[lkml]   [2002]   [Jul]   [8]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: BKL removal
On Sun, Jul 07, 2002 at 10:58:04PM -0400, Alexander Viro wrote:
> On Mon, 8 Jul 2002, Matthew Wilcox wrote:
> > one struct file per open(), yes. however, fork() shares a struct file,
> > as does unix domain fd passing. so we need protection between different
> > processes. there's some pretty good reasons to want to use a semaphore
> > to protect the struct file (see fasync code.. bleugh).
> ??? What exactly do you want to protect there?

andrea & i discussed this off-list a few days ago... see fs/fcntl.c

case F_SETFL:
err = setfl(fd, filp, arg);

setfl() does:

if ((arg ^ filp->f_flags) & FASYNC) {
if (filp->f_op && filp->f_op->fasync) {
error = filp->f_op->fasync(fd, filp, (arg & FASYNC) != 0);


if (arg & O_DIRECT) {
if (!filp->f_iobuf)
error = alloc_kiovec(1, &filp->f_iobuf);

and finally:

filp->f_flags = (arg & SETFL_MASK) | (filp->f_flags & ~SETFL_MASK);

i pointed out that if alloc_kiovec slept, the BKL provides no protection
against someone else doing a setfl at the same time, so we can get the
wrong number of fasync events sent. Marcus Alanen pointed out that
fasync can also sleep, so we're at risk anyway. i don't think that
abusing i_sem as andrea did is the Right Thing to do...

Revolutions do not require corporate support.
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
More majordomo info at
Please read the FAQ at

 \ /
  Last update: 2005-03-22 13:27    [W:0.119 / U:0.588 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site