[PATCH 2/2] staging/et131x: remove memset for the TCB

[PATCH 2/2] staging/et131x: remove memset for the TCB

[Devendra Naga]

TCB has kcalloc' ed so no need to call memset again on it.

Signed-off-by: Devendra Naga <devendra.aaru@gmail.com>
---
 drivers/staging/et131x/et131x.c |    2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/staging/et131x/et131x.c b/drivers/staging/et131x/et131x.c
index 227fca9..5010894 100644
--- a/drivers/staging/et131x/et131x.c
+++ b/drivers/staging/et131x/et131x.c
@@ -2131,8 +2131,6 @@ static void et131x_init_send(struct et131x_adapter *adapter)
 
 	tx_ring->tcb_qhead = tcb;
 
-	memset(tcb, 0, sizeof(struct tcb) * NUM_TCB);
-
 	/* Go through and set up each TCB */
 	for (ct = 0; ct++ < NUM_TCB; tcb++)
 		/* Set the link pointer in HW TCB to the next TCB in the
-- 
1.7.9.5

Re: [PATCH 2/2] staging/et131x: remove memset for the TCB

[Dan Carpenter]

On Thu, Jul 12, 2012 at 11:54:28AM +0545, Devendra Naga wrote:
> TCB has kcalloc' ed so no need to call memset again on it.
> 

Gar...  What about when this is called from et131x_adjust_link() and
et1310_disable_phy_coma()?

Really this changelog is annoying.  Based on the changelog I was
expecting the kcalloc() to be in the same function but it wasn't.
So then I have to redo all your work and check every code path.  You
could have said what the code paths were in your description so I
don't have to grovel through it all myself.

regards,
dan carpenter

Re: [PATCH 2/2] staging/et131x: remove memset for the TCB

[Alan Cox]

On Thu, 12 Jul 2012 11:54:28 +0545
Devendra Naga <devendra.aaru@gmail.com> wrote:

> TCB has kcalloc' ed so no need to call memset again on it.
> 
> Signed-off-by: Devendra Naga <devendra.aaru@gmail.com>
> ---
>  drivers/staging/et131x/et131x.c |    2 --
>  1 file changed, 2 deletions(-)
> 
> diff --git a/drivers/staging/et131x/et131x.c b/drivers/staging/et131x/et131x.c
> index 227fca9..5010894 100644
> --- a/drivers/staging/et131x/et131x.c
> +++ b/drivers/staging/et131x/et131x.c
> @@ -2131,8 +2131,6 @@ static void et131x_init_send(struct et131x_adapter *adapter)
>  
>  	tx_ring->tcb_qhead = tcb;
>  
> -	memset(tcb, 0, sizeof(struct tcb) * NUM_TCB);

Its such a rarely used path with several entry points that it's probably
safer kept as is.

Alan

Re: [PATCH 2/2] staging/et131x: remove memset for the TCB

[devendra.aaru]

On Thu, Jul 12, 2012 at 2:34 PM, Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:
> On Thu, 12 Jul 2012 11:54:28 +0545
> Devendra Naga <devendra.aaru@gmail.com> wrote:
>
>> TCB has kcalloc' ed so no need to call memset again on it.
>>
>> Signed-off-by: Devendra Naga <devendra.aaru@gmail.com>
>> ---
>>  drivers/staging/et131x/et131x.c |    2 --
>>  1 file changed, 2 deletions(-)
>>
>> diff --git a/drivers/staging/et131x/et131x.c b/drivers/staging/et131x/et131x.c
>> index 227fca9..5010894 100644
>> --- a/drivers/staging/et131x/et131x.c
>> +++ b/drivers/staging/et131x/et131x.c
>> @@ -2131,8 +2131,6 @@ static void et131x_init_send(struct et131x_adapter *adapter)
>>
>>       tx_ring->tcb_qhead = tcb;
>>
>> -     memset(tcb, 0, sizeof(struct tcb) * NUM_TCB);
>
> Its such a rarely used path with several entry points that it's probably
> safer kept as is.
>
Hi Alan Sir,

 Glad to see your reply, I really didn't saw the whole code paths, was
seeing the ring allocations, sorry for the noise.
> Alan

Thanks,
Devendra.

Re: [PATCH 2/2] staging/et131x: remove memset for the TCB

[devendra.aaru]

On Thu, Jul 12, 2012 at 2:21 PM, Dan Carpenter <dan.carpenter@oracle.com> wrote:
> On Thu, Jul 12, 2012 at 11:54:28AM +0545, Devendra Naga wrote:
>> TCB has kcalloc' ed so no need to call memset again on it.
>>
>
> Gar...  What about when this is called from et131x_adjust_link() and
> et1310_disable_phy_coma()?
>
Yeah, i didn't looked those paths, i am sorry.
> Really this changelog is annoying.  Based on the changelog I was
> expecting the kcalloc() to be in the same function but it wasn't.
> So then I have to redo all your work and check every code path.  You
> could have said what the code paths were in your description so I
> don't have to grovel through it all myself.
>
Ok. will try to put all the info i found while sending patches from next time
> regards,
> dan carpenter
>
>

Thanks,
Devendra.