DPDK patches and discussions
 help / color / mirror / Atom feed
* [dpdk-dev] [PATCH 0/2] fix of configuring inside NIC RX interrupt
@ 2014-10-29  3:42 Helin Zhang
  2014-10-29  3:42 ` [dpdk-dev] [PATCH 1/2] i40e: code style fix Helin Zhang
                   ` (2 more replies)
  0 siblings, 3 replies; 8+ messages in thread
From: Helin Zhang @ 2014-10-29  3:42 UTC (permalink / raw)
  To: dev

Inside NIC RX interrupt is needed for single RX descriptor
write back. The fix is to correct the wrong configuration
of register 'I40E_QINT_RQCTL'. In addition, several code
style fixes are added.
Note that interrupt will be inside NIC only, that means it
will never be reported outside NIC hardware.

Helin Zhang (2):
  i40e: code style fix
  i40e: fix of configuring inside NIC RX interrupt

 lib/librte_pmd_i40e/i40e_ethdev.c | 23 +++++++++++++----------
 1 file changed, 13 insertions(+), 10 deletions(-)

-- 
1.8.1.4

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

* [dpdk-dev] [PATCH 1/2] i40e: code style fix
  2014-10-29  3:42 [dpdk-dev] [PATCH 0/2] fix of configuring inside NIC RX interrupt Helin Zhang
@ 2014-10-29  3:42 ` Helin Zhang
  2014-10-29  6:11   ` Qiu, Michael
  2014-10-29  3:42 ` [dpdk-dev] [PATCH 2/2] i40e: fix of configuring inside NIC RX interrupt Helin Zhang
  2014-10-29  5:00 ` [dpdk-dev] [PATCH 0/2] " Chen, Jing D
  2 siblings, 1 reply; 8+ messages in thread
From: Helin Zhang @ 2014-10-29  3:42 UTC (permalink / raw)
  To: dev

Add several code style fixes.

Signed-off-by: Helin Zhang <helin.zhang@intel.com>
---
 lib/librte_pmd_i40e/i40e_ethdev.c | 20 +++++++++++---------
 1 file changed, 11 insertions(+), 9 deletions(-)

diff --git a/lib/librte_pmd_i40e/i40e_ethdev.c b/lib/librte_pmd_i40e/i40e_ethdev.c
index e1f581a..20c99a4 100644
--- a/lib/librte_pmd_i40e/i40e_ethdev.c
+++ b/lib/librte_pmd_i40e/i40e_ethdev.c
@@ -580,7 +580,8 @@ i40e_vsi_queues_bind_intr(struct i40e_vsi *vsi)
 	uint32_t val;
 	struct i40e_hw *hw = I40E_VSI_TO_HW(vsi);
 	uint16_t msix_vect = vsi->msix_intr;
-	uint16_t interval = i40e_calc_itr_interval(RTE_LIBRTE_I40E_ITR_INTERVAL);
+	uint16_t interval =
+		i40e_calc_itr_interval(RTE_LIBRTE_I40E_ITR_INTERVAL);
 	int i;
 
 	for (i = 0; i < vsi->nb_qps; i++)
@@ -603,25 +604,26 @@ i40e_vsi_queues_bind_intr(struct i40e_vsi *vsi)
 	/* Write first RX queue to Link list register as the head element */
 	if (vsi->type != I40E_VSI_SRIOV) {
 		I40E_WRITE_REG(hw, I40E_PFINT_LNKLSTN(msix_vect - 1),
-			(vsi->base_queue << I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
+						(vsi->base_queue <<
+				I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
 			(0x0 << I40E_PFINT_LNKLSTN_FIRSTQ_TYPE_SHIFT));
 
 		I40E_WRITE_REG(hw, I40E_PFINT_ITRN(I40E_ITR_INDEX_DEFAULT,
-				msix_vect - 1), interval);
+						msix_vect - 1), interval);
 
