[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAHp75Vejt2h_nnLwS+_hy1WYSgo_U7ry1vRAYksaCA4e002k4A@mail.gmail.com>
Date: Wed, 24 Jan 2018 18:40:32 +0200
From: Andy Shevchenko <andy.shevchenko@...il.com>
To: Alan Cox <gnomes@...rguk.ukuu.org.uk>
Cc: Shankara Pailoor <sp3485@...umbia.edu>,
Greg KH <gregkh@...uxfoundation.org>,
Jiri Slaby <jslaby@...e.com>,
LKML <linux-kernel@...r.kernel.org>,
syzkaller <syzkaller@...glegroups.com>,
linux-serial@...r.kernel.org
Subject: Re: RCU stall in 8250 serial driver Linux 4.15-rc1
On Tue, Jan 23, 2018 at 5:52 PM, Alan Cox <gnomes@...rguk.ukuu.org.uk> wrote:
> On Wed, 17 Jan 2018 09:24:32 -0800
> Shankara Pailoor <sp3485@...umbia.edu> wrote:
>
>> Hi Greg,
>>
>> Sorry for that. Here is the stack trace. C Program below
>
>
>
>> serial_in drivers/tty/serial/8250/8250.h:111 [inline]
>> wait_for_xmitr+0x8a/0x1d0 drivers/tty/serial/8250/8250_port.c:2033
>> serial8250_console_putchar+0x19/0x50 drivers/tty/serial/8250/8250_port.c:3170
>> uart_console_write+0x98/0xc0 drivers/tty/serial/serial_core.c:1858
>
> The console is spinning in polled mode trying to write data to the
> console port, which it seems has gotten a bit stuck. Could be the
> hypervisor hits some kind of buffering limit, could be hypervisor
> interface broke.
>
> Either way the console interface is supposed to stall the machine to
> ensure the bytes always get out and if your serial port jams or gets
> massively behind then this will happen.
+1
https://lkml.org/lkml/2017/12/21/480
--
With Best Regards,
Andy Shevchenko
Powered by blists - more mailing lists