From: Stephen Hemminger <[EMAIL PROTECTED]>
Date: Tue, 11 Dec 2007 15:42:22 -0800
> Perhaps we should change the warning to identify the guilty device.
Applied.
Stephen, you often don't supply a proper signoff line
for one-off changes like this and I find it very irritating.
It doesn't cost you
From: Stephen Hemminger [EMAIL PROTECTED]
Date: Tue, 11 Dec 2007 15:42:22 -0800
Perhaps we should change the warning to identify the guilty device.
Applied.
Stephen, you often don't supply a proper signoff line
for one-off changes like this and I find it very irritating.
It doesn't cost you
From: "Brandeburg, Jesse" <[EMAIL PROTECTED]>
Date: Tue, 11 Dec 2007 16:38:37 -0800
> @@ -3933,6 +3933,10 @@ quit_polling:
> e1000_set_itr(adapter);
> netif_rx_complete(poll_dev, napi);
> e1000_irq_enable(adapter);
> + if
From: Brandeburg, Jesse [EMAIL PROTECTED]
Date: Tue, 11 Dec 2007 16:38:37 -0800
@@ -3933,6 +3933,10 @@ quit_polling:
e1000_set_itr(adapter);
netif_rx_complete(poll_dev, napi);
e1000_irq_enable(adapter);
+ if (work_done ==
Joonwoo Park wrote:
> 2007/12/12, Brandeburg, Jesse <[EMAIL PROTECTED]>:
>>
>> all drivers using NAPI in 2.6.24+ (NNAPI??) must return zero here,
>> after calling netif_rx_complete. netif_rx_complete plus work_done
>> != 0 causes a bug.
>>
>
> Brandeburg,
> Don't we need to return non-zero
2007/12/12, Brandeburg, Jesse <[EMAIL PROTECTED]>:
>
> all drivers using NAPI in 2.6.24+ (NNAPI??) must return zero here, after
> calling netif_rx_complete. netif_rx_complete plus work_done != 0 causes
> a bug.
>
Brandeburg,
Don't we need to return non-zero work_done after netif_rx_complete if
Perhaps we should change the warning to identify the guilty device.
--- a/net/core/dev.c2007-11-19 09:09:57.0 -0800
+++ b/net/core/dev.c2007-12-07 15:54:03.0 -0800
@@ -2196,7 +2196,13 @@ static void net_rx_action(struct softirq
if
2007/12/12, Brandeburg, Jesse <[EMAIL PROTECTED]>:
> Joonwoo Park wrote:
> > /* If no Tx and not enough Rx work done, exit the polling mode */
> > if ((!tx_cleaned && (work_done == 0)) ||
> >!netif_running(poll_dev)) {
> > quit_polling:
> > if (likely(adapter->itr_setting &
Joonwoo Park wrote:
> /* If no Tx and not enough Rx work done, exit the polling mode */
> if ((!tx_cleaned && (work_done == 0)) ||
>!netif_running(poll_dev)) {
> quit_polling:
> if (likely(adapter->itr_setting & 3))
> e1000_set_itr(adapter);
>
2007/12/11, David Miller <[EMAIL PROTECTED]>:
> From: "Joonwoo Park" <[EMAIL PROTECTED]>
> Date: Tue, 11 Dec 2007 18:13:34 +0900
>
> Joonwoo-ssi annyoung haseyo,
Wow Great! :-)
> How can the NAPI_STATE_SCHED bit be cleared externally yet we take
> this list_move_tail() code path?
>
> If
From: Herbert Xu <[EMAIL PROTECTED]>
Date: Tue, 11 Dec 2007 20:36:21 +0800
> David Miller <[EMAIL PROTECTED]> wrote:
> >
> > How can the NAPI_STATE_SCHED bit be cleared externally yet we take
> > this list_move_tail() code path?
>
> His driver is probably buggy. When we had two drivers
David Miller <[EMAIL PROTECTED]> wrote:
>
> How can the NAPI_STATE_SCHED bit be cleared externally yet we take
> this list_move_tail() code path?
