linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Wei Yang <richard.weiyang@gmail.com>
To: Dave Hansen <dave.hansen@intel.com>
Cc: Wei Yang <richard.weiyang@gmail.com>,
	akpm@linux-foundation.org, mhocko@suse.com, rientjes@google.com,
	linux-mm@kvack.org, kirill.shutemov@linux.intel.com,
	bob.picco@hp.com
Subject: Re: [PATCH 3/3] mm/sparse: use __highest_present_section_nr as the boundary for pfn check
Date: Sat, 25 Aug 2018 02:11:27 +0800	[thread overview]
Message-ID: <20180824181127.GA11055@WeideMacBook-Pro.local> (raw)
In-Reply-To: <4bb956bd-a061-7122-2c2d-7c910a176681@intel.com>

On Thu, Aug 23, 2018 at 05:15:39PM -0700, Dave Hansen wrote:
>On 08/23/2018 06:07 AM, Wei Yang wrote:
>> And it is known, __highest_present_section_nr is a more strict boundary
>> than NR_MEM_SECTIONS.
>
>What is the benefit of this patch?
>

The original idea is simple: with a more strict boundary, it will has less
computation.

>You're adding a more "strict" boundary, but you're also adding a
>potential cacheline miss and removing optimizations that the compiler
>can make with a constant vs. a variable.  Providing absolute bounds
>limits that the compiler knows about can actually be pretty handy for it
>to optimize things

You are right.

I haven't thought about the compiler optimization case.

>
>Do you have *any* analysis to show that this has a benefit?  What does
>it do to text size, for instance?

I don't have more analysis about this. Originally, I thought a strict boundary
will have a benefit. But as you mentioned, it loose the compiler optimization.

BTW, I did some tests and found during a normal boot, all the section number
are within NR_MEM_SECTIONS. I am wondering in which case, this check will be
valid and return 0?

-- 
Wei Yang
Help you, Help me

      reply	other threads:[~2018-08-24 18:11 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-08-23 13:07 [PATCH 0/3] trivial code refine for sparsemem Wei Yang
2018-08-23 13:07 ` [PATCH 1/3] mm/sparse: add likely to mem_section[root] check in sparse_index_init() Wei Yang
2018-08-23 13:13   ` Michal Hocko
2018-08-23 22:57     ` Wei Yang
2018-08-24  7:31       ` Michal Hocko
2018-08-24  0:11   ` Dave Hansen
2018-08-24 15:07     ` Wei Yang
2018-09-03 22:27       ` Wei Yang
2018-09-09  1:38         ` Wei Yang
2018-09-10 20:30           ` Dave Hansen
2018-09-11 15:00             ` Wei Yang
2018-08-23 13:07 ` [PATCH 2/3] mm/sparse: expand the CONFIG_SPARSEMEM_EXTREME range in __nr_to_section() Wei Yang
2018-08-23 13:21   ` Michal Hocko
2018-08-23 23:03     ` Wei Yang
2018-08-24  0:09     ` Dave Hansen
2018-08-24 15:24       ` Wei Yang
2018-08-23 13:07 ` [PATCH 3/3] mm/sparse: use __highest_present_section_nr as the boundary for pfn check Wei Yang
2018-08-23 13:25   ` Michal Hocko
2018-08-23 14:00     ` Oscar Salvador
2018-08-23 19:17       ` Michal Hocko
2018-08-23 20:52         ` Oscar Salvador
2018-08-24  0:15   ` Dave Hansen
2018-08-24 18:11     ` Wei Yang [this message]

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=20180824181127.GA11055@WeideMacBook-Pro.local \
    --to=richard.weiyang@gmail.com \
    --cc=akpm@linux-foundation.org \
    --cc=bob.picco@hp.com \
    --cc=dave.hansen@intel.com \
    --cc=kirill.shutemov@linux.intel.com \
    --cc=linux-mm@kvack.org \
    --cc=mhocko@suse.com \
    --cc=rientjes@google.com \
    /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).