Re: [PATCH v2] of/address: Add of_iomap_nocache
From: Scott Wood <hidden>
Date: 2011-08-04 16:56:41
Also in:
linux-arm-msm, lkml
On 08/04/2011 05:36 AM, David Brown wrote:
quoted hunk ↗ jump to hunk
Add uncached mappings from devicetree nodes similar to regular io mappings. SPARC is coherent, so there this call is the same as regular of_iomap. Cc: David Miller<davem@davemloft.net> Signed-off-by: David Brown<redacted> --- v2 - Add implementation for SPARC drivers/of/address.c | 19 +++++++++++++++++++ include/linux/of_address.h | 10 ++++++++++ 2 files changed, 29 insertions(+), 0 deletions(-)diff --git a/drivers/of/address.c b/drivers/of/address.c index 72c33fb..9bee7f8 100644 --- a/drivers/of/address.c +++ b/drivers/of/address.c@@ -613,3 +613,22 @@ void __iomem *of_iomap(struct device_node *np, int index) return ioremap(res.start, resource_size(&res)); } EXPORT_SYMBOL(of_iomap); + +/** + * of_iomap_nocache - Maps the memory mapped IO for a given + * device_node, using ioremap_nocache. + * @device: the device whose io range will be mapped + * @index: index of the io range + * + * Returns a pointer to the mapped memory + */ +void __iomem *of_iomap_nocache(struct device_node *np, int index) +{ + struct resource res; + + if (of_address_to_resource(np, index,&res)) + return NULL; + + return ioremap_nocache(res.start, 1 + res.end - res.start); +}
resource_size()?
quoted hunk ↗ jump to hunk
+EXPORT_SYMBOL(of_iomap_nocache);diff --git a/include/linux/of_address.h b/include/linux/of_address.h index 3118623..0e4734b 100644 --- a/include/linux/of_address.h +++ b/include/linux/of_address.h@@ -13,6 +13,16 @@ extern struct device_node *of_find_matching_node_by_address( u64 base_address); extern void __iomem *of_iomap(struct device_node *device, int index); +#ifndef SPARC +extern void __iomem *of_iomap_nocache(struct device_node *device, int index); +#else +static inline void __iomem *of_iomap_nocache(struct device_node *device, + int index) +{ + return of_iomap(device, index); +} +#endif
Why is sparc special? It looks like it defines ioremap_nocache() as ioremap() just like powerpc and some others, so shouldn't the normal of_iomap_nocache just work? -Scott