[net] ixgbe: check return value of napi_complete_done()

2018-10-03 Thread Jeff Kirsher
From: Song Liu The NIC driver should only enable interrupts when napi_complete_done() returns true. This patch adds the check for ixgbe. Cc: sta...@vger.kernel.org # 4.10+ Suggested-by: Eric Dumazet Signed-off-by: Song Liu Tested-by: Andrew Bowers Signed-off-by: Jeff Kirsher --- drivers/net

[PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-20 Thread Song Liu
The NIC driver should only enable interrupts when napi_complete_done() returns true. This patch adds the check for ixgbe. Cc: sta...@vger.kernel.org # 4.10+ Cc: Jeff Kirsher Suggested-by: Eric Dumazet Signed-off-by: Song Liu --- drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 12 +++-

Re: [net] ixgbe: check return value of napi_complete_done()

2018-10-03 Thread David Miller
From: Jeff Kirsher Date: Wed, 3 Oct 2018 11:30:35 -0700 > From: Song Liu > > The NIC driver should only enable interrupts when napi_complete_done() > returns true. This patch adds the check for ixgbe. > > Cc: sta...@vger.kernel.org # 4.10+ > Suggested-by: Eric Dumazet > Signed-off-by: Song L

Re: [PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-20 Thread Eric Dumazet
On 09/20/2018 12:01 PM, Song Liu wrote: > The NIC driver should only enable interrupts when napi_complete_done() > returns true. This patch adds the check for ixgbe. > > Cc: sta...@vger.kernel.org # 4.10+ > Cc: Jeff Kirsher > Suggested-by: Eric Dumazet > Signed-off-by: Song Liu > --- Well,

Re: [PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-20 Thread Jeff Kirsher
On Thu, 2018-09-20 at 13:35 -0700, Eric Dumazet wrote: > On 09/20/2018 12:01 PM, Song Liu wrote: > > The NIC driver should only enable interrupts when napi_complete_done() > > returns true. This patch adds the check for ixgbe. > > > > Cc: sta...@vger.kernel.org # 4.10+ > > Cc: Jeff Kirsher > > Su

Re: [PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-20 Thread Song Liu
> On Sep 20, 2018, at 2:01 PM, Jeff Kirsher wrote: > > On Thu, 2018-09-20 at 13:35 -0700, Eric Dumazet wrote: >> On 09/20/2018 12:01 PM, Song Liu wrote: >>> The NIC driver should only enable interrupts when napi_complete_done() >>> returns true. This patch adds the check for ixgbe. >>> >>> Cc

Re: [PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-20 Thread Eric Dumazet
On 09/20/2018 03:42 PM, Song Liu wrote: > > >> On Sep 20, 2018, at 2:01 PM, Jeff Kirsher >> wrote: >> >> On Thu, 2018-09-20 at 13:35 -0700, Eric Dumazet wrote: >>> On 09/20/2018 12:01 PM, Song Liu wrote: The NIC driver should only enable interrupts when napi_complete_done() returns

Re: [PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-20 Thread Song Liu
> On Sep 20, 2018, at 4:22 PM, Eric Dumazet wrote: > > > > On 09/20/2018 03:42 PM, Song Liu wrote: >> >> >>> On Sep 20, 2018, at 2:01 PM, Jeff Kirsher >>> wrote: >>> >>> On Thu, 2018-09-20 at 13:35 -0700, Eric Dumazet wrote: On 09/20/2018 12:01 PM, Song Liu wrote: > The NIC dri

Re: [PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-20 Thread Eric Dumazet
On 09/20/2018 04:43 PM, Song Liu wrote: > > I tried to totally skip ndo_poll_controller() here. It did avoid hitting > the issue. However, netpoll will drop (fail to send) more packets. > Why is it failing ? If you are under high memory pressure, then maybe if you absolutely want memory to

Re: [PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-21 Thread Song Liu
> On Sep 20, 2018, at 4:49 PM, Eric Dumazet wrote: > > > > On 09/20/2018 04:43 PM, Song Liu wrote: >> > >> I tried to totally skip ndo_poll_controller() here. It did avoid hitting >> the issue. However, netpoll will drop (fail to send) more packets. >> > > Why is it failing ? > > If yo

Re: [PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-21 Thread Eric Dumazet
On 09/21/2018 12:17 AM, Song Liu wrote: > > >> On Sep 20, 2018, at 4:49 PM, Eric Dumazet wrote: >> >> >> >> On 09/20/2018 04:43 PM, Song Liu wrote: >>> >> >>> I tried to totally skip ndo_poll_controller() here. It did avoid hitting >>> the issue. However, netpoll will drop (fail to send) more

Re: [PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-21 Thread Alexei Starovoitov
On 9/21/18 6:33 AM, Eric Dumazet wrote: On 09/21/2018 12:17 AM, Song Liu wrote: On Sep 20, 2018, at 4:49 PM, Eric Dumazet wrote: On 09/20/2018 04:43 PM, Song Liu wrote: I tried to totally skip ndo_poll_controller() here. It did avoid hitting the issue. However, netpoll will drop (f

Re: [PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-21 Thread Eric Dumazet
On 09/21/2018 07:55 AM, Alexei Starovoitov wrote: > > should we remove ndo_poll_controller then? > My understanding that the patch helps by not letting > drivers do napi_schedule() for all queues into this_cpu, right? > But most of the drivers do exactly that in their ndo_poll_controller > imp

Re: [PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-21 Thread Alexei Starovoitov
On 9/21/18 7:59 AM, Eric Dumazet wrote: > > > On 09/21/2018 07:55 AM, Alexei Starovoitov wrote: > >> >> should we remove ndo_poll_controller then? >> My understanding that the patch helps by not letting >> drivers do napi_schedule() for all queues into this_cpu, right? >> But most of the drivers do

Re: [PATCH net] ixgbe: check return value of napi_complete_done()

2018-09-21 Thread Eric Dumazet
On 09/21/2018 08:14 AM, Alexei Starovoitov wrote: > On 9/21/18 7:59 AM, Eric Dumazet wrote: >> >> >> On 09/21/2018 07:55 AM, Alexei Starovoitov wrote: >> >>> >>> should we remove ndo_poll_controller then? >>> My understanding that the patch helps by not letting >>> drivers do napi_schedule() for