[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <ZHnHs+ro44SBS+lV@MiWiFi-R3L-srv>
Date: Fri, 2 Jun 2023 18:42:59 +0800
From: Baoquan He <bhe@...hat.com>
To: Christoph Hellwig <hch@...radead.org>
Cc: linux-kernel@...r.kernel.org, linux-arch@...r.kernel.org,
linux-mm@...ck.org, arnd@...db.de, christophe.leroy@...roup.eu,
agordeev@...ux.ibm.com, wangkefeng.wang@...wei.com,
schnelle@...ux.ibm.com, David.Laight@...lab.com, shorne@...il.com,
willy@...radead.org, deller@....de
Subject: Re: [PATCH v5 RESEND 14/17] mm/ioremap: Consider IOREMAP space in
generic ioremap
On 06/01/23 at 04:13am, Christoph Hellwig wrote:
> On Tue, May 30, 2023 at 05:37:23PM +0800, Baoquan He wrote:
> > If we want to consolidate code, we can move is_ioremap_addr() to
> > include/linux/mm.h libe below. Not sure if it's fine. With it,
> > both kernel/iomem.c and mm/ioremap.c can use is_ioremap_addr().
>
> Can we just add a ne header for this given that no one else really
> needs it?
Is it OK like below?
>From fe5d4d25afa1e989fa82877c8466a97fc8bd8c93 Mon Sep 17 00:00:00 2001
From: Baoquan He <bhe@...hat.com>
Date: Fri, 2 Jun 2023 18:36:48 +0800
Subject: [PATCH] mm: move is_ioremap_addr() into new header file
Content-type: text/plain
Now is_ioremap_addr() is only used in kernel/iomem.c and gonna be used
in mm/ioremap.c. Move it into its own new header file linux/ioremap.h.
Signed-off-by: Baoquan He <bhe@...hat.com>
---
include/linux/ioremap.h | 29 +++++++++++++++++++++++++++++
include/linux/mm.h | 5 -----
kernel/iomem.c | 1 +
mm/ioremap.c | 1 +
4 files changed, 31 insertions(+), 5 deletions(-)
create mode 100644 include/linux/ioremap.h
diff --git a/include/linux/ioremap.h b/include/linux/ioremap.h
new file mode 100644
index 000000000000..2fd51a77ebdc
--- /dev/null
+++ b/include/linux/ioremap.h
@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _LINUX_IOREMAP_H
+#define _LINUX_IOREMAP_H
+
+#include <linux/kasan.h>
+#include <asm/pgtable.h>
+
+#if defined(CONFIG_HAS_IOMEM) || defined(CONFIG_GENERIC_IOREMAP)
+/*
+ * Ioremap often, but not always uses the generic vmalloc area. E.g on
+ * Power ARCH, it could have different ioremap space.
+ */
+#ifndef IOREMAP_START
+#define IOREMAP_START VMALLOC_START
+#define IOREMAP_END VMALLOC_END
+#endif
+static inline bool is_ioremap_addr(const void *x)
+{
+ unsigned long addr = (unsigned long)kasan_reset_tag(x);
+ return addr >= IOREMAP_START && addr < IOREMAP_END;
+}
+#else
+static inline bool is_ioremap_addr(const void *x)
+{
+ return false;
+}
+#endif
+
+#endif /* _LINUX_IOREMAP_H */
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 27ce77080c79..7379f19768b4 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1041,11 +1041,6 @@ unsigned long vmalloc_to_pfn(const void *addr);
* On nommu, vmalloc/vfree wrap through kmalloc/kfree directly, so there
* is no special casing required.
*/
-
-#ifndef is_ioremap_addr
-#define is_ioremap_addr(x) is_vmalloc_addr(x)
-#endif
-
#ifdef CONFIG_MMU
extern bool is_vmalloc_addr(const void *x);
extern int is_vmalloc_or_module_addr(const void *x);
diff --git a/kernel/iomem.c b/kernel/iomem.c
index 62c92e43aa0d..9682471e6471 100644
--- a/kernel/iomem.c
+++ b/kernel/iomem.c
@@ -3,6 +3,7 @@
#include <linux/types.h>
#include <linux/io.h>
#include <linux/mm.h>
+#include <linux/ioremap.h>
#ifndef ioremap_cache
/* temporary while we convert existing ioremap_cache users to memremap */
diff --git a/mm/ioremap.c b/mm/ioremap.c
index 0248e630561b..3dede3302eba 100644
--- a/mm/ioremap.c
+++ b/mm/ioremap.c
@@ -10,6 +10,7 @@
#include <linux/mm.h>
#include <linux/io.h>
#include <linux/export.h>
+#include <linux/ioremap.h>
/*
* Ioremap often, but not always uses the generic vmalloc area. E.g on
--
2.34.1
>
>
Powered by blists - more mailing lists