, 0 replies; 8+ messages in thread
From: Linkai Gong @ 2026-09-08  6:02 UTC (permalink / raw)
  To: Heikki Krogerus; +Cc: Greg Kroah-Hartman, Xin Ji, linux-usb, linux-kernel

On Mon, Sep 07, 2026 at 12:53:35PM +0200, Heikki Krogerus wrote:
> Is the workqueu necessary at all? It's threaded irq.

Sorry, I spoke too soon. You are right that the workqueue looks
redundant, but I will keep this patch to the remove() race only.
I can drop the workqueue in a separate follow-up if you prefer.

Thanks,
Linkai

^ permalink raw reply	[flat|nested] 8+ messages in thread

* [PATCH v2] usb: typec: anx7411 - stop the IRQ before destroying the workqueue
  2026-09-03  8:22 [PATCH] usb: typec: anx7411 - stop the IRQ before destroying the workqueue Linkai Gong
  2026-09-07 10:53 ` Heikki Krogerus
@ 2026-09-08  6:03 ` Linkai Gong
  2026-09-09 14:23   ` Heikki Krogerus
  2026-09-10  1:11   ` [PATCH v3] " Linkai Gong
  1 sibling, 2 replies; 8+ messages in thread
From: Linkai Gong @ 2026-09-08  6:03 UTC (permalink / raw)
  To: Heikki Krogerus, Greg Kroah-Hartman; +Cc: Xin Ji, linux-usb, linux-kernel

The IRQ is devm-managed, so it can still queue plat->work while
remove() is tearing the queue down. Disable it first.

Fixes: fe6d8a9c8e64 ("usb: typec: anx7411: Add Analogix PD ANX7411 support")
Signed-off-by: Linkai Gong <gonglinkai@kylinos.cn>
---
v2:
- keep the minimal remove() race fix; drop the workqueue later if wanted

 drivers/usb/typec/anx7411.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/typec/anx7411.c b/drivers/usb/typec/anx7411.c
index 41df115912b9..c3d36da692e4 100644
--- a/drivers/usb/typec/anx7411.c
+++ b/drivers/usb/typec/anx7411.c
@@ -1566,8 +1566,11 @@ static void anx7411_i2c_remove(struct i2c_client *client)
 	anx7411_partner_unregister_altmode(plat);
 	anx7411_unregister_partner(plat);
 
-	if (plat->workqueue)
+	if (plat->workqueue) {
+		disable_irq(plat->intp_irq);
+		cancel_work_sync(&plat->work);
 		destroy_workqueue(plat->workqueue);
+	}
 
 	i2c_unregister_device(plat->spi_client);
 
-- 
2.25.1

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH v2] usb: typec: anx7411 - stop the IRQ before destroying the workqueue
  2026-09-08  6:03 ` [PATCH v2] " Linkai Gong
@ 2026-09-09 14:23   ` Heikki Krogerus
  2026-09-10  1:13     ` Linkai Gong
  2026-09-10  1:11   ` [PATCH v3] " Linkai Gong
  1 sibling, 1 reply; 8+ messages in thread
From: Heikki Krogerus @ 2026-09-09 14:23 UTC (permalink / raw)
  To: Linkai Gong; +Cc: Greg Kroah-Hartman, Xin Ji, linux-usb, linux-kernel

On Tue, Sep 08, 2026 at 02:03:33PM +0800, Linkai Gong wrote:
> The IRQ is devm-managed, so it can still queue plat->work while
> remove() is tearing the queue down. Disable it first.
> 
> Fixes: fe6d8a9c8e64 ("usb: typec: anx7411: Add Analogix PD ANX7411 support")

Not for stable?

