From: Thomas Gleixner <tglx@linutronix.de>
To: Vincenzo Frascino <vincenzo.frascino@arm.com>
Cc: linux-arch@vger.kernel.org, Shuah Khan <shuah@kernel.org>,
Arnd Bergmann <arnd@arndb.de>,
Catalin Marinas <catalin.marinas@arm.com>,
Daniel Lezcano <daniel.lezcano@linaro.org>,
Will Deacon <will.deacon@arm.com>,
Russell King <linux@armlinux.org.uk>,
Ralf Baechle <ralf@linux-mips.org>,
Mark Salyzyn <salyzyn@android.com>,
Paul Burton <paul.burton@mips.com>,
Dmitry Safonov <0x7f454c46@gmail.com>,
Rasmus Villemoes <linux@rasmusvillemoes.dk>,
Peter Collingbourne <pcc@google.com>,
linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH v5 02/23] kernel: Define gettimeofday vdso common code
Date: Sat, 23 Feb 2019 18:31:32 +0100 (CET) [thread overview]
Message-ID: <alpine.DEB.2.21.1902231751510.1666@nanos.tec.linutronix.de> (raw)
In-Reply-To: <20190222122430.21180-3-vincenzo.frascino@arm.com>
On Fri, 22 Feb 2019, Vincenzo Frascino wrote:
> +static notrace int do_hres(const struct vdso_data *vd,
> + clockid_t clk,
> + struct __vdso_timespec *ts)
> +{
> + const struct vdso_timestamp *vdso_ts = &vd->basetime[clk];
> + u64 cycles, last, sec, ns;
> + u32 seq, cs_index = CLOCKSOURCE_MONO;
> +
> + if (clk == CLOCK_MONOTONIC_RAW)
> + cs_index = CLOCKSOURCE_RAW;
Uuurgh. So you create an array with 16 members and then use two. This code
is really optimized and now you add not only the pointless array, you also
need the extra index plus another conditional. Not to talk about the cache
impact which makes things even worse. In the x86 implementation we have:
u32 seq; + 0
int mode; + 4
u64 mask; + 8
u32 mult; + 16
u32 shift; + 20
struct vgtod_ts basetimer[VGTOD_BASES]; + 24
Each basetime array member occupies 16 bytes. So
CLOCK_REALTIME + 24
CLOCK_MONOTONIC + 40
..
cacheline boundary
..
CLOCK_REALTIME_COARSE + 104
CLOCK_MONOTONIC_COARSE + 120 <- cacheline boundary
CLOCK_BOOTTIME + 136
CLOCK_REALTIME_ALARM + 152
CLOCK_BOOTTIME_ALARM + 168
So the most used clocks REALTIME/MONO are in the first cacheline.
So with your scheme the thing becomes
u32 seq; + 0
int mode; + 4
struct cs cs[16] + 8
struct vgtod_ts basetimer[VGTOD_BASES]; + 264
and
CLOCK_REALTIME + 264
CLOCK_MONOTONIC + 280
IOW, the most important clocks touch TWO cachelines now which are not even
adjacent. No, they are 256 bytes apart, which really sucks for prefetching.
We're surely not going to sacrify the performance which we carefully tuned
in that code just to support MONO_RAW. The solution I showed you in the
other reply does not have these problems at all.
It's easy enough to benchmark these implementations and without trying I'm
pretty sure that you can see the performance drop nicely. Please do so next
time and provide the numbers in the changelogs.
