Re: [PATCH 2/3 v2] futex: avoid double wake up in futex_wake() on -RT

From: Davidlohr Bueso
Date: Thu Apr 16 2015 - 12:23:08 EST


On Thu, 2015-04-16 at 17:54 +0200, Peter Zijlstra wrote:
> On Thu, Apr 16, 2015 at 07:42:55AM -0700, Davidlohr Bueso wrote:
> > On Thu, 2015-04-16 at 12:16 +0200, Peter Zijlstra wrote:
> > > But if we decide we want to go do this, I'd propose we reintroduce this
> > > delayed wake list thing again.
> >
> > Given that futexes aren't the only potential users, I definitely agree.
> > Lemme cleanup the patches and I'll resend. Now, one thing I wonder about
> > is if we should bother making it a delayed list a plist instead -- as
> > not all users would consider rt-tasks like futexes do.
>
> plist don't work and should not be used for tasks anymore. I suppose I
> should go rip them out of futexes too. If you want to make the thing
> priority aware we should probably abstract the rb-tree from rtmutex.c

Hmm yeah I noticed that, but a tree for this thing seems like an
overkill imho. I mean, at least wrt futexes, I don't think there would
ever be a nr_wake - within a single futex call as we use the list on the
stack - value large enough to even remotely justify the data structure.

Oh and that reminds me, we should update the rtmutex docs, it still
mentions plist -- but replacing those nice ascii art lists with trees is
evil ;)

Thanks,
Davidlohr

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/