 		/* Disable auto-mask on enabling of all none-zero  interrupt */
 		I40E_WRITE_REG(hw, I40E_GLINT_CTL,
-				I40E_GLINT_CTL_DIS_AUTOMASK_N_MASK);
-	}
-	else {
+			I40E_GLINT_CTL_DIS_AUTOMASK_N_MASK);
+	} else {
 		uint32_t reg;
+
 		/* num_msix_vectors_vf needs to minus irq0 */
 		reg = (hw->func_caps.num_msix_vectors_vf - 1) *
 			vsi->user_param + (msix_vect - 1);
 
-		I40E_WRITE_REG(hw, I40E_VPINT_LNKLSTN(reg),
-			(vsi->base_queue << I40E_VPINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
-			(0x0 << I40E_VPINT_LNKLSTN_FIRSTQ_TYPE_SHIFT));
+		I40E_WRITE_REG(hw, I40E_VPINT_LNKLSTN(reg), (vsi->base_queue <<
+					I40E_VPINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
+				(0x0 << I40E_VPINT_LNKLSTN_FIRSTQ_TYPE_SHIFT));
 	}
 
 	I40E_WRITE_FLUSH(hw);
-- 
1.8.1.4

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

* [dpdk-dev] [PATCH 2/2] i40e: fix of configuring inside NIC RX interrupt
  2014-10-29  3:42 [dpdk-dev] [PATCH 0/2] fix of configuring inside NIC RX interrupt Helin Zhang
  2014-10-29  3:42 ` [dpdk-dev] [PATCH 1/2] i40e: code style fix Helin Zhang
@ 2014-10-29  3:42 ` Helin Zhang
  2014-10-29  5:00 ` [dpdk-dev] [PATCH 0/2] " Chen, Jing D
  2 siblings, 0 replies; 8+ messages in thread
From: Helin Zhang @ 2014-10-29  3:42 UTC (permalink / raw)
  To: dev

Inside NIC RX interrupt is needed for single RX descriptor
write back. The fix is to correct the wrong configuration
of register 'I40E_QINT_RQCTL'.
Note that interrupt will be inside NIC only, that means it
will never be reported outside NIC hardware.

Signed-off-by: Helin Zhang <helin.zhang@intel.com>
---
 lib/librte_pmd_i40e/i40e_ethdev.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/lib/librte_pmd_i40e/i40e_ethdev.c b/lib/librte_pmd_i40e/i40e_ethdev.c
index 20c99a4..4614c3d 100644
--- a/lib/librte_pmd_i40e/i40e_ethdev.c
+++ b/lib/librte_pmd_i40e/i40e_ethdev.c
@@ -580,8 +580,6 @@ i40e_vsi_queues_bind_intr(struct i40e_vsi *vsi)
 	uint32_t val;
 	struct i40e_hw *hw = I40E_VSI_TO_HW(vsi);
 	uint16_t msix_vect = vsi->msix_intr;
-	uint16_t interval =
-		i40e_calc_itr_interval(RTE_LIBRTE_I40E_ITR_INTERVAL);
 	int i;
 
 	for (i = 0; i < vsi->nb_qps; i++)
@@ -590,7 +588,7 @@ i40e_vsi_queues_bind_intr(struct i40e_vsi *vsi)
 	/* Bind all RX queues to allocated MSIX interrupt */
 	for (i = 0; i < vsi->nb_qps; i++) {
 		val = (msix_vect << I40E_QINT_RQCTL_MSIX_INDX_SHIFT) |
-			(interval << I40E_QINT_RQCTL_ITR_INDX_SHIFT) |
+			I40E_QINT_RQCTL_ITR_INDX_MASK |
 			((vsi->base_queue + i + 1) <<
 			I40E_QINT_RQCTL_NEXTQ_INDX_SHIFT) |
 			(0 << I40E_QINT_RQCTL_NEXTQ_TYPE_SHIFT) |
@@ -603,6 +601,9 @@ i40e_vsi_queues_bind_intr(struct i40e_vsi *vsi)
 
