lkml.org 
[lkml]   [2007]   [May]   [22]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
SubjectRE: 2.6.22-rc1-mm1 - s390 vs. md
Date
From
> From: Cornelia Huck [mailto:cornelia.huck@de.ibm.com]
> On Fri, 18 May 2007 09:30:09 -0700,
> "Williams, Dan J" <dan.j.williams@intel.com> wrote:
>
> > When CONFIG_DMA_ENGINE=n async_tx_find_channel takes the form:
> > ... async_tx_find_channel( ... )
> > {
> > return NULL;
> > }
> >
> > So in the S390 case the entire asynchronous path will be compiled
away.
>
> Unfortunately, do_async_xor() (and others) is not ifdef'ed and
contains
> dma_map_page(), which led to the compile failure...

The approach I have taken is to add the missing definitions to
include/asm-s390/dma-mapping.h [ a non-outlook-mangled version of the
patch is pushed out in my rebased git tree ]. I was not able to fully
compile-test this change as the three s390-cross-toolchains I tried each
died early in the kernel build process. The most common error was:
"s390-unknown-linux-gnu-ld: unrecognised emulation mode: elf64_s390"

---
s390: add dma mapping api stub definitions for async_tx

From: Dan Williams <dan.j.williams@intel.com>

The asynchronous path in async_tx is meant to be compiled away on
platforms
like s390 with CONFIG_DMA_ENGINE=n. However, it is difficult to compile
something away if it does not compile in the first place. This patch
adds
the missing dma api definitions as BUG() stubs.

Cc: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
include/asm-s390/dma-mapping.h | 78
++++++++++++++++++++++++++++++++++++++++
1 files changed, 78 insertions(+), 0 deletions(-)
diff --git a/include/asm-s390/dma-mapping.h
b/include/asm-s390/dma-mapping.h
index 3f8c12f..33a3c82 100644
--- a/include/asm-s390/dma-mapping.h
+++ b/include/asm-s390/dma-mapping.h
@@ -4,9 +4,87 @@
* S390 version
*
* This file exists so that #include <dma-mapping.h> doesn't break
anything.
+ * It also includes stub definitions of the API so common code like
async_tx
+ * can compile.
*/

#ifndef _ASM_DMA_MAPPING_H
#define _ASM_DMA_MAPPING_H

+#include <linux/mm.h>
+#include <asm/scatterlist.h>
+
+static inline dma_addr_t
+dma_map_single(struct device *dev, void *cpu_addr, size_t size,
+ enum dma_data_direction dir)
+{
+ BUG();
+ return 0;
+}
+
+static inline dma_addr_t
+dma_map_page(struct device *dev, struct page *page,
+ unsigned long offset, size_t size,
+ enum dma_data_direction dir)
+{
+ BUG();
+ return 0;
+}
+
+static inline void
+dma_unmap_single(struct device *dev, dma_addr_t handle, size_t size,
+ enum dma_data_direction dir)
+{
+ BUG();
+}
+
+static inline void
+dma_unmap_page(struct device *dev, dma_addr_t handle, size_t size,
+ enum dma_data_direction dir)
+{
+ BUG();
+}
+
+static inline int
+dma_map_sg(struct device *dev, struct scatterlist *sg, int nents,
+ enum dma_data_direction dir)
+{
+ BUG();
+ return 0;
+}
+
+static inline void
+dma_unmap_sg(struct device *dev, struct scatterlist *sg, int nents,
+ enum dma_data_direction dir)
+{
+ BUG();
+}
+
+static inline void
+dma_sync_single_for_cpu(struct device *dev, dma_addr_t handle, size_t
size,
+ enum dma_data_direction dir)
+{
+ BUG();
+}
+
+static inline void
+dma_sync_single_for_device(struct device *dev, dma_addr_t handle,
size_t size,
+ enum dma_data_direction dir)
+{
+ BUG();
+}
+
+static inline void
+dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg, int
nents,
+ enum dma_data_direction dir)
+{
+ BUG();
+}
+
+static inline void
+dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg, int
nents,
+ enum dma_data_direction dir)
+{
+ BUG();
+}
#endif /* _ASM_DMA_MAPPING_H */
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
\
 
 \ /
  Last update: 2007-05-23 02:27    [from the cache]
©2003-2011 Jasper Spaans