[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20191210.172821.1406974012095303846.davem@davemloft.net>
Date: Tue, 10 Dec 2019 17:28:21 -0800 (PST)
From: David Miller <davem@...emloft.net>
To: bpf@...r.kernel.org, netdev@...r.kernel.org
Cc: ast@...nel.org, daniel@...earbox.net, tglx@...utronix.de
Subject: Re: [RFC v1 PATCH 0/7] bpf: Make RT friendly.
From: David Miller <davem@...emloft.net>
Date: Sat, 07 Dec 2019 16:03:57 -0800 (PST)
>
> The goal of this patch set is to make the BPF code friendly in RT
> configurations.
>
> The first step is eliminating preemption disable/enable and replacing
> it with local lock usage when full RT is enabled.
>
> Likewise we also need to elide usage of up_read_non_owner() in the
> stackmap code when full RT is turned on.
>
> Signed-off-by: David S. Miller <davem@...emloft.net>
Thomas can you please take a look at this patch series?
It eliminates all of the RT problems we were made aware of, and these
patches have been through the bpf test suite as well as gotten 0-day
testing.
The only major thing we needs ACK'd is the locallock stubs.
Thank you.
Powered by blists - more mailing lists