linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Bryan O'Sullivan" <bos@pathscale.com>
To: Andrew Morton <akpm@osdl.org>, Roland Dreier <rdreier@cisco.com>
Cc: sam@ravnborg.org, linux-kernel@vger.kernel.org
Subject: [PATCH] [RFC] Generic 32-bit MMIO copy, out of line
Date: Tue, 10 Jan 2006 08:56:05 -0800	[thread overview]
Message-ID: <1136912166.32183.45.camel@serpentine.pathscale.com> (raw)
In-Reply-To: <1136909276.32183.28.camel@serpentine.pathscale.com>

On Tue, 2006-01-10 at 08:07 -0800, Bryan O'Sullivan wrote:

> The only problem with this is that it's an unusual approach, so I don't
> have any obvious examples to copy.

It was easy to hack up a quick patch to try the idea out.

This single-arch patch is for review purposes only; I compile-tested it
on i386, and it has no obvious problems at compile- or link-time.

If it looks OK, I have a complete patch set that affects all arches.
The parts that affect arch/*/lib/Makefile and include/asm-*/io.h are
essentially identical in all cases.

	<b

diff -r 48616306e7bd include/asm-generic/raw_memcpy_toio32.h
--- /dev/null	Thu Jan  1 00:00:00 1970 +0000
+++ b/include/asm-generic/raw_memcpy_toio32.h	Tue Jan 10 08:44:30 2006 -0800
@@ -0,0 +1,16 @@
+#ifndef _ASM_GENERIC_RAW_MEMCPYTOIO32_H
+#define _ASM_GENERIC_RAW_MEMCPYTOIO32_H
+
+/*
+ * __raw_memcpy_toio32 - copy data to MMIO space, in 32-bit units
+ *
+ * Order of access is not guaranteed, nor is a memory barrier performed
+ * afterwards.  This is an arch-independent generic implementation.
+ *
+ * @to: destination, in MMIO space (must be 32-bit aligned)
+ * @from: source (must be 32-bit aligned)
+ * @count: number of 32-bit quantities to copy
+ */
+void __raw_memcpy_toio32(void __iomem *to, const void *from, size_t count);
+
+#endif // _ASM_GENERIC_RAW_MEMCPYTOIO32_H
--- /dev/null	Thu Jan  1 00:00:00 1970 +0000
+++ b/lib/raw_memcpy_toio32.c	Tue Jan 10 08:44:30 2006 -0800
@@ -0,0 +1,13 @@
+#include <linux/types.h>
+#include <asm/io.h>
+#include <asm-generic/raw_memcpy_toio32.h>
+
+void __raw_memcpy_toio32(void __iomem *to, const void *from, size_t count)
+{
+	u32 __iomem *dst = to;
+	const u32 *src = from;
+	size_t i;
+
+	for (i = 0; i < count; i++)
+		__raw_writel(*src++, dst++);
+}
--- a/arch/i386/lib/Makefile	Tue Jan 10 08:44:35 2006 -0800
+++ b/arch/i386/lib/Makefile	Tue Jan 10 08:44:44 2006 -0800
@@ -5,5 +5,6 @@
 
 lib-y = checksum.o delay.o usercopy.o getuser.o putuser.o memcpy.o strstr.o \
 	bitops.o
+lib-y += ../../../lib/raw_memcpy_toio32.o
 
 lib-$(CONFIG_X86_USE_3DNOW) += mmx.o
--- a/include/asm-i386/io.h	Tue Jan 10 08:44:35 2006 -0800
+++ b/include/asm-i386/io.h	Tue Jan 10 08:44:44 2006 -0800
@@ -203,6 +203,8 @@
 {
 	__memcpy((void __force *) dst, src, count);
 }
+
+#include <asm-generic/raw_memcpy_toio32.h>
 
 /*
  * ISA space is 'always mapped' on a typical x86 system, no need to



  reply	other threads:[~2006-01-10 16:56 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-01-06 20:26 [PATCH 0 of 3] 32-bit MMIO copy routine Bryan O'Sullivan
2006-01-06 20:26 ` [PATCH 1 of 3] Introduce __raw_memcpy_toio32 Bryan O'Sullivan
2006-01-10  9:18   ` Andrew Morton
2006-01-10 14:55     ` Roland Dreier
2006-01-10 16:07       ` Bryan O'Sullivan
2006-01-10 16:56         ` Bryan O'Sullivan [this message]
2006-01-10 17:07         ` Christoph Hellwig
2006-01-10 17:13           ` Bryan O'Sullivan
2006-01-10 17:49           ` Bryan O'Sullivan
2006-01-10 17:51             ` Christoph Hellwig
2006-01-10 17:55               ` Bryan O'Sullivan
2006-01-10 22:05                 ` Andrew Morton
2006-01-10 22:29                   ` Bryan O'Sullivan
2006-01-10 23:32                     ` Andrew Morton
2006-01-11 17:20                       ` Bryan O'Sullivan
2006-01-11 17:22                         ` Sam Ravnborg
2006-01-11 17:30                           ` Andrew Morton
2006-01-11 17:43                             ` Bryan O'Sullivan
2006-01-11 18:49                               ` Roland Dreier
2006-01-11 18:57                                 ` Bryan O'Sullivan
2006-01-11 19:01                                   ` Roland Dreier
2006-01-11 19:08                                     ` Bryan O'Sullivan
2006-01-13 15:19                       ` Adrian Bunk
2006-01-10 18:02               ` Randy.Dunlap
2006-01-10 20:04         ` Sam Ravnborg
2006-01-10 15:59     ` Bryan O'Sullivan
2006-01-06 20:26 ` [PATCH 2 of 3] memcpy32 for x86_64 Bryan O'Sullivan
2006-01-06 20:26 ` [PATCH 3 of 3] Add __raw_memcpy_toio32 to each arch Bryan O'Sullivan

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=1136912166.32183.45.camel@serpentine.pathscale.com \
    --to=bos@pathscale.com \
    --cc=akpm@osdl.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=rdreier@cisco.com \
    --cc=sam@ravnborg.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).