On Fri, Feb 08, 2008 at 02:56:09PM -0800, Arjan van de Ven wrote:
> Nick Piggin wrote:
> >>>Maybe cpus these days have so much store bandwith that doing
> >>>things like the above is OK, but I doubt it :-)
> >>on modern x86 cpus the memset may even be faster if the memory isn't in
> >>cache;
>
Nick Piggin wrote:
Maybe cpus these days have so much store bandwith that doing
things like the above is OK, but I doubt it :-)
on modern x86 cpus the memset may even be faster if the memory isn't in
cache;
the "explicit" method ends up doing Write Allocate on the cache lines
(so read them
On Fri, Feb 08, 2008 at 07:09:07AM -0800, Arjan van de Ven wrote:
> David Miller wrote:
> >From: Linus Torvalds <[EMAIL PROTECTED]>
> >Date: Thu, 7 Feb 2008 09:42:56 -0800 (PST)
> >
> >>Can we please just stop doing these one-by-one assignments, and just do
> >>something like
> >>
> >>
David Miller wrote:
From: Linus Torvalds <[EMAIL PROTECTED]>
Date: Thu, 7 Feb 2008 09:42:56 -0800 (PST)
Can we please just stop doing these one-by-one assignments, and just do
something like
memset(rq, 0, sizeof(*rq));
rq->q = q;
rq->ref_count = 1;
On Thu, Feb 07 2008, Linus Torvalds wrote:
>
>
> On Thu, 7 Feb 2008, Ingo Molnar wrote:
> > INIT_HLIST_NODE(>hash);
> > RB_CLEAR_NODE(>rb_node);
> > - rq->ioprio = 0;
> > - rq->buffer = NULL;
> > - rq->ref_count = 1;
> > - rq->q = q;
> > - rq->special = NULL;
> > -
On Fri, Feb 08, 2008 at 02:56:09PM -0800, Arjan van de Ven wrote:
Nick Piggin wrote:
Maybe cpus these days have so much store bandwith that doing
things like the above is OK, but I doubt it :-)
on modern x86 cpus the memset may even be faster if the memory isn't in
cache;
the explicit
On Thu, Feb 07 2008, Linus Torvalds wrote:
On Thu, 7 Feb 2008, Ingo Molnar wrote:
INIT_HLIST_NODE(rq-hash);
RB_CLEAR_NODE(rq-rb_node);
- rq-ioprio = 0;
- rq-buffer = NULL;
- rq-ref_count = 1;
- rq-q = q;
- rq-special = NULL;
- rq-data_len = 0;
- rq-data
David Miller wrote:
From: Linus Torvalds [EMAIL PROTECTED]
Date: Thu, 7 Feb 2008 09:42:56 -0800 (PST)
Can we please just stop doing these one-by-one assignments, and just do
something like
memset(rq, 0, sizeof(*rq));
rq-q = q;
rq-ref_count = 1;
On Fri, Feb 08, 2008 at 07:09:07AM -0800, Arjan van de Ven wrote:
David Miller wrote:
From: Linus Torvalds [EMAIL PROTECTED]
Date: Thu, 7 Feb 2008 09:42:56 -0800 (PST)
Can we please just stop doing these one-by-one assignments, and just do
something like
memset(rq, 0, sizeof(*rq));
Nick Piggin wrote:
Maybe cpus these days have so much store bandwith that doing
things like the above is OK, but I doubt it :-)
on modern x86 cpus the memset may even be faster if the memory isn't in
cache;
the explicit method ends up doing Write Allocate on the cache lines
(so read them from
On Thu, 7 Feb 2008, David Miller wrote:
>
> Maybe cpus these days have so much store bandwith that doing
> things like the above is OK, but I doubt it :-)
I seriously doubt the same is true for the IO requests (which are
