From: Nathan Zimmer <nzimmer@sgi.com>
To: Ingo Molnar <mingo@kernel.org>
Cc: Nathan Zimmer <nzimmer@sgi.com>,
holt@sgi.com, travis@sgi.com, rob@landley.net,
tglx@linutronix.de, mingo@redhat.com, hpa@zytor.com,
yinghai@kernel.org, akpm@linux-foundation.org,
gregkh@linuxfoundation.org, x86@kernel.org,
linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org,
Linus Torvalds <torvalds@linux-foundation.org>,
Peter Zijlstra <a.p.zijlstra@chello.nl>
Subject: Re: [RFC 2/2] x86_64, mm: Reinsert the absent memory
Date: Mon, 24 Jun 2013 15:36:57 -0500 [thread overview]
Message-ID: <20130624203657.GA107621@asylum.americas.sgi.com> (raw)
In-Reply-To: <20130623092840.GB13445@gmail.com>
On Sun, Jun 23, 2013 at 11:28:40AM +0200, Ingo Molnar wrote:
>
> That's 4.5 GB/sec initialization speed - that feels a bit slow and the
> boot time effect should be felt on smaller 'a couple of gigabytes' desktop
> boxes as well. Do we know exactly where the 2 hours of boot time on a 32
> TB system is spent?
>
There are other several spots that could be improved on a large system but
memory initialization is by far the biggest.
> While you cannot profile the boot process (yet), you could try your
> delayed patch and run a "perf record -g" call-graph profiling of the
> late-time initialization routines. What does 'perf report' show?
>
I have some data from earlier runs.
memmap_init_zone was the function that was the biggest hitter by far.
Parts of it could certianly are low hanging fruit, set_pageblock_migratetype
for example.
However it seems for a larger system SetPageReserved will be the largest
consumer of cycles. On a 1TB system I just booted it was around 50% of time
spent in memmap_init_zone.
perf seems to struggle with 512 cpus, but I did get some data.
It seems to indicate similar data to what I found in earlier experiments.
Lots of time in memmap_init_zone,
Some are waiting on locks, this guy seems to be representative of that.
- 0.14% kworker/160:1 [kernel.kallsyms] [k] mspin_lock ▒
+ mspin_lock ▒
+ __mutex_lock_slowpath ▒
- mutex_lock ▒
- 99.69% online_pages
> Delayed initialization makes sense I guess because 32 TB is a lot of
> memory - I'm just wondering whether there's some low hanging fruits left
> in the mem init code, that code is certainly not optimized for
> performance.
>
> Plus with a struct page size of around 64 bytes (?) 32 TB of RAM has 512
> GB of struct page arrays alone. Initializing those will take quite some
> time as well - and I suspect they are allocated via zeroing them first. If
> that memset() exists then getting rid of it might be a good move as well.
>
> Yet another thing to consider would be to implement an initialization
> speedup of 3 orders of magnitude: initialize on the large page (2MB)
> grandularity and on-demand delay the initialization of the 4K granular
> struct pages [but still allocating them] - which I suspect are a good
> chunk of the overhead? That way we could initialize in 2MB steps and speed
> up the 2 hours bootup of 32 TB of RAM to 14 seconds...
