Date: Tue, 21 Aug 2007 22:21:36 +0200 From: Pawel Jakub Dawidek <pjd@FreeBSD.org> To: John Baldwin <jhb@freebsd.org> Cc: Alfred Perlstein <alfred@freebsd.org>, freebsd-arch@freebsd.org Subject: Re: Lockless uidinfo. Message-ID: <20070821202136.GB4187@garage.freebsd.pl> In-Reply-To: <20070821191902.GA4187@garage.freebsd.pl> References: <20070818120056.GA6498@garage.freebsd.pl> <20070818155041.GY90381@elvis.mu.org> <20070818161449.GE6498@garage.freebsd.pl> <200708211403.29293.jhb@freebsd.org> <20070821191902.GA4187@garage.freebsd.pl>
next in thread | previous in thread | raw e-mail | index | archive | help
--3uo+9/B/ebqu+fSQ
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Tue, Aug 21, 2007 at 09:19:02PM +0200, Pawel Jakub Dawidek wrote:
> > Memory barriers on another CPU don't mean anything about the CPU thread=
2 is=20
> > on. Memory barriers do not flush caches on other CPUs, etc. Normally =
when=20
> > objects are refcounted in a table, the table holds a reference on the o=
bject,=20
> > but that doesn't seem to be the case here. [...]
>=20
> But the memory barrier from 'mtx_lock(&uihashtbl_mtx)' above
> 'if (uip->ui_ref > 0)' would do the trick and I can safely avoid using
> atomic read in this if statement, right?
>=20
> > [...] Have you tried doing something=20
> > very simple in uifree():
> >=20
> > {
> > mtx_lock(&uihashtbl_mtx);
> > if (refcount_release(...)) {
> > LIST_REMOVE();
> > mtx_unlock(&uihashtbl_mtx);
> > ...
> > free();
> > } else
> > mtx_unlock(&uihashtbl_mtx);
> > }
> >=20
> > I wouldn't use a more complex algo in uifree() unless the simple one is=
shown=20
> > to perform badly. Needless complexity is a hindrance to future mainten=
ance.
>=20
> Of coure we could do that, but I was trying really hard to remove
> contention in the common case. Before we used UIDINFO_LOCK() in the
> common case, now you suggesting using global lock here, and I'd really,
> really prefer using one atomic only.
>=20
> > Also, even if you do go with the more complex route, I'd rather you red=
uce=20
> > diffs with the current code by keeping the test as 'uip->ui_ref =3D=3D =
0' and=20
> > keeping the removal code in the if-block.
>=20
> Will do.
>=20
> > In chgproccnt() you should use atomic_fetchadd_long() to avoid a race w=
hen=20
> > reading ui_proccnt.
> >=20
> >=20
> > old =3D atomic_fetchadd_long(&uip->ui_proccnt, diff);
> > if (old + diff < 0)
> > printf("....");
>=20
> I'm aware of this race, but I don't find closing it that much important.
> We won't generate false positive here. My vote is to leave it as it is,
> because atomic_fetchadd_long() is slower on some archs than
> atomic_add_long(), ie. it is implemented using atomic_cmpset_long()
> loop, and as I checked by running 8 processes on 8way machine with
> older code that used atomic_cmpset_long() loop in 'diff > 0' case,
> there is almost one extra loop on every call, which makes it about 6%
> slower.
>=20
> > OTOH, atomic_fetchadd_long() doesn't yet exist, so you will need to fix=
that,=20
> > or just always use an atomic_cmpset() loop.
>=20
> I already implemented those.
New patch is here:
http://people.freebsd.org/~pjd/patches/uidinfo_waitfree.patch
--=20
Pawel Jakub Dawidek http://www.wheel.pl
pjd@FreeBSD.org http://www.FreeBSD.org
FreeBSD committer Am I Evil? Yes, I Am!
--3uo+9/B/ebqu+fSQ
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.4 (FreeBSD)
iD8DBQFGy0lQForvXbEpPzQRAgIlAKDIP5P5TWOBUAHsgLX7lIe4tYsgxwCgvmpD
d/K97EVO24dghgaNS0woxDA=
=rVvA
-----END PGP SIGNATURE-----
--3uo+9/B/ebqu+fSQ--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070821202136.GB4187>