| To: | Sam Ravnborg <sam@xxxxxxxxxxxx> |
|---|---|
| Subject: | Re: RFC: Fix f_flags races without the BKL |
| From: | Jonathan Corbet <corbet@xxxxxxx> |
| Date: | Tue, 30 Dec 2008 05:49:46 -0700 |
| Cc: | LKML <linux-kernel@xxxxxxxxxxxxxxx>, Andi Kleen <andi@xxxxxxxxxxxxxx>, Alan Cox <alan@xxxxxxxxxxxxxxxxxxx>, Al Viro <viro@xxxxxxxxxxxxxxxxxx>, Oleg Nesterov <oleg@xxxxxxxxxx>, bfields@xxxxxxxxxxxx, xfs-masters@xxxxxxxxxxx |
| In-reply-to: | <20081229115706.GA955@uranus.ravnborg.org> |
| Organization: | LWN.net |
| References: | <20081229041352.6bbdf57c@tpl> <20081229115706.GA955@uranus.ravnborg.org> |
On Mon, 29 Dec 2008 12:57:06 +0100 Sam Ravnborg <sam@xxxxxxxxxxxx> wrote: > Rather than open coded mutex how about adding a few helpers to > set and clear the flags and hide locking there? There's a couple of problems with that. One being that SETFL wants to manipulate a bunch of flags together, so a simple set_flag/clear_flag interface won't do it. Beyond that, though, calls to the ->fasync() function need to be atomic with respect to changes to the associated flag. Still, it seems that the global lock approach isn't too popular, so I'll get back to the drawing board once I'm theoretically not on vacation. Thanks, jon |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | [Bug 715] touchup AC_DEFINE(HAVE_VISIBILITY_ATTRIBUTE) in visibility_hidden.m4, bugzilla-daemon |
|---|---|
| Next by Date: | Re: RFC: Fix f_flags races without the BKL, Jonathan Corbet |
| Previous by Thread: | Re: RFC: Fix f_flags races without the BKL, Sam Ravnborg |
| Next by Thread: | Re: RFC: Fix f_flags races without the BKL, Oleg Nesterov |
| Indexes: | [Date] [Thread] [Top] [All Lists] |