 	/* Write first RX queue to Link list register as the head element */
 	if (vsi->type != I40E_VSI_SRIOV) {
+		uint16_t interval =
+			i40e_calc_itr_interval(RTE_LIBRTE_I40E_ITR_INTERVAL);
+
 		I40E_WRITE_REG(hw, I40E_PFINT_LNKLSTN(msix_vect - 1),
 						(vsi->base_queue <<
 				I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
-- 
1.8.1.4

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

* Re: [dpdk-dev] [PATCH 0/2] fix of configuring inside NIC RX interrupt
  2014-10-29  3:42 [dpdk-dev] [PATCH 0/2] fix of configuring inside NIC RX interrupt Helin Zhang
  2014-10-29  3:42 ` [dpdk-dev] [PATCH 1/2] i40e: code style fix Helin Zhang
  2014-10-29  3:42 ` [dpdk-dev] [PATCH 2/2] i40e: fix of configuring inside NIC RX interrupt Helin Zhang
@ 2014-10-29  5:00 ` Chen, Jing D
  2014-10-29 22:44   ` Thomas Monjalon
  2 siblings, 1 reply; 8+ messages in thread
From: Chen, Jing D @ 2014-10-29  5:00 UTC (permalink / raw)
  To: Zhang, Helin, dev



> -----Original Message-----
> From: Zhang, Helin
> Sent: Wednesday, October 29, 2014 11:43 AM
> To: dev@dpdk.org
> Cc: Cao, Waterman; Chen, Jing D; Zhang, Helin
> Subject: [PATCH 0/2] fix of configuring inside NIC RX interrupt
> 
> Inside NIC RX interrupt is needed for single RX descriptor
> write back. The fix is to correct the wrong configuration
> of register 'I40E_QINT_RQCTL'. In addition, several code
> style fixes are added.
> Note that interrupt will be inside NIC only, that means it
> will never be reported outside NIC hardware.
> 
> Helin Zhang (2):
>   i40e: code style fix
>   i40e: fix of configuring inside NIC RX interrupt
> 
>  lib/librte_pmd_i40e/i40e_ethdev.c | 23 +++++++++++++----------
>  1 file changed, 13 insertions(+), 10 deletions(-)
> 
> --
> 1.8.1.4

Acked-by: Jing Chen <jing.d.chen@intel.com>

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

* Re: [dpdk-dev] [PATCH 1/2] i40e: code style fix
  2014-10-29  3:42 ` [dpdk-dev] [PATCH 1/2] i40e: code style fix Helin Zhang
@ 2014-10-29  6:11   ` Qiu, Michael
  2014-10-29  6:26     ` Zhang, Helin
  0 siblings, 1 reply; 8+ messages in thread
From: Qiu, Michael @ 2014-10-29  6:11 UTC (permalink / raw)
  To: Zhang, Helin, dev

 10/29/2014 11:44 AM, Helin Zhang  :
> Add several code style fixes.
>
> Signed-off-by: Helin Zhang <helin.zhang@intel.com>
> ---
>  lib/librte_pmd_i40e/i40e_ethdev.c | 20 +++++++++++---------
>  1 file changed, 11 insertions(+), 9 deletions(-)
>
> diff --git a/lib/librte_pmd_i40e/i40e_ethdev.c b/lib/librte_pmd_i40e/i40e_ethdev.c
> index e1f581a..20c99a4 100644
> --- a/lib/librte_pmd_i40e/i40e_ethdev.c
> +++ b/lib/librte_pmd_i40e/i40e_ethdev.c
> @@ -580,7 +580,8 @@ i40e_vsi_queues_bind_intr(struct i40e_vsi *vsi)
>  	uint32_t val;
>  	struct i40e_hw *hw = I40E_VSI_TO_HW(vsi);
>  	uint16_t msix_vect = vsi->msix_intr;
> -	uint16_t interval = i40e_calc_itr_interval(RTE_LIBRTE_I40E_ITR_INTERVAL);
> +	uint16_t interval =
> +		i40e_calc_itr_interval(RTE_LIBRTE_I40E_ITR_INTERVAL);
>  	int i;
>  
>  	for (i = 0; i < vsi->nb_qps; i++)
> @@ -603,25 +604,26 @@ i40e_vsi_queues_bind_intr(struct i40e_vsi *vsi)
>  	/* Write first RX queue to Link list register as the head element */
>  	if (vsi->type != I40E_VSI_SRIOV) {
>  		I40E_WRITE_REG(hw, I40E_PFINT_LNKLSTN(msix_vect - 1),
> -			(vsi->base_queue << I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
> +						(vsi->base_queue <<
                                                                       
