[PATCH 1/2] usb: musb: Fix DMA for host mode

[PATCH 1/2] usb: musb: Fix DMA for host mode

[Joshua Henderson]

From: Cristian Birsan <cristian.birsan@microchip.com>

Commit ac33cdb166811223cc (usb: musb: Remove ifdefs for musb_host_rx in
musb_host.c part5) introduces a problem setting DMA host mode.

This fixes the done condition that advances the musb schedule. Without this
patch the the msub_advance_schedule() is called immediately after receiving an
endpoint rx interrupt without waiting for the dma transfer to complete. As a
consequence when the dma complete interrupt arrives the in_qh member of hw_ep
is already null an the musb_host_rx() exits on !urb error case.

Signed-off-by: Cristian Birsan <cristian.birsan@microchip.com>
Signed-off-by: Joshua Henderson <joshua.henderson@microchip.com>
---
 drivers/usb/musb/musb_host.c |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/usb/musb/musb_host.c b/drivers/usb/musb/musb_host.c
index 795a45b..3fdc99b 100644
--- a/drivers/usb/musb/musb_host.c
+++ b/drivers/usb/musb/musb_host.c
@@ -1825,6 +1825,7 @@ void musb_host_rx(struct musb *musb, u8 epnum)
 	u16			rx_csr, val;
 	bool			iso_err = false;
 	bool			done = false;
+	int			ret;
 	u32			status;
 	struct dma_channel	*dma;
 	unsigned int sg_flags = SG_MITER_ATOMIC | SG_MITER_TO_SG;
@@ -2003,10 +2004,10 @@ void musb_host_rx(struct musb *musb, u8 epnum)
 				qh->offset,
 				urb->transfer_buffer_length);
 
-			done = musb_rx_dma_in_inventra_cppi41(c, hw_ep, qh,
+			ret = musb_rx_dma_in_inventra_cppi41(c, hw_ep, qh,
 							      urb, xfer_len,
 							      iso_err);
-			if (done)
+			if (ret)
 				goto finish;
 			else
 				dev_err(musb->controller, "error: rx_dma failed\n");
-- 
1.7.9.5

[PATCH 2/2] usb: musb: Fix DMA desired mode for Mentor DMA engine

[Joshua Henderson]

From: Cristian Birsan <cristian.birsan@microchip.com>

Commit 6b6e97107f12f3a9f7 (USB: musb: fix isochronous TXDMA (take 2)) introduces
a problem setting the desired channel mode for the Mentor DMA engine.

There is a case where the pointer of the channel DMA mode is incorrectly
assigned to a pointer value, when it should be assigned the actual mode value.
This results in the value of channel->desired_mode not being correct.

Signed-off-by: Cristian Birsan <cristian.birsan@microchip.com>
Signed-off-by: Joshua Henderson <joshua.henderson@microchip.com>
---
 drivers/usb/musb/musb_host.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/usb/musb/musb_host.c b/drivers/usb/musb/musb_host.c
index 3fdc99b..748379e 100644
--- a/drivers/usb/musb/musb_host.c
+++ b/drivers/usb/musb/musb_host.c
@@ -662,7 +662,7 @@ static int musb_tx_dma_set_mode_mentor(struct dma_controller *dma,
 		csr &= ~(MUSB_TXCSR_AUTOSET | MUSB_TXCSR_DMAMODE);
 		csr |= MUSB_TXCSR_DMAENAB; /* against programmer's guide */
 	}
-	channel->desired_mode = mode;
+	channel->desired_mode = *mode;
 	musb_writew(epio, MUSB_TXCSR, csr);
 
 	return 0;
-- 
1.7.9.5

Re: [PATCH 2/2] usb: musb: Fix DMA desired mode for Mentor DMA engine

[Sergei Shtylyov]

On 2/2/2016 3:37 AM, Joshua Henderson wrote:

> From: Cristian Birsan <cristian.birsan@microchip.com>
>
> Commit 6b6e97107f12f3a9f7 (USB: musb: fix isochronous TXDMA (take 2)) introduces

    Again, wrong commit style. And I really doubt that blaming my commit was 
correct. :-)

> a problem setting the desired channel mode for the Mentor DMA engine.
>
> There is a case where the pointer of the channel DMA mode is incorrectly
> assigned to a pointer value, when it should be assigned the actual mode value.
> This results in the value of channel->desired_mode not being correct.
>
> Signed-off-by: Cristian Birsan <cristian.birsan@microchip.com>
> Signed-off-by: Joshua Henderson <joshua.henderson@microchip.com>
> ---
>   drivers/usb/musb/musb_host.c |    2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/usb/musb/musb_host.c b/drivers/usb/musb/musb_host.c
> index 3fdc99b..748379e 100644
> --- a/drivers/usb/musb/musb_host.c
> +++ b/drivers/usb/musb/musb_host.c
> @@ -662,7 +662,7 @@ static int musb_tx_dma_set_mode_mentor(struct dma_controller *dma,

    Hm, there was no such function at the time of my commit...

