ARC: Hardcode ARCH_DMA_MINALIGN to max line length we may have

Current implementation relies on L1 line length which might easily
be smaller than L2 line (which is usually the case BTW).

Imagine this typical case: L2 line is 128 bytes while L1 line is
64-bytes. Now we want to allocate small buffer and later use it for DMA
(consider IOC is not available).

kmalloc() allocates small KMALLOC_MIN_SIZE-sized, KMALLOC_MIN_SIZE-aligned
That way if buffer happens to be aligned to L1 line and not L2 line we'll be
flushing and invalidating extra portions of data from L2 which will cause
cache coherency issues.

And since KMALLOC_MIN_SIZE is bound to ARCH_DMA_MINALIGN the fix could
be simple - set ARCH_DMA_MINALIGN to the largest cache line we may ever
get. As of today neither L1 of ARC700 and ARC HS38 nor SLC might not be
longer than 128 bytes.

Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>

diff --git a/arch/arc/include/asm/cache.h b/arch/arc/include/asm/cache.h
index 02fd1ce..6349af7 100644 (file)
--- a/arch/arc/include/asm/cache.h
+++ b/arch/arc/include/asm/cache.h
@@ -47,7 +47,8 @@
        : "r"(data), "r"(ptr));         \
 })
 
-#define ARCH_DMA_MINALIGN      L1_CACHE_BYTES
+/* Largest line length for either L1 or L2 is 128 bytes */
+#define ARCH_DMA_MINALIGN      128
 
 extern void arc_cache_init(void);
 extern char *arc_cache_mumbojumbo(int cpu_id, char *buf, int len);