Thanks,
tglx
_______________________________________________
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:[~2019-02-23 17:31 UTC|newest]
Thread overview: 60+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-02-22 12:24 [PATCH v5 00/23] Unify vDSOs across more architectures Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 01/23] kernel: Standardize vdso_datapage Vincenzo Frascino
2019-02-22 12:58 ` Mark Rutland
2019-02-23 16:51 ` Thomas Gleixner
2019-02-27 14:23 ` Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 02/23] kernel: Define gettimeofday vdso common code Vincenzo Frascino
2019-02-22 13:34 ` Mark Rutland
2019-02-22 13:49 ` Arnd Bergmann
2019-02-22 14:36 ` Arnd Bergmann
2019-02-27 14:52 ` Vincenzo Frascino
2019-02-28 9:29 ` Arnd Bergmann
2019-02-28 11:58 ` [PATCH 1/2] vdso: use fixed-size time types Arnd Bergmann
2019-02-28 11:58 ` [PATCH 2/2] vdso: add clock_gettime64 Arnd Bergmann
2019-02-28 13:42 ` [PATCH 1/2] vdso: use fixed-size time types Thomas Gleixner
2019-02-28 13:45 ` Vincenzo Frascino
2019-02-23 10:34 ` [PATCH v5 02/23] kernel: Define gettimeofday vdso common code Thomas Gleixner
2019-02-25 14:09 ` Vincenzo Frascino
2019-02-23 17:31 ` Thomas Gleixner [this message]
2019-02-27 13:47 ` Vincenzo Frascino
2019-02-27 15:49 ` Thomas Gleixner
2019-02-27 16:06 ` Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 03/23] arm64: Build vDSO with -ffixed-x18 Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 04/23] arm64: Substitute gettimeofday with C implementation Vincenzo Frascino
2019-02-22 13:42 ` Mark Rutland
2019-02-22 12:24 ` [PATCH v5 05/23] arm64: compat: Alloc separate pages for vectors and sigpage Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 06/23] arm64: compat: Split kuser32 Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 07/23] arm64: compat: Refactor aarch32_alloc_vdso_pages() Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 08/23] arm64: compat: Add KUSER_HELPERS config option Vincenzo Frascino
2019-02-22 14:04 ` Mark Rutland
2019-02-22 14:09 ` Russell King - ARM Linux admin
2019-02-26 12:10 ` Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 09/23] arm64: compat: Add missing syscall numbers Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 10/23] arm64: compat: Expose signal related structures Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 11/23] arm64: compat: Generate asm offsets for signals Vincenzo Frascino
2019-02-22 14:22 ` Mark Rutland
2019-02-22 12:24 ` [PATCH v5 12/23] lib: vdso: Add compat support Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 13/23] arm64: compat: Add vDSO Vincenzo Frascino
2019-02-22 14:32 ` Mark Rutland
2019-02-22 12:24 ` [PATCH v5 14/23] arm64: Refactor vDSO code Vincenzo Frascino
2019-02-22 14:37 ` Mark Rutland
2019-02-22 12:24 ` [PATCH v5 15/23] arm64: compat: vDSO setup for compat layer Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 16/23] arm64: elf: vDSO code page discovery Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 17/23] arm64: compat: Get sigreturn trampolines from vDSO Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 18/23] arm64: Add vDSO compat support Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 19/23] arm64: Enable compat vDSO support Vincenzo Frascino
2019-02-22 14:39 ` Mark Rutland
2019-02-22 14:41 ` Mark Rutland
2019-02-22 12:24 ` [PATCH v5 20/23] arm: Add support for generic vDSO Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 21/23] mips: " Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 22/23] x86: " Vincenzo Frascino
2019-02-23 10:45 ` Thomas Gleixner
2019-02-27 16:18 ` Vincenzo Frascino
2019-02-22 12:24 ` [PATCH v5 23/23] kselftest: Extend vDSO selftest Vincenzo Frascino
2019-02-23 17:39 ` [PATCH v5 00/23] Unify vDSOs across more architectures Thomas Gleixner
2019-02-28 11:40 ` Arnd Bergmann
2019-02-28 12:09 ` Vincenzo Frascino
2019-02-28 12:38 ` Arnd Bergmann
2019-02-28 12:42 ` Vincenzo Frascino
2019-02-28 13:54 ` Arnd Bergmann
2019-02-28 15:51 ` Vincenzo Frascino
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=alpine.DEB.2.21.1902231751510.1666@nanos.tec.linutronix.de \
--to=tglx@linutronix.de \
--cc=0x7f454c46@gmail.com \
--cc=arnd@arndb.de \
--cc=catalin.marinas@arm.com \
--cc=daniel.lezcano@linaro.org \
--cc=linux-arch@vger.kernel.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux@armlinux.org.uk \
--cc=linux@rasmusvillemoes.dk \
--cc=paul.burton@mips.com \
--cc=pcc@google.com \
--cc=ralf@linux-mips.org \
--cc=salyzyn@android.com \
--cc=shuah@kernel.org \
--cc=vincenzo.frascino@arm.com \
--cc=will.deacon@arm.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).