>   		csr &= ~(MUSB_TXCSR_AUTOSET | MUSB_TXCSR_DMAMODE);
>   		csr |= MUSB_TXCSR_DMAENAB; /* against programmer's guide */
>   	}
> -	channel->desired_mode = mode;
> +	channel->desired_mode = *mode;

   'mode' was 'u8' at the time of my commit, see:

https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=6b6e97107f12f3a9f7b5b43a6c3b94409240bcff

    I think that the recent commit below should be blamed instead:

https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=754fe4a92c072a6e36d89fa328ed789c9ebc1af5

[...]

MBR, Sergei

Re: [PATCH 2/2] usb: musb: Fix DMA desired mode for Mentor DMA engine

[Joshua Henderson]

Sergei,

On 02/02/2016 06:32 AM, Sergei Shtylyov wrote:
> On 2/2/2016 3:37 AM, Joshua Henderson wrote:
> 
>> From: Cristian Birsan <cristian.birsan@microchip.com>
>>
>> Commit 6b6e97107f12f3a9f7 (USB: musb: fix isochronous TXDMA (take 2)) introduces
> 
>    Again, wrong commit style. And I really doubt that blaming my commit was correct. :-)

I apologize.  I noticed you added the line in question, but did not notice the variable type of "mode" changed *afterward*.  Completely my fault.  I shall also fix the commit message style.

> 
>> a problem setting the desired channel mode for the Mentor DMA engine.
>>
>> There is a case where the pointer of the channel DMA mode is incorrectly
>> assigned to a pointer value, when it should be assigned the actual mode value.
>> This results in the value of channel->desired_mode not being correct.
>>
>> Signed-off-by: Cristian Birsan <cristian.birsan@microchip.com>
>> Signed-off-by: Joshua Henderson <joshua.henderson@microchip.com>
>> ---
>>   drivers/usb/musb/musb_host.c |    2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/usb/musb/musb_host.c b/drivers/usb/musb/musb_host.c
>> index 3fdc99b..748379e 100644
>> --- a/drivers/usb/musb/musb_host.c
>> +++ b/drivers/usb/musb/musb_host.c
>> @@ -662,7 +662,7 @@ static int musb_tx_dma_set_mode_mentor(struct dma_controller *dma,
> 
>    Hm, there was no such function at the time of my commit...
> 
>>           csr &= ~(MUSB_TXCSR_AUTOSET | MUSB_TXCSR_DMAMODE);
>>           csr |= MUSB_TXCSR_DMAENAB; /* against programmer's guide */
>>       }
>> -    channel->desired_mode = mode;
>> +    channel->desired_mode = *mode;
> 
>   'mode' was 'u8' at the time of my commit, see:
> 
> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=6b6e97107f12f3a9f7b5b43a6c3b94409240bcff
> 
>    I think that the recent commit below should be blamed instead:
> 
> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=754fe4a92c072a6e36d89fa328ed789c9ebc1af5

I'll take a look at this and make sure.

Thanks,
Josh

Re: [PATCH 2/2] usb: musb: Fix DMA desired mode for Mentor DMA engine

[Sergei Shtylyov]

On 02/02/2016 07:36 PM, Joshua Henderson wrote:

>>> From: Cristian Birsan <cristian.birsan@microchip.com>
>>>
>>> Commit 6b6e97107f12f3a9f7 (USB: musb: fix isochronous TXDMA (take 2)) introduces
>>
>>     Again, wrong commit style.

    Commit citing style, I meant.

>> And I really doubt that blaming my commit was correct. :-)
>
> I apologize.  I noticed you added the line in question, but did not notice the variable type of "mode" changed *afterward*.

    So I thought. :-)

> Completely my fault.  I shall also fix the commit message style.

    TIA.

[...]

> Thanks,
> Josh

MBR, Sergei

Re: [PATCH 1/2] usb: musb: Fix DMA for host mode

[Sergei Shtylyov]

Hello.

On 2/2/2016 3:37 AM, Joshua Henderson wrote:

> From: Cristian Birsan <cristian.birsan@microchip.com>
>
> Commit ac33cdb166811223cc (usb: musb: Remove ifdefs for musb_host_rx in
> musb_host.c part5) introduces a problem setting DMA host mode.

    scripts/checkpatch.pl now enforces certain commit citing style, yours 
doesn't quite match it.

