Re: linux-kernel-digest V1 #2839 (writable /proc/<pid>/cmdli
kwrohrer@ce.mediaone.net
Wed, 25 Nov 1998 15:29:18 -0600 (EST)
And lo, Stephen C. Tweedie saith unto me:
> On Tue, 17 Nov 1998 05:48:43 +0100, Andi Kleen <ak@muc.de> said:
>
> > If the threads share VM (with CLONE_VM) they share all memory,
> > including the environment. argv[] is stored in the environment.
> > Changing Linux to support only partly shared memory in CLONE_VM (with
> > threads private pages like in Windows or OS/2) would be a fairly major
> > redesign of the kernel threads mechanism.
>
> Not to mention that the performance would suck: currently we avoid any
> flushing of the hardware task context when we switch between threads.
> That would not be possible if the threads maintained their own page
> tables.
>
> The best suggestion I've heard was to use the %fs register to maintain
> a user-space pointer to thread-local storage. However, any access to
> that space would be closely reliant on gcc extensions.
If processes *really* need thread-local storage, they can make an array of
pointers before they go thread-hot, then index on pthread_self(). So unless
processes *really* need *protected* thread-local storage, this is a userspace
problem anyway.
Keith
--
"The avalanche has already started; |Linux: http://www.linuxhq.com |"Zooty,
it is too late for the pebbles to |KDE: http://www.kde.org | zoot
vote." Kosh, "Believers", Babylon 5 |Keith: kwrohrer@enteract.com | zoot!"
www.midwinter.com/lurk/lurker.html |http://www.enteract.com/~kwrohrer | --Rebo
-
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/