[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAKgT0Uc88xF5AjEhpBnr7m_+gnH5WVAoJFC=AVEPY0+qN1BNsQ@mail.gmail.com>
Date: Thu, 26 Jan 2023 08:40:19 -0800
From: Alexander Duyck <alexander.duyck@...il.com>
To: Felix Fietkau <nbd@....name>
Cc: Ilias Apalodimas <ilias.apalodimas@...aro.org>,
netdev@...r.kernel.org, Jesper Dangaard Brouer <hawk@...nel.org>,
"David S. Miller" <davem@...emloft.net>,
Eric Dumazet <edumazet@...gle.com>,
Jakub Kicinski <kuba@...nel.org>,
Paolo Abeni <pabeni@...hat.com>,
Lorenzo Bianconi <lorenzo@...nel.org>,
linux-kernel@...r.kernel.org, Yunsheng Lin <linyunsheng@...wei.com>
Subject: Re: [PATCH] net: page_pool: fix refcounting issues with fragmented allocation
?
On Thu, Jan 26, 2023 at 8:08 AM Alexander Duyck
<alexander.duyck@...il.com> wrote:
>
> On Thu, Jan 26, 2023 at 1:15 AM Felix Fietkau <nbd@....name> wrote:
> >
...
> > - if I return false in skb_try_coalesce, it still crashes:
> > https://nbd.name/p/18cac078
>
> Yeah, I wasn't suspecting skb_try_coalesce since we have exercised the
> code. My thought was something like the function you mentioned above
> plus cloning or something else.
One question I would have. Is GRO happening after the call to
ieee80211_amsdu_to_8023s? If so we might want to try switching that
off to see if it might be aggregating page pool frames and non-page
pool frames together.
Powered by blists - more mailing lists