                                  ^^^^^^^^^^^^^^^^^^
> +				I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |

Here why so many indentations? or just my mail agent's display issue, if
it does pls ignor. anyway, it is not a beautiful style in my mind :)

 		I40E_WRITE_REG(hw, I40E_PFINT_LNKLSTN(msix_vect - 1),
-			(vsi->base_queue << I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
+			       (vsi->base_queue <<
+				I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |

BTW, for coding style, I think we can follow the linux kernel, for kernel indentations, it always use the format like: n"Tab"+m"space", 0 <= m <=7.
Another example is qemu, it is all "space" without "tab".

>  			(0x0 << I40E_PFINT_LNKLSTN_FIRSTQ_TYPE_SHIFT));
>  
>  		I40E_WRITE_REG(hw, I40E_PFINT_ITRN(I40E_ITR_INDEX_DEFAULT,
> -				msix_vect - 1), interval);
> +						msix_vect - 1), interval);
>  
>  		/* Disable auto-mask on enabling of all none-zero  interrupt */
>  		I40E_WRITE_REG(hw, I40E_GLINT_CTL,
> -				I40E_GLINT_CTL_DIS_AUTOMASK_N_MASK);
> -	}
> -	else {
> +			I40E_GLINT_CTL_DIS_AUTOMASK_N_MASK);
> +	} else {
>  		uint32_t reg;
> +
>  		/* num_msix_vectors_vf needs to minus irq0 */
>  		reg = (hw->func_caps.num_msix_vectors_vf - 1) *
>  			vsi->user_param + (msix_vect - 1);
>  
> -		I40E_WRITE_REG(hw, I40E_VPINT_LNKLSTN(reg),
> -			(vsi->base_queue << I40E_VPINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
> -			(0x0 << I40E_VPINT_LNKLSTN_FIRSTQ_TYPE_SHIFT));
> +		I40E_WRITE_REG(hw, I40E_VPINT_LNKLSTN(reg), (vsi->base_queue <<
> +					I40E_VPINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
> +				(0x0 << I40E_VPINT_LNKLSTN_FIRSTQ_TYPE_SHIFT));

Here the same I think.

Thanks,
Michael
>  	}
>  
>  	I40E_WRITE_FLUSH(hw);


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

* Re: [dpdk-dev] [PATCH 1/2] i40e: code style fix
  2014-10-29  6:11   ` Qiu, Michael
@ 2014-10-29  6:26     ` Zhang, Helin
  2014-10-29  8:06       ` Qiu, Michael
  0 siblings, 1 reply; 8+ messages in thread
From: Zhang, Helin @ 2014-10-29  6:26 UTC (permalink / raw)
  To: Qiu, Michael, dev