His driver is probably buggy. When we had two drivers beginning
with e100 we often forgot to apply fixes to the both of them. Now
that we have three
From: "Joonwoo Park" <[EMAIL PROTECTED]>
Date: Tue, 11 Dec 2007 18:13:34 +0900
Joonwoo-ssi annyoung haseyo,
> [NET]: Fix Ooops of napi net_rx_action.
> Before doing list_move_tail napi poll_list, it should be ensured
>
> Signed-off-by: Joonwoo Park <[EMAIL PROTECTED]>
> ---
> diff --git
[NET]: Fix Ooops of napi net_rx_action.
Before doing list_move_tail napi poll_list, it should be ensured
Signed-off-by: Joonwoo Park <[EMAIL PROTECTED]>
---
diff --git a/net/core/dev.c b/net/core/dev.c
index 86d6261..74bd5ab 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -2207,7 +2207,8 @@
[NET]: Fix Ooops of napi net_rx_action.
Before doing list_move_tail napi poll_list, it should be ensured
Signed-off-by: Joonwoo Park [EMAIL PROTECTED]
---
diff --git a/net/core/dev.c b/net/core/dev.c
index 86d6261..74bd5ab 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -2207,7 +2207,8 @@
From: Joonwoo Park [EMAIL PROTECTED]
Date: Tue, 11 Dec 2007 18:13:34 +0900
Joonwoo-ssi annyoung haseyo,
[NET]: Fix Ooops of napi net_rx_action.
Before doing list_move_tail napi poll_list, it should be ensured
Signed-off-by: Joonwoo Park [EMAIL PROTECTED]
---
diff --git a/net/core/dev.c
David Miller [EMAIL PROTECTED] wrote:
How can the NAPI_STATE_SCHED bit be cleared externally yet we take
this list_move_tail() code path?
His driver is probably buggy. When we had two drivers beginning
with e100 we often forgot to apply fixes to the both of them. Now
that we have three it's
From: Herbert Xu [EMAIL PROTECTED]
Date: Tue, 11 Dec 2007 20:36:21 +0800
David Miller [EMAIL PROTECTED] wrote:
How can the NAPI_STATE_SCHED bit be cleared externally yet we take
this list_move_tail() code path?
His driver is probably buggy. When we had two drivers beginning
with e100
2007/12/11, David Miller [EMAIL PROTECTED]:
From: Joonwoo Park [EMAIL PROTECTED]
Date: Tue, 11 Dec 2007 18:13:34 +0900
Joonwoo-ssi annyoung haseyo,
Wow Great! :-)
How can the NAPI_STATE_SCHED bit be cleared externally yet we take
this list_move_tail() code path?
If NAPI_STATE_SCHED is
Joonwoo Park wrote:
/* If no Tx and not enough Rx work done, exit the polling mode */
if ((!tx_cleaned (work_done == 0)) ||
!netif_running(poll_dev)) {
quit_polling:
if (likely(adapter-itr_setting 3))
e1000_set_itr(adapter);
2007/12/12, Brandeburg, Jesse [EMAIL PROTECTED]:
Joonwoo Park wrote:
/* If no Tx and not enough Rx work done, exit the polling mode */
if ((!tx_cleaned (work_done == 0)) ||
!netif_running(poll_dev)) {
quit_polling:
if (likely(adapter-itr_setting 3))
Perhaps we should change the warning to identify the guilty device.
--- a/net/core/dev.c2007-11-19 09:09:57.0 -0800
+++ b/net/core/dev.c2007-12-07 15:54:03.0 -0800
@@ -2196,7 +2196,13 @@ static void net_rx_action(struct softirq
if
2007/12/12, Brandeburg, Jesse [EMAIL PROTECTED]:
all drivers using NAPI in 2.6.24+ (NNAPI??) must return zero here, after
calling netif_rx_complete. netif_rx_complete plus work_done != 0 causes
a bug.
Brandeburg,
Don't we need to return non-zero work_done after netif_rx_complete if
Joonwoo Park wrote:
2007/12/12, Brandeburg, Jesse [EMAIL PROTECTED]:
all drivers using NAPI in 2.6.24+ (NNAPI??) must return zero here,
after calling netif_rx_complete. netif_rx_complete plus work_done
!= 0 causes a bug.
Brandeburg,
Don't we need to return non-zero work_done after
24 matches
Mail list logo