linux-kernel - kgdb debugger proxy question - using the same serial port for both serial console and kgdb

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <AANLkTinjVwjY88eb9IhxojuerroCY2EqnUy4-J4fQy0o@mail.gmail.com>
Date:	Mon, 31 May 2010 00:51:50 +0300
From:	Paul Fisher <r01.gjk@...il.com>
To:	linux-kernel@...r.kernel.org
Cc:	jason.wessel@...driver.com
Subject: kgdb debugger proxy question - using the same serial port for both 
	serial console and kgdb

Hello,

I'm trying to get kgdb setup working. I have two PCs but unfortunately there is
only one serial port in each of those. So far everything worked great, but
several days ago one of the monitors was taken away from my desk, and the second
machine is now headless.

So now I need some kind of splitter to split the traffic between the gdb and
the minicom (serial console). Since Jason wrote in the kgdb documentation:

> For kgdb/gdb, kgdboc is designed to work with a single serial port. It is
> intended to cover the circumstance where you want to use a serial console
> as our primary console as well as using it to perform kernel debugging.

I assume that such splitter should exist. Moreover, Jason mentions that
'debugger proxy' several times throughout the docs. So I searched the archives
but found only Ingo's mail from 2008 where he asked for a link to such a proxy
but without any response.

After a deeper research I found a kdmx (kgdb demux) script, and in fact it
works, but in a very limited and weird way. For one, you can't send a BRK
(break) from minicom (Ctrl-A f) because it operates on the system call level
(read, write, select syscalls) and the BRK is sent by means of ioctl(). Sending
a break is essential, because that's the way to stop kernel execution and break
into the debugger. Also detach gdb command freezes the whole serial line because
of the way this script is written. I assume that detach is not the only command
that does that...

So does a reliable splitter actually exist? If it doesn't, could you
please share
your experience - briefly describe your kgdb setup? I bet that someone already
had this problem...

Much appreciated,

--paul
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/