Re: [PATCH] f_op->poll() without lock_kernel()

From: David S. Miller (davem@redhat.com)
Date: Fri Apr 21 2000 - 22:52:31 EDT

  • Next message: Andrew Morton: "Re: [RFC] Flat /lib/modules/`uname -r`"

       Date: Fri, 21 Apr 2000 22:23:26 -0400 (EDT)
       From: Alexander Viro <viro@math.psu.edu>

    > Please no, the whole point of the sock_no_*() stuff was to _REMOVE_
    > the tests from the socket.c code.

       OK, what about the following:

               a) ->fcntl() is _always_ sock_no_fcntl().

    I think this can safely die. It is true even in 2.2.x

               b) ->socketpair() is nontrivial only for AF_UNIX.

    Some other protocol someday may want to use it too.
    It has to stay.

               c) ->mmap() is nontrivial only for AF_PACKET and I'm not
                  too happy about its abuse there.

    Some day it may be fun to mmap tcp sockets. It has to stay.

       [patch being tested right now; the question being whether we want to keep
       ->fcntl() or not]

    I think we don't want to keep it.

    Later,
    David S. Miller
    davem@redhat.com

    -
    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/

    This archive was generated by hypermail 2b29 : Fri Apr 21 2000 - 23:05:36 EDT