From: Bart Van Assche <bvanassche@acm.org>
To: Alexander Duyck <alexander.duyck@gmail.com>
Cc: linux@rasmusvillemoes.dk,
Andrew Morton <akpm@linux-foundation.org>,
LKML <linux-kernel@vger.kernel.org>,
Vlastimil Babka <vbabka@suse.cz>,
Mel Gorman <mgorman@techsingularity.net>,
Christoph Lameter <cl@linux.com>,
guro@fb.com, Pekka Enberg <penberg@kernel.org>,
David Rientjes <rientjes@google.com>,
Joonsoo Kim <iamjoonsoo.kim@lge.com>,
linux-mm <linux-mm@kvack.org>
Subject: Re: [PATCH] slab.h: Avoid using & for logical and of booleans
Date: Mon, 05 Nov 2018 16:01:06 -0800 [thread overview]
Message-ID: <1541462466.196084.163.camel@acm.org> (raw)
In-Reply-To: <CAKgT0Udci4Ai4OD20NSRuDckE_G4RHma3Bg6H1Um6N9Se_zPew@mail.gmail.com>
On Mon, 2018-11-05 at 14:48 -0800, Alexander Duyck wrote:
> On Mon, Nov 5, 2018 at 2:41 PM Bart Van Assche <bvanassche@acm.org> wrote:
> > How about this version, still untested? My compiler is able to evaluate
> > the switch expression if the argument is constant.
> >
> > static __always_inline enum kmalloc_cache_type kmalloc_type(gfp_t flags)
> > {
> > - int is_dma = 0;
> > - int type_dma = 0;
> > - int is_reclaimable;
> > + unsigned int dr = !!(flags & __GFP_RECLAIMABLE);
> >
> > #ifdef CONFIG_ZONE_DMA
> > - is_dma = !!(flags & __GFP_DMA);
> > - type_dma = is_dma * KMALLOC_DMA;
> > + dr |= !!(flags & __GFP_DMA) << 1;
> > #endif
> >
> > - is_reclaimable = !!(flags & __GFP_RECLAIMABLE);
> > -
> > /*
> > * If an allocation is both __GFP_DMA and __GFP_RECLAIMABLE, return
> > * KMALLOC_DMA and effectively ignore __GFP_RECLAIMABLE
> > */
> > - return type_dma + (is_reclaimable & !is_dma) * KMALLOC_RECLAIM;
> > + switch (dr) {
> > + default:
> > + case 0:
> > + return 0;
> > + case 1:
> > + return KMALLOC_RECLAIM;
> > + case 2:
> > + case 3:
> > + return KMALLOC_DMA;
> > + }
> > }
>
> Doesn't this defeat the whole point of the code which I thought was to
> avoid conditional jumps and branches? Also why would you bother with
> the "dr" value when you could just mask the flags value and switch on
> that directly?
Storing the relevant bits of 'flags' in the 'dr' variable avoids that the
bit selection expressions have to be repeated and allows to use a switch
statement instead of multiple if / else statements.
Most kmalloc() calls pass a constant to the gfp argument. That allows the
compiler to evaluate kmalloc_type() at compile time. So the conditional jumps
and branches only appear when the gfp argument is not a constant. What makes
you think it is important to optimize for that case?
Bart.
next prev parent reply other threads:[~2018-11-06 0:01 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-11-05 20:40 [PATCH] slab.h: Avoid using & for logical and of booleans Bart Van Assche
2018-11-05 21:13 ` Andrew Morton
2018-11-05 21:48 ` Bart Van Assche
2018-11-05 22:14 ` Rasmus Villemoes
2018-11-05 22:40 ` Bart Van Assche
2018-11-05 22:48 ` Alexander Duyck
2018-11-06 0:01 ` Bart Van Assche [this message]
2018-11-06 0:11 ` Alexander Duyck
2018-11-06 0:32 ` Bart Van Assche
2018-11-06 17:20 ` Alexander Duyck
2018-11-06 17:48 ` Bart Van Assche
2018-11-06 18:17 ` Alexander Duyck
2018-11-06 9:45 ` William Kucharski
2018-11-06 8:40 ` Vlastimil Babka
2018-11-06 10:08 ` David Laight
2018-11-06 10:22 ` Vlastimil Babka
2018-11-06 11:07 ` David Laight
2018-11-06 12:51 ` Vlastimil Babka
2018-11-07 10:41 ` David Laight
2018-11-09 8:12 ` Vlastimil Babka
2018-11-09 19:00 ` Andrew Morton
2018-11-09 19:16 ` Vlastimil Babka
2018-11-09 19:47 ` Darryl T. Agostinelli
2018-11-09 21:31 ` Vlastimil Babka
2018-11-12 9:55 ` David Laight
2018-11-13 18:22 ` Vlastimil Babka
2018-11-21 13:22 ` Vlastimil Babka
2018-11-19 11:04 ` Pavel Machek
2018-11-19 12:51 ` Vlastimil Babka
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=1541462466.196084.163.camel@acm.org \
--to=bvanassche@acm.org \
--cc=akpm@linux-foundation.org \
--cc=alexander.duyck@gmail.com \
--cc=cl@linux.com \
--cc=guro@fb.com \
--cc=iamjoonsoo.kim@lge.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux@rasmusvillemoes.dk \
--cc=mgorman@techsingularity.net \
--cc=penberg@kernel.org \
--cc=rientjes@google.com \
--cc=vbabka@suse.cz \
/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;
as well as URLs for NNTP newsgroup(s).