[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <0BA3FCBA62E2DC44AF3030971E174FB32E8264A2@hasmsx107.ger.corp.intel.com>
Date: Thu, 20 Aug 2015 13:27:48 +0000
From: "Grumbach, Emmanuel" <emmanuel.grumbach@...el.com>
To: Eric Dumazet <eric.dumazet@...il.com>
CC: "linux-wireless@...r.kernel.org" <linux-wireless@...r.kernel.org>,
"Sharon, Sara" <sara.sharon@...el.com>,
"ido@...ery.com" <ido@...ery.com>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>
Subject: Re: [RFC v3 1/3] iwlwifi: mvm: add real TSO implementation
On 08/20/2015 04:13 PM, Eric Dumazet wrote:
> On Thu, 2015-08-20 at 11:15 +0300, Emmanuel Grumbach wrote:
>> The segmentation is done completely in software. The
>> driver creates several MPDUs out of a single large send.
>> Each MPDU is a newly allocated SKB.
>> A page is allocated to create the headers that need to be
>> duplicated (SNAP / IP / TCP). The WiFi header is in the
>> header of the newly created SKBs.
>>
>> type=feature
>>
>> Change-Id: I238ffa79cacc5bbdacdfbf3e9673c8d4f02b462a
>> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@...el.com>
>> ---
>> drivers/net/wireless/iwlwifi/mvm/tx.c | 513 +++++++++++++++++++++++++++++++---
>> 1 file changed, 481 insertions(+), 32 deletions(-)
>
>
> Ouch again.
>
> This will be a NACK. Sorry.
>
>
To be honest, I don't see why. No matter how you look at it, you need to
add headers. And these headers need to sit on some memory that is not
yet allocated. skb_gso_segment allocates new skb which is a memory
allocation.
I looked at ethernet/marvell/mvneta.c (which uses net/core/tso.c) and
they pre-allocate DMA coherent memory for the additional headers they
create. I can have pre-allocated pages if you want.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists