Generic facilities for byteswapping and unaligned references

For future portability, add functions to access both littleendian and
bigendian numbers, and to perform unaligned memory references.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>

2 files changed, 141 insertions, 11 deletions

diff --git a/com32/include/byteswap.h b/com32/include/byteswap.h
index 11870e97..5ff2cd69 100644
--- a/com32/include/byteswap.h
+++ b/com32/include/byteswap.h
@@ -1,11 +1,12 @@
 #ifndef _BYTESWAP_H
 #define _BYTESWAP_H
 
-/* COM32 will be running on an i386 platform */
-
 #include <stdint.h>
+#include <klibc/endian.h>
 #include <klibc/compiler.h>
 
+/* This assumes an i386 platform */
+
 #define __bswap_16_macro(v) ((uint16_t)		  			\
 			     (((uint16_t)(v) << 8) | 			\
 			      ((uint16_t)(v) >> 8)))
@@ -26,8 +27,14 @@ static inline __constfunc uint16_t __bswap_16(uint16_t v)
 
 static inline __constfunc uint32_t __bswap_32(uint32_t v)
 {
+#if defined(__x86_64__)
+    asm("bswap %0" : "+r" (v));
+#elif defined(__i386__)
     asm("xchgb %h0,%b0 ; roll $16,%0 ; xchgb %h0,%b0"
 	: "+q" (v));
+#else
+    v = __bswap_32_macro(v);
+#endif
     return v;
 }
 
@@ -37,14 +44,137 @@ static inline __constfunc uint32_t __bswap_32(uint32_t v)
 
 #define __bswap_64_macro(v) ((uint64_t)					\
     (((uint64_t)__bswap_32_macro((uint32_t)(v)) << 32) |		\
-     (__bswap_32__macro((uint32_t)((uint64_t)(v) >> 32)))))
+     (__bswap_32_macro((uint32_t)((uint64_t)(v) >> 32)))))
 
 static inline __constfunc uint64_t __bswap_64(uint64_t v)
 {
-    return ((uint64_t)__bswap_32(v) << 32) | __bswap_32(v >> 32);
+#if defined(__x86_64__)
+    asm("bswap %0" : "+r" (v));
+#else
+    v = ((uint64_t)__bswap_32(v) << 32) | __bswap_32(v >> 32);
+#endif
+    return v;
 }
 
 #define bswap_64(x) (__builtin_constant_p(x) ? 				\
 			__bswap_64_macro(x) :  __bswap_64(x))
 
