From 0d74c42f788caf3cad727c61c490d9459bc8918b Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Thu, 5 Dec 2013 14:54:38 -0800 Subject: ether_addr_equal: Optimize implementation, remove unused compare_ether_addr Add a new check for CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS to reduce the number of or's used in the ether_addr_equal comparison to very slightly improve function performance. Simplify the ether_addr_equal_64bits implementation. Integrate and remove the zap_last_2bytes helper as it's now used only once. Remove the now unused compare_ether_addr function. Update the unaligned-memory-access documentation to remove the compare_ether_addr description and show how unaligned accesses could occur with ether_addr_equal. Signed-off-by: Joe Perches Signed-off-by: David S. Miller --- include/linux/etherdevice.h | 51 ++++++++++++++++----------------------------- 1 file changed, 18 insertions(+), 33 deletions(-) (limited to 'include/linux/etherdevice.h') diff --git a/include/linux/etherdevice.h b/include/linux/etherdevice.h index fc4a9aa7dd82..3526e819d7ae 100644 --- a/include/linux/etherdevice.h +++ b/include/linux/etherdevice.h @@ -26,6 +26,7 @@ #include #include #include +#include #ifdef __KERNEL__ __be16 eth_type_trans(struct sk_buff *skb, struct net_device *dev); @@ -210,41 +211,27 @@ static inline void eth_hw_addr_inherit(struct net_device *dst, memcpy(dst->dev_addr, src->dev_addr, ETH_ALEN); } -/** - * compare_ether_addr - Compare two Ethernet addresses - * @addr1: Pointer to a six-byte array containing the Ethernet address - * @addr2: Pointer other six-byte array containing the Ethernet address - * - * Compare two Ethernet addresses, returns 0 if equal, non-zero otherwise. - * Unlike memcmp(), it doesn't return a value suitable for sorting. - */ -static inline unsigned compare_ether_addr(const u8 *addr1, const u8 *addr2) -{ - const u16 *a = (const u16 *) addr1; - const u16 *b = (const u16 *) addr2; - - BUILD_BUG_ON(ETH_ALEN != 6); - return ((a[0] ^ b[0]) | (a[1] ^ b[1]) | (a[2] ^ b[2])) != 0; -} - /** * ether_addr_equal - Compare two Ethernet addresses * @addr1: Pointer to a six-byte array containing the Ethernet address * @addr2: Pointer other six-byte array containing the Ethernet address * * Compare two Ethernet addresses, returns true if equal + * + * Please note: addr1 & addr2 must both be aligned to u16. */ static inline bool ether_addr_equal(const u8 *addr1, const u8 *addr2) { - return !compare_ether_addr(addr1, addr2); -} +#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) + u32 fold = ((*(const u32 *)addr1) ^ (*(const u32 *)addr2)) | + ((*(const u16 *)(addr1 + 4)) ^ (*(const u16 *)(addr2 + 4))); -static inline unsigned long zap_last_2bytes(unsigned long value) -{ -#ifdef __BIG_ENDIAN - return value >> 16; + return fold == 0; #else - return value << 16; + const u16 *a = (const u16 *)addr1; + const u16 *b = (const u16 *)addr2; + + return ((a[0] ^ b[0]) | (a[1] ^ b[1]) | (a[2] ^ b[2])) == 0; #endif } @@ -265,16 +252,14 @@ static inline unsigned long zap_last_2bytes(unsigned long value) static inline bool ether_addr_equal_64bits(const u8 addr1[6+2], const u8 addr2[6+2]) { -#ifdef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS - unsigned long fold = ((*(unsigned long *)addr1) ^ - (*(unsigned long *)addr2)); - - if (sizeof(fold) == 8) - return zap_last_2bytes(fold) == 0; +#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) && BITS_PER_LONG == 64 + u64 fold = (*(const u64 *)addr1) ^ (*(const u64 *)addr2); - fold |= zap_last_2bytes((*(unsigned long *)(addr1 + 4)) ^ - (*(unsigned long *)(addr2 + 4))); - return fold == 0; +#ifdef __BIG_ENDIAN + return (fold >> 16) == 0; +#else + return (fold << 16) == 0; +#endif #else return ether_addr_equal(addr1, addr2); #endif -- cgit v1.2.3 From 73eaef87e98a96fe8b8a586f916b2721bf512efa Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Fri, 6 Dec 2013 14:21:01 -0800 Subject: etherdevice: Add ether_addr_equal_unaligned Add a generic routine to test if possibly unaligned to u16 Ethernet addresses are equal. If CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS is set, this uses the slightly faster generic routine ether_addr_equal, otherwise this uses memcmp. Signed-off-by: Joe Perches Signed-off-by: David S. Miller --- include/linux/etherdevice.h | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) (limited to 'include/linux/etherdevice.h') diff --git a/include/linux/etherdevice.h b/include/linux/etherdevice.h index 3526e819d7ae..2f0e3d0818bc 100644 --- a/include/linux/etherdevice.h +++ b/include/linux/etherdevice.h @@ -265,6 +265,24 @@ static inline bool ether_addr_equal_64bits(const u8 addr1[6+2], #endif } +/** + * ether_addr_equal_unaligned - Compare two not u16 aligned Ethernet addresses + * @addr1: Pointer to a six-byte array containing the Ethernet address + * @addr2: Pointer other six-byte array containing the Ethernet address + * + * Compare two Ethernet addresses, returns true if equal + * + * Please note: Use only when any Ethernet address may not be u16 aligned. + */ +static inline bool ether_addr_equal_unaligned(const u8 *addr1, const u8 *addr2) +{ +#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) + return