lkml.org 
[lkml]   [2013]   [Feb]   [7]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    SubjectRe: [PATCH v3 07/23] tty: Strengthen no-subsequent-use guarantee of tty_ldisc_halt()
    From
    Date
    On Thu, 2013-02-07 at 16:38 +0100, Jiri Slaby wrote:
    > On 02/05/2013 09:20 PM, Peter Hurley wrote:
    > > @@ -688,9 +702,9 @@ int tty_set_ldisc(struct tty_struct *tty, int ldisc)
    > > * parallel to the change and re-referencing the tty.
    > > */
    > >
    > > - work = tty_ldisc_halt(tty);
    > > + retval = tty_ldisc_halt(tty, &work, 5 * HZ);
    > > if (o_tty)
    > > - o_work = tty_ldisc_halt(o_tty);
    > > + tty_ldisc_halt(o_tty, &o_work, 0);
    >
    > Zero in there will cause the other end's work not to be cancelled and
    > TTY_LDISC_HALTED unset if there are two or more readers on the slave,
    > right? So there should be something like:
    > if (timeout)
    > retval = tty_ldisc_wait_idle(tty, timeout);
    > ...
    >
    > in tty_ldisc_halt.

    True.

    Fixed in 08/23, where both ldiscs are halted simultaneously, forced to
    wait for identical timeouts, and returns an error if either tty timed
    out. From 08/23

    in tty_ldisc_halt():

    retval = tty_ldisc_wait_idle(tty, timeout);
    + if (!retval && o_tty)
    + retval = tty_ldisc_wait_idle(o_tty, timeout);
    if (retval)
    return retval;

    in tty_set_ldisc():

    - retval = tty_ldisc_halt(tty, &work, 5 * HZ);
    - if (o_tty)
    - tty_ldisc_halt(o_tty, &o_work, 0);
    + retval = tty_ldisc_halt(tty, o_tty, &work, &o_work, 5 * HZ);


    If you'd prefer, I could put the 5 sec. wait in 7/23 as well, test the
    error and so forth.



    \
     
     \ /
      Last update: 2013-02-07 18:02    [W:4.134 / U:0.152 seconds]
    ©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site