lkml.org 
[lkml]   [2000]   [Mar]   [19]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH] unnecessary blocking interrupts in exit_notify()


On Fri, 17 Mar 2000, Linus Torvalds wrote:

>
>
> On Fri, 17 Mar 2000, Jun Sun wrote:
> >
> > Would you please take this patch? It improves interrupt responsiveness,
> > especially important for embedded systems.
>
> It has a very serious bug.
>
> It can deadlock due to an interrupt doing a read_lock() on the
> tasklist_lock while we hold the write lock.
>
> Linus

Ahh, this is a problem.

The only other fix I can think of is to put blocking interupts and
enabling interrupts inside loop, i.e., interrupts are opened and
closed for removing each child process.

The overhead is obvious. I am not sure if it is worth the benefit
anymore.

---

On the other hand, why would an interrupt routine need to acquire
the read lock on tasklist_lock at first place? For synchronizing
with other interrupt routines in SMP version? Or the same functions
acquiring the read lock need to be called by both interrupt handlers and
regular kernel routines?

Jun


-
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 13:57    [W:0.035 / U:0.280 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site