Hi Michael

> -----Original Message-----
> From: Qiu, Michael
> Sent: Wednesday, October 29, 2014 2:11 PM
> To: Zhang, Helin; dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH 1/2] i40e: code style fix
> 
>  10/29/2014 11:44 AM, Helin Zhang  :
> > Add several code style fixes.
> >
> > Signed-off-by: Helin Zhang <helin.zhang@intel.com>
> > ---
> >  lib/librte_pmd_i40e/i40e_ethdev.c | 20 +++++++++++---------
> >  1 file changed, 11 insertions(+), 9 deletions(-)
> >
> > diff --git a/lib/librte_pmd_i40e/i40e_ethdev.c
> > b/lib/librte_pmd_i40e/i40e_ethdev.c
> > index e1f581a..20c99a4 100644
> > --- a/lib/librte_pmd_i40e/i40e_ethdev.c
> > +++ b/lib/librte_pmd_i40e/i40e_ethdev.c
> > @@ -580,7 +580,8 @@ i40e_vsi_queues_bind_intr(struct i40e_vsi *vsi)
> >  	uint32_t val;
> >  	struct i40e_hw *hw = I40E_VSI_TO_HW(vsi);
> >  	uint16_t msix_vect = vsi->msix_intr;
> > -	uint16_t interval = i40e_calc_itr_interval(RTE_LIBRTE_I40E_ITR_INTERVAL);
> > +	uint16_t interval =
> > +		i40e_calc_itr_interval(RTE_LIBRTE_I40E_ITR_INTERVAL);
> >  	int i;
> >
> >  	for (i = 0; i < vsi->nb_qps; i++)
> > @@ -603,25 +604,26 @@ i40e_vsi_queues_bind_intr(struct i40e_vsi *vsi)
> >  	/* Write first RX queue to Link list register as the head element */
> >  	if (vsi->type != I40E_VSI_SRIOV) {
> >  		I40E_WRITE_REG(hw, I40E_PFINT_LNKLSTN(msix_vect - 1),
> > -			(vsi->base_queue << I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT)
> |
> > +						(vsi->base_queue <<
> 
>                                   ^^^^^^^^^^^^^^^^^^
> > +				I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
> 
> Here why so many indentations? or just my mail agent's display issue, if it does
> pls ignor. anyway, it is not a beautiful style in my mind :)
I do not know why. The right patch can be downloaded directly from dpdk.org. The email editor may not show the correct format.

> 
>  		I40E_WRITE_REG(hw, I40E_PFINT_LNKLSTN(msix_vect - 1),
> -			(vsi->base_queue << I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT)
> |
> +			       (vsi->base_queue <<
> +				I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
> 
> BTW, for coding style, I think we can follow the linux kernel, for kernel
> indentations, it always use the format like: n"Tab"+m"space", 0 <= m <=7.
> Another example is qemu, it is all "space" without "tab".
Yes, we have DPDK style, tab or tab + space. Space only is not allowed. Checkpatch.pl can check it for us.

> 
> >  			(0x0 << I40E_PFINT_LNKLSTN_FIRSTQ_TYPE_SHIFT));
> >
> >  		I40E_WRITE_REG(hw, I40E_PFINT_ITRN(I40E_ITR_INDEX_DEFAULT,
> > -				msix_vect - 1), interval);
> > +						msix_vect - 1), interval);
> >
> >  		/* Disable auto-mask on enabling of all none-zero  interrupt */
> >  		I40E_WRITE_REG(hw, I40E_GLINT_CTL,
> > -				I40E_GLINT_CTL_DIS_AUTOMASK_N_MASK);
> > -	}
> > -	else {
> > +			I40E_GLINT_CTL_DIS_AUTOMASK_N_MASK);
> > +	} else {
> >  		uint32_t reg;
> > +
> >  		/* num_msix_vectors_vf needs to minus irq0 */
> >  		reg = (hw->func_caps.num_msix_vectors_vf - 1) *
> >  			vsi->user_param + (msix_vect - 1);
> >
> > -		I40E_WRITE_REG(hw, I40E_VPINT_LNKLSTN(reg),
> > -			(vsi->base_queue << I40E_VPINT_LNKLSTN_FIRSTQ_INDX_SHIFT)
> |
> > -			(0x0 << I40E_VPINT_LNKLSTN_FIRSTQ_TYPE_SHIFT));
> > +		I40E_WRITE_REG(hw, I40E_VPINT_LNKLSTN(reg), (vsi->base_queue
> <<
> > +					I40E_VPINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
> > +				(0x0 << I40E_VPINT_LNKLSTN_FIRSTQ_TYPE_SHIFT));
> 
> Here the same I think.
> 
> Thanks,
> Michael
> >  	}
> >
> >  	I40E_WRITE_FLUSH(hw);

