[PATCH] usb: ehci-brcm: fix sleep during atomic

[PATCH] usb: ehci-brcm: fix sleep during atomic

[justin.chen]

From: Justin Chen <justin.chen@broadcom.com>

echi_brcm_wait_for_sof() gets called after disabling interrupts
in ehci_brcm_hub_control(). Use the atomic version of poll_timeout
to fix the warning.

Fixes: ("9df231511bd6 usb: ehci: Add new EHCI driver for Broadcom STB SoC's")
Signed-off-by: Justin Chen <justin.chen@broadcom.com>
---
 drivers/usb/host/ehci-brcm.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/usb/host/ehci-brcm.c b/drivers/usb/host/ehci-brcm.c
index 888e8f6670d2..5e3156f94cc6 100644
--- a/drivers/usb/host/ehci-brcm.c
+++ b/drivers/usb/host/ehci-brcm.c
@@ -31,8 +31,8 @@ static inline void ehci_brcm_wait_for_sof(struct ehci_hcd *ehci, u32 delay)
 	int res;
 
 	/* Wait for next microframe (every 125 usecs) */
-	res = readl_relaxed_poll_timeout(&ehci->regs->frame_index, val,
-					 val != frame_idx, 1, 130);
+	res = readl_relaxed_poll_timeout_atomic(&ehci->regs->frame_index,
+						val, val != frame_idx, 1, 130);
 	if (res)
 		ehci_err(ehci, "Error waiting for SOF\n");
 	udelay(delay);
-- 
2.34.1

Re: [PATCH] usb: ehci-brcm: fix sleep during atomic

[Florian Fainelli]

On 3/12/26 12:08, justin.chen@broadcom.com wrote:
> From: Justin Chen <justin.chen@broadcom.com>
> 
> echi_brcm_wait_for_sof() gets called after disabling interrupts
> in ehci_brcm_hub_control(). Use the atomic version of poll_timeout
> to fix the warning.
> 
> Fixes: ("9df231511bd6 usb: ehci: Add new EHCI driver for Broadcom STB SoC's")
> Signed-off-by: Justin Chen <justin.chen@broadcom.com>

Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
-- 
Florian

Re: [PATCH] usb: ehci-brcm: fix sleep during atomic

[Greg KH]

On Thu, Mar 12, 2026 at 12:08:25PM -0700, justin.chen@broadcom.com wrote:
> From: Justin Chen <justin.chen@broadcom.com>
> 
> echi_brcm_wait_for_sof() gets called after disabling interrupts
> in ehci_brcm_hub_control(). Use the atomic version of poll_timeout
> to fix the warning.
> 
> Fixes: ("9df231511bd6 usb: ehci: Add new EHCI driver for Broadcom STB SoC's")

Can you resend this with the proper format for how Fixes: lines should
be?  It's in the documentation, otherwise our scripts will reject this.

thanks,

greg k-h

Re: [PATCH] usb: ehci-brcm: fix sleep during atomic

[Greg KH]

On Thu, Mar 12, 2026 at 12:08:25PM -0700, justin.chen@broadcom.com wrote:
> From: Justin Chen <justin.chen@broadcom.com>
> 
> echi_brcm_wait_for_sof() gets called after disabling interrupts
> in ehci_brcm_hub_control(). Use the atomic version of poll_timeout
> to fix the warning.
> 
> Fixes: ("9df231511bd6 usb: ehci: Add new EHCI driver for Broadcom STB SoC's")
> Signed-off-by: Justin Chen <justin.chen@broadcom.com>
> ---
>  drivers/usb/host/ehci-brcm.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/usb/host/ehci-brcm.c b/drivers/usb/host/ehci-brcm.c
> index 888e8f6670d2..5e3156f94cc6 100644
> --- a/drivers/usb/host/ehci-brcm.c
> +++ b/drivers/usb/host/ehci-brcm.c
> @@ -31,8 +31,8 @@ static inline void ehci_brcm_wait_for_sof(struct ehci_hcd *ehci, u32 delay)
>  	int res;
>  
>  	/* Wait for next microframe (every 125 usecs) */
> -	res = readl_relaxed_poll_timeout(&ehci->regs->frame_index, val,
> -					 val != frame_idx, 1, 130);
> +	res = readl_relaxed_poll_timeout_atomic(&ehci->regs->frame_index,
> +						val, val != frame_idx, 1, 130);
>  	if (res)
>  		ehci_err(ehci, "Error waiting for SOF\n");
>  	udelay(delay);
> -- 
> 2.34.1
> 
> 

Why is something that has been around for so long just now suddenly
being seen?  What changed to trigger this?

thanks,

greg k-h

Re: [PATCH] usb: ehci-brcm: fix sleep during atomic

[Justin Chen]

