On Mon, 2008-12-15 at 12:18 -0500, Joel Granados wrote: > ----- "Hans de Goede" <hdegoede@xxxxxxxxxx> wrote: > Well, I spoke to my nearest kernel hacker and he said that the behavior with two arguments was undefined for rhel4 kernel. Moreover console seems to end up misconfigured after install. The behavior has been defined for as long as I can remember. Your nearest kernel hacker is mistaken ;-) What happens is that all consoles become print buffers (show messages) -- the last specified console= arguments becomes the "actual" console device which allows input as well as output. > > Hmm, I'm not terrible famliar with this code but why do we want to > > only have > > one console argument, according to the kernel docs it is perfectly > > fine to have > > can you point me to the docs, maybe this has some other cause. Maybe its user negligence, I see now that the second console argument was ttyS0 which is the default. Maybe what happened was that anaconda was installed without the console argument. Which is strange becuase that would mean that at installation time the person would have had to had access to the machine through other means. And I understand by report that that was most unlikely. > If you are correct this is most likely not a bug. :) Documentation/serial-console.txt in the kernel tree > > multiple console=foo options, any kernel output will be send to all of > > them, > > And will receive input from all of them as well. right? No, input is only from the last specified console. Jeremy _______________________________________________ Anaconda-devel-list mailing list Anaconda-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/anaconda-devel-list