different anyway, and tend to happen at a much lower frequency than
From: Linus Torvalds <[EMAIL PROTECTED]>
Date: Thu, 7 Feb 2008 09:42:56 -0800 (PST)
> Can we please just stop doing these one-by-one assignments, and just do
> something like
>
> memset(rq, 0, sizeof(*rq));
> rq->q = q;
> rq->ref_count = 1;
> INIT_HLIST_NODE(>hash);
>
On Thu, Feb 07 2008, Ingo Molnar wrote:
>
> * Linus Torvalds <[EMAIL PROTECTED]> wrote:
>
> > On Thu, 7 Feb 2008, Ingo Molnar wrote:
> > > INIT_HLIST_NODE(>hash);
> > > RB_CLEAR_NODE(>rb_node);
> > > - rq->ioprio = 0;
> > > - rq->buffer = NULL;
> > > - rq->ref_count = 1;
> > > - rq->q = q;
>
* Linus Torvalds <[EMAIL PROTECTED]> wrote:
> On Thu, 7 Feb 2008, Ingo Molnar wrote:
> > INIT_HLIST_NODE(>hash);
> > RB_CLEAR_NODE(>rb_node);
> > - rq->ioprio = 0;
> > - rq->buffer = NULL;
> > - rq->ref_count = 1;
> > - rq->q = q;
> > - rq->special = NULL;
> > - rq->data_len
On Thu, Feb 07 2008, Linus Torvalds wrote:
>
>
> On Thu, 7 Feb 2008, Ingo Molnar wrote:
> > INIT_HLIST_NODE(>hash);
> > RB_CLEAR_NODE(>rb_node);
> > - rq->ioprio = 0;
> > - rq->buffer = NULL;
> > - rq->ref_count = 1;
> > - rq->q = q;
> > - rq->special = NULL;
> > -
On Thu, 7 Feb 2008, Ingo Molnar wrote:
> INIT_HLIST_NODE(>hash);
> RB_CLEAR_NODE(>rb_node);
> - rq->ioprio = 0;
> - rq->buffer = NULL;
> - rq->ref_count = 1;
> - rq->q = q;
> - rq->special = NULL;
> - rq->data_len = 0;
> - rq->data = NULL;
> -
* Jens Axboe <[EMAIL PROTECTED]> wrote:
> [...] but may not post anything until after my vacation.
oh, you going on a vacation. I am sitting on a few block layer patches
you might be interested in :-)
i am playing with Vegard Nossum's kmemcheck on x86 (with much help from
Pekka Enberg for
On Thu, Feb 07 2008, Linus Torvalds wrote:
On Thu, 7 Feb 2008, Ingo Molnar wrote:
INIT_HLIST_NODE(rq-hash);
RB_CLEAR_NODE(rq-rb_node);
- rq-ioprio = 0;
- rq-buffer = NULL;
- rq-ref_count = 1;
- rq-q = q;
- rq-special = NULL;
- rq-data_len = 0;
- rq-data
On Thu, Feb 07 2008, Ingo Molnar wrote:
* Linus Torvalds [EMAIL PROTECTED] wrote:
On Thu, 7 Feb 2008, Ingo Molnar wrote:
INIT_HLIST_NODE(rq-hash);
RB_CLEAR_NODE(rq-rb_node);
- rq-ioprio = 0;
- rq-buffer = NULL;
- rq-ref_count = 1;
- rq-q = q;
- rq-special = NULL;
* Linus Torvalds [EMAIL PROTECTED] wrote:
On Thu, 7 Feb 2008, Ingo Molnar wrote:
INIT_HLIST_NODE(rq-hash);
RB_CLEAR_NODE(rq-rb_node);
- rq-ioprio = 0;
- rq-buffer = NULL;
- rq-ref_count = 1;
- rq-q = q;
- rq-special = NULL;
- rq-data_len = 0;
- rq-data =
From: Linus Torvalds [EMAIL PROTECTED]
Date: Thu, 7 Feb 2008 09:42:56 -0800 (PST)
Can we please just stop doing these one-by-one assignments, and just do
something like
memset(rq, 0, sizeof(*rq));
rq-q = q;
rq-ref_count = 1;
INIT_HLIST_NODE(rq-hash);
On Thu, 7 Feb 2008, David Miller wrote:
Maybe cpus these days have so much store bandwith that doing
things like the above is OK, but I doubt it :-)
I seriously doubt the same is true for the IO requests (which are
different anyway, and tend to happen at a much lower frequency than
22 matches
Mail list logo