>
> [ The cost would be one more branch in the buddy allocator, to detect
> not-yet-initialized 2 MB chunks as we encounter them. Acceptable I
> think. ]
>
> Thanks,
>
> Ingo
next prev parent reply other threads:[~2013-06-24 20:37 UTC|newest]
Thread overview: 66+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-06-21 16:25 [RFC 0/2] Delay initializing of large sections of memory Nathan Zimmer
2013-06-21 16:25 ` [RFC 1/2] x86_64, mm: Delay initializing large portion " Nathan Zimmer
2013-06-25 4:14 ` Rob Landley
2013-06-21 16:25 ` [RFC 2/2] x86_64, mm: Reinsert the absent memory Nathan Zimmer
2013-06-23 9:28 ` Ingo Molnar
2013-06-23 9:32 ` Ingo Molnar
2013-06-24 17:38 ` H. Peter Anvin
2013-06-24 19:39 ` Ingo Molnar
2013-06-24 20:08 ` H. Peter Anvin
2013-06-25 7:31 ` Ingo Molnar
2013-06-24 20:36 ` Nathan Zimmer [this message]
2013-06-25 7:38 ` Ingo Molnar
2013-06-25 15:07 ` H. Peter Anvin
2013-06-25 17:19 ` Mike Travis
2013-06-25 17:22 ` Mike Travis
2013-06-25 18:43 ` H. Peter Anvin
2013-06-25 18:51 ` Mike Travis
2013-06-26 9:22 ` [RFC] Transparent on-demand memory setup initialization embedded in the (GFP) buddy allocator Ingo Molnar
2013-06-26 13:28 ` Andrew Morton
2013-06-26 13:37 ` Ingo Molnar
2013-06-26 15:02 ` Nathan Zimmer
2013-06-26 16:15 ` Mike Travis
2013-06-26 12:14 ` [RFC 2/2] x86_64, mm: Reinsert the absent memory Ingo Molnar
2013-06-26 14:49 ` Nathan Zimmer
2013-06-26 15:12 ` Dave Hansen
2013-06-26 15:20 ` Nathan Zimmer
2013-06-26 15:58 ` Ingo Molnar
2013-06-26 16:11 ` Nathan Zimmer
2013-06-26 16:07 ` Mike Travis
2013-06-21 16:51 ` [RFC 0/2] Delay initializing of large sections of memory Greg KH
2013-06-21 17:03 ` H. Peter Anvin
2013-06-21 17:18 ` Nathan Zimmer
2013-06-21 17:28 ` H. Peter Anvin
2013-06-21 20:05 ` Nathan Zimmer
2013-06-21 20:08 ` H. Peter Anvin
2013-06-21 20:33 ` Nathan Zimmer
2013-06-21 21:36 ` Mike Travis
2013-06-21 21:07 ` Mike Travis
2013-06-21 18:44 ` Yinghai Lu
2013-06-21 18:50 ` Greg KH
2013-06-21 19:10 ` Yinghai Lu
2013-06-21 19:19 ` Nathan Zimmer
2013-06-21 20:28 ` Yinghai Lu
2013-06-21 20:40 ` Nathan Zimmer
2013-06-21 21:30 ` Mike Travis
2013-06-22 0:23 ` Yinghai Lu
2013-06-25 17:35 ` Mike Travis
2013-06-25 18:17 ` H. Peter Anvin
2013-06-25 18:40 ` Mike Travis
2013-06-25 18:40 ` Yinghai Lu
2013-06-25 18:44 ` H. Peter Anvin
2013-06-25 18:58 ` Mike Travis
2013-06-25 19:03 ` Yinghai Lu
2013-06-25 19:09 ` H. Peter Anvin
2013-06-25 19:28 ` Yinghai Lu
2013-06-27 6:37 ` Yinghai Lu
2013-06-27 11:05 ` Robin Holt
2013-06-27 15:50 ` Mike Travis
2013-06-26 9:23 ` Ingo Molnar
2013-06-25 18:38 ` Yinghai Lu
2013-06-25 18:42 ` Mike Travis
2013-06-21 18:36 ` Yinghai Lu
2013-06-21 18:44 ` Greg Kroah-Hartman
2013-06-21 19:00 ` Yinghai Lu
2013-06-21 21:28 ` Mike Travis
2013-06-21 21:19 ` Mike Travis
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=20130624203657.GA107621@asylum.americas.sgi.com \
--to=nzimmer@sgi.com \
--cc=a.p.zijlstra@chello.nl \
--cc=akpm@linux-foundation.org \
--cc=gregkh@linuxfoundation.org \
--cc=holt@sgi.com \
--cc=hpa@zytor.com \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=mingo@redhat.com \
--cc=rob@landley.net \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
--cc=travis@sgi.com \
--cc=x86@kernel.org \
--cc=yinghai@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;
as well as URLs for NNTP newsgroup(s).