ether_addr_equal(addr1, addr2); +#else + return memcmp(addr1, addr2, ETH_ALEN) == 0; +#endif +} + /** * is_etherdev_addr - Tell if given Ethernet address belongs to the device. * @dev: Pointer to a device structure -- cgit v1.2.3 From 2c722fe1c821a100ca87fcc36e90a9bceb497c7c Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Fri, 6 Dec 2013 15:44:21 -0800 Subject: etherdevice: Optimize a few is__ether_addr functions If CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS is set, several is__ether_addr functions can be slightly improved by using u32 dereferences. I believe all current uses of is_zero_ether_addr and is_broadcast_ether_addr are u16 aligned, so always use u16 references to improve those functions performance. Document the u16 alignment requirements. Signed-off-by: Joe Perches Signed-off-by: David S. Miller --- include/linux/etherdevice.h | 25 +++++++++++++++++++++++-- 1 file changed, 23 insertions(+), 2 deletions(-) (limited to 'include/linux/etherdevice.h') diff --git a/include/linux/etherdevice.h b/include/linux/etherdevice.h index 2f0e3d0818bc..f344ac04f858 100644 --- a/include/linux/etherdevice.h +++ b/include/linux/etherdevice.h @@ -61,6 +61,8 @@ static const u8 eth_reserved_addr_base[ETH_ALEN] __aligned(2) = * * Return true if address is link local reserved addr (01:80:c2:00:00:0X) per * IEEE 802.1Q 8.6.3 Frame filtering. + * + * Please note: addr must be aligned to u16. */ static inline bool is_link_local_ether_addr(const u8 *addr) { @@ -68,7 +70,12 @@ static inline bool is_link_local_ether_addr(const u8 *addr) static const __be16 *b = (const __be16 *)eth_reserved_addr_base; static const __be16 m = cpu_to_be16(0xfff0); +#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) + return (((*(const u32 *)addr) ^ (*(const u32 *)b)) | + ((a[2] ^ b[2]) & m)) == 0; +#else return ((a[0] ^ b[0]) | (a[1] ^ b[1]) | ((a[2] ^ b[2]) & m)) == 0; +#endif } /** @@ -76,10 +83,18 @@ static inline bool is_link_local_ether_addr(const u8 *addr) * @addr: Pointer to a six-byte array containing the Ethernet address * * Return true if the address is all zeroes. + * + * Please note: addr must be aligned to u16. */ static inline bool is_zero_ether_addr(const u8 *addr) { - return !(addr[0] | addr[1] | addr[2] | addr[3] | addr[4] | addr[5]); +#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) + return ((*(const u32 *)addr) | (*(const u16 *)(addr + 4))) == 0; +#else + return (*(const u16 *)(addr + 0) | + *(const u16 *)(addr + 2) | + *(const u16 *)(addr + 4)) == 0; +#endif } /** @@ -110,10 +125,14 @@ static inline bool is_local_ether_addr(const u8 *addr) * @addr: Pointer to a six-byte array containing the Ethernet address * * Return true if the address is the broadcast address. + * + * Please note: addr must be aligned to u16. */ static inline bool is_broadcast_ether_addr(const u8 *addr) { - return (addr[0] & addr[1] & addr[2] & addr[3] & addr[4] & addr[5]) == 0xff; + return (*(const u16 *)(addr + 0) & + *(const u16 *)(addr + 2) & + *(const u16 *)(addr + 4)) == 0xffff; } /** @@ -135,6 +154,8 @@ static inline bool is_unicast_ether_addr(const u8 *addr) * a multicast address, and is not FF:FF:FF:FF:FF:FF. * * Return true if the address is valid. + * + * Please note: addr must be aligned to u16. */ static inline bool is_valid_ether_addr(const u8 *addr) { -- cgit v1.2.3 From 286ab723d4b83d37deb4017008ef1444a95cfb0d Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Tue, 14 Jan 2014 15:18:47 -0800 Subject: etherdevice: Use ether_addr_copy to copy an Ethernet address Some systems can use the normally known u16 alignment of Ethernet addresses to save some code/text bytes and cycles. This does not change currently emitted code on x86 by gcc 4.8. Signed-off-by: Joe Perches Signed-off-by: David S. Miller --- include/linux/etherdevice.h | 24 +++++++++++++++++++++++- 1 file changed, 23 insertions(+), 1 deletion(-) (limited to 'include/linux/etherdevice.h') diff --git a/include/linux/etherdevice.h b/include/linux/etherdevice.h index f344ac04f858..9c5529dc6d07 100644 --- a/include/linux/etherdevice.h +++ b/include/linux/etherdevice.h @@ -217,6 +217,28 @@ static inline void eth_hw_addr_random(struct net_device *dev) eth_random_addr(dev->dev_addr); } +/** + * ether_addr_copy - Copy an Ethernet address + * @dst: Pointer to a six-byte array Ethernet address destination + * @src: Pointer to a six-byte array Ethernet address source + * + * Please note: dst & src must both be aligned to u16. + */ +static inline void ether_addr_copy(u8 *dst, const u8 *src) +{ +#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) + *(u32 *)dst = *(const u32 *)src; + *(u16 *)(dst + 4) = *(const u16 *)(src + 4); +#else + u16 *a = (u16 *)dst; + const u16 *b = (const u16 *)src; + + a[0] = b[0]; + a[1] = b[1]; + a[2] = b[2]; +#endif +} + /** * eth_hw_addr_inherit - Copy dev_addr from another net_device * @dst: pointer to net_device to copy dev_addr to @@ -229,7 +251,7 @@ static inline void eth_hw_addr_inherit(struct net_device *dst, struct net_device *src) { dst->addr_assign_type = src->addr_assign_type; - memcpy(dst->dev_addr, src->dev_addr, ETH_ALEN); + ether_addr_copy(dst->dev_addr, src->dev_addr); } /** -- cgit v1.2.3