lkml.org 
[lkml]   [1998]   [Jun]   [8]   [last100]   RSS Feed
Views: [more markup]  [less markup]  [headers]  [forward] 
 
Messages in this thread
/
From(david parsons)
SubjectRe: >256 ptys (previous subject line was garbage)
Date8 Jun 1998 19:03:27 -0700
In article <linux.kernel.199806090036.KAA08187@vindaloo.atnf.csiro.au>,
Richard Gooch  <Richard.Gooch@atnf.CSIRO.AU> wrote:
>tytso@mit.edu writes:
>>    Date: Fri, 5 Jun 1998 12:59:59 +0200 (MET DST)
>>    From: Peter Svensson <petersv@df.lth.se>
>> 
>>    I am not familiar with the reasons for using major/minor-pair for
>>    locking so I don't know if another solution is feasable. Do you have
>>    any pointers for additional reading? :-)
>> 
>> The problem is things like /dev/modem being a symlink (or perhaps even a
>> hard link) to /dev/ttyS0.  So it would be useful to use a lockfile that
>> includes the major and minor device number, in addition to using a
>> lockfile that is based on the device name.  The basic idea is that
>> people want to have different device names to refer to the same device,
>> so we need to lock based the major/minor devices.
>
>Well, I've seen one comment already questioning whether major/minor
>device locks are the better way of doing it, instead of flock(2).
>However, ignoring that, I think it would be simple enough to implement
>a non tty-specific locking scheme in devfs. I already have the auto
>ownership facility.
>What does this device locking need? Just limit the number of open(2)s
>to 1?

    More than that.

    You need a locking scheme that will support a getty listening on the
    port (one open) and will also support things that want to grab the
    port for outgoing traffic (the second open.)

                  ____
    david parsons \bi/ I [heart] mgetty
                   \/
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu

\
 
 \ /
  Last update: 2005-03-22 13:43    [from the cache]
©2003-2008