> Signed-off-by: Linkai Gong <gonglinkai@kylinos.cn>
> ---
> v2:
> - keep the minimal remove() race fix; drop the workqueue later if wanted

I don't see any difference compared to v1?

thanks,

>  drivers/usb/typec/anx7411.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/usb/typec/anx7411.c b/drivers/usb/typec/anx7411.c
> index 41df115912b9..c3d36da692e4 100644
> --- a/drivers/usb/typec/anx7411.c
> +++ b/drivers/usb/typec/anx7411.c
> @@ -1566,8 +1566,11 @@ static void anx7411_i2c_remove(struct i2c_client *client)
>  	anx7411_partner_unregister_altmode(plat);
>  	anx7411_unregister_partner(plat);
>  
> -	if (plat->workqueue)
> +	if (plat->workqueue) {
> +		disable_irq(plat->intp_irq);
> +		cancel_work_sync(&plat->work);
>  		destroy_workqueue(plat->workqueue);
> +	}
>  
>  	i2c_unregister_device(plat->spi_client);
>  
> -- 
> 2.25.1

-- 
heikki

^ permalink raw reply	[flat|nested] 8+ messages in thread

* [PATCH v3] usb: typec: anx7411 - stop the IRQ before destroying the workqueue
  2026-09-08  6:03 ` [PATCH v2] " Linkai Gong
  2026-09-09 14:23   ` Heikki Krogerus
@ 2026-09-10  1:11   ` Linkai Gong
  1 sibling, 0 replies; 8+ messages in thread
From: Linkai Gong @ 2026-09-10  1:11 UTC (permalink / raw)
  To: Heikki Krogerus, Greg Kroah-Hartman
  Cc: Xin Ji, linux-usb, linux-kernel, stable, gonglinkai

The IRQ is devm-managed, so it can still queue plat->work while
remove() is tearing the queue down. Disable it first.

Fixes: fe6d8a9c8e64 ("usb: typec: anx7411: Add Analogix PD ANX7411 support")
Cc: stable@vger.kernel.org
Signed-off-by: Linkai Gong <gonglinkai@kylinos.cn>
---
v3:
- add Cc: stable (Heikki)

v2:
- keep the minimal remove() race fix; drop the workqueue later if wanted
  (code unchanged from v1)

 drivers/usb/typec/anx7411.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/typec/anx7411.c b/drivers/usb/typec/anx7411.c
index 41df115912b9..c3d36da692e4 100644
--- a/drivers/usb/typec/anx7411.c
+++ b/drivers/usb/typec/anx7411.c
@@ -1566,8 +1566,11 @@ static void anx7411_i2c_remove(struct i2c_client *client)
 	anx7411_partner_unregister_altmode(plat);
 	anx7411_unregister_partner(plat);
 
-	if (plat->workqueue)
+	if (plat->workqueue) {
+		disable_irq(plat->intp_irq);
+		cancel_work_sync(&plat->work);
 		destroy_workqueue(plat->workqueue);
+	}
 
 	i2c_unregister_device(plat->spi_client);
 
-- 
2.25.1

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH v2] usb: typec: anx7411 - stop the IRQ before destroying the workqueue
  2026-09-09 14:23   ` Heikki Krogerus
@ 2026-09-10  1:13     ` Linkai Gong
  0 siblings, 0 replies; 8+ messages in thread
From: Linkai Gong @ 2026-09-10  1:13 UTC (permalink / raw)
  To: Heikki Krogerus
  Cc: Greg Kroah-Hartman, Xin Ji, linux-usb, linux-kernel, gonglinkai

On Wed, Sep 09, 2026 at 04:23:14PM +0200, Heikki Krogerus wrote:
> Not for stable?

Yes, it should go to stable. The race has been there since the driver
was added. I will send a v3 with Cc: stable.

> I don't see any difference compared to v1?

Correct, the code is the same as v1. After your question about the
workqueue I only wanted to say I am keeping this minimal remove()
fix for now; I should have replied without posting an identical
patch. Sorry for the noise.

Thanks,
Linkai

^ permalink raw reply	[flat|nested] 8+ messages in thread

end of thread, other threads:[~2026-09-10  1:13 UTC | newest]

Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-09-03  8:22 [PATCH] usb: typec: anx7411 - stop the IRQ before destroying the workqueue Linkai Gong
2026-09-07 10:53 ` Heikki Krogerus
2026-09-08  3:16   ` Linkai Gong
2026-09-08  6:02   ` Linkai Gong
2026-09-08  6:03 ` [PATCH v2] " Linkai Gong
2026-09-09 14:23   ` Heikki Krogerus
2026-09-10  1:13     ` Linkai Gong
2026-09-10  1:11   ` [PATCH v3] " Linkai Gong

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®