-#endif /* byteswap.h */
+/* This is generic */
+
+#if __BYTE_ORDER == __LITTLE_ENDIAN
+
+#define be16_to_cpu(x) bswap_16(x)
+#define cpu_to_be16(x) bswap_16(x)
+#define be32_to_cpu(x) bswap_32(x)
+#define cpu_to_be32(x) bswap_32(x)
+#define be64_to_cpu(x) bswap_64(x)
+#define cpu_to_be64(x) bswap_64(x)
+
+#define le16_to_cpu(x) (x)
+#define cpu_to_le16(x) (x)
+#define le32_to_cpu(x) (x)
+#define cpu_to_le32(x) (x)
+#define le64_to_cpu(x) (x)
+#define cpu_to_le64(x) (x)
+
+#elif __BYTE_ORDER == __BIG_ENDIAN
+
+#define le16_to_cpu(x) bswap_16(x)
+#define cpu_to_le16(x) bswap_16(x)
+#define le32_to_cpu(x) bswap_32(x)
+#define cpu_to_le32(x) bswap_32(x)
+#define le64_to_cpu(x) bswap_64(x)
+#define cpu_to_le64(x) bswap_64(x)
+
+#define be16_to_cpu(x) (x)
+#define cpu_to_be16(x) (x)
+#define be32_to_cpu(x) (x)
+#define cpu_to_be32(x) (x)
+#define be64_to_cpu(x) (x)
+#define cpu_to_be64(x) (x)
+
+#else
+
+#error "Unknown byte order!"
+
+#endif
+
+typedef struct { uint16_t x; } __attribute__((packed)) __ua_uint16_t;
+typedef struct { uint32_t x; } __attribute__((packed)) __ua_uint32_t;
+typedef struct { uint64_t x; } __attribute__((packed)) __ua_uint64_t;
+
+/* These are guaranteed to support unaligned accesses */
+static inline uint16_t get_le16(const uint16_t *p)
+{
+    const __ua_uint16_t *up = (const __ua_uint16_t *)p;
+    return le16_to_cpu(up->x);
+}
+
+static inline uint32_t get_le32(const uint32_t *p)
+{
+    const __ua_uint32_t *up = (const __ua_uint32_t *)p;
+    return le32_to_cpu(up->x);
+}    
+
+static inline uint64_t get_le64(const uint64_t *p)
+{
+    const __ua_uint64_t *up = (const __ua_uint64_t *)p;
+    return le64_to_cpu(up->x);
+}
+
+static inline uint16_t get_be16(const uint16_t *p)
+{
+    const __ua_uint16_t *up = (const __ua_uint16_t *)p;
+    return be16_to_cpu(up->x);
+}
+
+static inline uint32_t get_be32(const uint32_t *p)
+{
+    const __ua_uint32_t *up = (const __ua_uint32_t *)p;
+    return be32_to_cpu(up->x);
+}    
+
+static inline uint64_t get_be64(const uint64_t *p)
+{
+    const __ua_uint64_t *up = (const __ua_uint64_t *)p;
+    return be64_to_cpu(up->x);
+}
+
+static inline void put_le16(uint16_t v, uint16_t *p)
+{
+    __ua_uint16_t *up = (__ua_uint16_t *)p;
+    up->x = cpu_to_le16(v);
+}
+
+static inline void put_le32(uint32_t v, uint32_t *p)
+{
+    __ua_uint32_t *up = (__ua_uint32_t *)p;
+    up->x = cpu_to_le32(v);
+}
+
+static inline void put_le64(uint64_t v, uint64_t *p)
+{
+    __ua_uint64_t *up = (__ua_uint64_t *)p;
+    up->x = cpu_to_le64(v);
+}
+
+static inline void put_be16(uint16_t v, uint16_t *p)
+{
+    __ua_uint16_t *up = (__ua_uint16_t *)p;
+    up->x = cpu_to_be16(v);
+}
+
+static inline void put_be32(uint32_t v, uint32_t *p)
+{
+    __ua_uint32_t *up = (__ua_uint32_t *)p;
+    up->x = cpu_to_be32(v);
+}
+
+static inline void put_be64(uint64_t v, uint64_t *p)
+{
+    __ua_uint64_t *up = (__ua_uint64_t *)p;
+    up->x = cpu_to_be64(v);
+}
+
+#endif /* _BYTESWAP_H */
+
diff --git a/com32/include/netinet/in.h b/com32/include/netinet/in.h
index b24f8046..0a0049fd 100644
--- a/com32/include/netinet/in.h
+++ b/com32/include/netinet/in.h
@@ -8,12 +8,12 @@
 #include <stdint.h>
 #include <byteswap.h>
 
-#define htons(x) bswap_16(x)
-#define ntohs(x) bswap_16(x)
-#define htonl(x) bswap_32(x)
-#define ntohl(x) bswap_32(x)
-#define htonq(x) bswap_64(x)
-#define ntohq(x) bswap_64(x)
+#define htons(x) cpu_to_be16(x)
+#define ntohs(x) be16_to_cpu(x)
+#define htonl(x) cpu_to_be32(x)
+#define ntohl(x) be32_to_cpu(x)
+#define htonq(x) cpu_to_be64(x)
+#define ntohq(x) be64_to_cpu(x)
 
 typedef uint32_t in_addr_t;
 typedef uint16_t in_port_t;