From: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
To: Christoph Lameter <cl@linux-foundation.org>
Cc: Tejun Heo <tj@kernel.org>,
linux-kernel@vger.kernel.org, Mel Gorman <mel@csn.ul.ie>,
Pekka Enberg <penberg@cs.helsinki.fi>
Subject: Re: [this_cpu_xx V8 11/16] Generic support for this_cpu_cmpxchg
Date: Tue, 22 Dec 2009 12:24:36 -0500 [thread overview]
Message-ID: <20091222172436.GB12529@Krystal> (raw)
In-Reply-To: <alpine.DEB.2.00.0912220950030.12163@router.home>
* Christoph Lameter (cl@linux-foundation.org) wrote:
> On Sat, 19 Dec 2009, Mathieu Desnoyers wrote:
>
> > I am a bit concerned about the "generic" version of this_cpu_cmpxchg.
> > Given that what LTTng needs is basically an atomic, nmi-safe version of
> > the primitive (on all architectures that have something close to a NMI),
> > this means that it could not switch over to your primitives until we add
> > the equivalent support we currently have with local_t to all
> > architectures. The transition would be faster if we create an
> > atomic_cpu_*() variant which would map to local_t operations in the
> > initial version.
> >
> > Or maybe have I missed something in your patchset that address this ?
>
> NMI safeness is not covered by this_cpu operations.
>
> We could add nmi_safe_.... ops?
>
> The atomic_cpu reference make me think that you want full (LOCK)
> semantics? Then use the regular atomic ops?
nmi_safe would probably make sense here.
But given that we have to disable preemption to add precision in terms
of trace clock timestamp, I wonder if we would really gain something
considerable performance-wise.
I also thought about the design change this requires for the per-cpu
buffer commit count pointer which would have to become a per-cpu pointer
independent of the buffer structure, and I foresee a problem with
Steven's irq off tracing which need to perform buffer exchanges while
tracing is active. Basically, having only one top-level pointer for the
buffer makes it possible to exchange it atomically, but if we have to
have two separate pointers (one for per-cpu buffer, one for per-cpu
commit count array), then we are stucked.
So given that per-cpu ops limits us in terms of data structure layout, I
am less and less sure it's the best fit for ring buffers, especially if
we don't gain much performance-wise.
Thanks,
Mathieu
--
Mathieu Desnoyers
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F BA06 3F25 A8FE 3BAE 9A68
next prev parent reply other threads:[~2009-12-22 17:24 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-12-18 22:26 [this_cpu_xx V8 00/16] Per cpu atomics in core allocators, cleanup and more this_cpu_ops Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 01/16] this_cpu_ops: page allocator conversion Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 02/16] this_cpu ops: Remove pageset_notifier Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 03/16] Use this_cpu operations in slub Christoph Lameter
2009-12-20 9:11 ` Pekka Enberg
2009-12-18 22:26 ` [this_cpu_xx V8 04/16] SLUB: Get rid of dynamic DMA kmalloc cache allocation Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 05/16] this_cpu: Remove slub kmem_cache fields Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 06/16] Make slub statistics use this_cpu_inc Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 07/16] Module handling: Use this_cpu_xx to dynamically allocate counters Christoph Lameter
2009-12-21 7:47 ` Tejun Heo
2009-12-21 7:59 ` Tejun Heo
2009-12-21 8:19 ` Tejun Heo
2009-12-21 23:28 ` Rusty Russell
2009-12-22 0:02 ` Tejun Heo
2009-12-22 16:17 ` Christoph Lameter
2009-12-22 15:58 ` Christoph Lameter
2009-12-22 15:57 ` Christoph Lameter
2009-12-23 2:07 ` Tejun Heo
2010-01-04 17:22 ` Christoph Lameter
2010-01-04 17:51 ` Mathieu Desnoyers
2009-12-18 22:26 ` [this_cpu_xx V8 08/16] Remove cpu_local_xx macros Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 09/16] Allow arch to provide inc/dec functionality for each size separately Christoph Lameter
2009-12-21 7:25 ` Tejun Heo
2009-12-22 15:56 ` Christoph Lameter
2009-12-23 2:08 ` Tejun Heo
2009-12-18 22:26 ` [this_cpu_xx V8 10/16] Support generating inc/dec for this_cpu_inc/dec Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 11/16] Generic support for this_cpu_cmpxchg Christoph Lameter
2009-12-19 14:45 ` Mathieu Desnoyers
2009-12-22 15:54 ` Christoph Lameter
2009-12-22 17:24 ` Mathieu Desnoyers [this message]
2010-01-04 17:21 ` Christoph Lameter
2010-01-05 22:29 ` Mathieu Desnoyers
2010-01-05 22:35 ` Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 12/16] Add percpu cmpxchg operations Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 13/16] Generic support for this_cpu_xchg Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 14/16] x86 percpu xchg operation Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 15/16] Generic support for this_cpu_add_return() Christoph Lameter
2009-12-18 22:26 ` [this_cpu_xx V8 16/16] x86 support for this_cpu_add_return Christoph Lameter
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20091222172436.GB12529@Krystal \
--to=mathieu.desnoyers@polymtl.ca \
--cc=cl@linux-foundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mel@csn.ul.ie \
--cc=penberg@cs.helsinki.fi \
--cc=tj@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®