From: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
To: Lokesh Gidra <lokeshgidra@google.com>
Cc: Joel Fernandes <joelaf@google.com>,
Mark Rutland <mark.rutland@arm.com>,
Gavin Shan <gshan@redhat.com>, Brian Geffon <bgeffon@google.com>,
Peter Zijlstra <peterz@infradead.org>,
Catalin Marinas <catalin.marinas@arm.com>,
Kalesh Singh <kaleshsingh@google.com>,
Ram Pai <linuxram@us.ibm.com>,
Kamalesh Babulal <kamalesh@linux.vnet.ibm.com>,
"open list:MEMORY MANAGEMENT" <linux-mm@kvack.org>,
William Kucharski <william.kucharski@oracle.com>,
Sandipan Das <sandipan@linux.ibm.com>,
"open list:KERNEL SELFTEST FRAMEWORK"
<linux-kselftest@vger.kernel.org>,
"H. Peter Anvin" <hpa@zytor.com>,
Christian Brauner <christian.brauner@ubuntu.com>,
Shuah Khan <shuah@kernel.org>,
Mina Almasry <almasrymina@google.com>, Jia He <justin.he@arm.com>,
Arnd Bergmann <arnd@arndb.de>, Will Deacon <will@kernel.org>,
Masahiro Yamada <masahiroy@kernel.org>,
the arch/x86 maintainers <x86@kernel.org>,
Krzysztof Kozlowski <krzk@kernel.org>,
Ingo Molnar <mingo@redhat.com>,
Sami Tolvanen <samitolvanen@google.com>,
"Cc: Android Kernel" <kernel-team@android.com>,
Hassan Naveed <hnaveed@wavecomp.com>,
Ralph Campbell <rcampbell@nvidia.com>,
Kees Cook <keescook@chromium.org>,
Minchan Kim <minchan@google.com>,
Zhenyu Ye <yezhenyu2@huawei.com>,
John Hubbard <jhubbard@nvidia.com>,
Frederic Weisbecker <frederic@kernel.org>,
Borislav Petkov <bp@alien8.de>,
Thomas Gleixner <tglx@linutronix.de>,
Suren Baghdasaryan <surenb@google.com>,
"moderated list:ARM64 PORT \(AARCH64 ARCHITECTURE\)"
<linux-arm-kernel@lists.infradead.org>,
SeongJae Park <sjpark@amazon.de>,
Dave Hansen <dave.hansen@intel.com>,
linux-kernel <linux-kernel@vger.kernel.org>,
Masami Hiramatsu <mhiramat@kernel.org>,
"Aneesh Kumar K.V" <aneesh.kumar@linux.ibm.com>,
Andrew Morton <akpm@linux-foundation.org>,
Mike Rapoport <rppt@kernel.org>
Subject: Re: [PATCH 0/5] Speed up mremap on large regions
Date: Fri, 2 Oct 2020 08:35:47 +0300 [thread overview]
Message-ID: <20201002053547.7roe7b4mpamw4uk2@black.fi.intel.com> (raw)
In-Reply-To: <CA+EESO6Wyeht1GrboyZ5vW4E-DvjbWn=uOQM3ugAG=bRKMuigw@mail.gmail.com>
On Thu, Oct 01, 2020 at 05:09:02PM -0700, Lokesh Gidra wrote:
> On Thu, Oct 1, 2020 at 9:00 AM Kalesh Singh <kaleshsingh@google.com> wrote:
> >
> > On Thu, Oct 1, 2020 at 8:27 AM Kirill A. Shutemov
> > <kirill.shutemov@linux.intel.com> wrote:
> > >
> > > On Wed, Sep 30, 2020 at 03:42:17PM -0700, Lokesh Gidra wrote:
> > > > On Wed, Sep 30, 2020 at 3:32 PM Kirill A. Shutemov
> > > > <kirill.shutemov@linux.intel.com> wrote:
> > > > >
> > > > > On Wed, Sep 30, 2020 at 10:21:17PM +0000, Kalesh Singh wrote:
> > > > > > mremap time can be optimized by moving entries at the PMD/PUD level if
> > > > > > the source and destination addresses are PMD/PUD-aligned and
> > > > > > PMD/PUD-sized. Enable moving at the PMD and PUD levels on arm64 and
> > > > > > x86. Other architectures where this type of move is supported and known to
> > > > > > be safe can also opt-in to these optimizations by enabling HAVE_MOVE_PMD
> > > > > > and HAVE_MOVE_PUD.
> > > > > >
> > > > > > Observed Performance Improvements for remapping a PUD-aligned 1GB-sized
> > > > > > region on x86 and arm64:
> > > > > >
> > > > > > - HAVE_MOVE_PMD is already enabled on x86 : N/A
> > > > > > - Enabling HAVE_MOVE_PUD on x86 : ~13x speed up
> > > > > >
> > > > > > - Enabling HAVE_MOVE_PMD on arm64 : ~ 8x speed up
> > > > > > - Enabling HAVE_MOVE_PUD on arm64 : ~19x speed up
> > > > > >
> > > > > > Altogether, HAVE_MOVE_PMD and HAVE_MOVE_PUD
> > > > > > give a total of ~150x speed up on arm64.
> > > > >
> > > > > Is there a *real* workload that benefit from HAVE_MOVE_PUD?
> > > > >
> > > > We have a Java garbage collector under development which requires
> > > > moving physical pages of multi-gigabyte heap using mremap. During this
> > > > move, the application threads have to be paused for correctness. It is
> > > > critical to keep this pause as short as possible to avoid jitters
> > > > during user interaction. This is where HAVE_MOVE_PUD will greatly
> > > > help.
