[quoted lines by Mario Lang on 2007/11/13 at 15:30 +0100]
>No, the ramfs does not stay mounted, its gone after the real /sbin/init
>starts.
Then I don't see how brltty can chroot to the real root. It needs a valid path
to do that. Do you know of good documentation for exactly what pivot-root doe
Dave Mielke <[EMAIL PROTECTED]> writes:
>>"Shows up" is going to be complicated, because it might as well take some
>>time until the chroot actually happens and the root fs is fully populated
>>(e.g.,
>>I use a custom setup at work where the root fs is actually retrieved
>>with rsync, that can ta
[quoted lines by Mario Lang on 2007/11/08 at 20:11 +0100]
>>> So I could rpbobably also just set -W /dev ?
...
>For the record, I tried and it did not help.
Are you able to collect some debug logs so that we can try to figure out why it
doesn't work?
--
Dave Mielke | 2213 Fox Cresc
[quoted lines by Mario Lang on 2007/11/08 at 20:08 +0100]
>"Shows up" is going to be complicated, because it might as well take some
>time until the chroot actually happens and the root fs is fully populated
>(e.g.,
>I use a custom setup at work where the root fs is actually retrieved
>with rsync
Hi,
Mario Lang, le Thu 08 Nov 2007 20:08:36 +0100, a écrit :
> Could we perhaps use a special signal to implement this?
> We could have some parameter to brltty that tells it what the path
> of the new root fs is going to be, and if brltty receives this
> signal, it will reopen all its filehandles
Samuel Thibault <[EMAIL PROTECTED]> writes:
>> > Note that brltty does try to create them within the writable directory but
>> > it
>> > won't be able to if that directory either doesn't exist or isn't writable.
>>
>> So I could rpbobably also just set -W /dev ?
>
> That should probably work fo
Dave Mielke <[EMAIL PROTECTED]> writes:
>>Actually, initramfs moves the /dev hierarchy over to the
>>new root directory, so the /dev directory is actually not different,
>>but I guess it could happen that brltty still lives in the old root.
>
> Yes, the current root directory is part of a process'
Dave Mielke, le Thu 08 Nov 2007 08:33:53 -0500, a écrit :
> [quoted lines by Samuel Thibault on 2007/11/08 at 11:55 +]
>
> >udev creates them because gettys get launched and thus virtual consoles
> >get allocated.
>
> It has to be more complicated than that because there are vcsa devices for
[quoted lines by Samuel Thibault on 2007/11/08 at 11:55 +]
>udev creates them because gettys get launched and thus virtual consoles
>get allocated.
It has to be more complicated than that because there are vcsa devices for
which there aren't corresponding gettys. For example, a common case i
Dave Mielke, le Thu 08 Nov 2007 06:52:57 -0500, a écrit :
> >I just checked, tty exists, but vcsa1 is the only device
> >that exists (because init has not yet started and there are no gettys yet).
>
> I don't think gettys create those. I think its udev that does.
udev creates them because gettys
[quoted lines by Mario Lang on 2007/11/08 at 10:03 +0100]
>Actually, initramfs moves the /dev hierarchy over to the
>new root directory, so the /dev directory is actually not different,
>but I guess it could happen that brltty still lives in the old root.
Yes, the current root directory is part o
Mario Lang, le Thu 08 Nov 2007 10:03:24 +0100, a écrit :
> >>I guess this is because some device did not exist in my /dev hierarchy
> >>while brltty was started first. Is this right?
> >
> > Yes, it would be.
> >
> >>Can we perhaps fix this such that brltty does not need to be restarted?
> >
> > T
Dave Mielke <[EMAIL PROTECTED]> writes:
>>I guess this is because some device did not exist in my /dev hierarchy
>>while brltty was started first. Is this right?
>
> Yes, it would be.
>
>>Can we perhaps fix this such that brltty does not need to be restarted?
>
> That depends on why brltty needs
[quoted lines by Samuel Thibault on 2007/11/08 at 02:18 +]
>maybe we should try to chdir("/") in order to enter the real
>/ ?
That should indeed resolve the problem. Now where's the best place to do that?
We don't want to do it constantly. Is there a way we can query /, perhaps with
statfs(
Dave Mielke, le Wed 07 Nov 2007 08:13:25 -0500, a écrit :
> the problem is probably that brltty is running under a different
> root than later so it has a different /dev than later.
Indeed, maybe we should try to chdir("/") in order to enter the real
/ ?
> Note that brltty does try to create the
[quoted lines by Mario Lang on 2007/11/07 at 13:46 +0100]
>I guess this is because some device did not exist in my /dev hierarchy
>while brltty was started first. Is this right?
Yes, it would be.
>Can we perhaps fix this such that brltty does not need to be restarted?
That depends on why brltt
Hi.
I am getting the message from the subject if
I launch brltty from the initramfs (very early, before init is launched)
when I switch away from console 1 to something else (like 2, or 7).
If I restart brltty after runlevel 2 started up completely,
I can again switch to consoles other than 1. A
17 matches
Mail list logo
