Date: Tue, 16 Mar 2010 16:11:39 -0400 From: Charles Owens <cowens@greatbaysoftware.com> To: ticso@cicely.de Cc: freebsd-hardware@freebsd.org Subject: Re: ttyu1 stuck at 115200 Message-ID: <4B9FE5FB.8010201@greatbaysoftware.com> In-Reply-To: <20100316012827.GN39056@cicely7.cicely.de> References: <4B9EBB10.5050804@greatbaysoftware.com> <20100316012827.GN39056@cicely7.cicely.de>
next in thread | previous in thread | raw e-mail | index | archive | help
Bernd Walter wrote: > On Mon, Mar 15, 2010 at 06:56:16PM -0400, Charles Owens wrote: > >> Hello, >> >> Working with pretty vanilla hardware running 8.0-RELEASE-p2, I find that >> ttyu0 works just fine as a serial console, but ttyu1 wants to default to >> 115200 baud (ttyu0 defaults to 9600 as expected). I'm used having >> details like baud-rate handled by argument given to getty via the >> respective line in /etc/ttys... but that doesn't seem to be the behavior >> I'm seeing. I couldn't get a login prompt at 9600 baud until I added >> the following line to /boot/device.hints and rebooted: >> >> hint.uart.1.baud="9600" >> >> Is this supposed to be how this is done now, or is there something else >> going on here? Any thoughts as to why the two ports behave differently? >> > > This is automatically passed from earlier bootcode, but as long as > this code isn't compiled for uart1 it setups uart0. > Previously the kernel default was 9600, but it seems this has changed. > If you use boot2 with serial you need to recompile and reinstall it > with BOOT_COMCONSOLE_PORT and BOOT_COMCONSOLE_SPEED setup. > You need to recompile loader with at least BOOT_COMCONSOLE_PORT to > handle a different port instead, but you can change the speed with > comconsole_speed in loader.conf - nevertheless the default is also > set with BOOT_COMCONSOLE_SPEED at compiletime. > You should have at least the loader on serial as well - otherwise > you can't change boot parameters or boot another kernel. > I even prefer running boot2 on serial so I can switch loader in case > something breaks loader with an update. > That's helpful, thanks. The funny thing is that, with some systems (VMware, for example), both ports stay at 9600 baud (unlike this Intel server I'm wrestling with). Makes me think that maybe there's some kind of capabilities negotiation going on between the hardware and the driver. In any case, you've confirmed some things for me that hadn't quite sunk in. Thanks, Charles
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4B9FE5FB.8010201>