On 3/18/2026 7:14 AM, Greg KH wrote:
> On Thu, Mar 12, 2026 at 12:08:25PM -0700, justin.chen@broadcom.com wrote:
>> From: Justin Chen <justin.chen@broadcom.com>
>>
>> echi_brcm_wait_for_sof() gets called after disabling interrupts
>> in ehci_brcm_hub_control(). Use the atomic version of poll_timeout
>> to fix the warning.
>>
>> Fixes: ("9df231511bd6 usb: ehci: Add new EHCI driver for Broadcom STB SoC's")
>> Signed-off-by: Justin Chen <justin.chen@broadcom.com>
>> ---
>>   drivers/usb/host/ehci-brcm.c | 4 ++--
>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/usb/host/ehci-brcm.c b/drivers/usb/host/ehci-brcm.c
>> index 888e8f6670d2..5e3156f94cc6 100644
>> --- a/drivers/usb/host/ehci-brcm.c
>> +++ b/drivers/usb/host/ehci-brcm.c
>> @@ -31,8 +31,8 @@ static inline void ehci_brcm_wait_for_sof(struct ehci_hcd *ehci, u32 delay)
>>   	int res;
>>   
>>   	/* Wait for next microframe (every 125 usecs) */
>> -	res = readl_relaxed_poll_timeout(&ehci->regs->frame_index, val,
>> -					 val != frame_idx, 1, 130);
>> +	res = readl_relaxed_poll_timeout_atomic(&ehci->regs->frame_index,
>> +						val, val != frame_idx, 1, 130);
>>   	if (res)
>>   		ehci_err(ehci, "Error waiting for SOF\n");
>>   	udelay(delay);
>> -- 
>> 2.34.1
>>
>>
> 
> Why is something that has been around for so long just now suddenly
> being seen?  What changed to trigger this?
> 

The edge case occurs in rare occasions and predates my time maintaining 
this driver. So we just so happened to run into it recently in our 
tests. Guess in the past we just never hit the edge case with the 
warnings enabled. I had the same question when the warning was reported.

Will fix the Fixes tag in v2.

Thanks,
Justin

> thanks,
> 
> greg k-h

Re: [PATCH] usb: ehci-brcm: fix sleep during atomic

[Florian Fainelli]

On 3/18/26 09:31, Justin Chen wrote:
> 
> 
> On 3/18/2026 7:14 AM, Greg KH wrote:
>> On Thu, Mar 12, 2026 at 12:08:25PM -0700, justin.chen@broadcom.com wrote:
>>> From: Justin Chen <justin.chen@broadcom.com>
>>>
>>> echi_brcm_wait_for_sof() gets called after disabling interrupts
>>> in ehci_brcm_hub_control(). Use the atomic version of poll_timeout
>>> to fix the warning.
>>>
>>> Fixes: ("9df231511bd6 usb: ehci: Add new EHCI driver for Broadcom STB 
>>> SoC's")
>>> Signed-off-by: Justin Chen <justin.chen@broadcom.com>
>>> ---
>>>   drivers/usb/host/ehci-brcm.c | 4 ++--
>>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/usb/host/ehci-brcm.c b/drivers/usb/host/ehci-brcm.c
>>> index 888e8f6670d2..5e3156f94cc6 100644
>>> --- a/drivers/usb/host/ehci-brcm.c
>>> +++ b/drivers/usb/host/ehci-brcm.c
>>> @@ -31,8 +31,8 @@ static inline void ehci_brcm_wait_for_sof(struct 
>>> ehci_hcd *ehci, u32 delay)
>>>       int res;
>>>       /* Wait for next microframe (every 125 usecs) */
>>> -    res = readl_relaxed_poll_timeout(&ehci->regs->frame_index, val,
>>> -                     val != frame_idx, 1, 130);
>>> +    res = readl_relaxed_poll_timeout_atomic(&ehci->regs->frame_index,
>>> +                        val, val != frame_idx, 1, 130);
>>>       if (res)
>>>           ehci_err(ehci, "Error waiting for SOF\n");
>>>       udelay(delay);
>>> -- 
>>> 2.34.1
>>>
>>>
>>
>> Why is something that has been around for so long just now suddenly
>> being seen?  What changed to trigger this?
>>
> 
> The edge case occurs in rare occasions and predates my time maintaining 
> this driver. So we just so happened to run into it recently in our 
> tests. Guess in the past we just never hit the edge case with the 
> warnings enabled. I had the same question when the warning was reported.
> 
> Will fix the Fixes tag in v2.

The other thing is that recently enabled CONFIG_DEBUG_SLEEP_ATOMIC in 
our builds which is how this got caught, it was not the case before.
-- 
Florian