[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <ZKRAlJMi3tjOSrXD@gmail.com>
Date: Tue, 4 Jul 2023 08:53:56 -0700
From: Breno Leitao <leitao@...ian.org>
To: Andrew Lunn <andrew@...n.ch>
Cc: "David S. Miller" <davem@...emloft.net>,
Eric Dumazet <edumazet@...gle.com>,
Jakub Kicinski <kuba@...nel.org>,
Paolo Abeni <pabeni@...hat.com>, sergey.senozhatsky@...il.com,
pmladek@...e.com, Dave Jones <davej@...emonkey.org.uk>,
"open list:NETWORKING DRIVERS" <netdev@...r.kernel.org>,
open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] netconsole: Append kernel version to message
On Mon, Jul 03, 2023 at 06:46:25PM +0200, Andrew Lunn wrote:
> Hi Breno
Hello,
> Why not just send the message without uname appended. You probably
> want to see the OOM messages...
>
> Also, what context are we in here? Should that be GFP_ATOMIC, which
> net/core/netpoll.c is using to allocate the skbs?
Maybe this is not necessary anymore, since I might be using the buffer
already allocated.
> > +static inline void send_msg_udp(struct netconsole_target *nt,
> > + const char *msg, unsigned int len)
> > +{
> > +#ifdef CONFIG_NETCONSOLE_UNAME
> > + send_ext_msg_udp_uname(nt, msg, len);
> > +#else
> > + send_ext_msg_udp(nt, msg, len);
> > +#endif
>
> Please use
>
> if (IS_ENABLED(CONFIG_NETCONSOLE_UNAME)) {} else {}
>
> so the code is compiled and then thrown away. That nakes build testing
> more efficient.
Makes total sense, I am incorporating it into v2 now.
Thanks!
Powered by blists - more mailing lists