Regards,
Helin

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

* Re: [dpdk-dev] [PATCH 1/2] i40e: code style fix
  2014-10-29  6:26     ` Zhang, Helin
@ 2014-10-29  8:06       ` Qiu, Michael
  0 siblings, 0 replies; 8+ messages in thread
From: Qiu, Michael @ 2014-10-29  8:06 UTC (permalink / raw)
  To: Zhang, Helin, dev

10/29/2014 2:26 PM, Zhang, Helin:
> Hi Michael
>
>> -----Original Message-----
>> From: Qiu, Michael
>> Sent: Wednesday, October 29, 2014 2:11 PM
>> To: Zhang, Helin; dev@dpdk.org
>> Subject: Re: [dpdk-dev] [PATCH 1/2] i40e: code style fix
>>
>>  10/29/2014 11:44 AM, Helin Zhang  :
>>> Add several code style fixes.
>>>
>>> Signed-off-by: Helin Zhang <helin.zhang@intel.com>
>>> ---
>>>  lib/librte_pmd_i40e/i40e_ethdev.c | 20 +++++++++++---------
>>>  1 file changed, 11 insertions(+), 9 deletions(-)
>>>
>>> diff --git a/lib/librte_pmd_i40e/i40e_ethdev.c
>>> b/lib/librte_pmd_i40e/i40e_ethdev.c
>>> index e1f581a..20c99a4 100644
>>> --- a/lib/librte_pmd_i40e/i40e_ethdev.c
>>> +++ b/lib/librte_pmd_i40e/i40e_ethdev.c
>>> @@ -580,7 +580,8 @@ i40e_vsi_queues_bind_intr(struct i40e_vsi *vsi)
>>>  	uint32_t val;
>>>  	struct i40e_hw *hw = I40E_VSI_TO_HW(vsi);
>>>  	uint16_t msix_vect = vsi->msix_intr;
>>> -	uint16_t interval = i40e_calc_itr_interval(RTE_LIBRTE_I40E_ITR_INTERVAL);
>>> +	uint16_t interval =
>>> +		i40e_calc_itr_interval(RTE_LIBRTE_I40E_ITR_INTERVAL);
>>>  	int i;
>>>
>>>  	for (i = 0; i < vsi->nb_qps; i++)
>>> @@ -603,25 +604,26 @@ i40e_vsi_queues_bind_intr(struct i40e_vsi *vsi)
>>>  	/* Write first RX queue to Link list register as the head element */
>>>  	if (vsi->type != I40E_VSI_SRIOV) {
>>>  		I40E_WRITE_REG(hw, I40E_PFINT_LNKLSTN(msix_vect - 1),
>>> -			(vsi->base_queue << I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT)
>> |
>>> +						(vsi->base_queue <<
>>                                   ^^^^^^^^^^^^^^^^^^
>>> +				I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
>> Here why so many indentations? or just my mail agent's display issue, if it does
>> pls ignor. anyway, it is not a beautiful style in my mind :)
> I do not know why. The right patch can be downloaded directly from dpdk.org. The email editor may not show the correct format.

OK, if so, pls ignore my comment:)