> > >
> > > Any chance to quantify the effect of mremap() with and without
> > > HAVE_MOVE_PUD?
> > >
> > > I doubt it's a major contributor to the GC pause. I expect you need to
> > > move tens of gigs to get sizable effect. And if your GC routinely moves
> > > tens of gigs, maybe problem somewhere else?
> > >
> > > I'm asking for numbers, because increase in complexity comes with cost.
> > > If it doesn't provide an substantial benefit to a real workload
> > > maintaining the code forever doesn't make sense.
> >
> mremap is indeed the biggest contributor to the GC pause. It has to
> take place in what is typically known as a 'stop-the-world' pause,
> wherein all application threads are paused. During this pause the GC
> thread flips the GC roots (threads' stacks, globals etc.), and then
> resumes threads along with concurrent compaction of the heap.This
> GC-root flip differs depending on which compaction algorithm is being
> used.
>
> In our case it involves updating object references in threads' stacks
> and remapping java heap to a different location. The threads' stacks
> can be handled in parallel with the mremap. Therefore, the dominant
> factor is indeed the cost of mremap. From patches 2 and 4, it is clear
> that remapping 1GB without this optimization will take ~9ms on arm64.
>
> Although this mremap has to happen only once every GC cycle, and the
> typical size is also not going to be more than a GB or 2, pausing
> application threads for ~9ms is guaranteed to cause jitters. OTOH,
> with this optimization, mremap is reduced to ~60us, which is a totally
> acceptable pause time.
>
> Unfortunately, implementation of the new GC algorithm hasn't yet
> reached the point where I can quantify the effect of this
> optimization. But I can confirm that without this optimization the new
> GC will not be approved.
IIUC, the 9ms -> 90us improvement attributed to combination HAVE_MOVE_PMD
and HAVE_MOVE_PUD, right? I expect HAVE_MOVE_PMD to be reasonable for some
workloads, but marginal benefit of HAVE_MOVE_PUD is in doubt. Do you see
it's useful for your workload?
--
Kirill A. Shutemov
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
next prev parent reply other threads:[~2020-10-02 5:37 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-09-30 22:21 [PATCH 0/5] Speed up mremap on large regions Kalesh Singh
2020-09-30 22:21 ` [PATCH 1/5] kselftests: vm: Add mremap tests Kalesh Singh
2020-10-01 7:24 ` John Hubbard
2020-10-01 15:46 ` Kalesh Singh
2020-10-01 18:36 ` John Hubbard
2020-09-30 22:21 ` [PATCH 2/5] arm64: mremap speedup - Enable HAVE_MOVE_PMD Kalesh Singh
2020-09-30 22:21 ` [PATCH 3/5] mm: Speedup mremap on 1GB or larger regions Kalesh Singh
2020-10-01 12:36 ` Kirill A. Shutemov
2020-10-01 16:40 ` Kalesh Singh
2020-10-01 18:10 ` Kalesh Singh
2020-09-30 22:21 ` [PATCH 4/5] arm64: mremap speedup - Enable HAVE_MOVE_PUD Kalesh Singh
2020-09-30 22:21 ` [PATCH 5/5] x86: " Kalesh Singh
2020-09-30 22:32 ` [PATCH 0/5] Speed up mremap on large regions Kirill A. Shutemov
2020-09-30 22:42 ` Lokesh Gidra
2020-09-30 22:46 ` Joel Fernandes
2020-09-30 23:03 ` Kalesh Singh
2020-10-01 12:27 ` Kirill A. Shutemov
2020-10-01 15:59 ` Kalesh Singh
2020-10-02 0:09 ` Lokesh Gidra
2020-10-02 5:35 ` Kirill A. Shutemov [this message]
2020-10-02 6:39 ` Lokesh Gidra
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=20201002053547.7roe7b4mpamw4uk2@black.fi.intel.com \
--to=kirill.shutemov@linux.intel.com \
--cc=akpm@linux-foundation.org \
--cc=almasrymina@google.com \
--cc=aneesh.kumar@linux.ibm.com \
--cc=arnd@arndb.de \
--cc=bgeffon@google.com \
--cc=bp@alien8.de \
--cc=catalin.marinas@arm.com \
--cc=christian.brauner@ubuntu.com \
--cc=dave.hansen@intel.com \
--cc=frederic@kernel.org \
--cc=gshan@redhat.com \
--cc=hnaveed@wavecomp.com \
--cc=hpa@zytor.com \
--cc=jhubbard@nvidia.com \
--cc=joelaf@google.com \
--cc=justin.he@arm.com \
--cc=kaleshsingh@google.com \
--cc=kamalesh@linux.vnet.ibm.com \
--cc=keescook@chromium.org \
--cc=kernel-team@android.com \
--cc=krzk@kernel.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linuxram@us.ibm.com \
--cc=lokeshgidra@google.com \
--cc=mark.rutland@arm.com \
--cc=masahiroy@kernel.org \
--cc=mhiramat@kernel.org \
--cc=minchan@google.com \
--cc=mingo@redhat.com \
--cc=peterz@infradead.org \
--cc=rcampbell@nvidia.com \
--cc=rppt@kernel.org \
--cc=samitolvanen@google.com \
--cc=sandipan@linux.ibm.com \
--cc=shuah@kernel.org \
--cc=sjpark@amazon.de \
--cc=surenb@google.com \
--cc=tglx@linutronix.de \
--cc=will@kernel.org \
--cc=william.kucharski@oracle.com \
--cc=x86@kernel.org \
--cc=yezhenyu2@huawei.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).