From: arnd@arndb•de (Arnd Bergmann)
To: linux-arm-kernel@lists•infradead.org
Subject: [PATCH 13/15] ARM: make mach/io.h include optional
Date: Tue, 14 Feb 2012 18:12:38 +0000 [thread overview]
Message-ID: <201202141812.38847.arnd@arndb.de> (raw)
In-Reply-To: <20120214174035.GA29765@n2100.arm.linux.org.uk>
On Tuesday 14 February 2012, Russell King - ARM Linux wrote:
> > Yes, at least in the long run. Note that this should make no difference
> > at all from a performance point of view, but it does impact code size a bit.
>
> That depends whether the additional reloads of pci_io_base can be properly
> scheduled by the compiler, and experience shows that you tend to end up
> with the load delay slot not being filled on older processors.
>
> Not only that, but the compiler will evaluate the entire:
>
> pci_io_base + (addr & IO_SPACE_LIMIT)
>
> thing every time. With a 64K mask, that will include reloading the
> mask every single access.
>
> So, we'll probably end up with about three additional loads per IO
> operation, none of which would be scheduled particularly well.
>
> "Yuck" and "not in my kernel" comes to mind.
I totally agree with the code size point, but my point above was that
from performance perspective all that you mentioned should be dwarfed
by the overhead of actually doing a synchronous operation on an
external bus. writel may be reasonably fast on a CPU internal bus,
but inb/outb implies a full bus synchronization and is only
used on older PCI hardware, typically those that date back to
ISA in some form.
Arnd
next prev parent reply other threads:[~2012-02-14 18:12 UTC|newest]
Thread overview: 74+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-02-13 21:43 [PATCH 00/15] mach/io.h cleanup and removal Rob Herring
2012-02-13 21:43 ` [PATCH 01/15] usb: ohci-pxa27x: add explicit include of hardware.h Rob Herring
2012-02-13 21:43 ` [PATCH 02/15] ARM: add explicit include of system.h to processor.h Rob Herring
2012-02-13 22:14 ` H Hartley Sweeten
2012-02-13 21:43 ` [PATCH 03/15] ARM: provide runtime hook for ioremap Rob Herring
2012-02-13 22:13 ` H Hartley Sweeten
2012-02-13 22:30 ` Russell King - ARM Linux
2012-02-13 22:48 ` Rob Herring
2012-02-13 21:43 ` [PATCH 04/15] ARM: imx: convert to common runtime ioremap hook Rob Herring
2012-02-16 0:17 ` Shawn Guo
2012-02-13 21:43 ` [PATCH 05/15] ARM: msm: use " Rob Herring
2012-02-13 23:05 ` David Brown
2012-02-13 21:43 ` [PATCH 06/15] ARM: msm: clean-up mach/io.h Rob Herring
2012-02-13 21:43 ` [PATCH 07/15] ARM: at91: " Rob Herring
2012-02-14 9:21 ` Nicolas Ferre
2012-02-14 13:24 ` Rob Herring
2012-02-16 7:43 ` Jean-Christophe PLAGNIOL-VILLARD
2012-02-16 14:08 ` Rob Herring
2012-02-16 14:23 ` Jean-Christophe PLAGNIOL-VILLARD
2012-02-23 17:26 ` Nicolas Ferre
2012-02-27 16:55 ` Rob Herring
2012-02-27 17:27 ` Jean-Christophe PLAGNIOL-VILLARD
2012-02-13 21:43 ` [PATCH 08/15] ARM: davinci: remove unneeded mach/io.h include Rob Herring
2012-02-13 21:43 ` [PATCH 09/15] ARM: orion5x: clean-up mach/io.h Rob Herring
2012-02-13 21:43 ` [PATCH 10/15] ARM: tegra: " Rob Herring
2012-02-13 21:43 ` [PATCH 11/15] ARM: ep93xx: " Rob Herring
2012-02-13 21:52 ` Ryan Mallon
2012-02-13 22:15 ` Rob Herring
2012-02-13 22:16 ` H Hartley Sweeten
2012-02-27 15:17 ` [PATCH] " Rob Herring
2012-02-13 21:43 ` [PATCH 12/15] ARM: clps711x: remove unneeded include of mach/io.h Rob Herring
2012-02-13 21:43 ` [PATCH 13/15] ARM: make mach/io.h include optional Rob Herring
2012-02-13 22:14 ` H Hartley Sweeten
2012-02-13 22:36 ` Russell King - ARM Linux
2012-02-13 22:55 ` Rob Herring
2012-02-14 2:03 ` Arnd Bergmann
2012-02-14 2:54 ` Rob Herring
2012-02-14 8:04 ` Arnd Bergmann
2012-02-14 14:36 ` Rob Herring
2012-02-14 17:16 ` Arnd Bergmann
2012-02-14 17:40 ` Russell King - ARM Linux
2012-02-14 18:12 ` Arnd Bergmann [this message]
2012-02-14 23:09 ` Rob Herring
2012-02-14 23:43 ` Russell King - ARM Linux
2012-02-15 0:25 ` Nicolas Pitre
2012-02-15 14:14 ` Rob Herring
2012-02-15 0:57 ` Arnd Bergmann
2012-02-27 19:31 ` Rob Herring
2012-02-28 16:10 ` Arnd Bergmann
2012-02-27 22:31 ` Rob Herring
2012-02-28 16:32 ` Arnd Bergmann
2012-02-13 23:15 ` H Hartley Sweeten
2012-02-14 1:06 ` Arnd Bergmann
2012-02-14 17:38 ` H Hartley Sweeten
2012-02-14 18:20 ` Arnd Bergmann
2012-02-13 21:43 ` [PATCH 14/15] ARM: remove bunch of now unused mach/io.h files Rob Herring
2012-02-13 22:16 ` H Hartley Sweeten
2012-02-16 0:19 ` Shawn Guo
2012-02-16 18:57 ` Linus Walleij
2012-02-13 21:43 ` [PATCH 15/15] ARM: kill off __mem_pci Rob Herring
2012-02-13 22:22 ` [PATCH 00/15] mach/io.h cleanup and removal Tony Lindgren
2012-02-13 23:56 ` Tony Lindgren
2012-02-14 3:09 ` Rob Herring
2012-02-13 23:41 ` Tony Lindgren
2012-02-14 3:20 ` Rob Herring
2012-02-14 17:24 ` Tony Lindgren
2012-02-14 17:57 ` Arnd Bergmann
2012-02-14 18:28 ` Nicolas Pitre
2012-02-14 19:41 ` Rob Herring
2012-02-14 20:43 ` Tony Lindgren
2012-02-14 21:26 ` Arnd Bergmann
2012-02-14 21:54 ` Rob Herring
2012-02-14 22:38 ` Arnd Bergmann
2012-02-21 22:47 ` Stephen Warren
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=201202141812.38847.arnd@arndb.de \
--to=arnd@arndb$(echo .)de \
--cc=linux-arm-kernel@lists$(echo .)infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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