Thanks,
Michael
>
>>  		I40E_WRITE_REG(hw, I40E_PFINT_LNKLSTN(msix_vect - 1),
>> -			(vsi->base_queue << I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT)
>> |
>> +			       (vsi->base_queue <<
>> +				I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
>>
>> BTW, for coding style, I think we can follow the linux kernel, for kernel
>> indentations, it always use the format like: n"Tab"+m"space", 0 <= m <=7.
>> Another example is qemu, it is all "space" without "tab".
> Yes, we have DPDK style, tab or tab + space. Space only is not allowed. Checkpatch.pl can check it for us.
>
>>>  			(0x0 << I40E_PFINT_LNKLSTN_FIRSTQ_TYPE_SHIFT));
>>>
>>>  		I40E_WRITE_REG(hw, I40E_PFINT_ITRN(I40E_ITR_INDEX_DEFAULT,
>>> -				msix_vect - 1), interval);
>>> +						msix_vect - 1), interval);
>>>
>>>  		/* Disable auto-mask on enabling of all none-zero  interrupt */
>>>  		I40E_WRITE_REG(hw, I40E_GLINT_CTL,
>>> -				I40E_GLINT_CTL_DIS_AUTOMASK_N_MASK);
>>> -	}
>>> -	else {
>>> +			I40E_GLINT_CTL_DIS_AUTOMASK_N_MASK);
>>> +	} else {
>>>  		uint32_t reg;
>>> +
>>>  		/* num_msix_vectors_vf needs to minus irq0 */
>>>  		reg = (hw->func_caps.num_msix_vectors_vf - 1) *
>>>  			vsi->user_param + (msix_vect - 1);
>>>
>>> -		I40E_WRITE_REG(hw, I40E_VPINT_LNKLSTN(reg),
>>> -			(vsi->base_queue << I40E_VPINT_LNKLSTN_FIRSTQ_INDX_SHIFT)
>> |
>>> -			(0x0 << I40E_VPINT_LNKLSTN_FIRSTQ_TYPE_SHIFT));
>>> +		I40E_WRITE_REG(hw, I40E_VPINT_LNKLSTN(reg), (vsi->base_queue
>> <<
>>> +					I40E_VPINT_LNKLSTN_FIRSTQ_INDX_SHIFT) |
>>> +				(0x0 << I40E_VPINT_LNKLSTN_FIRSTQ_TYPE_SHIFT));
>> Here the same I think.
>>
>> Thanks,
>> Michael
>>>  	}
>>>
>>>  	I40E_WRITE_FLUSH(hw);
> Regards,
> Helin
>


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

* Re: [dpdk-dev] [PATCH 0/2] fix of configuring inside NIC RX interrupt
  2014-10-29  5:00 ` [dpdk-dev] [PATCH 0/2] " Chen, Jing D
@ 2014-10-29 22:44   ` Thomas Monjalon
  0 siblings, 0 replies; 8+ messages in thread
From: Thomas Monjalon @ 2014-10-29 22:44 UTC (permalink / raw)
  To: Zhang, Helin; +Cc: dev

2014-10-29 05:00, Chen, Jing D:
> From: Zhang, Helin
> > Inside NIC RX interrupt is needed for single RX descriptor
> > write back. The fix is to correct the wrong configuration
> > of register 'I40E_QINT_RQCTL'. In addition, several code
> > style fixes are added.
> > Note that interrupt will be inside NIC only, that means it
> > will never be reported outside NIC hardware.
> > 
> > Helin Zhang (2):
> >   i40e: code style fix
> >   i40e: fix of configuring inside NIC RX interrupt
> 
> Acked-by: Jing Chen <jing.d.chen@intel.com>

Applied

About commit log of the second patch:
- title of a fix should show which feature is fixed, not how it is fixed
- content of the commit log should explain what was broken

Thanks
-- 
Thomas

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

end of thread, other threads:[~2014-10-29 22:36 UTC | newest]

Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-10-29  3:42 [dpdk-dev] [PATCH 0/2] fix of configuring inside NIC RX interrupt Helin Zhang
2014-10-29  3:42 ` [dpdk-dev] [PATCH 1/2] i40e: code style fix Helin Zhang
2014-10-29  6:11   ` Qiu, Michael
2014-10-29  6:26     ` Zhang, Helin
2014-10-29  8:06       ` Qiu, Michael
2014-10-29  3:42 ` [dpdk-dev] [PATCH 2/2] i40e: fix of configuring inside NIC RX interrupt Helin Zhang
2014-10-29  5:00 ` [dpdk-dev] [PATCH 0/2] " Chen, Jing D
2014-10-29 22:44   ` Thomas Monjalon

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).