lkml.org 
[lkml]   [1999]   [Nov]   [8]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    From
    SubjectRe: [PATCH] POSIX timers for 2.3.25
    On Mon, 8 Nov 1999, Matthew Wilcox wrote:

    > On Sun, Nov 07, 1999 at 06:33:43PM +0100, Robert de Vries wrote:
    > > On Sun, 7 Nov 1999, Matthew Wilcox wrote:
    > >
    > > > On Sun, Nov 07, 1999 at 05:59:12PM +0100, Robert de Vries wrote:
    > > > > The siginfo structure is slightly changed. It moves si_value out of the
    > > > > _rt part of the union. si_value is used both by _rt and _timer and must
    > > > > map to the same field.
    > > >
    > > > You can't do this. You've just broken binary compatibility with any
    > > > application that uses POSIX RT signals. I can't see a nice way of doing
    > > > this though. Can anyone else?
    > >
    > > IMHO, NOT doing this now gives more problems that doing it "right" at this
    > > point. I must admit I am not very comfortable with this change either,
    > > but I don't see any other way, which is not kludgy.
    > > I'm open to suggestions. I would really not have two _sigval's in two
    > > structs in the same union. But then again, is it worth breaking a few
    > > apps at this point?
    >
    > If this _really_ needs doing then we need a PER_NEWLINUX, change the
    > kernel internal structures to the way you want, then translate the
    > structures to the appropriate form for old Linux binaries whenever
    > delivering a signal. It'll be unpleasant.

    This _really_ needs doing, because not only RT signals issued by sigqueue
    and the timers need this sigev_value, but also signals sent by completion
    of asynchronous I/O and by the arrival of a message on an empty POSIX
    message queue.

    The first two are the subject of this discussion now.

    The asynchronous I/O at kernel level is curently under development at SGI
    (see http://oss.sgi.com/projects/kaio) and the POSIX mq_* family will need
    it next.

    The SGI kaio patch currently does not copy the sigev_value into the signal
    structure, but that's another matter.

    I don't know if there are already loads of apps using RT signals out
    there. A simple recompilation would fix them anyway.

    Any RT signal users out there? :-)

    Robert


    --
    Robert de Vries
    rhdv@rhdv.cistron.nl


    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.rutgers.edu
    Please read the FAQ at http://www.tux.org/lkml/

    \
     
     \ /
      Last update: 2005-03-22 13:54    [W:0.023 / U:33.916 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site