> This fixes the done condition that advances the musb schedule. Without this
> patch the the msub_advance_schedule() is called immediately after receiving an
> endpoint rx interrupt without waiting for the dma transfer to complete. As a
> consequence when the dma complete interrupt arrives the in_qh member of hw_ep
> is already null an the musb_host_rx() exits on !urb error case.
>
> Signed-off-by: Cristian Birsan <cristian.birsan@microchip.com>
> Signed-off-by: Joshua Henderson <joshua.henderson@microchip.com>
> ---
>   drivers/usb/musb/musb_host.c |    5 +++--
>   1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/usb/musb/musb_host.c b/drivers/usb/musb/musb_host.c
> index 795a45b..3fdc99b 100644
> --- a/drivers/usb/musb/musb_host.c
> +++ b/drivers/usb/musb/musb_host.c
> @@ -1825,6 +1825,7 @@ void musb_host_rx(struct musb *musb, u8 epnum)
>   	u16			rx_csr, val;
>   	bool			iso_err = false;
>   	bool			done = false;
> +	int			ret;
>   	u32			status;
>   	struct dma_channel	*dma;
>   	unsigned int sg_flags = SG_MITER_ATOMIC | SG_MITER_TO_SG;
> @@ -2003,10 +2004,10 @@ void musb_host_rx(struct musb *musb, u8 epnum)
>   				qh->offset,
>   				urb->transfer_buffer_length);
>
> -			done = musb_rx_dma_in_inventra_cppi41(c, hw_ep, qh,
> +			ret = musb_rx_dma_in_inventra_cppi41(c, hw_ep, qh,
>   							      urb, xfer_len,
>   							      iso_err);
> -			if (done)
> +			if (ret)

    Hm, does that really fix anything?

>   				goto finish;
>   			else
>   				dev_err(musb->controller, "error: rx_dma failed\n");

MBR, Sergei

Re: [PATCH 1/2] usb: musb: Fix DMA for host mode

[Sergei Shtylyov]

On 02/02/2016 04:24 PM, Sergei Shtylyov wrote:

>> From: Cristian Birsan <cristian.birsan@microchip.com>
>>
>> Commit ac33cdb166811223cc (usb: musb: Remove ifdefs for musb_host_rx in
>> musb_host.c part5) introduces a problem setting DMA host mode.
>
>     scripts/checkpatch.pl now enforces certain commit citing style, yours
> doesn't quite match it.
>
>> This fixes the done condition that advances the musb schedule. Without this
>> patch the the msub_advance_schedule() is called immediately after receiving an
>> endpoint rx interrupt without waiting for the dma transfer to complete. As a
>> consequence when the dma complete interrupt arrives the in_qh member of hw_ep
>> is already null an the musb_host_rx() exits on !urb error case.
>>
>> Signed-off-by: Cristian Birsan <cristian.birsan@microchip.com>
>> Signed-off-by: Joshua Henderson <joshua.henderson@microchip.com>
>> ---
>>   drivers/usb/musb/musb_host.c |    5 +++--
>>   1 file changed, 3 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/usb/musb/musb_host.c b/drivers/usb/musb/musb_host.c
>> index 795a45b..3fdc99b 100644
>> --- a/drivers/usb/musb/musb_host.c
>> +++ b/drivers/usb/musb/musb_host.c
>> @@ -1825,6 +1825,7 @@ void musb_host_rx(struct musb *musb, u8 epnum)
>>       u16            rx_csr, val;
>>       bool            iso_err = false;
>>       bool            done = false;
>> +    int            ret;
>>       u32            status;
>>       struct dma_channel    *dma;
>>       unsigned int sg_flags = SG_MITER_ATOMIC | SG_MITER_TO_SG;
>> @@ -2003,10 +2004,10 @@ void musb_host_rx(struct musb *musb, u8 epnum)
>>                   qh->offset,
>>                   urb->transfer_buffer_length);
>>
>> -            done = musb_rx_dma_in_inventra_cppi41(c, hw_ep, qh,
>> +            ret = musb_rx_dma_in_inventra_cppi41(c, hw_ep, qh,
>>                                     urb, xfer_len,
>>                                     iso_err);
>> -            if (done)
>> +            if (ret)
>
>     Hm, does that really fix anything?

    Looking again, I got the idea. However, there's no dire need for the new 
variable...
    (Handling Inventra and CPPI 4.1 DMA in the same code looks fishy. Too bad 
I wasn't able to push the CPPI 4.1 drivers myself.)

MBR, Sergei